diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21978 @@
+{
+    "best_metric": null,
+    "best_model_checkpoint": null,
+    "epoch": 1.0,
+    "eval_steps": 500,
+    "global_step": 31359,
+    "is_hyper_param_search": false,
+    "is_local_process_zero": true,
+    "is_world_process_zero": true,
+    "log_history": [
+      {
+        "epoch": 0.0003188877196339169,
+        "grad_norm": 2.7591781616210938,
+        "learning_rate": 1.5944385981695843e-05,
+        "loss": 9.7119,
+        "step": 10
+      },
+      {
+        "epoch": 0.0006377754392678338,
+        "grad_norm": 2.3757214546203613,
+        "learning_rate": 3.1888771963391685e-05,
+        "loss": 8.9972,
+        "step": 20
+      },
+      {
+        "epoch": 0.0009566631589017507,
+        "grad_norm": 1.9443352222442627,
+        "learning_rate": 4.7833157945087525e-05,
+        "loss": 8.234,
+        "step": 30
+      },
+      {
+        "epoch": 0.0012755508785356677,
+        "grad_norm": 0.9958082437515259,
+        "learning_rate": 6.377754392678337e-05,
+        "loss": 7.4942,
+        "step": 40
+      },
+      {
+        "epoch": 0.0015944385981695845,
+        "grad_norm": 0.7051369547843933,
+        "learning_rate": 7.972192990847922e-05,
+        "loss": 6.9288,
+        "step": 50
+      },
+      {
+        "epoch": 0.0019133263178035013,
+        "grad_norm": 0.5652105808258057,
+        "learning_rate": 9.566631589017505e-05,
+        "loss": 6.5521,
+        "step": 60
+      },
+      {
+        "epoch": 0.0022322140374374183,
+        "grad_norm": 0.8407195806503296,
+        "learning_rate": 0.00011161070187187091,
+        "loss": 6.2748,
+        "step": 70
+      },
+      {
+        "epoch": 0.0025511017570713354,
+        "grad_norm": 0.7927109003067017,
+        "learning_rate": 0.00012755508785356674,
+        "loss": 6.0919,
+        "step": 80
+      },
+      {
+        "epoch": 0.002869989476705252,
+        "grad_norm": 0.647208571434021,
+        "learning_rate": 0.0001434994738352626,
+        "loss": 5.9766,
+        "step": 90
+      },
+      {
+        "epoch": 0.003188877196339169,
+        "grad_norm": 0.6491755843162537,
+        "learning_rate": 0.00015944385981695843,
+        "loss": 5.8506,
+        "step": 100
+      },
+      {
+        "epoch": 0.003507764915973086,
+        "grad_norm": 0.9509009718894958,
+        "learning_rate": 0.00017538824579865428,
+        "loss": 5.7138,
+        "step": 110
+      },
+      {
+        "epoch": 0.0038266526356070026,
+        "grad_norm": 0.7599376440048218,
+        "learning_rate": 0.0001913326317803501,
+        "loss": 5.5894,
+        "step": 120
+      },
+      {
+        "epoch": 0.00414554035524092,
+        "grad_norm": 0.8345359563827515,
+        "learning_rate": 0.00020727701776204597,
+        "loss": 5.4928,
+        "step": 130
+      },
+      {
+        "epoch": 0.004464428074874837,
+        "grad_norm": 0.9486031532287598,
+        "learning_rate": 0.00022322140374374182,
+        "loss": 5.3421,
+        "step": 140
+      },
+      {
+        "epoch": 0.004783315794508754,
+        "grad_norm": 0.7641623616218567,
+        "learning_rate": 0.00023916578972543764,
+        "loss": 5.2179,
+        "step": 150
+      },
+      {
+        "epoch": 0.005102203514142671,
+        "grad_norm": 0.8762153387069702,
+        "learning_rate": 0.0002551101757071335,
+        "loss": 5.0897,
+        "step": 160
+      },
+      {
+        "epoch": 0.005421091233776587,
+        "grad_norm": 0.9552673697471619,
+        "learning_rate": 0.00027105456168882936,
+        "loss": 5.0136,
+        "step": 170
+      },
+      {
+        "epoch": 0.005739978953410504,
+        "grad_norm": 0.8702957034111023,
+        "learning_rate": 0.0002869989476705252,
+        "loss": 4.8847,
+        "step": 180
+      },
+      {
+        "epoch": 0.006058866673044421,
+        "grad_norm": 1.1011112928390503,
+        "learning_rate": 0.00030294333365222105,
+        "loss": 4.793,
+        "step": 190
+      },
+      {
+        "epoch": 0.006377754392678338,
+        "grad_norm": 1.000261664390564,
+        "learning_rate": 0.00031888771963391687,
+        "loss": 4.6767,
+        "step": 200
+      },
+      {
+        "epoch": 0.006696642112312255,
+        "grad_norm": 0.9304866194725037,
+        "learning_rate": 0.0003348321056156127,
+        "loss": 4.5988,
+        "step": 210
+      },
+      {
+        "epoch": 0.007015529831946172,
+        "grad_norm": 1.0546783208847046,
+        "learning_rate": 0.00035077649159730856,
+        "loss": 4.4907,
+        "step": 220
+      },
+      {
+        "epoch": 0.007334417551580088,
+        "grad_norm": 0.9612847566604614,
+        "learning_rate": 0.00036672087757900443,
+        "loss": 4.4107,
+        "step": 230
+      },
+      {
+        "epoch": 0.007653305271214005,
+        "grad_norm": 0.9102572202682495,
+        "learning_rate": 0.0003826652635607002,
+        "loss": 4.3265,
+        "step": 240
+      },
+      {
+        "epoch": 0.007972192990847922,
+        "grad_norm": 0.9532923102378845,
+        "learning_rate": 0.00039860964954239607,
+        "loss": 4.2816,
+        "step": 250
+      },
+      {
+        "epoch": 0.00829108071048184,
+        "grad_norm": 1.0799651145935059,
+        "learning_rate": 0.00041455403552409194,
+        "loss": 4.1925,
+        "step": 260
+      },
+      {
+        "epoch": 0.008609968430115756,
+        "grad_norm": 0.8453513979911804,
+        "learning_rate": 0.0004304984215057878,
+        "loss": 4.1438,
+        "step": 270
+      },
+      {
+        "epoch": 0.008928856149749673,
+        "grad_norm": 0.9012154340744019,
+        "learning_rate": 0.00044644280748748364,
+        "loss": 4.0625,
+        "step": 280
+      },
+      {
+        "epoch": 0.00924774386938359,
+        "grad_norm": 0.9568570852279663,
+        "learning_rate": 0.0004623871934691794,
+        "loss": 4.0183,
+        "step": 290
+      },
+      {
+        "epoch": 0.009566631589017507,
+        "grad_norm": 0.9893571734428406,
+        "learning_rate": 0.0004783315794508753,
+        "loss": 3.9543,
+        "step": 300
+      },
+      {
+        "epoch": 0.009885519308651424,
+        "grad_norm": 1.001813530921936,
+        "learning_rate": 0.0004942759654325711,
+        "loss": 3.9078,
+        "step": 310
+      },
+      {
+        "epoch": 0.010204407028285341,
+        "grad_norm": 0.8830215930938721,
+        "learning_rate": 0.0005,
+        "loss": 3.8586,
+        "step": 320
+      },
+      {
+        "epoch": 0.010523294747919257,
+        "grad_norm": 1.0455135107040405,
+        "learning_rate": 0.0005,
+        "loss": 3.79,
+        "step": 330
+      },
+      {
+        "epoch": 0.010842182467553174,
+        "grad_norm": 0.891879677772522,
+        "learning_rate": 0.0005,
+        "loss": 3.7631,
+        "step": 340
+      },
+      {
+        "epoch": 0.01116107018718709,
+        "grad_norm": 0.8672872185707092,
+        "learning_rate": 0.0005,
+        "loss": 3.696,
+        "step": 350
+      },
+      {
+        "epoch": 0.011479957906821008,
+        "grad_norm": 0.9735280275344849,
+        "learning_rate": 0.0005,
+        "loss": 3.6841,
+        "step": 360
+      },
+      {
+        "epoch": 0.011798845626454925,
+        "grad_norm": 0.8796781301498413,
+        "learning_rate": 0.0005,
+        "loss": 3.6076,
+        "step": 370
+      },
+      {
+        "epoch": 0.012117733346088842,
+        "grad_norm": 0.7970229387283325,
+        "learning_rate": 0.0005,
+        "loss": 3.5803,
+        "step": 380
+      },
+      {
+        "epoch": 0.012436621065722759,
+        "grad_norm": 0.8341770768165588,
+        "learning_rate": 0.0005,
+        "loss": 3.5506,
+        "step": 390
+      },
+      {
+        "epoch": 0.012755508785356676,
+        "grad_norm": 1.0613290071487427,
+        "learning_rate": 0.0005,
+        "loss": 3.5163,
+        "step": 400
+      },
+      {
+        "epoch": 0.013074396504990593,
+        "grad_norm": 0.8499947786331177,
+        "learning_rate": 0.0005,
+        "loss": 3.4579,
+        "step": 410
+      },
+      {
+        "epoch": 0.01339328422462451,
+        "grad_norm": 0.868526816368103,
+        "learning_rate": 0.0005,
+        "loss": 3.4293,
+        "step": 420
+      },
+      {
+        "epoch": 0.013712171944258427,
+        "grad_norm": 1.0044291019439697,
+        "learning_rate": 0.0005,
+        "loss": 3.3861,
+        "step": 430
+      },
+      {
+        "epoch": 0.014031059663892344,
+        "grad_norm": 1.0256859064102173,
+        "learning_rate": 0.0005,
+        "loss": 3.3423,
+        "step": 440
+      },
+      {
+        "epoch": 0.014349947383526261,
+        "grad_norm": 1.0345772504806519,
+        "learning_rate": 0.0005,
+        "loss": 3.2853,
+        "step": 450
+      },
+      {
+        "epoch": 0.014668835103160176,
+        "grad_norm": 1.175614833831787,
+        "learning_rate": 0.0005,
+        "loss": 3.2852,
+        "step": 460
+      },
+      {
+        "epoch": 0.014987722822794093,
+        "grad_norm": 0.8701066970825195,
+        "learning_rate": 0.0005,
+        "loss": 3.2546,
+        "step": 470
+      },
+      {
+        "epoch": 0.01530661054242801,
+        "grad_norm": 0.8358347415924072,
+        "learning_rate": 0.0005,
+        "loss": 3.2179,
+        "step": 480
+      },
+      {
+        "epoch": 0.01562549826206193,
+        "grad_norm": 0.8820248246192932,
+        "learning_rate": 0.0005,
+        "loss": 3.1517,
+        "step": 490
+      },
+      {
+        "epoch": 0.015944385981695845,
+        "grad_norm": 0.9414588809013367,
+        "learning_rate": 0.0005,
+        "loss": 3.1477,
+        "step": 500
+      },
+      {
+        "epoch": 0.016263273701329763,
+        "grad_norm": 1.0134668350219727,
+        "learning_rate": 0.0005,
+        "loss": 3.1032,
+        "step": 510
+      },
+      {
+        "epoch": 0.01658216142096368,
+        "grad_norm": 1.2762371301651,
+        "learning_rate": 0.0005,
+        "loss": 3.077,
+        "step": 520
+      },
+      {
+        "epoch": 0.016901049140597594,
+        "grad_norm": 0.9885225892066956,
+        "learning_rate": 0.0005,
+        "loss": 3.0795,
+        "step": 530
+      },
+      {
+        "epoch": 0.017219936860231513,
+        "grad_norm": 1.1082136631011963,
+        "learning_rate": 0.0005,
+        "loss": 3.0401,
+        "step": 540
+      },
+      {
+        "epoch": 0.017538824579865428,
+        "grad_norm": 0.8814631700515747,
+        "learning_rate": 0.0005,
+        "loss": 2.9775,
+        "step": 550
+      },
+      {
+        "epoch": 0.017857712299499347,
+        "grad_norm": 0.8388587832450867,
+        "learning_rate": 0.0005,
+        "loss": 2.9665,
+        "step": 560
+      },
+      {
+        "epoch": 0.018176600019133262,
+        "grad_norm": 0.8874592781066895,
+        "learning_rate": 0.0005,
+        "loss": 2.9424,
+        "step": 570
+      },
+      {
+        "epoch": 0.01849548773876718,
+        "grad_norm": 0.8496912121772766,
+        "learning_rate": 0.0005,
+        "loss": 2.9261,
+        "step": 580
+      },
+      {
+        "epoch": 0.018814375458401096,
+        "grad_norm": 0.9394707679748535,
+        "learning_rate": 0.0005,
+        "loss": 2.8692,
+        "step": 590
+      },
+      {
+        "epoch": 0.019133263178035015,
+        "grad_norm": 0.9260649681091309,
+        "learning_rate": 0.0005,
+        "loss": 2.8651,
+        "step": 600
+      },
+      {
+        "epoch": 0.01945215089766893,
+        "grad_norm": 0.9702092409133911,
+        "learning_rate": 0.0005,
+        "loss": 2.8374,
+        "step": 610
+      },
+      {
+        "epoch": 0.01977103861730285,
+        "grad_norm": 0.9171802401542664,
+        "learning_rate": 0.0005,
+        "loss": 2.7758,
+        "step": 620
+      },
+      {
+        "epoch": 0.020089926336936764,
+        "grad_norm": 0.906156063079834,
+        "learning_rate": 0.0005,
+        "loss": 2.7732,
+        "step": 630
+      },
+      {
+        "epoch": 0.020408814056570683,
+        "grad_norm": 0.8847390413284302,
+        "learning_rate": 0.0005,
+        "loss": 2.7732,
+        "step": 640
+      },
+      {
+        "epoch": 0.0207277017762046,
+        "grad_norm": 0.8880306482315063,
+        "learning_rate": 0.0005,
+        "loss": 2.7348,
+        "step": 650
+      },
+      {
+        "epoch": 0.021046589495838514,
+        "grad_norm": 0.9086488485336304,
+        "learning_rate": 0.0005,
+        "loss": 2.7275,
+        "step": 660
+      },
+      {
+        "epoch": 0.021365477215472432,
+        "grad_norm": 0.8930760622024536,
+        "learning_rate": 0.0005,
+        "loss": 2.7,
+        "step": 670
+      },
+      {
+        "epoch": 0.021684364935106348,
+        "grad_norm": 0.8453980088233948,
+        "learning_rate": 0.0005,
+        "loss": 2.6833,
+        "step": 680
+      },
+      {
+        "epoch": 0.022003252654740266,
+        "grad_norm": 0.9567738771438599,
+        "learning_rate": 0.0005,
+        "loss": 2.6641,
+        "step": 690
+      },
+      {
+        "epoch": 0.02232214037437418,
+        "grad_norm": 0.8755207657814026,
+        "learning_rate": 0.0005,
+        "loss": 2.6219,
+        "step": 700
+      },
+      {
+        "epoch": 0.0226410280940081,
+        "grad_norm": 0.881864070892334,
+        "learning_rate": 0.0005,
+        "loss": 2.6393,
+        "step": 710
+      },
+      {
+        "epoch": 0.022959915813642016,
+        "grad_norm": 0.940329909324646,
+        "learning_rate": 0.0005,
+        "loss": 2.5864,
+        "step": 720
+      },
+      {
+        "epoch": 0.023278803533275935,
+        "grad_norm": 1.046134352684021,
+        "learning_rate": 0.0005,
+        "loss": 2.5888,
+        "step": 730
+      },
+      {
+        "epoch": 0.02359769125290985,
+        "grad_norm": 0.8772185444831848,
+        "learning_rate": 0.0005,
+        "loss": 2.5502,
+        "step": 740
+      },
+      {
+        "epoch": 0.02391657897254377,
+        "grad_norm": 0.865439236164093,
+        "learning_rate": 0.0005,
+        "loss": 2.5728,
+        "step": 750
+      },
+      {
+        "epoch": 0.024235466692177684,
+        "grad_norm": 0.8309611678123474,
+        "learning_rate": 0.0005,
+        "loss": 2.5484,
+        "step": 760
+      },
+      {
+        "epoch": 0.024554354411811603,
+        "grad_norm": 0.9572917819023132,
+        "learning_rate": 0.0005,
+        "loss": 2.5551,
+        "step": 770
+      },
+      {
+        "epoch": 0.024873242131445518,
+        "grad_norm": 0.8789076805114746,
+        "learning_rate": 0.0005,
+        "loss": 2.4962,
+        "step": 780
+      },
+      {
+        "epoch": 0.025192129851079433,
+        "grad_norm": 0.8763116002082825,
+        "learning_rate": 0.0005,
+        "loss": 2.5085,
+        "step": 790
+      },
+      {
+        "epoch": 0.025511017570713352,
+        "grad_norm": 0.8708914518356323,
+        "learning_rate": 0.0005,
+        "loss": 2.5152,
+        "step": 800
+      },
+      {
+        "epoch": 0.025829905290347267,
+        "grad_norm": 0.9939478039741516,
+        "learning_rate": 0.0005,
+        "loss": 2.4608,
+        "step": 810
+      },
+      {
+        "epoch": 0.026148793009981186,
+        "grad_norm": 0.9306032061576843,
+        "learning_rate": 0.0005,
+        "loss": 2.4971,
+        "step": 820
+      },
+      {
+        "epoch": 0.0264676807296151,
+        "grad_norm": 0.8735993504524231,
+        "learning_rate": 0.0005,
+        "loss": 2.441,
+        "step": 830
+      },
+      {
+        "epoch": 0.02678656844924902,
+        "grad_norm": 0.8575237393379211,
+        "learning_rate": 0.0005,
+        "loss": 2.4408,
+        "step": 840
+      },
+      {
+        "epoch": 0.027105456168882935,
+        "grad_norm": 0.8133852481842041,
+        "learning_rate": 0.0005,
+        "loss": 2.4189,
+        "step": 850
+      },
+      {
+        "epoch": 0.027424343888516854,
+        "grad_norm": 0.9014040231704712,
+        "learning_rate": 0.0005,
+        "loss": 2.4153,
+        "step": 860
+      },
+      {
+        "epoch": 0.02774323160815077,
+        "grad_norm": 0.801937460899353,
+        "learning_rate": 0.0005,
+        "loss": 2.3815,
+        "step": 870
+      },
+      {
+        "epoch": 0.028062119327784688,
+        "grad_norm": 0.8674513101577759,
+        "learning_rate": 0.0005,
+        "loss": 2.3742,
+        "step": 880
+      },
+      {
+        "epoch": 0.028381007047418604,
+        "grad_norm": 0.8789620399475098,
+        "learning_rate": 0.0005,
+        "loss": 2.3523,
+        "step": 890
+      },
+      {
+        "epoch": 0.028699894767052522,
+        "grad_norm": 1.5122326612472534,
+        "learning_rate": 0.0005,
+        "loss": 2.3561,
+        "step": 900
+      },
+      {
+        "epoch": 0.029018782486686438,
+        "grad_norm": 0.9573362469673157,
+        "learning_rate": 0.0005,
+        "loss": 2.3151,
+        "step": 910
+      },
+      {
+        "epoch": 0.029337670206320353,
+        "grad_norm": 0.9171894788742065,
+        "learning_rate": 0.0005,
+        "loss": 2.321,
+        "step": 920
+      },
+      {
+        "epoch": 0.02965655792595427,
+        "grad_norm": 0.9501611590385437,
+        "learning_rate": 0.0005,
+        "loss": 2.3244,
+        "step": 930
+      },
+      {
+        "epoch": 0.029975445645588187,
+        "grad_norm": 0.8088300228118896,
+        "learning_rate": 0.0005,
+        "loss": 2.3178,
+        "step": 940
+      },
+      {
+        "epoch": 0.030294333365222106,
+        "grad_norm": 0.8260691165924072,
+        "learning_rate": 0.0005,
+        "loss": 2.2892,
+        "step": 950
+      },
+      {
+        "epoch": 0.03061322108485602,
+        "grad_norm": 0.8876654505729675,
+        "learning_rate": 0.0005,
+        "loss": 2.2933,
+        "step": 960
+      },
+      {
+        "epoch": 0.03093210880448994,
+        "grad_norm": 0.8482714295387268,
+        "learning_rate": 0.0005,
+        "loss": 2.2748,
+        "step": 970
+      },
+      {
+        "epoch": 0.03125099652412386,
+        "grad_norm": 0.8355926275253296,
+        "learning_rate": 0.0005,
+        "loss": 2.286,
+        "step": 980
+      },
+      {
+        "epoch": 0.03156988424375777,
+        "grad_norm": 0.8398257493972778,
+        "learning_rate": 0.0005,
+        "loss": 2.2733,
+        "step": 990
+      },
+      {
+        "epoch": 0.03188877196339169,
+        "grad_norm": 0.9260374903678894,
+        "learning_rate": 0.0005,
+        "loss": 2.251,
+        "step": 1000
+      },
+      {
+        "epoch": 0.03220765968302561,
+        "grad_norm": 0.8768055438995361,
+        "learning_rate": 0.0005,
+        "loss": 2.2422,
+        "step": 1010
+      },
+      {
+        "epoch": 0.03252654740265953,
+        "grad_norm": 0.9418902397155762,
+        "learning_rate": 0.0005,
+        "loss": 2.254,
+        "step": 1020
+      },
+      {
+        "epoch": 0.03284543512229344,
+        "grad_norm": 0.9414481520652771,
+        "learning_rate": 0.0005,
+        "loss": 2.221,
+        "step": 1030
+      },
+      {
+        "epoch": 0.03316432284192736,
+        "grad_norm": 1.0012530088424683,
+        "learning_rate": 0.0005,
+        "loss": 2.1724,
+        "step": 1040
+      },
+      {
+        "epoch": 0.033483210561561276,
+        "grad_norm": 1.026558756828308,
+        "learning_rate": 0.0005,
+        "loss": 2.2094,
+        "step": 1050
+      },
+      {
+        "epoch": 0.03380209828119519,
+        "grad_norm": 0.8032659292221069,
+        "learning_rate": 0.0005,
+        "loss": 2.1611,
+        "step": 1060
+      },
+      {
+        "epoch": 0.03412098600082911,
+        "grad_norm": 0.8239543437957764,
+        "learning_rate": 0.0005,
+        "loss": 2.1877,
+        "step": 1070
+      },
+      {
+        "epoch": 0.034439873720463025,
+        "grad_norm": 0.8418667316436768,
+        "learning_rate": 0.0005,
+        "loss": 2.18,
+        "step": 1080
+      },
+      {
+        "epoch": 0.034758761440096944,
+        "grad_norm": 0.8027506470680237,
+        "learning_rate": 0.0005,
+        "loss": 2.177,
+        "step": 1090
+      },
+      {
+        "epoch": 0.035077649159730856,
+        "grad_norm": 0.7979521751403809,
+        "learning_rate": 0.0005,
+        "loss": 2.1828,
+        "step": 1100
+      },
+      {
+        "epoch": 0.035396536879364775,
+        "grad_norm": 0.8028553128242493,
+        "learning_rate": 0.0005,
+        "loss": 2.1322,
+        "step": 1110
+      },
+      {
+        "epoch": 0.035715424598998693,
+        "grad_norm": 0.7876245379447937,
+        "learning_rate": 0.0005,
+        "loss": 2.1393,
+        "step": 1120
+      },
+      {
+        "epoch": 0.03603431231863261,
+        "grad_norm": 0.8016355633735657,
+        "learning_rate": 0.0005,
+        "loss": 2.1355,
+        "step": 1130
+      },
+      {
+        "epoch": 0.036353200038266524,
+        "grad_norm": 0.8408504128456116,
+        "learning_rate": 0.0005,
+        "loss": 2.1464,
+        "step": 1140
+      },
+      {
+        "epoch": 0.03667208775790044,
+        "grad_norm": 0.9178126454353333,
+        "learning_rate": 0.0005,
+        "loss": 2.135,
+        "step": 1150
+      },
+      {
+        "epoch": 0.03699097547753436,
+        "grad_norm": 0.965160071849823,
+        "learning_rate": 0.0005,
+        "loss": 2.1258,
+        "step": 1160
+      },
+      {
+        "epoch": 0.03730986319716828,
+        "grad_norm": 0.7981363534927368,
+        "learning_rate": 0.0005,
+        "loss": 2.0838,
+        "step": 1170
+      },
+      {
+        "epoch": 0.03762875091680219,
+        "grad_norm": 0.894565999507904,
+        "learning_rate": 0.0005,
+        "loss": 2.0513,
+        "step": 1180
+      },
+      {
+        "epoch": 0.03794763863643611,
+        "grad_norm": 0.8362721800804138,
+        "learning_rate": 0.0005,
+        "loss": 2.0724,
+        "step": 1190
+      },
+      {
+        "epoch": 0.03826652635607003,
+        "grad_norm": 0.819324254989624,
+        "learning_rate": 0.0005,
+        "loss": 2.0788,
+        "step": 1200
+      },
+      {
+        "epoch": 0.03858541407570394,
+        "grad_norm": 0.7400639057159424,
+        "learning_rate": 0.0005,
+        "loss": 2.0486,
+        "step": 1210
+      },
+      {
+        "epoch": 0.03890430179533786,
+        "grad_norm": 0.9596002697944641,
+        "learning_rate": 0.0005,
+        "loss": 2.0635,
+        "step": 1220
+      },
+      {
+        "epoch": 0.03922318951497178,
+        "grad_norm": 0.7574199438095093,
+        "learning_rate": 0.0005,
+        "loss": 2.0509,
+        "step": 1230
+      },
+      {
+        "epoch": 0.0395420772346057,
+        "grad_norm": 0.7776308059692383,
+        "learning_rate": 0.0005,
+        "loss": 2.057,
+        "step": 1240
+      },
+      {
+        "epoch": 0.03986096495423961,
+        "grad_norm": 0.7979622483253479,
+        "learning_rate": 0.0005,
+        "loss": 2.0323,
+        "step": 1250
+      },
+      {
+        "epoch": 0.04017985267387353,
+        "grad_norm": 0.777959942817688,
+        "learning_rate": 0.0005,
+        "loss": 2.0241,
+        "step": 1260
+      },
+      {
+        "epoch": 0.04049874039350745,
+        "grad_norm": 0.7778024673461914,
+        "learning_rate": 0.0005,
+        "loss": 2.0401,
+        "step": 1270
+      },
+      {
+        "epoch": 0.040817628113141366,
+        "grad_norm": 0.7771192193031311,
+        "learning_rate": 0.0005,
+        "loss": 2.0219,
+        "step": 1280
+      },
+      {
+        "epoch": 0.04113651583277528,
+        "grad_norm": 0.7752227187156677,
+        "learning_rate": 0.0005,
+        "loss": 2.018,
+        "step": 1290
+      },
+      {
+        "epoch": 0.0414554035524092,
+        "grad_norm": 0.7638623118400574,
+        "learning_rate": 0.0005,
+        "loss": 1.986,
+        "step": 1300
+      },
+      {
+        "epoch": 0.041774291272043115,
+        "grad_norm": 0.7941219806671143,
+        "learning_rate": 0.0005,
+        "loss": 1.9811,
+        "step": 1310
+      },
+      {
+        "epoch": 0.04209317899167703,
+        "grad_norm": 0.7345162630081177,
+        "learning_rate": 0.0005,
+        "loss": 2.0212,
+        "step": 1320
+      },
+      {
+        "epoch": 0.042412066711310946,
+        "grad_norm": 0.8033922910690308,
+        "learning_rate": 0.0005,
+        "loss": 1.9976,
+        "step": 1330
+      },
+      {
+        "epoch": 0.042730954430944865,
+        "grad_norm": 0.7183955311775208,
+        "learning_rate": 0.0005,
+        "loss": 1.9659,
+        "step": 1340
+      },
+      {
+        "epoch": 0.04304984215057878,
+        "grad_norm": 0.7607952356338501,
+        "learning_rate": 0.0005,
+        "loss": 1.9921,
+        "step": 1350
+      },
+      {
+        "epoch": 0.043368729870212695,
+        "grad_norm": 0.8439708352088928,
+        "learning_rate": 0.0005,
+        "loss": 1.9723,
+        "step": 1360
+      },
+      {
+        "epoch": 0.043687617589846614,
+        "grad_norm": 0.7470654249191284,
+        "learning_rate": 0.0005,
+        "loss": 1.9738,
+        "step": 1370
+      },
+      {
+        "epoch": 0.04400650530948053,
+        "grad_norm": 0.7575402855873108,
+        "learning_rate": 0.0005,
+        "loss": 1.9791,
+        "step": 1380
+      },
+      {
+        "epoch": 0.04432539302911445,
+        "grad_norm": 0.7686700224876404,
+        "learning_rate": 0.0005,
+        "loss": 1.9262,
+        "step": 1390
+      },
+      {
+        "epoch": 0.04464428074874836,
+        "grad_norm": 0.956557035446167,
+        "learning_rate": 0.0005,
+        "loss": 1.9607,
+        "step": 1400
+      },
+      {
+        "epoch": 0.04496316846838228,
+        "grad_norm": 0.7784798741340637,
+        "learning_rate": 0.0005,
+        "loss": 1.9567,
+        "step": 1410
+      },
+      {
+        "epoch": 0.0452820561880162,
+        "grad_norm": 0.7413488626480103,
+        "learning_rate": 0.0005,
+        "loss": 1.9216,
+        "step": 1420
+      },
+      {
+        "epoch": 0.04560094390765012,
+        "grad_norm": 0.7566946148872375,
+        "learning_rate": 0.0005,
+        "loss": 1.9394,
+        "step": 1430
+      },
+      {
+        "epoch": 0.04591983162728403,
+        "grad_norm": 0.7943182587623596,
+        "learning_rate": 0.0005,
+        "loss": 1.914,
+        "step": 1440
+      },
+      {
+        "epoch": 0.04623871934691795,
+        "grad_norm": 0.8151775598526001,
+        "learning_rate": 0.0005,
+        "loss": 1.9598,
+        "step": 1450
+      },
+      {
+        "epoch": 0.04655760706655187,
+        "grad_norm": 0.7676635384559631,
+        "learning_rate": 0.0005,
+        "loss": 1.9039,
+        "step": 1460
+      },
+      {
+        "epoch": 0.04687649478618578,
+        "grad_norm": 0.7939006686210632,
+        "learning_rate": 0.0005,
+        "loss": 1.9082,
+        "step": 1470
+      },
+      {
+        "epoch": 0.0471953825058197,
+        "grad_norm": 0.7992001175880432,
+        "learning_rate": 0.0005,
+        "loss": 1.9097,
+        "step": 1480
+      },
+      {
+        "epoch": 0.04751427022545362,
+        "grad_norm": 0.8140571713447571,
+        "learning_rate": 0.0005,
+        "loss": 1.9024,
+        "step": 1490
+      },
+      {
+        "epoch": 0.04783315794508754,
+        "grad_norm": 0.717022180557251,
+        "learning_rate": 0.0005,
+        "loss": 1.8898,
+        "step": 1500
+      },
+      {
+        "epoch": 0.04815204566472145,
+        "grad_norm": 0.8296144604682922,
+        "learning_rate": 0.0005,
+        "loss": 1.8656,
+        "step": 1510
+      },
+      {
+        "epoch": 0.04847093338435537,
+        "grad_norm": 0.759627640247345,
+        "learning_rate": 0.0005,
+        "loss": 1.8984,
+        "step": 1520
+      },
+      {
+        "epoch": 0.048789821103989287,
+        "grad_norm": 0.6937538385391235,
+        "learning_rate": 0.0005,
+        "loss": 1.9031,
+        "step": 1530
+      },
+      {
+        "epoch": 0.049108708823623205,
+        "grad_norm": 0.8209205269813538,
+        "learning_rate": 0.0005,
+        "loss": 1.8583,
+        "step": 1540
+      },
+      {
+        "epoch": 0.04942759654325712,
+        "grad_norm": 0.7268592715263367,
+        "learning_rate": 0.0005,
+        "loss": 1.8808,
+        "step": 1550
+      },
+      {
+        "epoch": 0.049746484262891036,
+        "grad_norm": 0.7960999608039856,
+        "learning_rate": 0.0005,
+        "loss": 1.8662,
+        "step": 1560
+      },
+      {
+        "epoch": 0.050065371982524955,
+        "grad_norm": 0.7445297241210938,
+        "learning_rate": 0.0005,
+        "loss": 1.865,
+        "step": 1570
+      },
+      {
+        "epoch": 0.050384259702158866,
+        "grad_norm": 0.7228101491928101,
+        "learning_rate": 0.0005,
+        "loss": 1.842,
+        "step": 1580
+      },
+      {
+        "epoch": 0.050703147421792785,
+        "grad_norm": 0.7101908326148987,
+        "learning_rate": 0.0005,
+        "loss": 1.8432,
+        "step": 1590
+      },
+      {
+        "epoch": 0.051022035141426704,
+        "grad_norm": 0.817542552947998,
+        "learning_rate": 0.0005,
+        "loss": 1.8422,
+        "step": 1600
+      },
+      {
+        "epoch": 0.05134092286106062,
+        "grad_norm": 0.7786903977394104,
+        "learning_rate": 0.0005,
+        "loss": 1.8447,
+        "step": 1610
+      },
+      {
+        "epoch": 0.051659810580694535,
+        "grad_norm": 0.8136606216430664,
+        "learning_rate": 0.0005,
+        "loss": 1.8723,
+        "step": 1620
+      },
+      {
+        "epoch": 0.05197869830032845,
+        "grad_norm": 0.74000483751297,
+        "learning_rate": 0.0005,
+        "loss": 1.8278,
+        "step": 1630
+      },
+      {
+        "epoch": 0.05229758601996237,
+        "grad_norm": 0.6954797506332397,
+        "learning_rate": 0.0005,
+        "loss": 1.8355,
+        "step": 1640
+      },
+      {
+        "epoch": 0.05261647373959629,
+        "grad_norm": 0.6831514835357666,
+        "learning_rate": 0.0005,
+        "loss": 1.8538,
+        "step": 1650
+      },
+      {
+        "epoch": 0.0529353614592302,
+        "grad_norm": 0.9402697086334229,
+        "learning_rate": 0.0005,
+        "loss": 1.8373,
+        "step": 1660
+      },
+      {
+        "epoch": 0.05325424917886412,
+        "grad_norm": 0.6758236289024353,
+        "learning_rate": 0.0005,
+        "loss": 1.8319,
+        "step": 1670
+      },
+      {
+        "epoch": 0.05357313689849804,
+        "grad_norm": 0.763216495513916,
+        "learning_rate": 0.0005,
+        "loss": 1.7892,
+        "step": 1680
+      },
+      {
+        "epoch": 0.05389202461813196,
+        "grad_norm": 0.7123900055885315,
+        "learning_rate": 0.0005,
+        "loss": 1.8057,
+        "step": 1690
+      },
+      {
+        "epoch": 0.05421091233776587,
+        "grad_norm": 0.7843760848045349,
+        "learning_rate": 0.0005,
+        "loss": 1.8182,
+        "step": 1700
+      },
+      {
+        "epoch": 0.05452980005739979,
+        "grad_norm": 0.7198774218559265,
+        "learning_rate": 0.0005,
+        "loss": 1.8059,
+        "step": 1710
+      },
+      {
+        "epoch": 0.05484868777703371,
+        "grad_norm": 0.6590190529823303,
+        "learning_rate": 0.0005,
+        "loss": 1.8036,
+        "step": 1720
+      },
+      {
+        "epoch": 0.05516757549666762,
+        "grad_norm": 0.7103943228721619,
+        "learning_rate": 0.0005,
+        "loss": 1.8009,
+        "step": 1730
+      },
+      {
+        "epoch": 0.05548646321630154,
+        "grad_norm": 0.973406970500946,
+        "learning_rate": 0.0005,
+        "loss": 1.8117,
+        "step": 1740
+      },
+      {
+        "epoch": 0.05580535093593546,
+        "grad_norm": 0.7230905294418335,
+        "learning_rate": 0.0005,
+        "loss": 1.795,
+        "step": 1750
+      },
+      {
+        "epoch": 0.056124238655569376,
+        "grad_norm": 0.6838631629943848,
+        "learning_rate": 0.0005,
+        "loss": 1.7943,
+        "step": 1760
+      },
+      {
+        "epoch": 0.05644312637520329,
+        "grad_norm": 0.69809889793396,
+        "learning_rate": 0.0005,
+        "loss": 1.7839,
+        "step": 1770
+      },
+      {
+        "epoch": 0.05676201409483721,
+        "grad_norm": 0.6809301376342773,
+        "learning_rate": 0.0005,
+        "loss": 1.8128,
+        "step": 1780
+      },
+      {
+        "epoch": 0.057080901814471126,
+        "grad_norm": 0.7151408791542053,
+        "learning_rate": 0.0005,
+        "loss": 1.776,
+        "step": 1790
+      },
+      {
+        "epoch": 0.057399789534105045,
+        "grad_norm": 0.7247176766395569,
+        "learning_rate": 0.0005,
+        "loss": 1.776,
+        "step": 1800
+      },
+      {
+        "epoch": 0.057718677253738956,
+        "grad_norm": 0.7167004346847534,
+        "learning_rate": 0.0005,
+        "loss": 1.8057,
+        "step": 1810
+      },
+      {
+        "epoch": 0.058037564973372875,
+        "grad_norm": 0.6864603757858276,
+        "learning_rate": 0.0005,
+        "loss": 1.7778,
+        "step": 1820
+      },
+      {
+        "epoch": 0.058356452693006794,
+        "grad_norm": 0.7611004710197449,
+        "learning_rate": 0.0005,
+        "loss": 1.7702,
+        "step": 1830
+      },
+      {
+        "epoch": 0.058675340412640706,
+        "grad_norm": 0.7142066359519958,
+        "learning_rate": 0.0005,
+        "loss": 1.7642,
+        "step": 1840
+      },
+      {
+        "epoch": 0.058994228132274625,
+        "grad_norm": 0.6733101010322571,
+        "learning_rate": 0.0005,
+        "loss": 1.7448,
+        "step": 1850
+      },
+      {
+        "epoch": 0.05931311585190854,
+        "grad_norm": 0.6972571611404419,
+        "learning_rate": 0.0005,
+        "loss": 1.7637,
+        "step": 1860
+      },
+      {
+        "epoch": 0.05963200357154246,
+        "grad_norm": 0.6883410811424255,
+        "learning_rate": 0.0005,
+        "loss": 1.7597,
+        "step": 1870
+      },
+      {
+        "epoch": 0.059950891291176374,
+        "grad_norm": 0.6904264092445374,
+        "learning_rate": 0.0005,
+        "loss": 1.7673,
+        "step": 1880
+      },
+      {
+        "epoch": 0.06026977901081029,
+        "grad_norm": 0.7625157237052917,
+        "learning_rate": 0.0005,
+        "loss": 1.7568,
+        "step": 1890
+      },
+      {
+        "epoch": 0.06058866673044421,
+        "grad_norm": 0.7773521542549133,
+        "learning_rate": 0.0005,
+        "loss": 1.7628,
+        "step": 1900
+      },
+      {
+        "epoch": 0.06090755445007813,
+        "grad_norm": 0.6850663423538208,
+        "learning_rate": 0.0005,
+        "loss": 1.7486,
+        "step": 1910
+      },
+      {
+        "epoch": 0.06122644216971204,
+        "grad_norm": 0.829558253288269,
+        "learning_rate": 0.0005,
+        "loss": 1.7295,
+        "step": 1920
+      },
+      {
+        "epoch": 0.06154532988934596,
+        "grad_norm": 0.6841607689857483,
+        "learning_rate": 0.0005,
+        "loss": 1.7603,
+        "step": 1930
+      },
+      {
+        "epoch": 0.06186421760897988,
+        "grad_norm": 0.8563835620880127,
+        "learning_rate": 0.0005,
+        "loss": 1.7298,
+        "step": 1940
+      },
+      {
+        "epoch": 0.0621831053286138,
+        "grad_norm": 0.6699181795120239,
+        "learning_rate": 0.0005,
+        "loss": 1.7362,
+        "step": 1950
+      },
+      {
+        "epoch": 0.06250199304824772,
+        "grad_norm": 0.6846010088920593,
+        "learning_rate": 0.0005,
+        "loss": 1.7306,
+        "step": 1960
+      },
+      {
+        "epoch": 0.06282088076788163,
+        "grad_norm": 0.7223895788192749,
+        "learning_rate": 0.0005,
+        "loss": 1.7413,
+        "step": 1970
+      },
+      {
+        "epoch": 0.06313976848751554,
+        "grad_norm": 0.6938334107398987,
+        "learning_rate": 0.0005,
+        "loss": 1.7191,
+        "step": 1980
+      },
+      {
+        "epoch": 0.06345865620714947,
+        "grad_norm": 0.677183210849762,
+        "learning_rate": 0.0005,
+        "loss": 1.7135,
+        "step": 1990
+      },
+      {
+        "epoch": 0.06377754392678338,
+        "grad_norm": 0.6501774787902832,
+        "learning_rate": 0.0005,
+        "loss": 1.7226,
+        "step": 2000
+      },
+      {
+        "epoch": 0.06409643164641729,
+        "grad_norm": 0.68111252784729,
+        "learning_rate": 0.0005,
+        "loss": 1.7158,
+        "step": 2010
+      },
+      {
+        "epoch": 0.06441531936605122,
+        "grad_norm": 0.6771100163459778,
+        "learning_rate": 0.0005,
+        "loss": 1.7362,
+        "step": 2020
+      },
+      {
+        "epoch": 0.06473420708568513,
+        "grad_norm": 0.9300406575202942,
+        "learning_rate": 0.0005,
+        "loss": 1.7184,
+        "step": 2030
+      },
+      {
+        "epoch": 0.06505309480531905,
+        "grad_norm": 0.6919383406639099,
+        "learning_rate": 0.0005,
+        "loss": 1.712,
+        "step": 2040
+      },
+      {
+        "epoch": 0.06537198252495297,
+        "grad_norm": 0.6832332611083984,
+        "learning_rate": 0.0005,
+        "loss": 1.6954,
+        "step": 2050
+      },
+      {
+        "epoch": 0.06569087024458688,
+        "grad_norm": 0.6679097414016724,
+        "learning_rate": 0.0005,
+        "loss": 1.6775,
+        "step": 2060
+      },
+      {
+        "epoch": 0.0660097579642208,
+        "grad_norm": 0.6957271695137024,
+        "learning_rate": 0.0005,
+        "loss": 1.7101,
+        "step": 2070
+      },
+      {
+        "epoch": 0.06632864568385471,
+        "grad_norm": 0.6392918825149536,
+        "learning_rate": 0.0005,
+        "loss": 1.6956,
+        "step": 2080
+      },
+      {
+        "epoch": 0.06664753340348863,
+        "grad_norm": 0.6429009437561035,
+        "learning_rate": 0.0005,
+        "loss": 1.712,
+        "step": 2090
+      },
+      {
+        "epoch": 0.06696642112312255,
+        "grad_norm": 0.696384847164154,
+        "learning_rate": 0.0005,
+        "loss": 1.7035,
+        "step": 2100
+      },
+      {
+        "epoch": 0.06728530884275646,
+        "grad_norm": 0.6736230254173279,
+        "learning_rate": 0.0005,
+        "loss": 1.6815,
+        "step": 2110
+      },
+      {
+        "epoch": 0.06760419656239038,
+        "grad_norm": 0.6436002254486084,
+        "learning_rate": 0.0005,
+        "loss": 1.683,
+        "step": 2120
+      },
+      {
+        "epoch": 0.0679230842820243,
+        "grad_norm": 0.6723842620849609,
+        "learning_rate": 0.0005,
+        "loss": 1.6912,
+        "step": 2130
+      },
+      {
+        "epoch": 0.06824197200165821,
+        "grad_norm": 0.6679061055183411,
+        "learning_rate": 0.0005,
+        "loss": 1.6814,
+        "step": 2140
+      },
+      {
+        "epoch": 0.06856085972129214,
+        "grad_norm": 0.6640651822090149,
+        "learning_rate": 0.0005,
+        "loss": 1.6863,
+        "step": 2150
+      },
+      {
+        "epoch": 0.06887974744092605,
+        "grad_norm": 0.6838980317115784,
+        "learning_rate": 0.0005,
+        "loss": 1.6938,
+        "step": 2160
+      },
+      {
+        "epoch": 0.06919863516055996,
+        "grad_norm": 0.6549293994903564,
+        "learning_rate": 0.0005,
+        "loss": 1.6792,
+        "step": 2170
+      },
+      {
+        "epoch": 0.06951752288019389,
+        "grad_norm": 0.6822001338005066,
+        "learning_rate": 0.0005,
+        "loss": 1.6701,
+        "step": 2180
+      },
+      {
+        "epoch": 0.0698364105998278,
+        "grad_norm": 0.7280387282371521,
+        "learning_rate": 0.0005,
+        "loss": 1.6589,
+        "step": 2190
+      },
+      {
+        "epoch": 0.07015529831946171,
+        "grad_norm": 0.6739123463630676,
+        "learning_rate": 0.0005,
+        "loss": 1.6739,
+        "step": 2200
+      },
+      {
+        "epoch": 0.07047418603909564,
+        "grad_norm": 0.6753026843070984,
+        "learning_rate": 0.0005,
+        "loss": 1.6546,
+        "step": 2210
+      },
+      {
+        "epoch": 0.07079307375872955,
+        "grad_norm": 0.7010917663574219,
+        "learning_rate": 0.0005,
+        "loss": 1.6658,
+        "step": 2220
+      },
+      {
+        "epoch": 0.07111196147836348,
+        "grad_norm": 0.6419041752815247,
+        "learning_rate": 0.0005,
+        "loss": 1.6547,
+        "step": 2230
+      },
+      {
+        "epoch": 0.07143084919799739,
+        "grad_norm": 0.7313352823257446,
+        "learning_rate": 0.0005,
+        "loss": 1.6504,
+        "step": 2240
+      },
+      {
+        "epoch": 0.0717497369176313,
+        "grad_norm": 0.6374030709266663,
+        "learning_rate": 0.0005,
+        "loss": 1.6372,
+        "step": 2250
+      },
+      {
+        "epoch": 0.07206862463726522,
+        "grad_norm": 0.6675490140914917,
+        "learning_rate": 0.0005,
+        "loss": 1.6238,
+        "step": 2260
+      },
+      {
+        "epoch": 0.07238751235689914,
+        "grad_norm": 0.7854442000389099,
+        "learning_rate": 0.0005,
+        "loss": 1.6574,
+        "step": 2270
+      },
+      {
+        "epoch": 0.07270640007653305,
+        "grad_norm": 0.6852051615715027,
+        "learning_rate": 0.0005,
+        "loss": 1.647,
+        "step": 2280
+      },
+      {
+        "epoch": 0.07302528779616697,
+        "grad_norm": 0.7064688205718994,
+        "learning_rate": 0.0005,
+        "loss": 1.6258,
+        "step": 2290
+      },
+      {
+        "epoch": 0.07334417551580089,
+        "grad_norm": 0.6328973770141602,
+        "learning_rate": 0.0005,
+        "loss": 1.6332,
+        "step": 2300
+      },
+      {
+        "epoch": 0.0736630632354348,
+        "grad_norm": 0.6078288555145264,
+        "learning_rate": 0.0005,
+        "loss": 1.6222,
+        "step": 2310
+      },
+      {
+        "epoch": 0.07398195095506872,
+        "grad_norm": 0.6709187626838684,
+        "learning_rate": 0.0005,
+        "loss": 1.6442,
+        "step": 2320
+      },
+      {
+        "epoch": 0.07430083867470264,
+        "grad_norm": 0.6568571925163269,
+        "learning_rate": 0.0005,
+        "loss": 1.6406,
+        "step": 2330
+      },
+      {
+        "epoch": 0.07461972639433656,
+        "grad_norm": 0.6580377817153931,
+        "learning_rate": 0.0005,
+        "loss": 1.6228,
+        "step": 2340
+      },
+      {
+        "epoch": 0.07493861411397047,
+        "grad_norm": 0.6596248745918274,
+        "learning_rate": 0.0005,
+        "loss": 1.6444,
+        "step": 2350
+      },
+      {
+        "epoch": 0.07525750183360438,
+        "grad_norm": 0.6527488231658936,
+        "learning_rate": 0.0005,
+        "loss": 1.6392,
+        "step": 2360
+      },
+      {
+        "epoch": 0.07557638955323831,
+        "grad_norm": 0.6278038024902344,
+        "learning_rate": 0.0005,
+        "loss": 1.5949,
+        "step": 2370
+      },
+      {
+        "epoch": 0.07589527727287222,
+        "grad_norm": 0.6202778816223145,
+        "learning_rate": 0.0005,
+        "loss": 1.603,
+        "step": 2380
+      },
+      {
+        "epoch": 0.07621416499250613,
+        "grad_norm": 0.739552915096283,
+        "learning_rate": 0.0005,
+        "loss": 1.6397,
+        "step": 2390
+      },
+      {
+        "epoch": 0.07653305271214006,
+        "grad_norm": 0.6613728404045105,
+        "learning_rate": 0.0005,
+        "loss": 1.6224,
+        "step": 2400
+      },
+      {
+        "epoch": 0.07685194043177397,
+        "grad_norm": 0.6121648550033569,
+        "learning_rate": 0.0005,
+        "loss": 1.6353,
+        "step": 2410
+      },
+      {
+        "epoch": 0.07717082815140788,
+        "grad_norm": 0.6781162619590759,
+        "learning_rate": 0.0005,
+        "loss": 1.6094,
+        "step": 2420
+      },
+      {
+        "epoch": 0.07748971587104181,
+        "grad_norm": 0.6234122514724731,
+        "learning_rate": 0.0005,
+        "loss": 1.6106,
+        "step": 2430
+      },
+      {
+        "epoch": 0.07780860359067572,
+        "grad_norm": 0.6485508680343628,
+        "learning_rate": 0.0005,
+        "loss": 1.6038,
+        "step": 2440
+      },
+      {
+        "epoch": 0.07812749131030965,
+        "grad_norm": 0.7261289954185486,
+        "learning_rate": 0.0005,
+        "loss": 1.5956,
+        "step": 2450
+      },
+      {
+        "epoch": 0.07844637902994356,
+        "grad_norm": 0.6383315920829773,
+        "learning_rate": 0.0005,
+        "loss": 1.6133,
+        "step": 2460
+      },
+      {
+        "epoch": 0.07876526674957747,
+        "grad_norm": 0.6283700466156006,
+        "learning_rate": 0.0005,
+        "loss": 1.6076,
+        "step": 2470
+      },
+      {
+        "epoch": 0.0790841544692114,
+        "grad_norm": 0.6034983396530151,
+        "learning_rate": 0.0005,
+        "loss": 1.5841,
+        "step": 2480
+      },
+      {
+        "epoch": 0.07940304218884531,
+        "grad_norm": 0.6377360820770264,
+        "learning_rate": 0.0005,
+        "loss": 1.6206,
+        "step": 2490
+      },
+      {
+        "epoch": 0.07972192990847922,
+        "grad_norm": 0.631747841835022,
+        "learning_rate": 0.0005,
+        "loss": 1.6042,
+        "step": 2500
+      },
+      {
+        "epoch": 0.08004081762811315,
+        "grad_norm": 0.6026962995529175,
+        "learning_rate": 0.0005,
+        "loss": 1.5796,
+        "step": 2510
+      },
+      {
+        "epoch": 0.08035970534774706,
+        "grad_norm": 0.6390537619590759,
+        "learning_rate": 0.0005,
+        "loss": 1.6024,
+        "step": 2520
+      },
+      {
+        "epoch": 0.08067859306738097,
+        "grad_norm": 0.605514407157898,
+        "learning_rate": 0.0005,
+        "loss": 1.6008,
+        "step": 2530
+      },
+      {
+        "epoch": 0.0809974807870149,
+        "grad_norm": 0.5983066558837891,
+        "learning_rate": 0.0005,
+        "loss": 1.5949,
+        "step": 2540
+      },
+      {
+        "epoch": 0.0813163685066488,
+        "grad_norm": 0.6076896786689758,
+        "learning_rate": 0.0005,
+        "loss": 1.5942,
+        "step": 2550
+      },
+      {
+        "epoch": 0.08163525622628273,
+        "grad_norm": 0.6317102313041687,
+        "learning_rate": 0.0005,
+        "loss": 1.5887,
+        "step": 2560
+      },
+      {
+        "epoch": 0.08195414394591664,
+        "grad_norm": 0.6394833922386169,
+        "learning_rate": 0.0005,
+        "loss": 1.5628,
+        "step": 2570
+      },
+      {
+        "epoch": 0.08227303166555056,
+        "grad_norm": 0.6671807169914246,
+        "learning_rate": 0.0005,
+        "loss": 1.5965,
+        "step": 2580
+      },
+      {
+        "epoch": 0.08259191938518448,
+        "grad_norm": 0.6070294380187988,
+        "learning_rate": 0.0005,
+        "loss": 1.5699,
+        "step": 2590
+      },
+      {
+        "epoch": 0.0829108071048184,
+        "grad_norm": 0.6343675255775452,
+        "learning_rate": 0.0005,
+        "loss": 1.5775,
+        "step": 2600
+      },
+      {
+        "epoch": 0.0832296948244523,
+        "grad_norm": 0.6360995769500732,
+        "learning_rate": 0.0005,
+        "loss": 1.5782,
+        "step": 2610
+      },
+      {
+        "epoch": 0.08354858254408623,
+        "grad_norm": 0.6294621229171753,
+        "learning_rate": 0.0005,
+        "loss": 1.5628,
+        "step": 2620
+      },
+      {
+        "epoch": 0.08386747026372014,
+        "grad_norm": 0.6626117825508118,
+        "learning_rate": 0.0005,
+        "loss": 1.5692,
+        "step": 2630
+      },
+      {
+        "epoch": 0.08418635798335405,
+        "grad_norm": 0.5887053608894348,
+        "learning_rate": 0.0005,
+        "loss": 1.5651,
+        "step": 2640
+      },
+      {
+        "epoch": 0.08450524570298798,
+        "grad_norm": 0.5864725708961487,
+        "learning_rate": 0.0005,
+        "loss": 1.5682,
+        "step": 2650
+      },
+      {
+        "epoch": 0.08482413342262189,
+        "grad_norm": 0.6141831278800964,
+        "learning_rate": 0.0005,
+        "loss": 1.581,
+        "step": 2660
+      },
+      {
+        "epoch": 0.08514302114225582,
+        "grad_norm": 0.5977876782417297,
+        "learning_rate": 0.0005,
+        "loss": 1.5545,
+        "step": 2670
+      },
+      {
+        "epoch": 0.08546190886188973,
+        "grad_norm": 0.6059004068374634,
+        "learning_rate": 0.0005,
+        "loss": 1.5912,
+        "step": 2680
+      },
+      {
+        "epoch": 0.08578079658152364,
+        "grad_norm": 0.5856339931488037,
+        "learning_rate": 0.0005,
+        "loss": 1.5646,
+        "step": 2690
+      },
+      {
+        "epoch": 0.08609968430115757,
+        "grad_norm": 0.5980954766273499,
+        "learning_rate": 0.0005,
+        "loss": 1.5682,
+        "step": 2700
+      },
+      {
+        "epoch": 0.08641857202079148,
+        "grad_norm": 0.6472553014755249,
+        "learning_rate": 0.0005,
+        "loss": 1.5604,
+        "step": 2710
+      },
+      {
+        "epoch": 0.08673745974042539,
+        "grad_norm": 0.6788700222969055,
+        "learning_rate": 0.0005,
+        "loss": 1.5745,
+        "step": 2720
+      },
+      {
+        "epoch": 0.08705634746005932,
+        "grad_norm": 0.5948100686073303,
+        "learning_rate": 0.0005,
+        "loss": 1.5789,
+        "step": 2730
+      },
+      {
+        "epoch": 0.08737523517969323,
+        "grad_norm": 0.6130886673927307,
+        "learning_rate": 0.0005,
+        "loss": 1.5899,
+        "step": 2740
+      },
+      {
+        "epoch": 0.08769412289932715,
+        "grad_norm": 0.6039873361587524,
+        "learning_rate": 0.0005,
+        "loss": 1.553,
+        "step": 2750
+      },
+      {
+        "epoch": 0.08801301061896107,
+        "grad_norm": 0.6165336966514587,
+        "learning_rate": 0.0005,
+        "loss": 1.5493,
+        "step": 2760
+      },
+      {
+        "epoch": 0.08833189833859498,
+        "grad_norm": 0.6436738967895508,
+        "learning_rate": 0.0005,
+        "loss": 1.5479,
+        "step": 2770
+      },
+      {
+        "epoch": 0.0886507860582289,
+        "grad_norm": 0.6035174131393433,
+        "learning_rate": 0.0005,
+        "loss": 1.5779,
+        "step": 2780
+      },
+      {
+        "epoch": 0.08896967377786281,
+        "grad_norm": 0.5844959616661072,
+        "learning_rate": 0.0005,
+        "loss": 1.5297,
+        "step": 2790
+      },
+      {
+        "epoch": 0.08928856149749673,
+        "grad_norm": 0.5872601866722107,
+        "learning_rate": 0.0005,
+        "loss": 1.5617,
+        "step": 2800
+      },
+      {
+        "epoch": 0.08960744921713065,
+        "grad_norm": 0.6345117688179016,
+        "learning_rate": 0.0005,
+        "loss": 1.5247,
+        "step": 2810
+      },
+      {
+        "epoch": 0.08992633693676456,
+        "grad_norm": 0.6308701634407043,
+        "learning_rate": 0.0005,
+        "loss": 1.5495,
+        "step": 2820
+      },
+      {
+        "epoch": 0.09024522465639848,
+        "grad_norm": 0.6402453184127808,
+        "learning_rate": 0.0005,
+        "loss": 1.5494,
+        "step": 2830
+      },
+      {
+        "epoch": 0.0905641123760324,
+        "grad_norm": 0.5856583714485168,
+        "learning_rate": 0.0005,
+        "loss": 1.5478,
+        "step": 2840
+      },
+      {
+        "epoch": 0.09088300009566631,
+        "grad_norm": 0.5717542767524719,
+        "learning_rate": 0.0005,
+        "loss": 1.5255,
+        "step": 2850
+      },
+      {
+        "epoch": 0.09120188781530024,
+        "grad_norm": 0.5698398947715759,
+        "learning_rate": 0.0005,
+        "loss": 1.5312,
+        "step": 2860
+      },
+      {
+        "epoch": 0.09152077553493415,
+        "grad_norm": 0.6083899140357971,
+        "learning_rate": 0.0005,
+        "loss": 1.5315,
+        "step": 2870
+      },
+      {
+        "epoch": 0.09183966325456806,
+        "grad_norm": 0.6005008220672607,
+        "learning_rate": 0.0005,
+        "loss": 1.5453,
+        "step": 2880
+      },
+      {
+        "epoch": 0.09215855097420199,
+        "grad_norm": 0.5992957353591919,
+        "learning_rate": 0.0005,
+        "loss": 1.5357,
+        "step": 2890
+      },
+      {
+        "epoch": 0.0924774386938359,
+        "grad_norm": 0.5881666541099548,
+        "learning_rate": 0.0005,
+        "loss": 1.5308,
+        "step": 2900
+      },
+      {
+        "epoch": 0.09279632641346981,
+        "grad_norm": 0.6062901616096497,
+        "learning_rate": 0.0005,
+        "loss": 1.5388,
+        "step": 2910
+      },
+      {
+        "epoch": 0.09311521413310374,
+        "grad_norm": 0.5917524695396423,
+        "learning_rate": 0.0005,
+        "loss": 1.5409,
+        "step": 2920
+      },
+      {
+        "epoch": 0.09343410185273765,
+        "grad_norm": 0.6036853790283203,
+        "learning_rate": 0.0005,
+        "loss": 1.5436,
+        "step": 2930
+      },
+      {
+        "epoch": 0.09375298957237156,
+        "grad_norm": 0.5984651446342468,
+        "learning_rate": 0.0005,
+        "loss": 1.5526,
+        "step": 2940
+      },
+      {
+        "epoch": 0.09407187729200549,
+        "grad_norm": 0.6358986496925354,
+        "learning_rate": 0.0005,
+        "loss": 1.5239,
+        "step": 2950
+      },
+      {
+        "epoch": 0.0943907650116394,
+        "grad_norm": 0.6184073686599731,
+        "learning_rate": 0.0005,
+        "loss": 1.5247,
+        "step": 2960
+      },
+      {
+        "epoch": 0.09470965273127332,
+        "grad_norm": 0.6121427416801453,
+        "learning_rate": 0.0005,
+        "loss": 1.5344,
+        "step": 2970
+      },
+      {
+        "epoch": 0.09502854045090724,
+        "grad_norm": 0.5836875438690186,
+        "learning_rate": 0.0005,
+        "loss": 1.5249,
+        "step": 2980
+      },
+      {
+        "epoch": 0.09534742817054115,
+        "grad_norm": 0.5729764103889465,
+        "learning_rate": 0.0005,
+        "loss": 1.5302,
+        "step": 2990
+      },
+      {
+        "epoch": 0.09566631589017507,
+        "grad_norm": 0.5892743468284607,
+        "learning_rate": 0.0005,
+        "loss": 1.5433,
+        "step": 3000
+      },
+      {
+        "epoch": 0.09598520360980899,
+        "grad_norm": 0.593152642250061,
+        "learning_rate": 0.0005,
+        "loss": 1.5232,
+        "step": 3010
+      },
+      {
+        "epoch": 0.0963040913294429,
+        "grad_norm": 0.5753911733627319,
+        "learning_rate": 0.0005,
+        "loss": 1.5133,
+        "step": 3020
+      },
+      {
+        "epoch": 0.09662297904907682,
+        "grad_norm": 0.6225734353065491,
+        "learning_rate": 0.0005,
+        "loss": 1.5195,
+        "step": 3030
+      },
+      {
+        "epoch": 0.09694186676871074,
+        "grad_norm": 0.5819618105888367,
+        "learning_rate": 0.0005,
+        "loss": 1.5238,
+        "step": 3040
+      },
+      {
+        "epoch": 0.09726075448834465,
+        "grad_norm": 0.6067165732383728,
+        "learning_rate": 0.0005,
+        "loss": 1.5342,
+        "step": 3050
+      },
+      {
+        "epoch": 0.09757964220797857,
+        "grad_norm": 0.5918219089508057,
+        "learning_rate": 0.0005,
+        "loss": 1.531,
+        "step": 3060
+      },
+      {
+        "epoch": 0.09789852992761248,
+        "grad_norm": 0.5920441150665283,
+        "learning_rate": 0.0005,
+        "loss": 1.5125,
+        "step": 3070
+      },
+      {
+        "epoch": 0.09821741764724641,
+        "grad_norm": 0.5718549489974976,
+        "learning_rate": 0.0005,
+        "loss": 1.5217,
+        "step": 3080
+      },
+      {
+        "epoch": 0.09853630536688032,
+        "grad_norm": 0.5798651576042175,
+        "learning_rate": 0.0005,
+        "loss": 1.5092,
+        "step": 3090
+      },
+      {
+        "epoch": 0.09885519308651423,
+        "grad_norm": 0.5711941123008728,
+        "learning_rate": 0.0005,
+        "loss": 1.4999,
+        "step": 3100
+      },
+      {
+        "epoch": 0.09917408080614816,
+        "grad_norm": 0.6470732688903809,
+        "learning_rate": 0.0005,
+        "loss": 1.4928,
+        "step": 3110
+      },
+      {
+        "epoch": 0.09949296852578207,
+        "grad_norm": 0.5650610327720642,
+        "learning_rate": 0.0005,
+        "loss": 1.5182,
+        "step": 3120
+      },
+      {
+        "epoch": 0.09981185624541598,
+        "grad_norm": 0.6110008358955383,
+        "learning_rate": 0.0005,
+        "loss": 1.5261,
+        "step": 3130
+      },
+      {
+        "epoch": 0.10013074396504991,
+        "grad_norm": 0.6006331443786621,
+        "learning_rate": 0.0005,
+        "loss": 1.5124,
+        "step": 3140
+      },
+      {
+        "epoch": 0.10044963168468382,
+        "grad_norm": 0.5928711295127869,
+        "learning_rate": 0.0005,
+        "loss": 1.5159,
+        "step": 3150
+      },
+      {
+        "epoch": 0.10076851940431773,
+        "grad_norm": 0.5740655064582825,
+        "learning_rate": 0.0005,
+        "loss": 1.4932,
+        "step": 3160
+      },
+      {
+        "epoch": 0.10108740712395166,
+        "grad_norm": 0.5549852252006531,
+        "learning_rate": 0.0005,
+        "loss": 1.5022,
+        "step": 3170
+      },
+      {
+        "epoch": 0.10140629484358557,
+        "grad_norm": 0.5529758930206299,
+        "learning_rate": 0.0005,
+        "loss": 1.5011,
+        "step": 3180
+      },
+      {
+        "epoch": 0.1017251825632195,
+        "grad_norm": 0.5959564447402954,
+        "learning_rate": 0.0005,
+        "loss": 1.4831,
+        "step": 3190
+      },
+      {
+        "epoch": 0.10204407028285341,
+        "grad_norm": 0.5632524490356445,
+        "learning_rate": 0.0005,
+        "loss": 1.4886,
+        "step": 3200
+      },
+      {
+        "epoch": 0.10236295800248732,
+        "grad_norm": 0.5667331218719482,
+        "learning_rate": 0.0005,
+        "loss": 1.5205,
+        "step": 3210
+      },
+      {
+        "epoch": 0.10268184572212125,
+        "grad_norm": 0.561392605304718,
+        "learning_rate": 0.0005,
+        "loss": 1.508,
+        "step": 3220
+      },
+      {
+        "epoch": 0.10300073344175516,
+        "grad_norm": 0.5860125422477722,
+        "learning_rate": 0.0005,
+        "loss": 1.4788,
+        "step": 3230
+      },
+      {
+        "epoch": 0.10331962116138907,
+        "grad_norm": 0.5863109230995178,
+        "learning_rate": 0.0005,
+        "loss": 1.461,
+        "step": 3240
+      },
+      {
+        "epoch": 0.103638508881023,
+        "grad_norm": 0.5454897880554199,
+        "learning_rate": 0.0005,
+        "loss": 1.491,
+        "step": 3250
+      },
+      {
+        "epoch": 0.1039573966006569,
+        "grad_norm": 0.5635510087013245,
+        "learning_rate": 0.0005,
+        "loss": 1.5087,
+        "step": 3260
+      },
+      {
+        "epoch": 0.10427628432029083,
+        "grad_norm": 0.5486317276954651,
+        "learning_rate": 0.0005,
+        "loss": 1.4587,
+        "step": 3270
+      },
+      {
+        "epoch": 0.10459517203992474,
+        "grad_norm": 0.620488166809082,
+        "learning_rate": 0.0005,
+        "loss": 1.5002,
+        "step": 3280
+      },
+      {
+        "epoch": 0.10491405975955866,
+        "grad_norm": 0.5548288226127625,
+        "learning_rate": 0.0005,
+        "loss": 1.464,
+        "step": 3290
+      },
+      {
+        "epoch": 0.10523294747919258,
+        "grad_norm": 0.5835471749305725,
+        "learning_rate": 0.0005,
+        "loss": 1.492,
+        "step": 3300
+      },
+      {
+        "epoch": 0.1055518351988265,
+        "grad_norm": 0.585202693939209,
+        "learning_rate": 0.0005,
+        "loss": 1.48,
+        "step": 3310
+      },
+      {
+        "epoch": 0.1058707229184604,
+        "grad_norm": 0.5698010325431824,
+        "learning_rate": 0.0005,
+        "loss": 1.4633,
+        "step": 3320
+      },
+      {
+        "epoch": 0.10618961063809433,
+        "grad_norm": 0.5747951865196228,
+        "learning_rate": 0.0005,
+        "loss": 1.531,
+        "step": 3330
+      },
+      {
+        "epoch": 0.10650849835772824,
+        "grad_norm": 0.5524262189865112,
+        "learning_rate": 0.0005,
+        "loss": 1.4961,
+        "step": 3340
+      },
+      {
+        "epoch": 0.10682738607736215,
+        "grad_norm": 0.5779598355293274,
+        "learning_rate": 0.0005,
+        "loss": 1.5045,
+        "step": 3350
+      },
+      {
+        "epoch": 0.10714627379699608,
+        "grad_norm": 0.6228440999984741,
+        "learning_rate": 0.0005,
+        "loss": 1.4862,
+        "step": 3360
+      },
+      {
+        "epoch": 0.10746516151662999,
+        "grad_norm": 0.5359426140785217,
+        "learning_rate": 0.0005,
+        "loss": 1.4967,
+        "step": 3370
+      },
+      {
+        "epoch": 0.10778404923626392,
+        "grad_norm": 0.6059668660163879,
+        "learning_rate": 0.0005,
+        "loss": 1.4805,
+        "step": 3380
+      },
+      {
+        "epoch": 0.10810293695589783,
+        "grad_norm": 0.5769729018211365,
+        "learning_rate": 0.0005,
+        "loss": 1.4502,
+        "step": 3390
+      },
+      {
+        "epoch": 0.10842182467553174,
+        "grad_norm": 0.5658615231513977,
+        "learning_rate": 0.0005,
+        "loss": 1.4685,
+        "step": 3400
+      },
+      {
+        "epoch": 0.10874071239516567,
+        "grad_norm": 0.5631580352783203,
+        "learning_rate": 0.0005,
+        "loss": 1.4805,
+        "step": 3410
+      },
+      {
+        "epoch": 0.10905960011479958,
+        "grad_norm": 0.554395854473114,
+        "learning_rate": 0.0005,
+        "loss": 1.487,
+        "step": 3420
+      },
+      {
+        "epoch": 0.10937848783443349,
+        "grad_norm": 0.5861183404922485,
+        "learning_rate": 0.0005,
+        "loss": 1.4826,
+        "step": 3430
+      },
+      {
+        "epoch": 0.10969737555406742,
+        "grad_norm": 0.5401641726493835,
+        "learning_rate": 0.0005,
+        "loss": 1.4861,
+        "step": 3440
+      },
+      {
+        "epoch": 0.11001626327370133,
+        "grad_norm": 0.5490972399711609,
+        "learning_rate": 0.0005,
+        "loss": 1.4907,
+        "step": 3450
+      },
+      {
+        "epoch": 0.11033515099333524,
+        "grad_norm": 0.5460464954376221,
+        "learning_rate": 0.0005,
+        "loss": 1.4537,
+        "step": 3460
+      },
+      {
+        "epoch": 0.11065403871296917,
+        "grad_norm": 0.5482505559921265,
+        "learning_rate": 0.0005,
+        "loss": 1.477,
+        "step": 3470
+      },
+      {
+        "epoch": 0.11097292643260308,
+        "grad_norm": 0.5561589598655701,
+        "learning_rate": 0.0005,
+        "loss": 1.4591,
+        "step": 3480
+      },
+      {
+        "epoch": 0.111291814152237,
+        "grad_norm": 0.5551013350486755,
+        "learning_rate": 0.0005,
+        "loss": 1.4544,
+        "step": 3490
+      },
+      {
+        "epoch": 0.11161070187187092,
+        "grad_norm": 0.5580571293830872,
+        "learning_rate": 0.0005,
+        "loss": 1.468,
+        "step": 3500
+      },
+      {
+        "epoch": 0.11192958959150483,
+        "grad_norm": 0.583152174949646,
+        "learning_rate": 0.0005,
+        "loss": 1.4704,
+        "step": 3510
+      },
+      {
+        "epoch": 0.11224847731113875,
+        "grad_norm": 0.5443835258483887,
+        "learning_rate": 0.0005,
+        "loss": 1.462,
+        "step": 3520
+      },
+      {
+        "epoch": 0.11256736503077266,
+        "grad_norm": 0.5560067296028137,
+        "learning_rate": 0.0005,
+        "loss": 1.4542,
+        "step": 3530
+      },
+      {
+        "epoch": 0.11288625275040658,
+        "grad_norm": 0.5389274954795837,
+        "learning_rate": 0.0005,
+        "loss": 1.4681,
+        "step": 3540
+      },
+      {
+        "epoch": 0.1132051404700405,
+        "grad_norm": 0.5621799826622009,
+        "learning_rate": 0.0005,
+        "loss": 1.4688,
+        "step": 3550
+      },
+      {
+        "epoch": 0.11352402818967441,
+        "grad_norm": 0.5864140391349792,
+        "learning_rate": 0.0005,
+        "loss": 1.4607,
+        "step": 3560
+      },
+      {
+        "epoch": 0.11384291590930833,
+        "grad_norm": 0.5543456077575684,
+        "learning_rate": 0.0005,
+        "loss": 1.4383,
+        "step": 3570
+      },
+      {
+        "epoch": 0.11416180362894225,
+        "grad_norm": 0.556922972202301,
+        "learning_rate": 0.0005,
+        "loss": 1.4362,
+        "step": 3580
+      },
+      {
+        "epoch": 0.11448069134857616,
+        "grad_norm": 0.5521618127822876,
+        "learning_rate": 0.0005,
+        "loss": 1.4752,
+        "step": 3590
+      },
+      {
+        "epoch": 0.11479957906821009,
+        "grad_norm": 0.5324291586875916,
+        "learning_rate": 0.0005,
+        "loss": 1.4639,
+        "step": 3600
+      },
+      {
+        "epoch": 0.115118466787844,
+        "grad_norm": 0.5443735122680664,
+        "learning_rate": 0.0005,
+        "loss": 1.4357,
+        "step": 3610
+      },
+      {
+        "epoch": 0.11543735450747791,
+        "grad_norm": 0.5763177275657654,
+        "learning_rate": 0.0005,
+        "loss": 1.4309,
+        "step": 3620
+      },
+      {
+        "epoch": 0.11575624222711184,
+        "grad_norm": 0.5425290465354919,
+        "learning_rate": 0.0005,
+        "loss": 1.4659,
+        "step": 3630
+      },
+      {
+        "epoch": 0.11607512994674575,
+        "grad_norm": 0.598491907119751,
+        "learning_rate": 0.0005,
+        "loss": 1.477,
+        "step": 3640
+      },
+      {
+        "epoch": 0.11639401766637966,
+        "grad_norm": 0.5747283101081848,
+        "learning_rate": 0.0005,
+        "loss": 1.4482,
+        "step": 3650
+      },
+      {
+        "epoch": 0.11671290538601359,
+        "grad_norm": 0.5188102722167969,
+        "learning_rate": 0.0005,
+        "loss": 1.4331,
+        "step": 3660
+      },
+      {
+        "epoch": 0.1170317931056475,
+        "grad_norm": 0.5427939295768738,
+        "learning_rate": 0.0005,
+        "loss": 1.4646,
+        "step": 3670
+      },
+      {
+        "epoch": 0.11735068082528141,
+        "grad_norm": 0.5608155131340027,
+        "learning_rate": 0.0005,
+        "loss": 1.4812,
+        "step": 3680
+      },
+      {
+        "epoch": 0.11766956854491534,
+        "grad_norm": 0.5341671705245972,
+        "learning_rate": 0.0005,
+        "loss": 1.4434,
+        "step": 3690
+      },
+      {
+        "epoch": 0.11798845626454925,
+        "grad_norm": 0.5669570565223694,
+        "learning_rate": 0.0005,
+        "loss": 1.4547,
+        "step": 3700
+      },
+      {
+        "epoch": 0.11830734398418317,
+        "grad_norm": 0.5232808589935303,
+        "learning_rate": 0.0005,
+        "loss": 1.4501,
+        "step": 3710
+      },
+      {
+        "epoch": 0.11862623170381709,
+        "grad_norm": 0.5229708552360535,
+        "learning_rate": 0.0005,
+        "loss": 1.4325,
+        "step": 3720
+      },
+      {
+        "epoch": 0.118945119423451,
+        "grad_norm": 0.5569901466369629,
+        "learning_rate": 0.0005,
+        "loss": 1.4504,
+        "step": 3730
+      },
+      {
+        "epoch": 0.11926400714308492,
+        "grad_norm": 0.5216270685195923,
+        "learning_rate": 0.0005,
+        "loss": 1.4587,
+        "step": 3740
+      },
+      {
+        "epoch": 0.11958289486271884,
+        "grad_norm": 0.5247334837913513,
+        "learning_rate": 0.0005,
+        "loss": 1.4313,
+        "step": 3750
+      },
+      {
+        "epoch": 0.11990178258235275,
+        "grad_norm": 0.5420272946357727,
+        "learning_rate": 0.0005,
+        "loss": 1.439,
+        "step": 3760
+      },
+      {
+        "epoch": 0.12022067030198667,
+        "grad_norm": 0.5527462363243103,
+        "learning_rate": 0.0005,
+        "loss": 1.4538,
+        "step": 3770
+      },
+      {
+        "epoch": 0.12053955802162059,
+        "grad_norm": 0.5519984364509583,
+        "learning_rate": 0.0005,
+        "loss": 1.4697,
+        "step": 3780
+      },
+      {
+        "epoch": 0.12085844574125451,
+        "grad_norm": 0.5728175044059753,
+        "learning_rate": 0.0005,
+        "loss": 1.4228,
+        "step": 3790
+      },
+      {
+        "epoch": 0.12117733346088842,
+        "grad_norm": 0.5295974016189575,
+        "learning_rate": 0.0005,
+        "loss": 1.4217,
+        "step": 3800
+      },
+      {
+        "epoch": 0.12149622118052233,
+        "grad_norm": 0.5916241407394409,
+        "learning_rate": 0.0005,
+        "loss": 1.426,
+        "step": 3810
+      },
+      {
+        "epoch": 0.12181510890015626,
+        "grad_norm": 0.5520297288894653,
+        "learning_rate": 0.0005,
+        "loss": 1.4343,
+        "step": 3820
+      },
+      {
+        "epoch": 0.12213399661979017,
+        "grad_norm": 0.547126054763794,
+        "learning_rate": 0.0005,
+        "loss": 1.4223,
+        "step": 3830
+      },
+      {
+        "epoch": 0.12245288433942408,
+        "grad_norm": 0.5996837019920349,
+        "learning_rate": 0.0005,
+        "loss": 1.4485,
+        "step": 3840
+      },
+      {
+        "epoch": 0.12277177205905801,
+        "grad_norm": 0.5237704515457153,
+        "learning_rate": 0.0005,
+        "loss": 1.4456,
+        "step": 3850
+      },
+      {
+        "epoch": 0.12309065977869192,
+        "grad_norm": 0.5590718984603882,
+        "learning_rate": 0.0005,
+        "loss": 1.4472,
+        "step": 3860
+      },
+      {
+        "epoch": 0.12340954749832583,
+        "grad_norm": 0.5487239360809326,
+        "learning_rate": 0.0005,
+        "loss": 1.457,
+        "step": 3870
+      },
+      {
+        "epoch": 0.12372843521795976,
+        "grad_norm": 0.5274037718772888,
+        "learning_rate": 0.0005,
+        "loss": 1.4231,
+        "step": 3880
+      },
+      {
+        "epoch": 0.12404732293759367,
+        "grad_norm": 0.5649983882904053,
+        "learning_rate": 0.0005,
+        "loss": 1.4431,
+        "step": 3890
+      },
+      {
+        "epoch": 0.1243662106572276,
+        "grad_norm": 0.5578463077545166,
+        "learning_rate": 0.0005,
+        "loss": 1.4381,
+        "step": 3900
+      },
+      {
+        "epoch": 0.12468509837686151,
+        "grad_norm": 0.531040608882904,
+        "learning_rate": 0.0005,
+        "loss": 1.4313,
+        "step": 3910
+      },
+      {
+        "epoch": 0.12500398609649543,
+        "grad_norm": 0.5263800024986267,
+        "learning_rate": 0.0005,
+        "loss": 1.4373,
+        "step": 3920
+      },
+      {
+        "epoch": 0.12532287381612933,
+        "grad_norm": 0.5514016151428223,
+        "learning_rate": 0.0005,
+        "loss": 1.4105,
+        "step": 3930
+      },
+      {
+        "epoch": 0.12564176153576326,
+        "grad_norm": 0.5430245399475098,
+        "learning_rate": 0.0005,
+        "loss": 1.4072,
+        "step": 3940
+      },
+      {
+        "epoch": 0.12596064925539718,
+        "grad_norm": 0.49897560477256775,
+        "learning_rate": 0.0005,
+        "loss": 1.4286,
+        "step": 3950
+      },
+      {
+        "epoch": 0.12627953697503108,
+        "grad_norm": 0.5384120941162109,
+        "learning_rate": 0.0005,
+        "loss": 1.411,
+        "step": 3960
+      },
+      {
+        "epoch": 0.126598424694665,
+        "grad_norm": 0.511913537979126,
+        "learning_rate": 0.0005,
+        "loss": 1.4332,
+        "step": 3970
+      },
+      {
+        "epoch": 0.12691731241429893,
+        "grad_norm": 0.5535169839859009,
+        "learning_rate": 0.0005,
+        "loss": 1.4108,
+        "step": 3980
+      },
+      {
+        "epoch": 0.12723620013393283,
+        "grad_norm": 0.5501640439033508,
+        "learning_rate": 0.0005,
+        "loss": 1.4184,
+        "step": 3990
+      },
+      {
+        "epoch": 0.12755508785356676,
+        "grad_norm": 0.525273859500885,
+        "learning_rate": 0.0005,
+        "loss": 1.4155,
+        "step": 4000
+      },
+      {
+        "epoch": 0.12787397557320068,
+        "grad_norm": 0.5309306383132935,
+        "learning_rate": 0.0005,
+        "loss": 1.43,
+        "step": 4010
+      },
+      {
+        "epoch": 0.12819286329283458,
+        "grad_norm": 0.590304970741272,
+        "learning_rate": 0.0005,
+        "loss": 1.421,
+        "step": 4020
+      },
+      {
+        "epoch": 0.1285117510124685,
+        "grad_norm": 0.5559628009796143,
+        "learning_rate": 0.0005,
+        "loss": 1.428,
+        "step": 4030
+      },
+      {
+        "epoch": 0.12883063873210243,
+        "grad_norm": 0.523764967918396,
+        "learning_rate": 0.0005,
+        "loss": 1.4189,
+        "step": 4040
+      },
+      {
+        "epoch": 0.12914952645173636,
+        "grad_norm": 0.5560243129730225,
+        "learning_rate": 0.0005,
+        "loss": 1.4279,
+        "step": 4050
+      },
+      {
+        "epoch": 0.12946841417137026,
+        "grad_norm": 0.5219590663909912,
+        "learning_rate": 0.0005,
+        "loss": 1.4343,
+        "step": 4060
+      },
+      {
+        "epoch": 0.12978730189100418,
+        "grad_norm": 0.5206665396690369,
+        "learning_rate": 0.0005,
+        "loss": 1.4299,
+        "step": 4070
+      },
+      {
+        "epoch": 0.1301061896106381,
+        "grad_norm": 0.5236951112747192,
+        "learning_rate": 0.0005,
+        "loss": 1.3991,
+        "step": 4080
+      },
+      {
+        "epoch": 0.130425077330272,
+        "grad_norm": 0.5283119678497314,
+        "learning_rate": 0.0005,
+        "loss": 1.4192,
+        "step": 4090
+      },
+      {
+        "epoch": 0.13074396504990593,
+        "grad_norm": 0.5292713642120361,
+        "learning_rate": 0.0005,
+        "loss": 1.4188,
+        "step": 4100
+      },
+      {
+        "epoch": 0.13106285276953986,
+        "grad_norm": 0.5801336169242859,
+        "learning_rate": 0.0005,
+        "loss": 1.4115,
+        "step": 4110
+      },
+      {
+        "epoch": 0.13138174048917375,
+        "grad_norm": 0.5187575817108154,
+        "learning_rate": 0.0005,
+        "loss": 1.3969,
+        "step": 4120
+      },
+      {
+        "epoch": 0.13170062820880768,
+        "grad_norm": 0.541106641292572,
+        "learning_rate": 0.0005,
+        "loss": 1.4133,
+        "step": 4130
+      },
+      {
+        "epoch": 0.1320195159284416,
+        "grad_norm": 0.557482898235321,
+        "learning_rate": 0.0005,
+        "loss": 1.4106,
+        "step": 4140
+      },
+      {
+        "epoch": 0.1323384036480755,
+        "grad_norm": 0.5167677402496338,
+        "learning_rate": 0.0005,
+        "loss": 1.3988,
+        "step": 4150
+      },
+      {
+        "epoch": 0.13265729136770943,
+        "grad_norm": 0.5109487175941467,
+        "learning_rate": 0.0005,
+        "loss": 1.4044,
+        "step": 4160
+      },
+      {
+        "epoch": 0.13297617908734335,
+        "grad_norm": 0.5093071460723877,
+        "learning_rate": 0.0005,
+        "loss": 1.4067,
+        "step": 4170
+      },
+      {
+        "epoch": 0.13329506680697725,
+        "grad_norm": 0.5161386728286743,
+        "learning_rate": 0.0005,
+        "loss": 1.3862,
+        "step": 4180
+      },
+      {
+        "epoch": 0.13361395452661118,
+        "grad_norm": 0.5066111087799072,
+        "learning_rate": 0.0005,
+        "loss": 1.3957,
+        "step": 4190
+      },
+      {
+        "epoch": 0.1339328422462451,
+        "grad_norm": 0.5192795991897583,
+        "learning_rate": 0.0005,
+        "loss": 1.4062,
+        "step": 4200
+      },
+      {
+        "epoch": 0.134251729965879,
+        "grad_norm": 0.504766047000885,
+        "learning_rate": 0.0005,
+        "loss": 1.4001,
+        "step": 4210
+      },
+      {
+        "epoch": 0.13457061768551293,
+        "grad_norm": 0.5010153651237488,
+        "learning_rate": 0.0005,
+        "loss": 1.4002,
+        "step": 4220
+      },
+      {
+        "epoch": 0.13488950540514685,
+        "grad_norm": 0.5043392777442932,
+        "learning_rate": 0.0005,
+        "loss": 1.3885,
+        "step": 4230
+      },
+      {
+        "epoch": 0.13520839312478075,
+        "grad_norm": 0.5130932927131653,
+        "learning_rate": 0.0005,
+        "loss": 1.4271,
+        "step": 4240
+      },
+      {
+        "epoch": 0.13552728084441468,
+        "grad_norm": 0.4850212633609772,
+        "learning_rate": 0.0005,
+        "loss": 1.4028,
+        "step": 4250
+      },
+      {
+        "epoch": 0.1358461685640486,
+        "grad_norm": 0.5103309750556946,
+        "learning_rate": 0.0005,
+        "loss": 1.4137,
+        "step": 4260
+      },
+      {
+        "epoch": 0.13616505628368253,
+        "grad_norm": 0.5363282561302185,
+        "learning_rate": 0.0005,
+        "loss": 1.3568,
+        "step": 4270
+      },
+      {
+        "epoch": 0.13648394400331643,
+        "grad_norm": 0.5205022692680359,
+        "learning_rate": 0.0005,
+        "loss": 1.3999,
+        "step": 4280
+      },
+      {
+        "epoch": 0.13680283172295035,
+        "grad_norm": 0.5299255847930908,
+        "learning_rate": 0.0005,
+        "loss": 1.3793,
+        "step": 4290
+      },
+      {
+        "epoch": 0.13712171944258428,
+        "grad_norm": 0.5065796971321106,
+        "learning_rate": 0.0005,
+        "loss": 1.3932,
+        "step": 4300
+      },
+      {
+        "epoch": 0.13744060716221818,
+        "grad_norm": 0.5404181480407715,
+        "learning_rate": 0.0005,
+        "loss": 1.4033,
+        "step": 4310
+      },
+      {
+        "epoch": 0.1377594948818521,
+        "grad_norm": 0.495960533618927,
+        "learning_rate": 0.0005,
+        "loss": 1.3815,
+        "step": 4320
+      },
+      {
+        "epoch": 0.13807838260148603,
+        "grad_norm": 0.4996244013309479,
+        "learning_rate": 0.0005,
+        "loss": 1.3929,
+        "step": 4330
+      },
+      {
+        "epoch": 0.13839727032111993,
+        "grad_norm": 0.496735543012619,
+        "learning_rate": 0.0005,
+        "loss": 1.3939,
+        "step": 4340
+      },
+      {
+        "epoch": 0.13871615804075385,
+        "grad_norm": 0.5102179646492004,
+        "learning_rate": 0.0005,
+        "loss": 1.3932,
+        "step": 4350
+      },
+      {
+        "epoch": 0.13903504576038778,
+        "grad_norm": 0.515885055065155,
+        "learning_rate": 0.0005,
+        "loss": 1.3931,
+        "step": 4360
+      },
+      {
+        "epoch": 0.13935393348002167,
+        "grad_norm": 0.5257781744003296,
+        "learning_rate": 0.0005,
+        "loss": 1.4075,
+        "step": 4370
+      },
+      {
+        "epoch": 0.1396728211996556,
+        "grad_norm": 0.5537557601928711,
+        "learning_rate": 0.0005,
+        "loss": 1.3857,
+        "step": 4380
+      },
+      {
+        "epoch": 0.13999170891928953,
+        "grad_norm": 0.5099586844444275,
+        "learning_rate": 0.0005,
+        "loss": 1.3906,
+        "step": 4390
+      },
+      {
+        "epoch": 0.14031059663892342,
+        "grad_norm": 0.5115305185317993,
+        "learning_rate": 0.0005,
+        "loss": 1.4028,
+        "step": 4400
+      },
+      {
+        "epoch": 0.14062948435855735,
+        "grad_norm": 0.5236173868179321,
+        "learning_rate": 0.0005,
+        "loss": 1.4064,
+        "step": 4410
+      },
+      {
+        "epoch": 0.14094837207819128,
+        "grad_norm": 0.49574723839759827,
+        "learning_rate": 0.0005,
+        "loss": 1.3687,
+        "step": 4420
+      },
+      {
+        "epoch": 0.14126725979782517,
+        "grad_norm": 0.512898862361908,
+        "learning_rate": 0.0005,
+        "loss": 1.385,
+        "step": 4430
+      },
+      {
+        "epoch": 0.1415861475174591,
+        "grad_norm": 0.49689918756484985,
+        "learning_rate": 0.0005,
+        "loss": 1.3695,
+        "step": 4440
+      },
+      {
+        "epoch": 0.14190503523709302,
+        "grad_norm": 0.5022764801979065,
+        "learning_rate": 0.0005,
+        "loss": 1.3606,
+        "step": 4450
+      },
+      {
+        "epoch": 0.14222392295672695,
+        "grad_norm": 0.5280899405479431,
+        "learning_rate": 0.0005,
+        "loss": 1.3651,
+        "step": 4460
+      },
+      {
+        "epoch": 0.14254281067636085,
+        "grad_norm": 0.5257755517959595,
+        "learning_rate": 0.0005,
+        "loss": 1.3761,
+        "step": 4470
+      },
+      {
+        "epoch": 0.14286169839599477,
+        "grad_norm": 0.5031253099441528,
+        "learning_rate": 0.0005,
+        "loss": 1.3864,
+        "step": 4480
+      },
+      {
+        "epoch": 0.1431805861156287,
+        "grad_norm": 0.5129061341285706,
+        "learning_rate": 0.0005,
+        "loss": 1.3759,
+        "step": 4490
+      },
+      {
+        "epoch": 0.1434994738352626,
+        "grad_norm": 0.5187103152275085,
+        "learning_rate": 0.0005,
+        "loss": 1.4037,
+        "step": 4500
+      },
+      {
+        "epoch": 0.14381836155489652,
+        "grad_norm": 0.5009415745735168,
+        "learning_rate": 0.0005,
+        "loss": 1.4097,
+        "step": 4510
+      },
+      {
+        "epoch": 0.14413724927453045,
+        "grad_norm": 0.5013829469680786,
+        "learning_rate": 0.0005,
+        "loss": 1.3811,
+        "step": 4520
+      },
+      {
+        "epoch": 0.14445613699416435,
+        "grad_norm": 0.4731661081314087,
+        "learning_rate": 0.0005,
+        "loss": 1.3755,
+        "step": 4530
+      },
+      {
+        "epoch": 0.14477502471379827,
+        "grad_norm": 0.5249818563461304,
+        "learning_rate": 0.0005,
+        "loss": 1.3786,
+        "step": 4540
+      },
+      {
+        "epoch": 0.1450939124334322,
+        "grad_norm": 0.49896419048309326,
+        "learning_rate": 0.0005,
+        "loss": 1.38,
+        "step": 4550
+      },
+      {
+        "epoch": 0.1454128001530661,
+        "grad_norm": 0.5124298930168152,
+        "learning_rate": 0.0005,
+        "loss": 1.3999,
+        "step": 4560
+      },
+      {
+        "epoch": 0.14573168787270002,
+        "grad_norm": 0.4988538324832916,
+        "learning_rate": 0.0005,
+        "loss": 1.3761,
+        "step": 4570
+      },
+      {
+        "epoch": 0.14605057559233395,
+        "grad_norm": 0.49036043882369995,
+        "learning_rate": 0.0005,
+        "loss": 1.3684,
+        "step": 4580
+      },
+      {
+        "epoch": 0.14636946331196785,
+        "grad_norm": 0.5022983551025391,
+        "learning_rate": 0.0005,
+        "loss": 1.3694,
+        "step": 4590
+      },
+      {
+        "epoch": 0.14668835103160177,
+        "grad_norm": 0.5071865320205688,
+        "learning_rate": 0.0005,
+        "loss": 1.3891,
+        "step": 4600
+      },
+      {
+        "epoch": 0.1470072387512357,
+        "grad_norm": 0.4856301546096802,
+        "learning_rate": 0.0005,
+        "loss": 1.3796,
+        "step": 4610
+      },
+      {
+        "epoch": 0.1473261264708696,
+        "grad_norm": 0.4836377501487732,
+        "learning_rate": 0.0005,
+        "loss": 1.3829,
+        "step": 4620
+      },
+      {
+        "epoch": 0.14764501419050352,
+        "grad_norm": 0.4946225583553314,
+        "learning_rate": 0.0005,
+        "loss": 1.3766,
+        "step": 4630
+      },
+      {
+        "epoch": 0.14796390191013745,
+        "grad_norm": 0.4711849093437195,
+        "learning_rate": 0.0005,
+        "loss": 1.3866,
+        "step": 4640
+      },
+      {
+        "epoch": 0.14828278962977134,
+        "grad_norm": 0.5586421489715576,
+        "learning_rate": 0.0005,
+        "loss": 1.383,
+        "step": 4650
+      },
+      {
+        "epoch": 0.14860167734940527,
+        "grad_norm": 0.5114756226539612,
+        "learning_rate": 0.0005,
+        "loss": 1.381,
+        "step": 4660
+      },
+      {
+        "epoch": 0.1489205650690392,
+        "grad_norm": 0.526611864566803,
+        "learning_rate": 0.0005,
+        "loss": 1.3658,
+        "step": 4670
+      },
+      {
+        "epoch": 0.14923945278867312,
+        "grad_norm": 0.4793586730957031,
+        "learning_rate": 0.0005,
+        "loss": 1.3788,
+        "step": 4680
+      },
+      {
+        "epoch": 0.14955834050830702,
+        "grad_norm": 0.5218145251274109,
+        "learning_rate": 0.0005,
+        "loss": 1.3764,
+        "step": 4690
+      },
+      {
+        "epoch": 0.14987722822794095,
+        "grad_norm": 0.4978809058666229,
+        "learning_rate": 0.0005,
+        "loss": 1.3516,
+        "step": 4700
+      },
+      {
+        "epoch": 0.15019611594757487,
+        "grad_norm": 0.4808703362941742,
+        "learning_rate": 0.0005,
+        "loss": 1.3671,
+        "step": 4710
+      },
+      {
+        "epoch": 0.15051500366720877,
+        "grad_norm": 0.5049744248390198,
+        "learning_rate": 0.0005,
+        "loss": 1.3551,
+        "step": 4720
+      },
+      {
+        "epoch": 0.1508338913868427,
+        "grad_norm": 0.4735998213291168,
+        "learning_rate": 0.0005,
+        "loss": 1.3686,
+        "step": 4730
+      },
+      {
+        "epoch": 0.15115277910647662,
+        "grad_norm": 0.5061535835266113,
+        "learning_rate": 0.0005,
+        "loss": 1.3678,
+        "step": 4740
+      },
+      {
+        "epoch": 0.15147166682611052,
+        "grad_norm": 0.5199115872383118,
+        "learning_rate": 0.0005,
+        "loss": 1.3485,
+        "step": 4750
+      },
+      {
+        "epoch": 0.15179055454574444,
+        "grad_norm": 0.484674334526062,
+        "learning_rate": 0.0005,
+        "loss": 1.3521,
+        "step": 4760
+      },
+      {
+        "epoch": 0.15210944226537837,
+        "grad_norm": 0.5155425071716309,
+        "learning_rate": 0.0005,
+        "loss": 1.3646,
+        "step": 4770
+      },
+      {
+        "epoch": 0.15242832998501227,
+        "grad_norm": 0.514416515827179,
+        "learning_rate": 0.0005,
+        "loss": 1.3557,
+        "step": 4780
+      },
+      {
+        "epoch": 0.1527472177046462,
+        "grad_norm": 0.5200762152671814,
+        "learning_rate": 0.0005,
+        "loss": 1.3765,
+        "step": 4790
+      },
+      {
+        "epoch": 0.15306610542428012,
+        "grad_norm": 0.507409393787384,
+        "learning_rate": 0.0005,
+        "loss": 1.3558,
+        "step": 4800
+      },
+      {
+        "epoch": 0.15338499314391402,
+        "grad_norm": 0.47993403673171997,
+        "learning_rate": 0.0005,
+        "loss": 1.3741,
+        "step": 4810
+      },
+      {
+        "epoch": 0.15370388086354794,
+        "grad_norm": 0.48537927865982056,
+        "learning_rate": 0.0005,
+        "loss": 1.3639,
+        "step": 4820
+      },
+      {
+        "epoch": 0.15402276858318187,
+        "grad_norm": 0.47542354464530945,
+        "learning_rate": 0.0005,
+        "loss": 1.3495,
+        "step": 4830
+      },
+      {
+        "epoch": 0.15434165630281577,
+        "grad_norm": 0.4669359624385834,
+        "learning_rate": 0.0005,
+        "loss": 1.3683,
+        "step": 4840
+      },
+      {
+        "epoch": 0.1546605440224497,
+        "grad_norm": 0.4673759937286377,
+        "learning_rate": 0.0005,
+        "loss": 1.3621,
+        "step": 4850
+      },
+      {
+        "epoch": 0.15497943174208362,
+        "grad_norm": 0.47257837653160095,
+        "learning_rate": 0.0005,
+        "loss": 1.3612,
+        "step": 4860
+      },
+      {
+        "epoch": 0.15529831946171752,
+        "grad_norm": 0.5429351925849915,
+        "learning_rate": 0.0005,
+        "loss": 1.3419,
+        "step": 4870
+      },
+      {
+        "epoch": 0.15561720718135144,
+        "grad_norm": 0.48105567693710327,
+        "learning_rate": 0.0005,
+        "loss": 1.3404,
+        "step": 4880
+      },
+      {
+        "epoch": 0.15593609490098537,
+        "grad_norm": 0.49502962827682495,
+        "learning_rate": 0.0005,
+        "loss": 1.3621,
+        "step": 4890
+      },
+      {
+        "epoch": 0.1562549826206193,
+        "grad_norm": 0.4787370562553406,
+        "learning_rate": 0.0005,
+        "loss": 1.3659,
+        "step": 4900
+      },
+      {
+        "epoch": 0.1565738703402532,
+        "grad_norm": 0.5000290870666504,
+        "learning_rate": 0.0005,
+        "loss": 1.3689,
+        "step": 4910
+      },
+      {
+        "epoch": 0.15689275805988712,
+        "grad_norm": 0.48688530921936035,
+        "learning_rate": 0.0005,
+        "loss": 1.3592,
+        "step": 4920
+      },
+      {
+        "epoch": 0.15721164577952104,
+        "grad_norm": 0.49029314517974854,
+        "learning_rate": 0.0005,
+        "loss": 1.3672,
+        "step": 4930
+      },
+      {
+        "epoch": 0.15753053349915494,
+        "grad_norm": 0.4770297408103943,
+        "learning_rate": 0.0005,
+        "loss": 1.3401,
+        "step": 4940
+      },
+      {
+        "epoch": 0.15784942121878887,
+        "grad_norm": 0.485786497592926,
+        "learning_rate": 0.0005,
+        "loss": 1.3362,
+        "step": 4950
+      },
+      {
+        "epoch": 0.1581683089384228,
+        "grad_norm": 0.51774001121521,
+        "learning_rate": 0.0005,
+        "loss": 1.3643,
+        "step": 4960
+      },
+      {
+        "epoch": 0.1584871966580567,
+        "grad_norm": 0.4883527159690857,
+        "learning_rate": 0.0005,
+        "loss": 1.3578,
+        "step": 4970
+      },
+      {
+        "epoch": 0.15880608437769062,
+        "grad_norm": 0.4808884859085083,
+        "learning_rate": 0.0005,
+        "loss": 1.3452,
+        "step": 4980
+      },
+      {
+        "epoch": 0.15912497209732454,
+        "grad_norm": 0.48574575781822205,
+        "learning_rate": 0.0005,
+        "loss": 1.3456,
+        "step": 4990
+      },
+      {
+        "epoch": 0.15944385981695844,
+        "grad_norm": 0.4806765615940094,
+        "learning_rate": 0.0005,
+        "loss": 1.3183,
+        "step": 5000
+      },
+      {
+        "epoch": 0.15976274753659236,
+        "grad_norm": 0.47405245900154114,
+        "learning_rate": 0.0005,
+        "loss": 1.3571,
+        "step": 5010
+      },
+      {
+        "epoch": 0.1600816352562263,
+        "grad_norm": 0.5027293562889099,
+        "learning_rate": 0.0005,
+        "loss": 1.3413,
+        "step": 5020
+      },
+      {
+        "epoch": 0.1604005229758602,
+        "grad_norm": 0.5056706666946411,
+        "learning_rate": 0.0005,
+        "loss": 1.3807,
+        "step": 5030
+      },
+      {
+        "epoch": 0.1607194106954941,
+        "grad_norm": 0.47428926825523376,
+        "learning_rate": 0.0005,
+        "loss": 1.3603,
+        "step": 5040
+      },
+      {
+        "epoch": 0.16103829841512804,
+        "grad_norm": 0.49223700165748596,
+        "learning_rate": 0.0005,
+        "loss": 1.343,
+        "step": 5050
+      },
+      {
+        "epoch": 0.16135718613476194,
+        "grad_norm": 0.4843858778476715,
+        "learning_rate": 0.0005,
+        "loss": 1.3467,
+        "step": 5060
+      },
+      {
+        "epoch": 0.16167607385439586,
+        "grad_norm": 0.4874281585216522,
+        "learning_rate": 0.0005,
+        "loss": 1.3374,
+        "step": 5070
+      },
+      {
+        "epoch": 0.1619949615740298,
+        "grad_norm": 0.5039829611778259,
+        "learning_rate": 0.0005,
+        "loss": 1.3454,
+        "step": 5080
+      },
+      {
+        "epoch": 0.16231384929366371,
+        "grad_norm": 0.4736151099205017,
+        "learning_rate": 0.0005,
+        "loss": 1.366,
+        "step": 5090
+      },
+      {
+        "epoch": 0.1626327370132976,
+        "grad_norm": 0.4735645651817322,
+        "learning_rate": 0.0005,
+        "loss": 1.3539,
+        "step": 5100
+      },
+      {
+        "epoch": 0.16295162473293154,
+        "grad_norm": 0.4740479588508606,
+        "learning_rate": 0.0005,
+        "loss": 1.3478,
+        "step": 5110
+      },
+      {
+        "epoch": 0.16327051245256546,
+        "grad_norm": 0.4920395314693451,
+        "learning_rate": 0.0005,
+        "loss": 1.3257,
+        "step": 5120
+      },
+      {
+        "epoch": 0.16358940017219936,
+        "grad_norm": 0.4727405607700348,
+        "learning_rate": 0.0005,
+        "loss": 1.346,
+        "step": 5130
+      },
+      {
+        "epoch": 0.1639082878918333,
+        "grad_norm": 0.4696497619152069,
+        "learning_rate": 0.0005,
+        "loss": 1.34,
+        "step": 5140
+      },
+      {
+        "epoch": 0.1642271756114672,
+        "grad_norm": 0.48133373260498047,
+        "learning_rate": 0.0005,
+        "loss": 1.3443,
+        "step": 5150
+      },
+      {
+        "epoch": 0.1645460633311011,
+        "grad_norm": 0.5079322457313538,
+        "learning_rate": 0.0005,
+        "loss": 1.3423,
+        "step": 5160
+      },
+      {
+        "epoch": 0.16486495105073504,
+        "grad_norm": 0.4736444056034088,
+        "learning_rate": 0.0005,
+        "loss": 1.3488,
+        "step": 5170
+      },
+      {
+        "epoch": 0.16518383877036896,
+        "grad_norm": 0.4801396131515503,
+        "learning_rate": 0.0005,
+        "loss": 1.3514,
+        "step": 5180
+      },
+      {
+        "epoch": 0.16550272649000286,
+        "grad_norm": 0.4806341230869293,
+        "learning_rate": 0.0005,
+        "loss": 1.3682,
+        "step": 5190
+      },
+      {
+        "epoch": 0.1658216142096368,
+        "grad_norm": 0.46117836236953735,
+        "learning_rate": 0.0005,
+        "loss": 1.3248,
+        "step": 5200
+      },
+      {
+        "epoch": 0.1661405019292707,
+        "grad_norm": 0.47865962982177734,
+        "learning_rate": 0.0005,
+        "loss": 1.3268,
+        "step": 5210
+      },
+      {
+        "epoch": 0.1664593896489046,
+        "grad_norm": 0.47703343629837036,
+        "learning_rate": 0.0005,
+        "loss": 1.3303,
+        "step": 5220
+      },
+      {
+        "epoch": 0.16677827736853854,
+        "grad_norm": 0.5019536018371582,
+        "learning_rate": 0.0005,
+        "loss": 1.3262,
+        "step": 5230
+      },
+      {
+        "epoch": 0.16709716508817246,
+        "grad_norm": 0.4716322124004364,
+        "learning_rate": 0.0005,
+        "loss": 1.3411,
+        "step": 5240
+      },
+      {
+        "epoch": 0.16741605280780636,
+        "grad_norm": 0.4730423092842102,
+        "learning_rate": 0.0005,
+        "loss": 1.3355,
+        "step": 5250
+      },
+      {
+        "epoch": 0.16773494052744028,
+        "grad_norm": 0.4925325810909271,
+        "learning_rate": 0.0005,
+        "loss": 1.3315,
+        "step": 5260
+      },
+      {
+        "epoch": 0.1680538282470742,
+        "grad_norm": 0.48779094219207764,
+        "learning_rate": 0.0005,
+        "loss": 1.3214,
+        "step": 5270
+      },
+      {
+        "epoch": 0.1683727159667081,
+        "grad_norm": 0.4637846350669861,
+        "learning_rate": 0.0005,
+        "loss": 1.3366,
+        "step": 5280
+      },
+      {
+        "epoch": 0.16869160368634203,
+        "grad_norm": 0.48783788084983826,
+        "learning_rate": 0.0005,
+        "loss": 1.3301,
+        "step": 5290
+      },
+      {
+        "epoch": 0.16901049140597596,
+        "grad_norm": 0.47115302085876465,
+        "learning_rate": 0.0005,
+        "loss": 1.3293,
+        "step": 5300
+      },
+      {
+        "epoch": 0.16932937912560989,
+        "grad_norm": 0.4882822334766388,
+        "learning_rate": 0.0005,
+        "loss": 1.3382,
+        "step": 5310
+      },
+      {
+        "epoch": 0.16964826684524378,
+        "grad_norm": 0.4872843027114868,
+        "learning_rate": 0.0005,
+        "loss": 1.331,
+        "step": 5320
+      },
+      {
+        "epoch": 0.1699671545648777,
+        "grad_norm": 0.46673768758773804,
+        "learning_rate": 0.0005,
+        "loss": 1.3439,
+        "step": 5330
+      },
+      {
+        "epoch": 0.17028604228451164,
+        "grad_norm": 0.4746752083301544,
+        "learning_rate": 0.0005,
+        "loss": 1.3559,
+        "step": 5340
+      },
+      {
+        "epoch": 0.17060493000414553,
+        "grad_norm": 0.48173803091049194,
+        "learning_rate": 0.0005,
+        "loss": 1.3302,
+        "step": 5350
+      },
+      {
+        "epoch": 0.17092381772377946,
+        "grad_norm": 0.4860802888870239,
+        "learning_rate": 0.0005,
+        "loss": 1.3315,
+        "step": 5360
+      },
+      {
+        "epoch": 0.17124270544341338,
+        "grad_norm": 0.4615638256072998,
+        "learning_rate": 0.0005,
+        "loss": 1.3175,
+        "step": 5370
+      },
+      {
+        "epoch": 0.17156159316304728,
+        "grad_norm": 0.4676917791366577,
+        "learning_rate": 0.0005,
+        "loss": 1.3037,
+        "step": 5380
+      },
+      {
+        "epoch": 0.1718804808826812,
+        "grad_norm": 0.4855146110057831,
+        "learning_rate": 0.0005,
+        "loss": 1.3231,
+        "step": 5390
+      },
+      {
+        "epoch": 0.17219936860231513,
+        "grad_norm": 0.48758402466773987,
+        "learning_rate": 0.0005,
+        "loss": 1.3519,
+        "step": 5400
+      },
+      {
+        "epoch": 0.17251825632194903,
+        "grad_norm": 0.4420989751815796,
+        "learning_rate": 0.0005,
+        "loss": 1.317,
+        "step": 5410
+      },
+      {
+        "epoch": 0.17283714404158296,
+        "grad_norm": 0.4432215690612793,
+        "learning_rate": 0.0005,
+        "loss": 1.3363,
+        "step": 5420
+      },
+      {
+        "epoch": 0.17315603176121688,
+        "grad_norm": 0.5140141844749451,
+        "learning_rate": 0.0005,
+        "loss": 1.3301,
+        "step": 5430
+      },
+      {
+        "epoch": 0.17347491948085078,
+        "grad_norm": 0.4605754613876343,
+        "learning_rate": 0.0005,
+        "loss": 1.3343,
+        "step": 5440
+      },
+      {
+        "epoch": 0.1737938072004847,
+        "grad_norm": 0.46923375129699707,
+        "learning_rate": 0.0005,
+        "loss": 1.3341,
+        "step": 5450
+      },
+      {
+        "epoch": 0.17411269492011863,
+        "grad_norm": 0.4824751019477844,
+        "learning_rate": 0.0005,
+        "loss": 1.3169,
+        "step": 5460
+      },
+      {
+        "epoch": 0.17443158263975253,
+        "grad_norm": 0.4664471745491028,
+        "learning_rate": 0.0005,
+        "loss": 1.3375,
+        "step": 5470
+      },
+      {
+        "epoch": 0.17475047035938646,
+        "grad_norm": 0.48443225026130676,
+        "learning_rate": 0.0005,
+        "loss": 1.3362,
+        "step": 5480
+      },
+      {
+        "epoch": 0.17506935807902038,
+        "grad_norm": 0.46908897161483765,
+        "learning_rate": 0.0005,
+        "loss": 1.3,
+        "step": 5490
+      },
+      {
+        "epoch": 0.1753882457986543,
+        "grad_norm": 0.4562680125236511,
+        "learning_rate": 0.0005,
+        "loss": 1.2999,
+        "step": 5500
+      },
+      {
+        "epoch": 0.1757071335182882,
+        "grad_norm": 0.46892642974853516,
+        "learning_rate": 0.0005,
+        "loss": 1.3357,
+        "step": 5510
+      },
+      {
+        "epoch": 0.17602602123792213,
+        "grad_norm": 0.5013911724090576,
+        "learning_rate": 0.0005,
+        "loss": 1.3326,
+        "step": 5520
+      },
+      {
+        "epoch": 0.17634490895755606,
+        "grad_norm": 0.45573681592941284,
+        "learning_rate": 0.0005,
+        "loss": 1.3149,
+        "step": 5530
+      },
+      {
+        "epoch": 0.17666379667718995,
+        "grad_norm": 0.4677056074142456,
+        "learning_rate": 0.0005,
+        "loss": 1.3216,
+        "step": 5540
+      },
+      {
+        "epoch": 0.17698268439682388,
+        "grad_norm": 0.4879295527935028,
+        "learning_rate": 0.0005,
+        "loss": 1.3337,
+        "step": 5550
+      },
+      {
+        "epoch": 0.1773015721164578,
+        "grad_norm": 0.4852617084980011,
+        "learning_rate": 0.0005,
+        "loss": 1.3217,
+        "step": 5560
+      },
+      {
+        "epoch": 0.1776204598360917,
+        "grad_norm": 0.4770294725894928,
+        "learning_rate": 0.0005,
+        "loss": 1.3143,
+        "step": 5570
+      },
+      {
+        "epoch": 0.17793934755572563,
+        "grad_norm": 0.4696557819843292,
+        "learning_rate": 0.0005,
+        "loss": 1.3031,
+        "step": 5580
+      },
+      {
+        "epoch": 0.17825823527535956,
+        "grad_norm": 0.4751479923725128,
+        "learning_rate": 0.0005,
+        "loss": 1.3371,
+        "step": 5590
+      },
+      {
+        "epoch": 0.17857712299499345,
+        "grad_norm": 0.4551768898963928,
+        "learning_rate": 0.0005,
+        "loss": 1.3293,
+        "step": 5600
+      },
+      {
+        "epoch": 0.17889601071462738,
+        "grad_norm": 0.4669037461280823,
+        "learning_rate": 0.0005,
+        "loss": 1.3207,
+        "step": 5610
+      },
+      {
+        "epoch": 0.1792148984342613,
+        "grad_norm": 0.4847627878189087,
+        "learning_rate": 0.0005,
+        "loss": 1.3393,
+        "step": 5620
+      },
+      {
+        "epoch": 0.1795337861538952,
+        "grad_norm": 0.4655326008796692,
+        "learning_rate": 0.0005,
+        "loss": 1.3244,
+        "step": 5630
+      },
+      {
+        "epoch": 0.17985267387352913,
+        "grad_norm": 0.447279691696167,
+        "learning_rate": 0.0005,
+        "loss": 1.3149,
+        "step": 5640
+      },
+      {
+        "epoch": 0.18017156159316305,
+        "grad_norm": 0.45796069502830505,
+        "learning_rate": 0.0005,
+        "loss": 1.3346,
+        "step": 5650
+      },
+      {
+        "epoch": 0.18049044931279695,
+        "grad_norm": 0.4523908197879791,
+        "learning_rate": 0.0005,
+        "loss": 1.3209,
+        "step": 5660
+      },
+      {
+        "epoch": 0.18080933703243088,
+        "grad_norm": 0.46900004148483276,
+        "learning_rate": 0.0005,
+        "loss": 1.325,
+        "step": 5670
+      },
+      {
+        "epoch": 0.1811282247520648,
+        "grad_norm": 0.457965612411499,
+        "learning_rate": 0.0005,
+        "loss": 1.3031,
+        "step": 5680
+      },
+      {
+        "epoch": 0.1814471124716987,
+        "grad_norm": 0.4391346871852875,
+        "learning_rate": 0.0005,
+        "loss": 1.3066,
+        "step": 5690
+      },
+      {
+        "epoch": 0.18176600019133263,
+        "grad_norm": 0.44087544083595276,
+        "learning_rate": 0.0005,
+        "loss": 1.3129,
+        "step": 5700
+      },
+      {
+        "epoch": 0.18208488791096655,
+        "grad_norm": 0.46599528193473816,
+        "learning_rate": 0.0005,
+        "loss": 1.3201,
+        "step": 5710
+      },
+      {
+        "epoch": 0.18240377563060048,
+        "grad_norm": 0.44314372539520264,
+        "learning_rate": 0.0005,
+        "loss": 1.3357,
+        "step": 5720
+      },
+      {
+        "epoch": 0.18272266335023438,
+        "grad_norm": 0.4564199447631836,
+        "learning_rate": 0.0005,
+        "loss": 1.2977,
+        "step": 5730
+      },
+      {
+        "epoch": 0.1830415510698683,
+        "grad_norm": 0.45973026752471924,
+        "learning_rate": 0.0005,
+        "loss": 1.2927,
+        "step": 5740
+      },
+      {
+        "epoch": 0.18336043878950223,
+        "grad_norm": 0.4481637477874756,
+        "learning_rate": 0.0005,
+        "loss": 1.33,
+        "step": 5750
+      },
+      {
+        "epoch": 0.18367932650913613,
+        "grad_norm": 0.4800594747066498,
+        "learning_rate": 0.0005,
+        "loss": 1.313,
+        "step": 5760
+      },
+      {
+        "epoch": 0.18399821422877005,
+        "grad_norm": 0.49500760436058044,
+        "learning_rate": 0.0005,
+        "loss": 1.3104,
+        "step": 5770
+      },
+      {
+        "epoch": 0.18431710194840398,
+        "grad_norm": 0.4503342807292938,
+        "learning_rate": 0.0005,
+        "loss": 1.3013,
+        "step": 5780
+      },
+      {
+        "epoch": 0.18463598966803788,
+        "grad_norm": 0.451404869556427,
+        "learning_rate": 0.0005,
+        "loss": 1.3019,
+        "step": 5790
+      },
+      {
+        "epoch": 0.1849548773876718,
+        "grad_norm": 0.4840630888938904,
+        "learning_rate": 0.0005,
+        "loss": 1.3129,
+        "step": 5800
+      },
+      {
+        "epoch": 0.18527376510730573,
+        "grad_norm": 0.45992785692214966,
+        "learning_rate": 0.0005,
+        "loss": 1.3124,
+        "step": 5810
+      },
+      {
+        "epoch": 0.18559265282693962,
+        "grad_norm": 0.44602009654045105,
+        "learning_rate": 0.0005,
+        "loss": 1.3026,
+        "step": 5820
+      },
+      {
+        "epoch": 0.18591154054657355,
+        "grad_norm": 0.44860562682151794,
+        "learning_rate": 0.0005,
+        "loss": 1.3054,
+        "step": 5830
+      },
+      {
+        "epoch": 0.18623042826620748,
+        "grad_norm": 0.5289995074272156,
+        "learning_rate": 0.0005,
+        "loss": 1.306,
+        "step": 5840
+      },
+      {
+        "epoch": 0.18654931598584137,
+        "grad_norm": 0.4493541717529297,
+        "learning_rate": 0.0005,
+        "loss": 1.3108,
+        "step": 5850
+      },
+      {
+        "epoch": 0.1868682037054753,
+        "grad_norm": 0.454628586769104,
+        "learning_rate": 0.0005,
+        "loss": 1.3098,
+        "step": 5860
+      },
+      {
+        "epoch": 0.18718709142510923,
+        "grad_norm": 0.46552345156669617,
+        "learning_rate": 0.0005,
+        "loss": 1.2977,
+        "step": 5870
+      },
+      {
+        "epoch": 0.18750597914474312,
+        "grad_norm": 0.45460376143455505,
+        "learning_rate": 0.0005,
+        "loss": 1.3215,
+        "step": 5880
+      },
+      {
+        "epoch": 0.18782486686437705,
+        "grad_norm": 0.46688956022262573,
+        "learning_rate": 0.0005,
+        "loss": 1.3116,
+        "step": 5890
+      },
+      {
+        "epoch": 0.18814375458401097,
+        "grad_norm": 0.45595839619636536,
+        "learning_rate": 0.0005,
+        "loss": 1.3025,
+        "step": 5900
+      },
+      {
+        "epoch": 0.18846264230364487,
+        "grad_norm": 0.457356333732605,
+        "learning_rate": 0.0005,
+        "loss": 1.2995,
+        "step": 5910
+      },
+      {
+        "epoch": 0.1887815300232788,
+        "grad_norm": 0.44302088022232056,
+        "learning_rate": 0.0005,
+        "loss": 1.3031,
+        "step": 5920
+      },
+      {
+        "epoch": 0.18910041774291272,
+        "grad_norm": 0.46125978231430054,
+        "learning_rate": 0.0005,
+        "loss": 1.3006,
+        "step": 5930
+      },
+      {
+        "epoch": 0.18941930546254665,
+        "grad_norm": 0.45474714040756226,
+        "learning_rate": 0.0005,
+        "loss": 1.3132,
+        "step": 5940
+      },
+      {
+        "epoch": 0.18973819318218055,
+        "grad_norm": 0.4790920913219452,
+        "learning_rate": 0.0005,
+        "loss": 1.2941,
+        "step": 5950
+      },
+      {
+        "epoch": 0.19005708090181447,
+        "grad_norm": 0.4711848795413971,
+        "learning_rate": 0.0005,
+        "loss": 1.3296,
+        "step": 5960
+      },
+      {
+        "epoch": 0.1903759686214484,
+        "grad_norm": 0.45707839727401733,
+        "learning_rate": 0.0005,
+        "loss": 1.3104,
+        "step": 5970
+      },
+      {
+        "epoch": 0.1906948563410823,
+        "grad_norm": 0.47939345240592957,
+        "learning_rate": 0.0005,
+        "loss": 1.2997,
+        "step": 5980
+      },
+      {
+        "epoch": 0.19101374406071622,
+        "grad_norm": 0.4554172158241272,
+        "learning_rate": 0.0005,
+        "loss": 1.3091,
+        "step": 5990
+      },
+      {
+        "epoch": 0.19133263178035015,
+        "grad_norm": 0.45858779549598694,
+        "learning_rate": 0.0005,
+        "loss": 1.2891,
+        "step": 6000
+      },
+      {
+        "epoch": 0.19165151949998405,
+        "grad_norm": 0.4406243860721588,
+        "learning_rate": 0.0005,
+        "loss": 1.3123,
+        "step": 6010
+      },
+      {
+        "epoch": 0.19197040721961797,
+        "grad_norm": 0.4473991096019745,
+        "learning_rate": 0.0005,
+        "loss": 1.2913,
+        "step": 6020
+      },
+      {
+        "epoch": 0.1922892949392519,
+        "grad_norm": 0.4342772960662842,
+        "learning_rate": 0.0005,
+        "loss": 1.2756,
+        "step": 6030
+      },
+      {
+        "epoch": 0.1926081826588858,
+        "grad_norm": 0.4456859827041626,
+        "learning_rate": 0.0005,
+        "loss": 1.2947,
+        "step": 6040
+      },
+      {
+        "epoch": 0.19292707037851972,
+        "grad_norm": 0.443399578332901,
+        "learning_rate": 0.0005,
+        "loss": 1.2933,
+        "step": 6050
+      },
+      {
+        "epoch": 0.19324595809815365,
+        "grad_norm": 0.45340052247047424,
+        "learning_rate": 0.0005,
+        "loss": 1.2876,
+        "step": 6060
+      },
+      {
+        "epoch": 0.19356484581778755,
+        "grad_norm": 0.48477765917778015,
+        "learning_rate": 0.0005,
+        "loss": 1.2951,
+        "step": 6070
+      },
+      {
+        "epoch": 0.19388373353742147,
+        "grad_norm": 0.4576895534992218,
+        "learning_rate": 0.0005,
+        "loss": 1.2933,
+        "step": 6080
+      },
+      {
+        "epoch": 0.1942026212570554,
+        "grad_norm": 0.46342259645462036,
+        "learning_rate": 0.0005,
+        "loss": 1.2931,
+        "step": 6090
+      },
+      {
+        "epoch": 0.1945215089766893,
+        "grad_norm": 0.4377509355545044,
+        "learning_rate": 0.0005,
+        "loss": 1.3003,
+        "step": 6100
+      },
+      {
+        "epoch": 0.19484039669632322,
+        "grad_norm": 0.47648733854293823,
+        "learning_rate": 0.0005,
+        "loss": 1.3055,
+        "step": 6110
+      },
+      {
+        "epoch": 0.19515928441595715,
+        "grad_norm": 0.47392523288726807,
+        "learning_rate": 0.0005,
+        "loss": 1.2955,
+        "step": 6120
+      },
+      {
+        "epoch": 0.19547817213559107,
+        "grad_norm": 0.44861629605293274,
+        "learning_rate": 0.0005,
+        "loss": 1.2736,
+        "step": 6130
+      },
+      {
+        "epoch": 0.19579705985522497,
+        "grad_norm": 0.45019152760505676,
+        "learning_rate": 0.0005,
+        "loss": 1.3001,
+        "step": 6140
+      },
+      {
+        "epoch": 0.1961159475748589,
+        "grad_norm": 0.43096092343330383,
+        "learning_rate": 0.0005,
+        "loss": 1.2678,
+        "step": 6150
+      },
+      {
+        "epoch": 0.19643483529449282,
+        "grad_norm": 0.4468885064125061,
+        "learning_rate": 0.0005,
+        "loss": 1.2742,
+        "step": 6160
+      },
+      {
+        "epoch": 0.19675372301412672,
+        "grad_norm": 0.44142505526542664,
+        "learning_rate": 0.0005,
+        "loss": 1.3079,
+        "step": 6170
+      },
+      {
+        "epoch": 0.19707261073376064,
+        "grad_norm": 0.4425607919692993,
+        "learning_rate": 0.0005,
+        "loss": 1.2885,
+        "step": 6180
+      },
+      {
+        "epoch": 0.19739149845339457,
+        "grad_norm": 0.45576775074005127,
+        "learning_rate": 0.0005,
+        "loss": 1.3119,
+        "step": 6190
+      },
+      {
+        "epoch": 0.19771038617302847,
+        "grad_norm": 0.44537460803985596,
+        "learning_rate": 0.0005,
+        "loss": 1.2937,
+        "step": 6200
+      },
+      {
+        "epoch": 0.1980292738926624,
+        "grad_norm": 0.46090060472488403,
+        "learning_rate": 0.0005,
+        "loss": 1.2971,
+        "step": 6210
+      },
+      {
+        "epoch": 0.19834816161229632,
+        "grad_norm": 0.44849535822868347,
+        "learning_rate": 0.0005,
+        "loss": 1.2699,
+        "step": 6220
+      },
+      {
+        "epoch": 0.19866704933193022,
+        "grad_norm": 0.4581367075443268,
+        "learning_rate": 0.0005,
+        "loss": 1.276,
+        "step": 6230
+      },
+      {
+        "epoch": 0.19898593705156414,
+        "grad_norm": 0.4342295825481415,
+        "learning_rate": 0.0005,
+        "loss": 1.284,
+        "step": 6240
+      },
+      {
+        "epoch": 0.19930482477119807,
+        "grad_norm": 0.4359763264656067,
+        "learning_rate": 0.0005,
+        "loss": 1.3001,
+        "step": 6250
+      },
+      {
+        "epoch": 0.19962371249083197,
+        "grad_norm": 0.4430510103702545,
+        "learning_rate": 0.0005,
+        "loss": 1.2748,
+        "step": 6260
+      },
+      {
+        "epoch": 0.1999426002104659,
+        "grad_norm": 0.4533330202102661,
+        "learning_rate": 0.0005,
+        "loss": 1.2926,
+        "step": 6270
+      },
+      {
+        "epoch": 0.20026148793009982,
+        "grad_norm": 0.45962098240852356,
+        "learning_rate": 0.0005,
+        "loss": 1.2872,
+        "step": 6280
+      },
+      {
+        "epoch": 0.20058037564973372,
+        "grad_norm": 0.4355621039867401,
+        "learning_rate": 0.0005,
+        "loss": 1.3021,
+        "step": 6290
+      },
+      {
+        "epoch": 0.20089926336936764,
+        "grad_norm": 0.4176916182041168,
+        "learning_rate": 0.0005,
+        "loss": 1.2885,
+        "step": 6300
+      },
+      {
+        "epoch": 0.20121815108900157,
+        "grad_norm": 0.4410490393638611,
+        "learning_rate": 0.0005,
+        "loss": 1.2827,
+        "step": 6310
+      },
+      {
+        "epoch": 0.20153703880863547,
+        "grad_norm": 0.45107796788215637,
+        "learning_rate": 0.0005,
+        "loss": 1.2835,
+        "step": 6320
+      },
+      {
+        "epoch": 0.2018559265282694,
+        "grad_norm": 0.44999945163726807,
+        "learning_rate": 0.0005,
+        "loss": 1.2805,
+        "step": 6330
+      },
+      {
+        "epoch": 0.20217481424790332,
+        "grad_norm": 0.46005144715309143,
+        "learning_rate": 0.0005,
+        "loss": 1.2857,
+        "step": 6340
+      },
+      {
+        "epoch": 0.20249370196753724,
+        "grad_norm": 0.44353771209716797,
+        "learning_rate": 0.0005,
+        "loss": 1.2774,
+        "step": 6350
+      },
+      {
+        "epoch": 0.20281258968717114,
+        "grad_norm": 0.4384526014328003,
+        "learning_rate": 0.0005,
+        "loss": 1.2871,
+        "step": 6360
+      },
+      {
+        "epoch": 0.20313147740680507,
+        "grad_norm": 0.43561914563179016,
+        "learning_rate": 0.0005,
+        "loss": 1.2985,
+        "step": 6370
+      },
+      {
+        "epoch": 0.203450365126439,
+        "grad_norm": 0.4324412941932678,
+        "learning_rate": 0.0005,
+        "loss": 1.2948,
+        "step": 6380
+      },
+      {
+        "epoch": 0.2037692528460729,
+        "grad_norm": 0.4382753372192383,
+        "learning_rate": 0.0005,
+        "loss": 1.2745,
+        "step": 6390
+      },
+      {
+        "epoch": 0.20408814056570682,
+        "grad_norm": 0.4647562503814697,
+        "learning_rate": 0.0005,
+        "loss": 1.2706,
+        "step": 6400
+      },
+      {
+        "epoch": 0.20440702828534074,
+        "grad_norm": 0.4480821490287781,
+        "learning_rate": 0.0005,
+        "loss": 1.2923,
+        "step": 6410
+      },
+      {
+        "epoch": 0.20472591600497464,
+        "grad_norm": 0.4176822304725647,
+        "learning_rate": 0.0005,
+        "loss": 1.2749,
+        "step": 6420
+      },
+      {
+        "epoch": 0.20504480372460857,
+        "grad_norm": 0.4394923448562622,
+        "learning_rate": 0.0005,
+        "loss": 1.301,
+        "step": 6430
+      },
+      {
+        "epoch": 0.2053636914442425,
+        "grad_norm": 0.43086984753608704,
+        "learning_rate": 0.0005,
+        "loss": 1.2803,
+        "step": 6440
+      },
+      {
+        "epoch": 0.2056825791638764,
+        "grad_norm": 0.4294069707393646,
+        "learning_rate": 0.0005,
+        "loss": 1.2733,
+        "step": 6450
+      },
+      {
+        "epoch": 0.20600146688351031,
+        "grad_norm": 0.42091456055641174,
+        "learning_rate": 0.0005,
+        "loss": 1.2691,
+        "step": 6460
+      },
+      {
+        "epoch": 0.20632035460314424,
+        "grad_norm": 0.45179837942123413,
+        "learning_rate": 0.0005,
+        "loss": 1.2874,
+        "step": 6470
+      },
+      {
+        "epoch": 0.20663924232277814,
+        "grad_norm": 0.4361703395843506,
+        "learning_rate": 0.0005,
+        "loss": 1.284,
+        "step": 6480
+      },
+      {
+        "epoch": 0.20695813004241206,
+        "grad_norm": 0.45856398344039917,
+        "learning_rate": 0.0005,
+        "loss": 1.2815,
+        "step": 6490
+      },
+      {
+        "epoch": 0.207277017762046,
+        "grad_norm": 0.44898521900177,
+        "learning_rate": 0.0005,
+        "loss": 1.2737,
+        "step": 6500
+      },
+      {
+        "epoch": 0.2075959054816799,
+        "grad_norm": 0.44692954421043396,
+        "learning_rate": 0.0005,
+        "loss": 1.2723,
+        "step": 6510
+      },
+      {
+        "epoch": 0.2079147932013138,
+        "grad_norm": 0.43304726481437683,
+        "learning_rate": 0.0005,
+        "loss": 1.2761,
+        "step": 6520
+      },
+      {
+        "epoch": 0.20823368092094774,
+        "grad_norm": 0.4356973469257355,
+        "learning_rate": 0.0005,
+        "loss": 1.2845,
+        "step": 6530
+      },
+      {
+        "epoch": 0.20855256864058166,
+        "grad_norm": 0.4488913416862488,
+        "learning_rate": 0.0005,
+        "loss": 1.3023,
+        "step": 6540
+      },
+      {
+        "epoch": 0.20887145636021556,
+        "grad_norm": 0.43722909688949585,
+        "learning_rate": 0.0005,
+        "loss": 1.2822,
+        "step": 6550
+      },
+      {
+        "epoch": 0.2091903440798495,
+        "grad_norm": 0.4295201301574707,
+        "learning_rate": 0.0005,
+        "loss": 1.2736,
+        "step": 6560
+      },
+      {
+        "epoch": 0.20950923179948341,
+        "grad_norm": 0.44216859340667725,
+        "learning_rate": 0.0005,
+        "loss": 1.2815,
+        "step": 6570
+      },
+      {
+        "epoch": 0.2098281195191173,
+        "grad_norm": 0.4219614267349243,
+        "learning_rate": 0.0005,
+        "loss": 1.2776,
+        "step": 6580
+      },
+      {
+        "epoch": 0.21014700723875124,
+        "grad_norm": 0.4117146134376526,
+        "learning_rate": 0.0005,
+        "loss": 1.2856,
+        "step": 6590
+      },
+      {
+        "epoch": 0.21046589495838516,
+        "grad_norm": 0.4342212975025177,
+        "learning_rate": 0.0005,
+        "loss": 1.2647,
+        "step": 6600
+      },
+      {
+        "epoch": 0.21078478267801906,
+        "grad_norm": 0.4356221556663513,
+        "learning_rate": 0.0005,
+        "loss": 1.2861,
+        "step": 6610
+      },
+      {
+        "epoch": 0.211103670397653,
+        "grad_norm": 0.4763599932193756,
+        "learning_rate": 0.0005,
+        "loss": 1.2701,
+        "step": 6620
+      },
+      {
+        "epoch": 0.2114225581172869,
+        "grad_norm": 0.44057396054267883,
+        "learning_rate": 0.0005,
+        "loss": 1.2817,
+        "step": 6630
+      },
+      {
+        "epoch": 0.2117414458369208,
+        "grad_norm": 0.42277610301971436,
+        "learning_rate": 0.0005,
+        "loss": 1.2917,
+        "step": 6640
+      },
+      {
+        "epoch": 0.21206033355655474,
+        "grad_norm": 0.42706525325775146,
+        "learning_rate": 0.0005,
+        "loss": 1.2664,
+        "step": 6650
+      },
+      {
+        "epoch": 0.21237922127618866,
+        "grad_norm": 0.43516626954078674,
+        "learning_rate": 0.0005,
+        "loss": 1.2785,
+        "step": 6660
+      },
+      {
+        "epoch": 0.21269810899582256,
+        "grad_norm": 0.4840092957019806,
+        "learning_rate": 0.0005,
+        "loss": 1.2756,
+        "step": 6670
+      },
+      {
+        "epoch": 0.21301699671545649,
+        "grad_norm": 0.44978827238082886,
+        "learning_rate": 0.0005,
+        "loss": 1.2698,
+        "step": 6680
+      },
+      {
+        "epoch": 0.2133358844350904,
+        "grad_norm": 0.45999205112457275,
+        "learning_rate": 0.0005,
+        "loss": 1.2845,
+        "step": 6690
+      },
+      {
+        "epoch": 0.2136547721547243,
+        "grad_norm": 0.45006537437438965,
+        "learning_rate": 0.0005,
+        "loss": 1.2754,
+        "step": 6700
+      },
+      {
+        "epoch": 0.21397365987435824,
+        "grad_norm": 0.4400864541530609,
+        "learning_rate": 0.0005,
+        "loss": 1.2611,
+        "step": 6710
+      },
+      {
+        "epoch": 0.21429254759399216,
+        "grad_norm": 0.44918856024742126,
+        "learning_rate": 0.0005,
+        "loss": 1.2746,
+        "step": 6720
+      },
+      {
+        "epoch": 0.21461143531362606,
+        "grad_norm": 0.4447942078113556,
+        "learning_rate": 0.0005,
+        "loss": 1.2667,
+        "step": 6730
+      },
+      {
+        "epoch": 0.21493032303325998,
+        "grad_norm": 0.41364383697509766,
+        "learning_rate": 0.0005,
+        "loss": 1.2737,
+        "step": 6740
+      },
+      {
+        "epoch": 0.2152492107528939,
+        "grad_norm": 0.43288302421569824,
+        "learning_rate": 0.0005,
+        "loss": 1.2889,
+        "step": 6750
+      },
+      {
+        "epoch": 0.21556809847252784,
+        "grad_norm": 0.4335830509662628,
+        "learning_rate": 0.0005,
+        "loss": 1.2803,
+        "step": 6760
+      },
+      {
+        "epoch": 0.21588698619216173,
+        "grad_norm": 0.4247308373451233,
+        "learning_rate": 0.0005,
+        "loss": 1.2649,
+        "step": 6770
+      },
+      {
+        "epoch": 0.21620587391179566,
+        "grad_norm": 0.4426180422306061,
+        "learning_rate": 0.0005,
+        "loss": 1.2554,
+        "step": 6780
+      },
+      {
+        "epoch": 0.21652476163142959,
+        "grad_norm": 0.43544715642929077,
+        "learning_rate": 0.0005,
+        "loss": 1.2535,
+        "step": 6790
+      },
+      {
+        "epoch": 0.21684364935106348,
+        "grad_norm": 0.46235576272010803,
+        "learning_rate": 0.0005,
+        "loss": 1.2466,
+        "step": 6800
+      },
+      {
+        "epoch": 0.2171625370706974,
+        "grad_norm": 0.41557857394218445,
+        "learning_rate": 0.0005,
+        "loss": 1.2806,
+        "step": 6810
+      },
+      {
+        "epoch": 0.21748142479033133,
+        "grad_norm": 0.4564860463142395,
+        "learning_rate": 0.0005,
+        "loss": 1.2688,
+        "step": 6820
+      },
+      {
+        "epoch": 0.21780031250996523,
+        "grad_norm": 0.4496130645275116,
+        "learning_rate": 0.0005,
+        "loss": 1.2671,
+        "step": 6830
+      },
+      {
+        "epoch": 0.21811920022959916,
+        "grad_norm": 0.4237511157989502,
+        "learning_rate": 0.0005,
+        "loss": 1.2684,
+        "step": 6840
+      },
+      {
+        "epoch": 0.21843808794923308,
+        "grad_norm": 0.4281541109085083,
+        "learning_rate": 0.0005,
+        "loss": 1.2486,
+        "step": 6850
+      },
+      {
+        "epoch": 0.21875697566886698,
+        "grad_norm": 0.41945523023605347,
+        "learning_rate": 0.0005,
+        "loss": 1.2485,
+        "step": 6860
+      },
+      {
+        "epoch": 0.2190758633885009,
+        "grad_norm": 0.4274255037307739,
+        "learning_rate": 0.0005,
+        "loss": 1.2698,
+        "step": 6870
+      },
+      {
+        "epoch": 0.21939475110813483,
+        "grad_norm": 0.4264500141143799,
+        "learning_rate": 0.0005,
+        "loss": 1.2686,
+        "step": 6880
+      },
+      {
+        "epoch": 0.21971363882776873,
+        "grad_norm": 0.41923102736473083,
+        "learning_rate": 0.0005,
+        "loss": 1.2623,
+        "step": 6890
+      },
+      {
+        "epoch": 0.22003252654740266,
+        "grad_norm": 0.438001424074173,
+        "learning_rate": 0.0005,
+        "loss": 1.2769,
+        "step": 6900
+      },
+      {
+        "epoch": 0.22035141426703658,
+        "grad_norm": 0.4238506853580475,
+        "learning_rate": 0.0005,
+        "loss": 1.2699,
+        "step": 6910
+      },
+      {
+        "epoch": 0.22067030198667048,
+        "grad_norm": 0.4327387511730194,
+        "learning_rate": 0.0005,
+        "loss": 1.2808,
+        "step": 6920
+      },
+      {
+        "epoch": 0.2209891897063044,
+        "grad_norm": 0.429081529378891,
+        "learning_rate": 0.0005,
+        "loss": 1.2726,
+        "step": 6930
+      },
+      {
+        "epoch": 0.22130807742593833,
+        "grad_norm": 0.4404042363166809,
+        "learning_rate": 0.0005,
+        "loss": 1.2592,
+        "step": 6940
+      },
+      {
+        "epoch": 0.22162696514557226,
+        "grad_norm": 0.43843477964401245,
+        "learning_rate": 0.0005,
+        "loss": 1.2564,
+        "step": 6950
+      },
+      {
+        "epoch": 0.22194585286520616,
+        "grad_norm": 0.42280009388923645,
+        "learning_rate": 0.0005,
+        "loss": 1.249,
+        "step": 6960
+      },
+      {
+        "epoch": 0.22226474058484008,
+        "grad_norm": 0.4233909547328949,
+        "learning_rate": 0.0005,
+        "loss": 1.2641,
+        "step": 6970
+      },
+      {
+        "epoch": 0.222583628304474,
+        "grad_norm": 0.4561818540096283,
+        "learning_rate": 0.0005,
+        "loss": 1.2642,
+        "step": 6980
+      },
+      {
+        "epoch": 0.2229025160241079,
+        "grad_norm": 0.4259558320045471,
+        "learning_rate": 0.0005,
+        "loss": 1.2718,
+        "step": 6990
+      },
+      {
+        "epoch": 0.22322140374374183,
+        "grad_norm": 0.42075279355049133,
+        "learning_rate": 0.0005,
+        "loss": 1.2893,
+        "step": 7000
+      },
+      {
+        "epoch": 0.22354029146337576,
+        "grad_norm": 0.43573033809661865,
+        "learning_rate": 0.0005,
+        "loss": 1.2479,
+        "step": 7010
+      },
+      {
+        "epoch": 0.22385917918300965,
+        "grad_norm": 0.4520786702632904,
+        "learning_rate": 0.0005,
+        "loss": 1.2683,
+        "step": 7020
+      },
+      {
+        "epoch": 0.22417806690264358,
+        "grad_norm": 0.4258021116256714,
+        "learning_rate": 0.0005,
+        "loss": 1.2382,
+        "step": 7030
+      },
+      {
+        "epoch": 0.2244969546222775,
+        "grad_norm": 0.4456654489040375,
+        "learning_rate": 0.0005,
+        "loss": 1.2436,
+        "step": 7040
+      },
+      {
+        "epoch": 0.2248158423419114,
+        "grad_norm": 0.432917058467865,
+        "learning_rate": 0.0005,
+        "loss": 1.2592,
+        "step": 7050
+      },
+      {
+        "epoch": 0.22513473006154533,
+        "grad_norm": 0.41865724325180054,
+        "learning_rate": 0.0005,
+        "loss": 1.253,
+        "step": 7060
+      },
+      {
+        "epoch": 0.22545361778117926,
+        "grad_norm": 0.4204801321029663,
+        "learning_rate": 0.0005,
+        "loss": 1.2539,
+        "step": 7070
+      },
+      {
+        "epoch": 0.22577250550081315,
+        "grad_norm": 0.41760504245758057,
+        "learning_rate": 0.0005,
+        "loss": 1.2786,
+        "step": 7080
+      },
+      {
+        "epoch": 0.22609139322044708,
+        "grad_norm": 0.421063631772995,
+        "learning_rate": 0.0005,
+        "loss": 1.2795,
+        "step": 7090
+      },
+      {
+        "epoch": 0.226410280940081,
+        "grad_norm": 0.46195104718208313,
+        "learning_rate": 0.0005,
+        "loss": 1.2607,
+        "step": 7100
+      },
+      {
+        "epoch": 0.2267291686597149,
+        "grad_norm": 0.43076223134994507,
+        "learning_rate": 0.0005,
+        "loss": 1.2782,
+        "step": 7110
+      },
+      {
+        "epoch": 0.22704805637934883,
+        "grad_norm": 0.40276387333869934,
+        "learning_rate": 0.0005,
+        "loss": 1.2509,
+        "step": 7120
+      },
+      {
+        "epoch": 0.22736694409898275,
+        "grad_norm": 0.41084590554237366,
+        "learning_rate": 0.0005,
+        "loss": 1.2662,
+        "step": 7130
+      },
+      {
+        "epoch": 0.22768583181861665,
+        "grad_norm": 0.4378090500831604,
+        "learning_rate": 0.0005,
+        "loss": 1.2637,
+        "step": 7140
+      },
+      {
+        "epoch": 0.22800471953825058,
+        "grad_norm": 0.42991676926612854,
+        "learning_rate": 0.0005,
+        "loss": 1.2598,
+        "step": 7150
+      },
+      {
+        "epoch": 0.2283236072578845,
+        "grad_norm": 0.4566963016986847,
+        "learning_rate": 0.0005,
+        "loss": 1.2652,
+        "step": 7160
+      },
+      {
+        "epoch": 0.22864249497751843,
+        "grad_norm": 0.4291926920413971,
+        "learning_rate": 0.0005,
+        "loss": 1.2762,
+        "step": 7170
+      },
+      {
+        "epoch": 0.22896138269715233,
+        "grad_norm": 0.4488089084625244,
+        "learning_rate": 0.0005,
+        "loss": 1.2427,
+        "step": 7180
+      },
+      {
+        "epoch": 0.22928027041678625,
+        "grad_norm": 0.4147830903530121,
+        "learning_rate": 0.0005,
+        "loss": 1.2498,
+        "step": 7190
+      },
+      {
+        "epoch": 0.22959915813642018,
+        "grad_norm": 0.44582098722457886,
+        "learning_rate": 0.0005,
+        "loss": 1.2518,
+        "step": 7200
+      },
+      {
+        "epoch": 0.22991804585605408,
+        "grad_norm": 0.41427379846572876,
+        "learning_rate": 0.0005,
+        "loss": 1.2484,
+        "step": 7210
+      },
+      {
+        "epoch": 0.230236933575688,
+        "grad_norm": 0.41920924186706543,
+        "learning_rate": 0.0005,
+        "loss": 1.2295,
+        "step": 7220
+      },
+      {
+        "epoch": 0.23055582129532193,
+        "grad_norm": 0.4119439721107483,
+        "learning_rate": 0.0005,
+        "loss": 1.2607,
+        "step": 7230
+      },
+      {
+        "epoch": 0.23087470901495583,
+        "grad_norm": 0.42621639370918274,
+        "learning_rate": 0.0005,
+        "loss": 1.269,
+        "step": 7240
+      },
+      {
+        "epoch": 0.23119359673458975,
+        "grad_norm": 0.42913585901260376,
+        "learning_rate": 0.0005,
+        "loss": 1.2534,
+        "step": 7250
+      },
+      {
+        "epoch": 0.23151248445422368,
+        "grad_norm": 0.43496614694595337,
+        "learning_rate": 0.0005,
+        "loss": 1.2442,
+        "step": 7260
+      },
+      {
+        "epoch": 0.23183137217385758,
+        "grad_norm": 0.4075222909450531,
+        "learning_rate": 0.0005,
+        "loss": 1.2465,
+        "step": 7270
+      },
+      {
+        "epoch": 0.2321502598934915,
+        "grad_norm": 0.4576127231121063,
+        "learning_rate": 0.0005,
+        "loss": 1.269,
+        "step": 7280
+      },
+      {
+        "epoch": 0.23246914761312543,
+        "grad_norm": 0.42538854479789734,
+        "learning_rate": 0.0005,
+        "loss": 1.2565,
+        "step": 7290
+      },
+      {
+        "epoch": 0.23278803533275932,
+        "grad_norm": 0.42042431235313416,
+        "learning_rate": 0.0005,
+        "loss": 1.251,
+        "step": 7300
+      },
+      {
+        "epoch": 0.23310692305239325,
+        "grad_norm": 0.4098883271217346,
+        "learning_rate": 0.0005,
+        "loss": 1.2665,
+        "step": 7310
+      },
+      {
+        "epoch": 0.23342581077202718,
+        "grad_norm": 0.43415921926498413,
+        "learning_rate": 0.0005,
+        "loss": 1.251,
+        "step": 7320
+      },
+      {
+        "epoch": 0.23374469849166107,
+        "grad_norm": 0.4317355155944824,
+        "learning_rate": 0.0005,
+        "loss": 1.2345,
+        "step": 7330
+      },
+      {
+        "epoch": 0.234063586211295,
+        "grad_norm": 0.4150109589099884,
+        "learning_rate": 0.0005,
+        "loss": 1.2605,
+        "step": 7340
+      },
+      {
+        "epoch": 0.23438247393092893,
+        "grad_norm": 0.39927196502685547,
+        "learning_rate": 0.0005,
+        "loss": 1.2532,
+        "step": 7350
+      },
+      {
+        "epoch": 0.23470136165056282,
+        "grad_norm": 0.41622141003608704,
+        "learning_rate": 0.0005,
+        "loss": 1.244,
+        "step": 7360
+      },
+      {
+        "epoch": 0.23502024937019675,
+        "grad_norm": 0.40569183230400085,
+        "learning_rate": 0.0005,
+        "loss": 1.2433,
+        "step": 7370
+      },
+      {
+        "epoch": 0.23533913708983067,
+        "grad_norm": 0.4543701410293579,
+        "learning_rate": 0.0005,
+        "loss": 1.2545,
+        "step": 7380
+      },
+      {
+        "epoch": 0.2356580248094646,
+        "grad_norm": 0.41043969988822937,
+        "learning_rate": 0.0005,
+        "loss": 1.254,
+        "step": 7390
+      },
+      {
+        "epoch": 0.2359769125290985,
+        "grad_norm": 0.4187447726726532,
+        "learning_rate": 0.0005,
+        "loss": 1.2401,
+        "step": 7400
+      },
+      {
+        "epoch": 0.23629580024873242,
+        "grad_norm": 0.4198475778102875,
+        "learning_rate": 0.0005,
+        "loss": 1.245,
+        "step": 7410
+      },
+      {
+        "epoch": 0.23661468796836635,
+        "grad_norm": 0.42908045649528503,
+        "learning_rate": 0.0005,
+        "loss": 1.2683,
+        "step": 7420
+      },
+      {
+        "epoch": 0.23693357568800025,
+        "grad_norm": 0.4434865117073059,
+        "learning_rate": 0.0005,
+        "loss": 1.229,
+        "step": 7430
+      },
+      {
+        "epoch": 0.23725246340763417,
+        "grad_norm": 0.4296889901161194,
+        "learning_rate": 0.0005,
+        "loss": 1.2417,
+        "step": 7440
+      },
+      {
+        "epoch": 0.2375713511272681,
+        "grad_norm": 0.42803293466567993,
+        "learning_rate": 0.0005,
+        "loss": 1.2622,
+        "step": 7450
+      },
+      {
+        "epoch": 0.237890238846902,
+        "grad_norm": 0.4073934853076935,
+        "learning_rate": 0.0005,
+        "loss": 1.2612,
+        "step": 7460
+      },
+      {
+        "epoch": 0.23820912656653592,
+        "grad_norm": 0.4153025448322296,
+        "learning_rate": 0.0005,
+        "loss": 1.2498,
+        "step": 7470
+      },
+      {
+        "epoch": 0.23852801428616985,
+        "grad_norm": 0.42079266905784607,
+        "learning_rate": 0.0005,
+        "loss": 1.2667,
+        "step": 7480
+      },
+      {
+        "epoch": 0.23884690200580375,
+        "grad_norm": 0.400680810213089,
+        "learning_rate": 0.0005,
+        "loss": 1.2381,
+        "step": 7490
+      },
+      {
+        "epoch": 0.23916578972543767,
+        "grad_norm": 0.41702139377593994,
+        "learning_rate": 0.0005,
+        "loss": 1.2692,
+        "step": 7500
+      },
+      {
+        "epoch": 0.2394846774450716,
+        "grad_norm": 0.4298720955848694,
+        "learning_rate": 0.0005,
+        "loss": 1.2466,
+        "step": 7510
+      },
+      {
+        "epoch": 0.2398035651647055,
+        "grad_norm": 0.3950182795524597,
+        "learning_rate": 0.0005,
+        "loss": 1.2521,
+        "step": 7520
+      },
+      {
+        "epoch": 0.24012245288433942,
+        "grad_norm": 0.3977687954902649,
+        "learning_rate": 0.0005,
+        "loss": 1.2533,
+        "step": 7530
+      },
+      {
+        "epoch": 0.24044134060397335,
+        "grad_norm": 0.40387430787086487,
+        "learning_rate": 0.0005,
+        "loss": 1.2465,
+        "step": 7540
+      },
+      {
+        "epoch": 0.24076022832360724,
+        "grad_norm": 0.4266010522842407,
+        "learning_rate": 0.0005,
+        "loss": 1.2316,
+        "step": 7550
+      },
+      {
+        "epoch": 0.24107911604324117,
+        "grad_norm": 0.3930842876434326,
+        "learning_rate": 0.0005,
+        "loss": 1.2391,
+        "step": 7560
+      },
+      {
+        "epoch": 0.2413980037628751,
+        "grad_norm": 0.41269761323928833,
+        "learning_rate": 0.0005,
+        "loss": 1.2631,
+        "step": 7570
+      },
+      {
+        "epoch": 0.24171689148250902,
+        "grad_norm": 0.4143700897693634,
+        "learning_rate": 0.0005,
+        "loss": 1.2409,
+        "step": 7580
+      },
+      {
+        "epoch": 0.24203577920214292,
+        "grad_norm": 0.40925079584121704,
+        "learning_rate": 0.0005,
+        "loss": 1.248,
+        "step": 7590
+      },
+      {
+        "epoch": 0.24235466692177685,
+        "grad_norm": 0.4155504107475281,
+        "learning_rate": 0.0005,
+        "loss": 1.2536,
+        "step": 7600
+      },
+      {
+        "epoch": 0.24267355464141077,
+        "grad_norm": 0.41415005922317505,
+        "learning_rate": 0.0005,
+        "loss": 1.2212,
+        "step": 7610
+      },
+      {
+        "epoch": 0.24299244236104467,
+        "grad_norm": 0.4322291612625122,
+        "learning_rate": 0.0005,
+        "loss": 1.2285,
+        "step": 7620
+      },
+      {
+        "epoch": 0.2433113300806786,
+        "grad_norm": 0.4126437306404114,
+        "learning_rate": 0.0005,
+        "loss": 1.2521,
+        "step": 7630
+      },
+      {
+        "epoch": 0.24363021780031252,
+        "grad_norm": 0.42201003432273865,
+        "learning_rate": 0.0005,
+        "loss": 1.2684,
+        "step": 7640
+      },
+      {
+        "epoch": 0.24394910551994642,
+        "grad_norm": 0.4242708086967468,
+        "learning_rate": 0.0005,
+        "loss": 1.2413,
+        "step": 7650
+      },
+      {
+        "epoch": 0.24426799323958034,
+        "grad_norm": 0.4040601849555969,
+        "learning_rate": 0.0005,
+        "loss": 1.2337,
+        "step": 7660
+      },
+      {
+        "epoch": 0.24458688095921427,
+        "grad_norm": 0.4113594889640808,
+        "learning_rate": 0.0005,
+        "loss": 1.2357,
+        "step": 7670
+      },
+      {
+        "epoch": 0.24490576867884817,
+        "grad_norm": 0.39296871423721313,
+        "learning_rate": 0.0005,
+        "loss": 1.2275,
+        "step": 7680
+      },
+      {
+        "epoch": 0.2452246563984821,
+        "grad_norm": 0.42583292722702026,
+        "learning_rate": 0.0005,
+        "loss": 1.2515,
+        "step": 7690
+      },
+      {
+        "epoch": 0.24554354411811602,
+        "grad_norm": 0.40814805030822754,
+        "learning_rate": 0.0005,
+        "loss": 1.2325,
+        "step": 7700
+      },
+      {
+        "epoch": 0.24586243183774992,
+        "grad_norm": 0.40266889333724976,
+        "learning_rate": 0.0005,
+        "loss": 1.2177,
+        "step": 7710
+      },
+      {
+        "epoch": 0.24618131955738384,
+        "grad_norm": 0.4071027636528015,
+        "learning_rate": 0.0005,
+        "loss": 1.2353,
+        "step": 7720
+      },
+      {
+        "epoch": 0.24650020727701777,
+        "grad_norm": 0.4114187955856323,
+        "learning_rate": 0.0005,
+        "loss": 1.2439,
+        "step": 7730
+      },
+      {
+        "epoch": 0.24681909499665167,
+        "grad_norm": 0.41335058212280273,
+        "learning_rate": 0.0005,
+        "loss": 1.2357,
+        "step": 7740
+      },
+      {
+        "epoch": 0.2471379827162856,
+        "grad_norm": 0.41387081146240234,
+        "learning_rate": 0.0005,
+        "loss": 1.2565,
+        "step": 7750
+      },
+      {
+        "epoch": 0.24745687043591952,
+        "grad_norm": 0.40741318464279175,
+        "learning_rate": 0.0005,
+        "loss": 1.2342,
+        "step": 7760
+      },
+      {
+        "epoch": 0.24777575815555342,
+        "grad_norm": 0.4092518091201782,
+        "learning_rate": 0.0005,
+        "loss": 1.2448,
+        "step": 7770
+      },
+      {
+        "epoch": 0.24809464587518734,
+        "grad_norm": 0.40899863839149475,
+        "learning_rate": 0.0005,
+        "loss": 1.2185,
+        "step": 7780
+      },
+      {
+        "epoch": 0.24841353359482127,
+        "grad_norm": 0.42302390933036804,
+        "learning_rate": 0.0005,
+        "loss": 1.2528,
+        "step": 7790
+      },
+      {
+        "epoch": 0.2487324213144552,
+        "grad_norm": 0.41404077410697937,
+        "learning_rate": 0.0005,
+        "loss": 1.2351,
+        "step": 7800
+      },
+      {
+        "epoch": 0.2490513090340891,
+        "grad_norm": 0.4217240512371063,
+        "learning_rate": 0.0005,
+        "loss": 1.2472,
+        "step": 7810
+      },
+      {
+        "epoch": 0.24937019675372302,
+        "grad_norm": 0.4106791019439697,
+        "learning_rate": 0.0005,
+        "loss": 1.2335,
+        "step": 7820
+      },
+      {
+        "epoch": 0.24968908447335694,
+        "grad_norm": 0.39354920387268066,
+        "learning_rate": 0.0005,
+        "loss": 1.2208,
+        "step": 7830
+      },
+      {
+        "epoch": 0.25000797219299087,
+        "grad_norm": 0.4039877951145172,
+        "learning_rate": 0.0005,
+        "loss": 1.2217,
+        "step": 7840
+      },
+      {
+        "epoch": 0.25032685991262477,
+        "grad_norm": 0.4242667853832245,
+        "learning_rate": 0.0005,
+        "loss": 1.2093,
+        "step": 7850
+      },
+      {
+        "epoch": 0.25064574763225866,
+        "grad_norm": 0.4013097584247589,
+        "learning_rate": 0.0005,
+        "loss": 1.2115,
+        "step": 7860
+      },
+      {
+        "epoch": 0.2509646353518926,
+        "grad_norm": 0.4191020727157593,
+        "learning_rate": 0.0005,
+        "loss": 1.22,
+        "step": 7870
+      },
+      {
+        "epoch": 0.2512835230715265,
+        "grad_norm": 0.39803019165992737,
+        "learning_rate": 0.0005,
+        "loss": 1.2185,
+        "step": 7880
+      },
+      {
+        "epoch": 0.2516024107911604,
+        "grad_norm": 0.41739556193351746,
+        "learning_rate": 0.0005,
+        "loss": 1.2619,
+        "step": 7890
+      },
+      {
+        "epoch": 0.25192129851079437,
+        "grad_norm": 0.3922768533229828,
+        "learning_rate": 0.0005,
+        "loss": 1.2372,
+        "step": 7900
+      },
+      {
+        "epoch": 0.25224018623042826,
+        "grad_norm": 0.4070080816745758,
+        "learning_rate": 0.0005,
+        "loss": 1.2311,
+        "step": 7910
+      },
+      {
+        "epoch": 0.25255907395006216,
+        "grad_norm": 0.4060456454753876,
+        "learning_rate": 0.0005,
+        "loss": 1.2383,
+        "step": 7920
+      },
+      {
+        "epoch": 0.2528779616696961,
+        "grad_norm": 0.42530322074890137,
+        "learning_rate": 0.0005,
+        "loss": 1.2259,
+        "step": 7930
+      },
+      {
+        "epoch": 0.25319684938933,
+        "grad_norm": 0.41797924041748047,
+        "learning_rate": 0.0005,
+        "loss": 1.228,
+        "step": 7940
+      },
+      {
+        "epoch": 0.2535157371089639,
+        "grad_norm": 0.3849104940891266,
+        "learning_rate": 0.0005,
+        "loss": 1.2296,
+        "step": 7950
+      },
+      {
+        "epoch": 0.25383462482859787,
+        "grad_norm": 0.41749095916748047,
+        "learning_rate": 0.0005,
+        "loss": 1.2441,
+        "step": 7960
+      },
+      {
+        "epoch": 0.25415351254823176,
+        "grad_norm": 0.40891966223716736,
+        "learning_rate": 0.0005,
+        "loss": 1.2323,
+        "step": 7970
+      },
+      {
+        "epoch": 0.25447240026786566,
+        "grad_norm": 0.42078697681427,
+        "learning_rate": 0.0005,
+        "loss": 1.2602,
+        "step": 7980
+      },
+      {
+        "epoch": 0.2547912879874996,
+        "grad_norm": 0.4090920686721802,
+        "learning_rate": 0.0005,
+        "loss": 1.2506,
+        "step": 7990
+      },
+      {
+        "epoch": 0.2551101757071335,
+        "grad_norm": 0.41705599427223206,
+        "learning_rate": 0.0005,
+        "loss": 1.2238,
+        "step": 8000
+      },
+      {
+        "epoch": 0.2554290634267674,
+        "grad_norm": 0.40100833773612976,
+        "learning_rate": 0.0005,
+        "loss": 1.2228,
+        "step": 8010
+      },
+      {
+        "epoch": 0.25574795114640136,
+        "grad_norm": 0.396261602640152,
+        "learning_rate": 0.0005,
+        "loss": 1.2328,
+        "step": 8020
+      },
+      {
+        "epoch": 0.25606683886603526,
+        "grad_norm": 0.3865589499473572,
+        "learning_rate": 0.0005,
+        "loss": 1.2191,
+        "step": 8030
+      },
+      {
+        "epoch": 0.25638572658566916,
+        "grad_norm": 0.4185122549533844,
+        "learning_rate": 0.0005,
+        "loss": 1.2351,
+        "step": 8040
+      },
+      {
+        "epoch": 0.2567046143053031,
+        "grad_norm": 0.40291154384613037,
+        "learning_rate": 0.0005,
+        "loss": 1.2455,
+        "step": 8050
+      },
+      {
+        "epoch": 0.257023502024937,
+        "grad_norm": 0.4228131175041199,
+        "learning_rate": 0.0005,
+        "loss": 1.2309,
+        "step": 8060
+      },
+      {
+        "epoch": 0.2573423897445709,
+        "grad_norm": 0.4108589291572571,
+        "learning_rate": 0.0005,
+        "loss": 1.2433,
+        "step": 8070
+      },
+      {
+        "epoch": 0.25766127746420486,
+        "grad_norm": 0.4041846692562103,
+        "learning_rate": 0.0005,
+        "loss": 1.2269,
+        "step": 8080
+      },
+      {
+        "epoch": 0.25798016518383876,
+        "grad_norm": 0.40096405148506165,
+        "learning_rate": 0.0005,
+        "loss": 1.2629,
+        "step": 8090
+      },
+      {
+        "epoch": 0.2582990529034727,
+        "grad_norm": 0.39418691396713257,
+        "learning_rate": 0.0005,
+        "loss": 1.2421,
+        "step": 8100
+      },
+      {
+        "epoch": 0.2586179406231066,
+        "grad_norm": 0.42769312858581543,
+        "learning_rate": 0.0005,
+        "loss": 1.2407,
+        "step": 8110
+      },
+      {
+        "epoch": 0.2589368283427405,
+        "grad_norm": 0.40227988362312317,
+        "learning_rate": 0.0005,
+        "loss": 1.2128,
+        "step": 8120
+      },
+      {
+        "epoch": 0.25925571606237446,
+        "grad_norm": 0.388641893863678,
+        "learning_rate": 0.0005,
+        "loss": 1.2208,
+        "step": 8130
+      },
+      {
+        "epoch": 0.25957460378200836,
+        "grad_norm": 0.39878180623054504,
+        "learning_rate": 0.0005,
+        "loss": 1.1998,
+        "step": 8140
+      },
+      {
+        "epoch": 0.25989349150164226,
+        "grad_norm": 0.3936919569969177,
+        "learning_rate": 0.0005,
+        "loss": 1.2275,
+        "step": 8150
+      },
+      {
+        "epoch": 0.2602123792212762,
+        "grad_norm": 0.40643203258514404,
+        "learning_rate": 0.0005,
+        "loss": 1.2206,
+        "step": 8160
+      },
+      {
+        "epoch": 0.2605312669409101,
+        "grad_norm": 0.40189582109451294,
+        "learning_rate": 0.0005,
+        "loss": 1.212,
+        "step": 8170
+      },
+      {
+        "epoch": 0.260850154660544,
+        "grad_norm": 0.4173663556575775,
+        "learning_rate": 0.0005,
+        "loss": 1.2346,
+        "step": 8180
+      },
+      {
+        "epoch": 0.26116904238017796,
+        "grad_norm": 0.39820632338523865,
+        "learning_rate": 0.0005,
+        "loss": 1.2181,
+        "step": 8190
+      },
+      {
+        "epoch": 0.26148793009981186,
+        "grad_norm": 0.40434184670448303,
+        "learning_rate": 0.0005,
+        "loss": 1.2237,
+        "step": 8200
+      },
+      {
+        "epoch": 0.26180681781944576,
+        "grad_norm": 0.3958496153354645,
+        "learning_rate": 0.0005,
+        "loss": 1.2447,
+        "step": 8210
+      },
+      {
+        "epoch": 0.2621257055390797,
+        "grad_norm": 0.40260443091392517,
+        "learning_rate": 0.0005,
+        "loss": 1.2424,
+        "step": 8220
+      },
+      {
+        "epoch": 0.2624445932587136,
+        "grad_norm": 0.40898144245147705,
+        "learning_rate": 0.0005,
+        "loss": 1.2212,
+        "step": 8230
+      },
+      {
+        "epoch": 0.2627634809783475,
+        "grad_norm": 0.43052124977111816,
+        "learning_rate": 0.0005,
+        "loss": 1.2143,
+        "step": 8240
+      },
+      {
+        "epoch": 0.26308236869798146,
+        "grad_norm": 0.3874588906764984,
+        "learning_rate": 0.0005,
+        "loss": 1.2101,
+        "step": 8250
+      },
+      {
+        "epoch": 0.26340125641761536,
+        "grad_norm": 0.3905537724494934,
+        "learning_rate": 0.0005,
+        "loss": 1.2357,
+        "step": 8260
+      },
+      {
+        "epoch": 0.26372014413724926,
+        "grad_norm": 0.3907853662967682,
+        "learning_rate": 0.0005,
+        "loss": 1.2286,
+        "step": 8270
+      },
+      {
+        "epoch": 0.2640390318568832,
+        "grad_norm": 0.3982638716697693,
+        "learning_rate": 0.0005,
+        "loss": 1.2296,
+        "step": 8280
+      },
+      {
+        "epoch": 0.2643579195765171,
+        "grad_norm": 0.4627874195575714,
+        "learning_rate": 0.0005,
+        "loss": 1.2305,
+        "step": 8290
+      },
+      {
+        "epoch": 0.264676807296151,
+        "grad_norm": 0.40264928340911865,
+        "learning_rate": 0.0005,
+        "loss": 1.2188,
+        "step": 8300
+      },
+      {
+        "epoch": 0.26499569501578496,
+        "grad_norm": 0.3911851942539215,
+        "learning_rate": 0.0005,
+        "loss": 1.2174,
+        "step": 8310
+      },
+      {
+        "epoch": 0.26531458273541886,
+        "grad_norm": 0.37752363085746765,
+        "learning_rate": 0.0005,
+        "loss": 1.21,
+        "step": 8320
+      },
+      {
+        "epoch": 0.26563347045505276,
+        "grad_norm": 0.42199021577835083,
+        "learning_rate": 0.0005,
+        "loss": 1.2042,
+        "step": 8330
+      },
+      {
+        "epoch": 0.2659523581746867,
+        "grad_norm": 0.41406044363975525,
+        "learning_rate": 0.0005,
+        "loss": 1.2338,
+        "step": 8340
+      },
+      {
+        "epoch": 0.2662712458943206,
+        "grad_norm": 0.42696836590766907,
+        "learning_rate": 0.0005,
+        "loss": 1.2209,
+        "step": 8350
+      },
+      {
+        "epoch": 0.2665901336139545,
+        "grad_norm": 0.39001408219337463,
+        "learning_rate": 0.0005,
+        "loss": 1.2496,
+        "step": 8360
+      },
+      {
+        "epoch": 0.26690902133358846,
+        "grad_norm": 0.4086891710758209,
+        "learning_rate": 0.0005,
+        "loss": 1.2394,
+        "step": 8370
+      },
+      {
+        "epoch": 0.26722790905322236,
+        "grad_norm": 0.3911346197128296,
+        "learning_rate": 0.0005,
+        "loss": 1.2266,
+        "step": 8380
+      },
+      {
+        "epoch": 0.26754679677285625,
+        "grad_norm": 0.39909061789512634,
+        "learning_rate": 0.0005,
+        "loss": 1.2261,
+        "step": 8390
+      },
+      {
+        "epoch": 0.2678656844924902,
+        "grad_norm": 0.4086940288543701,
+        "learning_rate": 0.0005,
+        "loss": 1.2203,
+        "step": 8400
+      },
+      {
+        "epoch": 0.2681845722121241,
+        "grad_norm": 0.40768542885780334,
+        "learning_rate": 0.0005,
+        "loss": 1.215,
+        "step": 8410
+      },
+      {
+        "epoch": 0.268503459931758,
+        "grad_norm": 0.39183276891708374,
+        "learning_rate": 0.0005,
+        "loss": 1.2101,
+        "step": 8420
+      },
+      {
+        "epoch": 0.26882234765139196,
+        "grad_norm": 0.383931040763855,
+        "learning_rate": 0.0005,
+        "loss": 1.2115,
+        "step": 8430
+      },
+      {
+        "epoch": 0.26914123537102586,
+        "grad_norm": 0.39705631136894226,
+        "learning_rate": 0.0005,
+        "loss": 1.2093,
+        "step": 8440
+      },
+      {
+        "epoch": 0.26946012309065975,
+        "grad_norm": 0.39195719361305237,
+        "learning_rate": 0.0005,
+        "loss": 1.2219,
+        "step": 8450
+      },
+      {
+        "epoch": 0.2697790108102937,
+        "grad_norm": 0.40420153737068176,
+        "learning_rate": 0.0005,
+        "loss": 1.2152,
+        "step": 8460
+      },
+      {
+        "epoch": 0.2700978985299276,
+        "grad_norm": 0.4257016181945801,
+        "learning_rate": 0.0005,
+        "loss": 1.2303,
+        "step": 8470
+      },
+      {
+        "epoch": 0.2704167862495615,
+        "grad_norm": 0.40116095542907715,
+        "learning_rate": 0.0005,
+        "loss": 1.2122,
+        "step": 8480
+      },
+      {
+        "epoch": 0.27073567396919546,
+        "grad_norm": 0.42069900035858154,
+        "learning_rate": 0.0005,
+        "loss": 1.2106,
+        "step": 8490
+      },
+      {
+        "epoch": 0.27105456168882935,
+        "grad_norm": 0.38822034001350403,
+        "learning_rate": 0.0005,
+        "loss": 1.2378,
+        "step": 8500
+      },
+      {
+        "epoch": 0.2713734494084633,
+        "grad_norm": 0.38975998759269714,
+        "learning_rate": 0.0005,
+        "loss": 1.2138,
+        "step": 8510
+      },
+      {
+        "epoch": 0.2716923371280972,
+        "grad_norm": 0.4007030129432678,
+        "learning_rate": 0.0005,
+        "loss": 1.2051,
+        "step": 8520
+      },
+      {
+        "epoch": 0.2720112248477311,
+        "grad_norm": 0.39984530210494995,
+        "learning_rate": 0.0005,
+        "loss": 1.2053,
+        "step": 8530
+      },
+      {
+        "epoch": 0.27233011256736506,
+        "grad_norm": 0.3883681893348694,
+        "learning_rate": 0.0005,
+        "loss": 1.2196,
+        "step": 8540
+      },
+      {
+        "epoch": 0.27264900028699895,
+        "grad_norm": 0.40165334939956665,
+        "learning_rate": 0.0005,
+        "loss": 1.2413,
+        "step": 8550
+      },
+      {
+        "epoch": 0.27296788800663285,
+        "grad_norm": 0.3896324634552002,
+        "learning_rate": 0.0005,
+        "loss": 1.2273,
+        "step": 8560
+      },
+      {
+        "epoch": 0.2732867757262668,
+        "grad_norm": 0.3914828896522522,
+        "learning_rate": 0.0005,
+        "loss": 1.2383,
+        "step": 8570
+      },
+      {
+        "epoch": 0.2736056634459007,
+        "grad_norm": 0.39543914794921875,
+        "learning_rate": 0.0005,
+        "loss": 1.2028,
+        "step": 8580
+      },
+      {
+        "epoch": 0.2739245511655346,
+        "grad_norm": 0.38763898611068726,
+        "learning_rate": 0.0005,
+        "loss": 1.2109,
+        "step": 8590
+      },
+      {
+        "epoch": 0.27424343888516856,
+        "grad_norm": 0.37673720717430115,
+        "learning_rate": 0.0005,
+        "loss": 1.2265,
+        "step": 8600
+      },
+      {
+        "epoch": 0.27456232660480245,
+        "grad_norm": 0.3957275152206421,
+        "learning_rate": 0.0005,
+        "loss": 1.2341,
+        "step": 8610
+      },
+      {
+        "epoch": 0.27488121432443635,
+        "grad_norm": 0.4029931426048279,
+        "learning_rate": 0.0005,
+        "loss": 1.1993,
+        "step": 8620
+      },
+      {
+        "epoch": 0.2752001020440703,
+        "grad_norm": 0.3820151686668396,
+        "learning_rate": 0.0005,
+        "loss": 1.2055,
+        "step": 8630
+      },
+      {
+        "epoch": 0.2755189897637042,
+        "grad_norm": 0.39909201860427856,
+        "learning_rate": 0.0005,
+        "loss": 1.2057,
+        "step": 8640
+      },
+      {
+        "epoch": 0.2758378774833381,
+        "grad_norm": 0.3830402195453644,
+        "learning_rate": 0.0005,
+        "loss": 1.2292,
+        "step": 8650
+      },
+      {
+        "epoch": 0.27615676520297205,
+        "grad_norm": 0.38620173931121826,
+        "learning_rate": 0.0005,
+        "loss": 1.2111,
+        "step": 8660
+      },
+      {
+        "epoch": 0.27647565292260595,
+        "grad_norm": 0.39842554926872253,
+        "learning_rate": 0.0005,
+        "loss": 1.2183,
+        "step": 8670
+      },
+      {
+        "epoch": 0.27679454064223985,
+        "grad_norm": 0.3806487023830414,
+        "learning_rate": 0.0005,
+        "loss": 1.2022,
+        "step": 8680
+      },
+      {
+        "epoch": 0.2771134283618738,
+        "grad_norm": 0.4076787829399109,
+        "learning_rate": 0.0005,
+        "loss": 1.2007,
+        "step": 8690
+      },
+      {
+        "epoch": 0.2774323160815077,
+        "grad_norm": 0.3915601968765259,
+        "learning_rate": 0.0005,
+        "loss": 1.2146,
+        "step": 8700
+      },
+      {
+        "epoch": 0.2777512038011416,
+        "grad_norm": 0.38058754801750183,
+        "learning_rate": 0.0005,
+        "loss": 1.2189,
+        "step": 8710
+      },
+      {
+        "epoch": 0.27807009152077555,
+        "grad_norm": 0.39291903376579285,
+        "learning_rate": 0.0005,
+        "loss": 1.2177,
+        "step": 8720
+      },
+      {
+        "epoch": 0.27838897924040945,
+        "grad_norm": 0.3893909454345703,
+        "learning_rate": 0.0005,
+        "loss": 1.1909,
+        "step": 8730
+      },
+      {
+        "epoch": 0.27870786696004335,
+        "grad_norm": 0.400829017162323,
+        "learning_rate": 0.0005,
+        "loss": 1.212,
+        "step": 8740
+      },
+      {
+        "epoch": 0.2790267546796773,
+        "grad_norm": 0.40721598267555237,
+        "learning_rate": 0.0005,
+        "loss": 1.2168,
+        "step": 8750
+      },
+      {
+        "epoch": 0.2793456423993112,
+        "grad_norm": 0.4186878204345703,
+        "learning_rate": 0.0005,
+        "loss": 1.2119,
+        "step": 8760
+      },
+      {
+        "epoch": 0.2796645301189451,
+        "grad_norm": 0.3906959593296051,
+        "learning_rate": 0.0005,
+        "loss": 1.217,
+        "step": 8770
+      },
+      {
+        "epoch": 0.27998341783857905,
+        "grad_norm": 0.3925018906593323,
+        "learning_rate": 0.0005,
+        "loss": 1.221,
+        "step": 8780
+      },
+      {
+        "epoch": 0.28030230555821295,
+        "grad_norm": 0.39244839549064636,
+        "learning_rate": 0.0005,
+        "loss": 1.2088,
+        "step": 8790
+      },
+      {
+        "epoch": 0.28062119327784685,
+        "grad_norm": 0.3988844156265259,
+        "learning_rate": 0.0005,
+        "loss": 1.1995,
+        "step": 8800
+      },
+      {
+        "epoch": 0.2809400809974808,
+        "grad_norm": 0.38632071018218994,
+        "learning_rate": 0.0005,
+        "loss": 1.2249,
+        "step": 8810
+      },
+      {
+        "epoch": 0.2812589687171147,
+        "grad_norm": 0.3827994167804718,
+        "learning_rate": 0.0005,
+        "loss": 1.2106,
+        "step": 8820
+      },
+      {
+        "epoch": 0.2815778564367486,
+        "grad_norm": 0.4094688594341278,
+        "learning_rate": 0.0005,
+        "loss": 1.2355,
+        "step": 8830
+      },
+      {
+        "epoch": 0.28189674415638255,
+        "grad_norm": 0.3843954801559448,
+        "learning_rate": 0.0005,
+        "loss": 1.2099,
+        "step": 8840
+      },
+      {
+        "epoch": 0.28221563187601645,
+        "grad_norm": 0.3869307041168213,
+        "learning_rate": 0.0005,
+        "loss": 1.2141,
+        "step": 8850
+      },
+      {
+        "epoch": 0.28253451959565035,
+        "grad_norm": 0.38374051451683044,
+        "learning_rate": 0.0005,
+        "loss": 1.2162,
+        "step": 8860
+      },
+      {
+        "epoch": 0.2828534073152843,
+        "grad_norm": 0.39685219526290894,
+        "learning_rate": 0.0005,
+        "loss": 1.2,
+        "step": 8870
+      },
+      {
+        "epoch": 0.2831722950349182,
+        "grad_norm": 0.3965778052806854,
+        "learning_rate": 0.0005,
+        "loss": 1.2194,
+        "step": 8880
+      },
+      {
+        "epoch": 0.2834911827545521,
+        "grad_norm": 0.390626460313797,
+        "learning_rate": 0.0005,
+        "loss": 1.1979,
+        "step": 8890
+      },
+      {
+        "epoch": 0.28381007047418605,
+        "grad_norm": 0.40536853671073914,
+        "learning_rate": 0.0005,
+        "loss": 1.2207,
+        "step": 8900
+      },
+      {
+        "epoch": 0.28412895819381995,
+        "grad_norm": 0.40149563550949097,
+        "learning_rate": 0.0005,
+        "loss": 1.2049,
+        "step": 8910
+      },
+      {
+        "epoch": 0.2844478459134539,
+        "grad_norm": 0.40702196955680847,
+        "learning_rate": 0.0005,
+        "loss": 1.205,
+        "step": 8920
+      },
+      {
+        "epoch": 0.2847667336330878,
+        "grad_norm": 0.3878912925720215,
+        "learning_rate": 0.0005,
+        "loss": 1.2151,
+        "step": 8930
+      },
+      {
+        "epoch": 0.2850856213527217,
+        "grad_norm": 0.407080739736557,
+        "learning_rate": 0.0005,
+        "loss": 1.2083,
+        "step": 8940
+      },
+      {
+        "epoch": 0.28540450907235565,
+        "grad_norm": 0.40740883350372314,
+        "learning_rate": 0.0005,
+        "loss": 1.2198,
+        "step": 8950
+      },
+      {
+        "epoch": 0.28572339679198955,
+        "grad_norm": 0.4337395131587982,
+        "learning_rate": 0.0005,
+        "loss": 1.2008,
+        "step": 8960
+      },
+      {
+        "epoch": 0.28604228451162345,
+        "grad_norm": 0.3903014063835144,
+        "learning_rate": 0.0005,
+        "loss": 1.2072,
+        "step": 8970
+      },
+      {
+        "epoch": 0.2863611722312574,
+        "grad_norm": 0.38654598593711853,
+        "learning_rate": 0.0005,
+        "loss": 1.2098,
+        "step": 8980
+      },
+      {
+        "epoch": 0.2866800599508913,
+        "grad_norm": 0.3934592306613922,
+        "learning_rate": 0.0005,
+        "loss": 1.2082,
+        "step": 8990
+      },
+      {
+        "epoch": 0.2869989476705252,
+        "grad_norm": 0.3974146246910095,
+        "learning_rate": 0.0005,
+        "loss": 1.234,
+        "step": 9000
+      },
+      {
+        "epoch": 0.28731783539015915,
+        "grad_norm": 0.39898407459259033,
+        "learning_rate": 0.0005,
+        "loss": 1.1914,
+        "step": 9010
+      },
+      {
+        "epoch": 0.28763672310979305,
+        "grad_norm": 0.39268526434898376,
+        "learning_rate": 0.0005,
+        "loss": 1.2091,
+        "step": 9020
+      },
+      {
+        "epoch": 0.28795561082942694,
+        "grad_norm": 0.3900041878223419,
+        "learning_rate": 0.0005,
+        "loss": 1.207,
+        "step": 9030
+      },
+      {
+        "epoch": 0.2882744985490609,
+        "grad_norm": 0.3990797996520996,
+        "learning_rate": 0.0005,
+        "loss": 1.2143,
+        "step": 9040
+      },
+      {
+        "epoch": 0.2885933862686948,
+        "grad_norm": 0.37181395292282104,
+        "learning_rate": 0.0005,
+        "loss": 1.1998,
+        "step": 9050
+      },
+      {
+        "epoch": 0.2889122739883287,
+        "grad_norm": 0.3809095323085785,
+        "learning_rate": 0.0005,
+        "loss": 1.1981,
+        "step": 9060
+      },
+      {
+        "epoch": 0.28923116170796265,
+        "grad_norm": 0.3911319673061371,
+        "learning_rate": 0.0005,
+        "loss": 1.2012,
+        "step": 9070
+      },
+      {
+        "epoch": 0.28955004942759655,
+        "grad_norm": 0.40503358840942383,
+        "learning_rate": 0.0005,
+        "loss": 1.2069,
+        "step": 9080
+      },
+      {
+        "epoch": 0.28986893714723044,
+        "grad_norm": 0.40217137336730957,
+        "learning_rate": 0.0005,
+        "loss": 1.2307,
+        "step": 9090
+      },
+      {
+        "epoch": 0.2901878248668644,
+        "grad_norm": 0.39365431666374207,
+        "learning_rate": 0.0005,
+        "loss": 1.2027,
+        "step": 9100
+      },
+      {
+        "epoch": 0.2905067125864983,
+        "grad_norm": 0.3819471597671509,
+        "learning_rate": 0.0005,
+        "loss": 1.2018,
+        "step": 9110
+      },
+      {
+        "epoch": 0.2908256003061322,
+        "grad_norm": 0.41711509227752686,
+        "learning_rate": 0.0005,
+        "loss": 1.2114,
+        "step": 9120
+      },
+      {
+        "epoch": 0.29114448802576615,
+        "grad_norm": 0.38931921124458313,
+        "learning_rate": 0.0005,
+        "loss": 1.2114,
+        "step": 9130
+      },
+      {
+        "epoch": 0.29146337574540004,
+        "grad_norm": 0.4078792929649353,
+        "learning_rate": 0.0005,
+        "loss": 1.1975,
+        "step": 9140
+      },
+      {
+        "epoch": 0.29178226346503394,
+        "grad_norm": 0.3806973397731781,
+        "learning_rate": 0.0005,
+        "loss": 1.193,
+        "step": 9150
+      },
+      {
+        "epoch": 0.2921011511846679,
+        "grad_norm": 0.37954655289649963,
+        "learning_rate": 0.0005,
+        "loss": 1.2152,
+        "step": 9160
+      },
+      {
+        "epoch": 0.2924200389043018,
+        "grad_norm": 0.38385963439941406,
+        "learning_rate": 0.0005,
+        "loss": 1.2195,
+        "step": 9170
+      },
+      {
+        "epoch": 0.2927389266239357,
+        "grad_norm": 0.3808298408985138,
+        "learning_rate": 0.0005,
+        "loss": 1.2128,
+        "step": 9180
+      },
+      {
+        "epoch": 0.29305781434356964,
+        "grad_norm": 0.3786800503730774,
+        "learning_rate": 0.0005,
+        "loss": 1.1958,
+        "step": 9190
+      },
+      {
+        "epoch": 0.29337670206320354,
+        "grad_norm": 0.3799290359020233,
+        "learning_rate": 0.0005,
+        "loss": 1.2006,
+        "step": 9200
+      },
+      {
+        "epoch": 0.29369558978283744,
+        "grad_norm": 0.37467071413993835,
+        "learning_rate": 0.0005,
+        "loss": 1.1909,
+        "step": 9210
+      },
+      {
+        "epoch": 0.2940144775024714,
+        "grad_norm": 0.3991237282752991,
+        "learning_rate": 0.0005,
+        "loss": 1.1974,
+        "step": 9220
+      },
+      {
+        "epoch": 0.2943333652221053,
+        "grad_norm": 0.38379865884780884,
+        "learning_rate": 0.0005,
+        "loss": 1.2086,
+        "step": 9230
+      },
+      {
+        "epoch": 0.2946522529417392,
+        "grad_norm": 0.3732675313949585,
+        "learning_rate": 0.0005,
+        "loss": 1.1895,
+        "step": 9240
+      },
+      {
+        "epoch": 0.29497114066137314,
+        "grad_norm": 0.3875043988227844,
+        "learning_rate": 0.0005,
+        "loss": 1.2025,
+        "step": 9250
+      },
+      {
+        "epoch": 0.29529002838100704,
+        "grad_norm": 0.41846537590026855,
+        "learning_rate": 0.0005,
+        "loss": 1.1986,
+        "step": 9260
+      },
+      {
+        "epoch": 0.29560891610064094,
+        "grad_norm": 0.37385421991348267,
+        "learning_rate": 0.0005,
+        "loss": 1.1868,
+        "step": 9270
+      },
+      {
+        "epoch": 0.2959278038202749,
+        "grad_norm": 0.3830406069755554,
+        "learning_rate": 0.0005,
+        "loss": 1.198,
+        "step": 9280
+      },
+      {
+        "epoch": 0.2962466915399088,
+        "grad_norm": 0.3768352270126343,
+        "learning_rate": 0.0005,
+        "loss": 1.1808,
+        "step": 9290
+      },
+      {
+        "epoch": 0.2965655792595427,
+        "grad_norm": 0.37796854972839355,
+        "learning_rate": 0.0005,
+        "loss": 1.2005,
+        "step": 9300
+      },
+      {
+        "epoch": 0.29688446697917664,
+        "grad_norm": 0.3770412802696228,
+        "learning_rate": 0.0005,
+        "loss": 1.2082,
+        "step": 9310
+      },
+      {
+        "epoch": 0.29720335469881054,
+        "grad_norm": 0.37880739569664,
+        "learning_rate": 0.0005,
+        "loss": 1.1779,
+        "step": 9320
+      },
+      {
+        "epoch": 0.29752224241844444,
+        "grad_norm": 0.38096004724502563,
+        "learning_rate": 0.0005,
+        "loss": 1.2064,
+        "step": 9330
+      },
+      {
+        "epoch": 0.2978411301380784,
+        "grad_norm": 0.39088889956474304,
+        "learning_rate": 0.0005,
+        "loss": 1.1833,
+        "step": 9340
+      },
+      {
+        "epoch": 0.2981600178577123,
+        "grad_norm": 0.3733189105987549,
+        "learning_rate": 0.0005,
+        "loss": 1.1898,
+        "step": 9350
+      },
+      {
+        "epoch": 0.29847890557734624,
+        "grad_norm": 0.3921329081058502,
+        "learning_rate": 0.0005,
+        "loss": 1.1951,
+        "step": 9360
+      },
+      {
+        "epoch": 0.29879779329698014,
+        "grad_norm": 0.38713183999061584,
+        "learning_rate": 0.0005,
+        "loss": 1.2039,
+        "step": 9370
+      },
+      {
+        "epoch": 0.29911668101661404,
+        "grad_norm": 0.3787401616573334,
+        "learning_rate": 0.0005,
+        "loss": 1.2016,
+        "step": 9380
+      },
+      {
+        "epoch": 0.299435568736248,
+        "grad_norm": 0.36601531505584717,
+        "learning_rate": 0.0005,
+        "loss": 1.1808,
+        "step": 9390
+      },
+      {
+        "epoch": 0.2997544564558819,
+        "grad_norm": 0.383737176656723,
+        "learning_rate": 0.0005,
+        "loss": 1.2,
+        "step": 9400
+      },
+      {
+        "epoch": 0.3000733441755158,
+        "grad_norm": 0.3672783076763153,
+        "learning_rate": 0.0005,
+        "loss": 1.2034,
+        "step": 9410
+      },
+      {
+        "epoch": 0.30039223189514974,
+        "grad_norm": 0.37648218870162964,
+        "learning_rate": 0.0005,
+        "loss": 1.1984,
+        "step": 9420
+      },
+      {
+        "epoch": 0.30071111961478364,
+        "grad_norm": 0.39211979508399963,
+        "learning_rate": 0.0005,
+        "loss": 1.1931,
+        "step": 9430
+      },
+      {
+        "epoch": 0.30103000733441754,
+        "grad_norm": 0.4010789394378662,
+        "learning_rate": 0.0005,
+        "loss": 1.2021,
+        "step": 9440
+      },
+      {
+        "epoch": 0.3013488950540515,
+        "grad_norm": 0.37380537390708923,
+        "learning_rate": 0.0005,
+        "loss": 1.2043,
+        "step": 9450
+      },
+      {
+        "epoch": 0.3016677827736854,
+        "grad_norm": 0.399393230676651,
+        "learning_rate": 0.0005,
+        "loss": 1.2053,
+        "step": 9460
+      },
+      {
+        "epoch": 0.3019866704933193,
+        "grad_norm": 0.3876519203186035,
+        "learning_rate": 0.0005,
+        "loss": 1.2181,
+        "step": 9470
+      },
+      {
+        "epoch": 0.30230555821295324,
+        "grad_norm": 0.3972403109073639,
+        "learning_rate": 0.0005,
+        "loss": 1.1972,
+        "step": 9480
+      },
+      {
+        "epoch": 0.30262444593258714,
+        "grad_norm": 0.383587121963501,
+        "learning_rate": 0.0005,
+        "loss": 1.1982,
+        "step": 9490
+      },
+      {
+        "epoch": 0.30294333365222104,
+        "grad_norm": 0.38184329867362976,
+        "learning_rate": 0.0005,
+        "loss": 1.2009,
+        "step": 9500
+      },
+      {
+        "epoch": 0.303262221371855,
+        "grad_norm": 0.3812735378742218,
+        "learning_rate": 0.0005,
+        "loss": 1.1845,
+        "step": 9510
+      },
+      {
+        "epoch": 0.3035811090914889,
+        "grad_norm": 0.3952953815460205,
+        "learning_rate": 0.0005,
+        "loss": 1.193,
+        "step": 9520
+      },
+      {
+        "epoch": 0.3038999968111228,
+        "grad_norm": 0.39169639348983765,
+        "learning_rate": 0.0005,
+        "loss": 1.1801,
+        "step": 9530
+      },
+      {
+        "epoch": 0.30421888453075674,
+        "grad_norm": 0.3926263749599457,
+        "learning_rate": 0.0005,
+        "loss": 1.1917,
+        "step": 9540
+      },
+      {
+        "epoch": 0.30453777225039064,
+        "grad_norm": 0.38617074489593506,
+        "learning_rate": 0.0005,
+        "loss": 1.1705,
+        "step": 9550
+      },
+      {
+        "epoch": 0.30485665997002453,
+        "grad_norm": 0.37101539969444275,
+        "learning_rate": 0.0005,
+        "loss": 1.1953,
+        "step": 9560
+      },
+      {
+        "epoch": 0.3051755476896585,
+        "grad_norm": 0.3760269582271576,
+        "learning_rate": 0.0005,
+        "loss": 1.1884,
+        "step": 9570
+      },
+      {
+        "epoch": 0.3054944354092924,
+        "grad_norm": 0.37347906827926636,
+        "learning_rate": 0.0005,
+        "loss": 1.2076,
+        "step": 9580
+      },
+      {
+        "epoch": 0.3058133231289263,
+        "grad_norm": 0.3571949601173401,
+        "learning_rate": 0.0005,
+        "loss": 1.1914,
+        "step": 9590
+      },
+      {
+        "epoch": 0.30613221084856024,
+        "grad_norm": 0.3888949751853943,
+        "learning_rate": 0.0005,
+        "loss": 1.1841,
+        "step": 9600
+      },
+      {
+        "epoch": 0.30645109856819414,
+        "grad_norm": 0.3792676627635956,
+        "learning_rate": 0.0005,
+        "loss": 1.1905,
+        "step": 9610
+      },
+      {
+        "epoch": 0.30676998628782803,
+        "grad_norm": 0.38213565945625305,
+        "learning_rate": 0.0005,
+        "loss": 1.1943,
+        "step": 9620
+      },
+      {
+        "epoch": 0.307088874007462,
+        "grad_norm": 0.38365045189857483,
+        "learning_rate": 0.0005,
+        "loss": 1.1645,
+        "step": 9630
+      },
+      {
+        "epoch": 0.3074077617270959,
+        "grad_norm": 0.37328922748565674,
+        "learning_rate": 0.0005,
+        "loss": 1.198,
+        "step": 9640
+      },
+      {
+        "epoch": 0.3077266494467298,
+        "grad_norm": 0.3832704424858093,
+        "learning_rate": 0.0005,
+        "loss": 1.1884,
+        "step": 9650
+      },
+      {
+        "epoch": 0.30804553716636374,
+        "grad_norm": 0.3734091818332672,
+        "learning_rate": 0.0005,
+        "loss": 1.1795,
+        "step": 9660
+      },
+      {
+        "epoch": 0.30836442488599763,
+        "grad_norm": 0.4023909568786621,
+        "learning_rate": 0.0005,
+        "loss": 1.1808,
+        "step": 9670
+      },
+      {
+        "epoch": 0.30868331260563153,
+        "grad_norm": 0.3570982813835144,
+        "learning_rate": 0.0005,
+        "loss": 1.1923,
+        "step": 9680
+      },
+      {
+        "epoch": 0.3090022003252655,
+        "grad_norm": 0.3908848166465759,
+        "learning_rate": 0.0005,
+        "loss": 1.2007,
+        "step": 9690
+      },
+      {
+        "epoch": 0.3093210880448994,
+        "grad_norm": 0.36542585492134094,
+        "learning_rate": 0.0005,
+        "loss": 1.1934,
+        "step": 9700
+      },
+      {
+        "epoch": 0.3096399757645333,
+        "grad_norm": 0.3709806799888611,
+        "learning_rate": 0.0005,
+        "loss": 1.2026,
+        "step": 9710
+      },
+      {
+        "epoch": 0.30995886348416724,
+        "grad_norm": 0.41271722316741943,
+        "learning_rate": 0.0005,
+        "loss": 1.2015,
+        "step": 9720
+      },
+      {
+        "epoch": 0.31027775120380113,
+        "grad_norm": 0.40473729372024536,
+        "learning_rate": 0.0005,
+        "loss": 1.2006,
+        "step": 9730
+      },
+      {
+        "epoch": 0.31059663892343503,
+        "grad_norm": 0.38929319381713867,
+        "learning_rate": 0.0005,
+        "loss": 1.1976,
+        "step": 9740
+      },
+      {
+        "epoch": 0.310915526643069,
+        "grad_norm": 0.3830064833164215,
+        "learning_rate": 0.0005,
+        "loss": 1.1798,
+        "step": 9750
+      },
+      {
+        "epoch": 0.3112344143627029,
+        "grad_norm": 0.3693712651729584,
+        "learning_rate": 0.0005,
+        "loss": 1.1867,
+        "step": 9760
+      },
+      {
+        "epoch": 0.31155330208233684,
+        "grad_norm": 0.3764219880104065,
+        "learning_rate": 0.0005,
+        "loss": 1.2072,
+        "step": 9770
+      },
+      {
+        "epoch": 0.31187218980197073,
+        "grad_norm": 0.3684931993484497,
+        "learning_rate": 0.0005,
+        "loss": 1.175,
+        "step": 9780
+      },
+      {
+        "epoch": 0.31219107752160463,
+        "grad_norm": 0.3662542402744293,
+        "learning_rate": 0.0005,
+        "loss": 1.1829,
+        "step": 9790
+      },
+      {
+        "epoch": 0.3125099652412386,
+        "grad_norm": 0.3679318428039551,
+        "learning_rate": 0.0005,
+        "loss": 1.1931,
+        "step": 9800
+      },
+      {
+        "epoch": 0.3128288529608725,
+        "grad_norm": 0.36969873309135437,
+        "learning_rate": 0.0005,
+        "loss": 1.1973,
+        "step": 9810
+      },
+      {
+        "epoch": 0.3131477406805064,
+        "grad_norm": 0.3816441595554352,
+        "learning_rate": 0.0005,
+        "loss": 1.1991,
+        "step": 9820
+      },
+      {
+        "epoch": 0.31346662840014033,
+        "grad_norm": 0.3715498149394989,
+        "learning_rate": 0.0005,
+        "loss": 1.1844,
+        "step": 9830
+      },
+      {
+        "epoch": 0.31378551611977423,
+        "grad_norm": 0.3715684413909912,
+        "learning_rate": 0.0005,
+        "loss": 1.1899,
+        "step": 9840
+      },
+      {
+        "epoch": 0.31410440383940813,
+        "grad_norm": 0.3811728358268738,
+        "learning_rate": 0.0005,
+        "loss": 1.1711,
+        "step": 9850
+      },
+      {
+        "epoch": 0.3144232915590421,
+        "grad_norm": 0.3746051490306854,
+        "learning_rate": 0.0005,
+        "loss": 1.1945,
+        "step": 9860
+      },
+      {
+        "epoch": 0.314742179278676,
+        "grad_norm": 0.36689209938049316,
+        "learning_rate": 0.0005,
+        "loss": 1.1963,
+        "step": 9870
+      },
+      {
+        "epoch": 0.3150610669983099,
+        "grad_norm": 0.3704276382923126,
+        "learning_rate": 0.0005,
+        "loss": 1.1974,
+        "step": 9880
+      },
+      {
+        "epoch": 0.31537995471794383,
+        "grad_norm": 0.3665519952774048,
+        "learning_rate": 0.0005,
+        "loss": 1.1768,
+        "step": 9890
+      },
+      {
+        "epoch": 0.31569884243757773,
+        "grad_norm": 0.3684837222099304,
+        "learning_rate": 0.0005,
+        "loss": 1.1824,
+        "step": 9900
+      },
+      {
+        "epoch": 0.31601773015721163,
+        "grad_norm": 0.37614014744758606,
+        "learning_rate": 0.0005,
+        "loss": 1.1725,
+        "step": 9910
+      },
+      {
+        "epoch": 0.3163366178768456,
+        "grad_norm": 0.3925752341747284,
+        "learning_rate": 0.0005,
+        "loss": 1.1831,
+        "step": 9920
+      },
+      {
+        "epoch": 0.3166555055964795,
+        "grad_norm": 0.3735015094280243,
+        "learning_rate": 0.0005,
+        "loss": 1.1932,
+        "step": 9930
+      },
+      {
+        "epoch": 0.3169743933161134,
+        "grad_norm": 0.37925341725349426,
+        "learning_rate": 0.0005,
+        "loss": 1.1925,
+        "step": 9940
+      },
+      {
+        "epoch": 0.31729328103574733,
+        "grad_norm": 0.3753606677055359,
+        "learning_rate": 0.0005,
+        "loss": 1.1986,
+        "step": 9950
+      },
+      {
+        "epoch": 0.31761216875538123,
+        "grad_norm": 0.3785763680934906,
+        "learning_rate": 0.0005,
+        "loss": 1.184,
+        "step": 9960
+      },
+      {
+        "epoch": 0.31793105647501513,
+        "grad_norm": 0.37484627962112427,
+        "learning_rate": 0.0005,
+        "loss": 1.2061,
+        "step": 9970
+      },
+      {
+        "epoch": 0.3182499441946491,
+        "grad_norm": 0.371409147977829,
+        "learning_rate": 0.0005,
+        "loss": 1.1843,
+        "step": 9980
+      },
+      {
+        "epoch": 0.318568831914283,
+        "grad_norm": 0.37367263436317444,
+        "learning_rate": 0.0005,
+        "loss": 1.1872,
+        "step": 9990
+      },
+      {
+        "epoch": 0.3188877196339169,
+        "grad_norm": 0.38979968428611755,
+        "learning_rate": 0.0005,
+        "loss": 1.1923,
+        "step": 10000
+      },
+      {
+        "epoch": 0.31920660735355083,
+        "grad_norm": 0.3798271119594574,
+        "learning_rate": 0.0005,
+        "loss": 1.194,
+        "step": 10010
+      },
+      {
+        "epoch": 0.31952549507318473,
+        "grad_norm": 0.38313472270965576,
+        "learning_rate": 0.0005,
+        "loss": 1.1788,
+        "step": 10020
+      },
+      {
+        "epoch": 0.3198443827928186,
+        "grad_norm": 0.372326135635376,
+        "learning_rate": 0.0005,
+        "loss": 1.1815,
+        "step": 10030
+      },
+      {
+        "epoch": 0.3201632705124526,
+        "grad_norm": 0.3908335566520691,
+        "learning_rate": 0.0005,
+        "loss": 1.1869,
+        "step": 10040
+      },
+      {
+        "epoch": 0.3204821582320865,
+        "grad_norm": 0.40376147627830505,
+        "learning_rate": 0.0005,
+        "loss": 1.2042,
+        "step": 10050
+      },
+      {
+        "epoch": 0.3208010459517204,
+        "grad_norm": 0.3750164806842804,
+        "learning_rate": 0.0005,
+        "loss": 1.1887,
+        "step": 10060
+      },
+      {
+        "epoch": 0.32111993367135433,
+        "grad_norm": 0.37391674518585205,
+        "learning_rate": 0.0005,
+        "loss": 1.1732,
+        "step": 10070
+      },
+      {
+        "epoch": 0.3214388213909882,
+        "grad_norm": 0.37153011560440063,
+        "learning_rate": 0.0005,
+        "loss": 1.1878,
+        "step": 10080
+      },
+      {
+        "epoch": 0.3217577091106221,
+        "grad_norm": 0.3877953886985779,
+        "learning_rate": 0.0005,
+        "loss": 1.1831,
+        "step": 10090
+      },
+      {
+        "epoch": 0.3220765968302561,
+        "grad_norm": 0.3789294958114624,
+        "learning_rate": 0.0005,
+        "loss": 1.1929,
+        "step": 10100
+      },
+      {
+        "epoch": 0.32239548454989,
+        "grad_norm": 0.3732856810092926,
+        "learning_rate": 0.0005,
+        "loss": 1.1915,
+        "step": 10110
+      },
+      {
+        "epoch": 0.3227143722695239,
+        "grad_norm": 0.35892724990844727,
+        "learning_rate": 0.0005,
+        "loss": 1.1653,
+        "step": 10120
+      },
+      {
+        "epoch": 0.32303325998915783,
+        "grad_norm": 0.39610254764556885,
+        "learning_rate": 0.0005,
+        "loss": 1.1752,
+        "step": 10130
+      },
+      {
+        "epoch": 0.3233521477087917,
+        "grad_norm": 0.3713800311088562,
+        "learning_rate": 0.0005,
+        "loss": 1.1784,
+        "step": 10140
+      },
+      {
+        "epoch": 0.3236710354284256,
+        "grad_norm": 0.36443793773651123,
+        "learning_rate": 0.0005,
+        "loss": 1.1908,
+        "step": 10150
+      },
+      {
+        "epoch": 0.3239899231480596,
+        "grad_norm": 0.3666205108165741,
+        "learning_rate": 0.0005,
+        "loss": 1.1808,
+        "step": 10160
+      },
+      {
+        "epoch": 0.3243088108676935,
+        "grad_norm": 0.3741157650947571,
+        "learning_rate": 0.0005,
+        "loss": 1.1761,
+        "step": 10170
+      },
+      {
+        "epoch": 0.32462769858732743,
+        "grad_norm": 0.37408700585365295,
+        "learning_rate": 0.0005,
+        "loss": 1.1757,
+        "step": 10180
+      },
+      {
+        "epoch": 0.3249465863069613,
+        "grad_norm": 0.3955936133861542,
+        "learning_rate": 0.0005,
+        "loss": 1.1673,
+        "step": 10190
+      },
+      {
+        "epoch": 0.3252654740265952,
+        "grad_norm": 0.36712345480918884,
+        "learning_rate": 0.0005,
+        "loss": 1.1777,
+        "step": 10200
+      },
+      {
+        "epoch": 0.3255843617462292,
+        "grad_norm": 0.37262213230133057,
+        "learning_rate": 0.0005,
+        "loss": 1.1777,
+        "step": 10210
+      },
+      {
+        "epoch": 0.3259032494658631,
+        "grad_norm": 0.37740543484687805,
+        "learning_rate": 0.0005,
+        "loss": 1.1778,
+        "step": 10220
+      },
+      {
+        "epoch": 0.326222137185497,
+        "grad_norm": 0.37197840213775635,
+        "learning_rate": 0.0005,
+        "loss": 1.2135,
+        "step": 10230
+      },
+      {
+        "epoch": 0.3265410249051309,
+        "grad_norm": 0.3649875521659851,
+        "learning_rate": 0.0005,
+        "loss": 1.1988,
+        "step": 10240
+      },
+      {
+        "epoch": 0.3268599126247648,
+        "grad_norm": 0.37759947776794434,
+        "learning_rate": 0.0005,
+        "loss": 1.1572,
+        "step": 10250
+      },
+      {
+        "epoch": 0.3271788003443987,
+        "grad_norm": 0.3873971104621887,
+        "learning_rate": 0.0005,
+        "loss": 1.1837,
+        "step": 10260
+      },
+      {
+        "epoch": 0.3274976880640327,
+        "grad_norm": 0.37649694085121155,
+        "learning_rate": 0.0005,
+        "loss": 1.1821,
+        "step": 10270
+      },
+      {
+        "epoch": 0.3278165757836666,
+        "grad_norm": 0.3752191662788391,
+        "learning_rate": 0.0005,
+        "loss": 1.1841,
+        "step": 10280
+      },
+      {
+        "epoch": 0.3281354635033005,
+        "grad_norm": 0.36976438760757446,
+        "learning_rate": 0.0005,
+        "loss": 1.1849,
+        "step": 10290
+      },
+      {
+        "epoch": 0.3284543512229344,
+        "grad_norm": 0.3746798038482666,
+        "learning_rate": 0.0005,
+        "loss": 1.1784,
+        "step": 10300
+      },
+      {
+        "epoch": 0.3287732389425683,
+        "grad_norm": 0.39386677742004395,
+        "learning_rate": 0.0005,
+        "loss": 1.2019,
+        "step": 10310
+      },
+      {
+        "epoch": 0.3290921266622022,
+        "grad_norm": 0.39480507373809814,
+        "learning_rate": 0.0005,
+        "loss": 1.1712,
+        "step": 10320
+      },
+      {
+        "epoch": 0.3294110143818362,
+        "grad_norm": 0.3793785572052002,
+        "learning_rate": 0.0005,
+        "loss": 1.2034,
+        "step": 10330
+      },
+      {
+        "epoch": 0.3297299021014701,
+        "grad_norm": 0.3793615698814392,
+        "learning_rate": 0.0005,
+        "loss": 1.1909,
+        "step": 10340
+      },
+      {
+        "epoch": 0.33004878982110397,
+        "grad_norm": 0.3702448308467865,
+        "learning_rate": 0.0005,
+        "loss": 1.1725,
+        "step": 10350
+      },
+      {
+        "epoch": 0.3303676775407379,
+        "grad_norm": 0.3776017725467682,
+        "learning_rate": 0.0005,
+        "loss": 1.1788,
+        "step": 10360
+      },
+      {
+        "epoch": 0.3306865652603718,
+        "grad_norm": 0.3725394606590271,
+        "learning_rate": 0.0005,
+        "loss": 1.2005,
+        "step": 10370
+      },
+      {
+        "epoch": 0.3310054529800057,
+        "grad_norm": 0.3663187623023987,
+        "learning_rate": 0.0005,
+        "loss": 1.1725,
+        "step": 10380
+      },
+      {
+        "epoch": 0.3313243406996397,
+        "grad_norm": 0.38458192348480225,
+        "learning_rate": 0.0005,
+        "loss": 1.1662,
+        "step": 10390
+      },
+      {
+        "epoch": 0.3316432284192736,
+        "grad_norm": 0.35204780101776123,
+        "learning_rate": 0.0005,
+        "loss": 1.184,
+        "step": 10400
+      },
+      {
+        "epoch": 0.33196211613890747,
+        "grad_norm": 0.35831499099731445,
+        "learning_rate": 0.0005,
+        "loss": 1.1646,
+        "step": 10410
+      },
+      {
+        "epoch": 0.3322810038585414,
+        "grad_norm": 0.36791619658470154,
+        "learning_rate": 0.0005,
+        "loss": 1.1792,
+        "step": 10420
+      },
+      {
+        "epoch": 0.3325998915781753,
+        "grad_norm": 0.3646932542324066,
+        "learning_rate": 0.0005,
+        "loss": 1.1671,
+        "step": 10430
+      },
+      {
+        "epoch": 0.3329187792978092,
+        "grad_norm": 0.36495405435562134,
+        "learning_rate": 0.0005,
+        "loss": 1.2007,
+        "step": 10440
+      },
+      {
+        "epoch": 0.3332376670174432,
+        "grad_norm": 0.3621331751346588,
+        "learning_rate": 0.0005,
+        "loss": 1.171,
+        "step": 10450
+      },
+      {
+        "epoch": 0.33355655473707707,
+        "grad_norm": 0.3788112998008728,
+        "learning_rate": 0.0005,
+        "loss": 1.1649,
+        "step": 10460
+      },
+      {
+        "epoch": 0.33387544245671097,
+        "grad_norm": 0.38030800223350525,
+        "learning_rate": 0.0005,
+        "loss": 1.1813,
+        "step": 10470
+      },
+      {
+        "epoch": 0.3341943301763449,
+        "grad_norm": 0.3575531840324402,
+        "learning_rate": 0.0005,
+        "loss": 1.1739,
+        "step": 10480
+      },
+      {
+        "epoch": 0.3345132178959788,
+        "grad_norm": 0.3673001527786255,
+        "learning_rate": 0.0005,
+        "loss": 1.1621,
+        "step": 10490
+      },
+      {
+        "epoch": 0.3348321056156127,
+        "grad_norm": 0.3933727443218231,
+        "learning_rate": 0.0005,
+        "loss": 1.1723,
+        "step": 10500
+      },
+      {
+        "epoch": 0.33515099333524667,
+        "grad_norm": 0.3708011209964752,
+        "learning_rate": 0.0005,
+        "loss": 1.1836,
+        "step": 10510
+      },
+      {
+        "epoch": 0.33546988105488057,
+        "grad_norm": 0.3957275450229645,
+        "learning_rate": 0.0005,
+        "loss": 1.1937,
+        "step": 10520
+      },
+      {
+        "epoch": 0.33578876877451447,
+        "grad_norm": 0.37122806906700134,
+        "learning_rate": 0.0005,
+        "loss": 1.1811,
+        "step": 10530
+      },
+      {
+        "epoch": 0.3361076564941484,
+        "grad_norm": 0.3752969205379486,
+        "learning_rate": 0.0005,
+        "loss": 1.1578,
+        "step": 10540
+      },
+      {
+        "epoch": 0.3364265442137823,
+        "grad_norm": 0.3687181770801544,
+        "learning_rate": 0.0005,
+        "loss": 1.1646,
+        "step": 10550
+      },
+      {
+        "epoch": 0.3367454319334162,
+        "grad_norm": 0.36669522523880005,
+        "learning_rate": 0.0005,
+        "loss": 1.1773,
+        "step": 10560
+      },
+      {
+        "epoch": 0.33706431965305017,
+        "grad_norm": 0.37474456429481506,
+        "learning_rate": 0.0005,
+        "loss": 1.1659,
+        "step": 10570
+      },
+      {
+        "epoch": 0.33738320737268407,
+        "grad_norm": 0.39351922273635864,
+        "learning_rate": 0.0005,
+        "loss": 1.1668,
+        "step": 10580
+      },
+      {
+        "epoch": 0.337702095092318,
+        "grad_norm": 0.3685235381126404,
+        "learning_rate": 0.0005,
+        "loss": 1.1431,
+        "step": 10590
+      },
+      {
+        "epoch": 0.3380209828119519,
+        "grad_norm": 0.3805023431777954,
+        "learning_rate": 0.0005,
+        "loss": 1.165,
+        "step": 10600
+      },
+      {
+        "epoch": 0.3383398705315858,
+        "grad_norm": 0.38222357630729675,
+        "learning_rate": 0.0005,
+        "loss": 1.1732,
+        "step": 10610
+      },
+      {
+        "epoch": 0.33865875825121977,
+        "grad_norm": 0.36740052700042725,
+        "learning_rate": 0.0005,
+        "loss": 1.1966,
+        "step": 10620
+      },
+      {
+        "epoch": 0.33897764597085367,
+        "grad_norm": 0.3713676929473877,
+        "learning_rate": 0.0005,
+        "loss": 1.1754,
+        "step": 10630
+      },
+      {
+        "epoch": 0.33929653369048757,
+        "grad_norm": 0.3694130480289459,
+        "learning_rate": 0.0005,
+        "loss": 1.1816,
+        "step": 10640
+      },
+      {
+        "epoch": 0.3396154214101215,
+        "grad_norm": 0.36890512704849243,
+        "learning_rate": 0.0005,
+        "loss": 1.1737,
+        "step": 10650
+      },
+      {
+        "epoch": 0.3399343091297554,
+        "grad_norm": 0.3650340735912323,
+        "learning_rate": 0.0005,
+        "loss": 1.1748,
+        "step": 10660
+      },
+      {
+        "epoch": 0.3402531968493893,
+        "grad_norm": 0.3503739833831787,
+        "learning_rate": 0.0005,
+        "loss": 1.1781,
+        "step": 10670
+      },
+      {
+        "epoch": 0.34057208456902327,
+        "grad_norm": 0.3604639172554016,
+        "learning_rate": 0.0005,
+        "loss": 1.1828,
+        "step": 10680
+      },
+      {
+        "epoch": 0.34089097228865717,
+        "grad_norm": 0.3638792932033539,
+        "learning_rate": 0.0005,
+        "loss": 1.175,
+        "step": 10690
+      },
+      {
+        "epoch": 0.34120986000829107,
+        "grad_norm": 0.36150917410850525,
+        "learning_rate": 0.0005,
+        "loss": 1.176,
+        "step": 10700
+      },
+      {
+        "epoch": 0.341528747727925,
+        "grad_norm": 0.35987234115600586,
+        "learning_rate": 0.0005,
+        "loss": 1.1682,
+        "step": 10710
+      },
+      {
+        "epoch": 0.3418476354475589,
+        "grad_norm": 0.36292168498039246,
+        "learning_rate": 0.0005,
+        "loss": 1.1788,
+        "step": 10720
+      },
+      {
+        "epoch": 0.3421665231671928,
+        "grad_norm": 0.38154059648513794,
+        "learning_rate": 0.0005,
+        "loss": 1.1536,
+        "step": 10730
+      },
+      {
+        "epoch": 0.34248541088682677,
+        "grad_norm": 0.37422358989715576,
+        "learning_rate": 0.0005,
+        "loss": 1.1845,
+        "step": 10740
+      },
+      {
+        "epoch": 0.34280429860646067,
+        "grad_norm": 0.35607483983039856,
+        "learning_rate": 0.0005,
+        "loss": 1.1679,
+        "step": 10750
+      },
+      {
+        "epoch": 0.34312318632609456,
+        "grad_norm": 0.3682827651500702,
+        "learning_rate": 0.0005,
+        "loss": 1.1579,
+        "step": 10760
+      },
+      {
+        "epoch": 0.3434420740457285,
+        "grad_norm": 0.3609450161457062,
+        "learning_rate": 0.0005,
+        "loss": 1.1527,
+        "step": 10770
+      },
+      {
+        "epoch": 0.3437609617653624,
+        "grad_norm": 0.3711645305156708,
+        "learning_rate": 0.0005,
+        "loss": 1.1739,
+        "step": 10780
+      },
+      {
+        "epoch": 0.3440798494849963,
+        "grad_norm": 0.3735518157482147,
+        "learning_rate": 0.0005,
+        "loss": 1.1652,
+        "step": 10790
+      },
+      {
+        "epoch": 0.34439873720463027,
+        "grad_norm": 0.3735860586166382,
+        "learning_rate": 0.0005,
+        "loss": 1.1748,
+        "step": 10800
+      },
+      {
+        "epoch": 0.34471762492426417,
+        "grad_norm": 0.3775559663772583,
+        "learning_rate": 0.0005,
+        "loss": 1.1929,
+        "step": 10810
+      },
+      {
+        "epoch": 0.34503651264389806,
+        "grad_norm": 0.3657887279987335,
+        "learning_rate": 0.0005,
+        "loss": 1.1805,
+        "step": 10820
+      },
+      {
+        "epoch": 0.345355400363532,
+        "grad_norm": 0.3744265139102936,
+        "learning_rate": 0.0005,
+        "loss": 1.1694,
+        "step": 10830
+      },
+      {
+        "epoch": 0.3456742880831659,
+        "grad_norm": 0.36719611287117004,
+        "learning_rate": 0.0005,
+        "loss": 1.1576,
+        "step": 10840
+      },
+      {
+        "epoch": 0.3459931758027998,
+        "grad_norm": 0.35522088408470154,
+        "learning_rate": 0.0005,
+        "loss": 1.1597,
+        "step": 10850
+      },
+      {
+        "epoch": 0.34631206352243377,
+        "grad_norm": 0.34695446491241455,
+        "learning_rate": 0.0005,
+        "loss": 1.1587,
+        "step": 10860
+      },
+      {
+        "epoch": 0.34663095124206766,
+        "grad_norm": 0.35940998792648315,
+        "learning_rate": 0.0005,
+        "loss": 1.1787,
+        "step": 10870
+      },
+      {
+        "epoch": 0.34694983896170156,
+        "grad_norm": 0.3649153411388397,
+        "learning_rate": 0.0005,
+        "loss": 1.1883,
+        "step": 10880
+      },
+      {
+        "epoch": 0.3472687266813355,
+        "grad_norm": 0.373657763004303,
+        "learning_rate": 0.0005,
+        "loss": 1.1751,
+        "step": 10890
+      },
+      {
+        "epoch": 0.3475876144009694,
+        "grad_norm": 0.38196080923080444,
+        "learning_rate": 0.0005,
+        "loss": 1.1819,
+        "step": 10900
+      },
+      {
+        "epoch": 0.3479065021206033,
+        "grad_norm": 0.3717299997806549,
+        "learning_rate": 0.0005,
+        "loss": 1.1535,
+        "step": 10910
+      },
+      {
+        "epoch": 0.34822538984023726,
+        "grad_norm": 0.3629402816295624,
+        "learning_rate": 0.0005,
+        "loss": 1.1781,
+        "step": 10920
+      },
+      {
+        "epoch": 0.34854427755987116,
+        "grad_norm": 0.3839849829673767,
+        "learning_rate": 0.0005,
+        "loss": 1.1799,
+        "step": 10930
+      },
+      {
+        "epoch": 0.34886316527950506,
+        "grad_norm": 0.3889947831630707,
+        "learning_rate": 0.0005,
+        "loss": 1.1562,
+        "step": 10940
+      },
+      {
+        "epoch": 0.349182052999139,
+        "grad_norm": 0.3629694879055023,
+        "learning_rate": 0.0005,
+        "loss": 1.1859,
+        "step": 10950
+      },
+      {
+        "epoch": 0.3495009407187729,
+        "grad_norm": 0.3563177287578583,
+        "learning_rate": 0.0005,
+        "loss": 1.1712,
+        "step": 10960
+      },
+      {
+        "epoch": 0.3498198284384068,
+        "grad_norm": 0.3640787601470947,
+        "learning_rate": 0.0005,
+        "loss": 1.1764,
+        "step": 10970
+      },
+      {
+        "epoch": 0.35013871615804076,
+        "grad_norm": 0.36180850863456726,
+        "learning_rate": 0.0005,
+        "loss": 1.1752,
+        "step": 10980
+      },
+      {
+        "epoch": 0.35045760387767466,
+        "grad_norm": 0.36016371846199036,
+        "learning_rate": 0.0005,
+        "loss": 1.1529,
+        "step": 10990
+      },
+      {
+        "epoch": 0.3507764915973086,
+        "grad_norm": 0.37246090173721313,
+        "learning_rate": 0.0005,
+        "loss": 1.1658,
+        "step": 11000
+      },
+      {
+        "epoch": 0.3510953793169425,
+        "grad_norm": 0.38452988862991333,
+        "learning_rate": 0.0005,
+        "loss": 1.1807,
+        "step": 11010
+      },
+      {
+        "epoch": 0.3514142670365764,
+        "grad_norm": 0.3714504539966583,
+        "learning_rate": 0.0005,
+        "loss": 1.1618,
+        "step": 11020
+      },
+      {
+        "epoch": 0.35173315475621036,
+        "grad_norm": 0.3711185157299042,
+        "learning_rate": 0.0005,
+        "loss": 1.1423,
+        "step": 11030
+      },
+      {
+        "epoch": 0.35205204247584426,
+        "grad_norm": 0.36155474185943604,
+        "learning_rate": 0.0005,
+        "loss": 1.1618,
+        "step": 11040
+      },
+      {
+        "epoch": 0.35237093019547816,
+        "grad_norm": 0.36459237337112427,
+        "learning_rate": 0.0005,
+        "loss": 1.1434,
+        "step": 11050
+      },
+      {
+        "epoch": 0.3526898179151121,
+        "grad_norm": 0.3625276982784271,
+        "learning_rate": 0.0005,
+        "loss": 1.1632,
+        "step": 11060
+      },
+      {
+        "epoch": 0.353008705634746,
+        "grad_norm": 0.35428375005722046,
+        "learning_rate": 0.0005,
+        "loss": 1.1604,
+        "step": 11070
+      },
+      {
+        "epoch": 0.3533275933543799,
+        "grad_norm": 0.36311075091362,
+        "learning_rate": 0.0005,
+        "loss": 1.1588,
+        "step": 11080
+      },
+      {
+        "epoch": 0.35364648107401386,
+        "grad_norm": 0.357806921005249,
+        "learning_rate": 0.0005,
+        "loss": 1.1489,
+        "step": 11090
+      },
+      {
+        "epoch": 0.35396536879364776,
+        "grad_norm": 0.3589811623096466,
+        "learning_rate": 0.0005,
+        "loss": 1.1464,
+        "step": 11100
+      },
+      {
+        "epoch": 0.35428425651328166,
+        "grad_norm": 0.3645656704902649,
+        "learning_rate": 0.0005,
+        "loss": 1.1691,
+        "step": 11110
+      },
+      {
+        "epoch": 0.3546031442329156,
+        "grad_norm": 0.37154215574264526,
+        "learning_rate": 0.0005,
+        "loss": 1.1766,
+        "step": 11120
+      },
+      {
+        "epoch": 0.3549220319525495,
+        "grad_norm": 0.36165502667427063,
+        "learning_rate": 0.0005,
+        "loss": 1.1706,
+        "step": 11130
+      },
+      {
+        "epoch": 0.3552409196721834,
+        "grad_norm": 0.36878177523612976,
+        "learning_rate": 0.0005,
+        "loss": 1.1641,
+        "step": 11140
+      },
+      {
+        "epoch": 0.35555980739181736,
+        "grad_norm": 0.3841949701309204,
+        "learning_rate": 0.0005,
+        "loss": 1.1614,
+        "step": 11150
+      },
+      {
+        "epoch": 0.35587869511145126,
+        "grad_norm": 0.36055484414100647,
+        "learning_rate": 0.0005,
+        "loss": 1.1699,
+        "step": 11160
+      },
+      {
+        "epoch": 0.35619758283108516,
+        "grad_norm": 0.3562716841697693,
+        "learning_rate": 0.0005,
+        "loss": 1.1683,
+        "step": 11170
+      },
+      {
+        "epoch": 0.3565164705507191,
+        "grad_norm": 0.37487974762916565,
+        "learning_rate": 0.0005,
+        "loss": 1.1617,
+        "step": 11180
+      },
+      {
+        "epoch": 0.356835358270353,
+        "grad_norm": 0.37108105421066284,
+        "learning_rate": 0.0005,
+        "loss": 1.1582,
+        "step": 11190
+      },
+      {
+        "epoch": 0.3571542459899869,
+        "grad_norm": 0.35356369614601135,
+        "learning_rate": 0.0005,
+        "loss": 1.1382,
+        "step": 11200
+      },
+      {
+        "epoch": 0.35747313370962086,
+        "grad_norm": 0.3768700361251831,
+        "learning_rate": 0.0005,
+        "loss": 1.1722,
+        "step": 11210
+      },
+      {
+        "epoch": 0.35779202142925476,
+        "grad_norm": 0.3485049605369568,
+        "learning_rate": 0.0005,
+        "loss": 1.1617,
+        "step": 11220
+      },
+      {
+        "epoch": 0.35811090914888866,
+        "grad_norm": 0.36197659373283386,
+        "learning_rate": 0.0005,
+        "loss": 1.181,
+        "step": 11230
+      },
+      {
+        "epoch": 0.3584297968685226,
+        "grad_norm": 0.37221449613571167,
+        "learning_rate": 0.0005,
+        "loss": 1.1772,
+        "step": 11240
+      },
+      {
+        "epoch": 0.3587486845881565,
+        "grad_norm": 0.36957746744155884,
+        "learning_rate": 0.0005,
+        "loss": 1.154,
+        "step": 11250
+      },
+      {
+        "epoch": 0.3590675723077904,
+        "grad_norm": 0.35753747820854187,
+        "learning_rate": 0.0005,
+        "loss": 1.1469,
+        "step": 11260
+      },
+      {
+        "epoch": 0.35938646002742436,
+        "grad_norm": 0.36781930923461914,
+        "learning_rate": 0.0005,
+        "loss": 1.1735,
+        "step": 11270
+      },
+      {
+        "epoch": 0.35970534774705826,
+        "grad_norm": 0.3567204177379608,
+        "learning_rate": 0.0005,
+        "loss": 1.1451,
+        "step": 11280
+      },
+      {
+        "epoch": 0.36002423546669216,
+        "grad_norm": 0.3888096809387207,
+        "learning_rate": 0.0005,
+        "loss": 1.1547,
+        "step": 11290
+      },
+      {
+        "epoch": 0.3603431231863261,
+        "grad_norm": 0.3750171959400177,
+        "learning_rate": 0.0005,
+        "loss": 1.1634,
+        "step": 11300
+      },
+      {
+        "epoch": 0.36066201090596,
+        "grad_norm": 0.3716561496257782,
+        "learning_rate": 0.0005,
+        "loss": 1.1548,
+        "step": 11310
+      },
+      {
+        "epoch": 0.3609808986255939,
+        "grad_norm": 0.3608260452747345,
+        "learning_rate": 0.0005,
+        "loss": 1.1478,
+        "step": 11320
+      },
+      {
+        "epoch": 0.36129978634522786,
+        "grad_norm": 0.3671262264251709,
+        "learning_rate": 0.0005,
+        "loss": 1.149,
+        "step": 11330
+      },
+      {
+        "epoch": 0.36161867406486176,
+        "grad_norm": 0.3494800627231598,
+        "learning_rate": 0.0005,
+        "loss": 1.1707,
+        "step": 11340
+      },
+      {
+        "epoch": 0.36193756178449565,
+        "grad_norm": 0.35910022258758545,
+        "learning_rate": 0.0005,
+        "loss": 1.1607,
+        "step": 11350
+      },
+      {
+        "epoch": 0.3622564495041296,
+        "grad_norm": 0.35987815260887146,
+        "learning_rate": 0.0005,
+        "loss": 1.17,
+        "step": 11360
+      },
+      {
+        "epoch": 0.3625753372237635,
+        "grad_norm": 0.36094966530799866,
+        "learning_rate": 0.0005,
+        "loss": 1.157,
+        "step": 11370
+      },
+      {
+        "epoch": 0.3628942249433974,
+        "grad_norm": 0.3670837879180908,
+        "learning_rate": 0.0005,
+        "loss": 1.1536,
+        "step": 11380
+      },
+      {
+        "epoch": 0.36321311266303136,
+        "grad_norm": 0.3617798984050751,
+        "learning_rate": 0.0005,
+        "loss": 1.1596,
+        "step": 11390
+      },
+      {
+        "epoch": 0.36353200038266525,
+        "grad_norm": 0.3594600260257721,
+        "learning_rate": 0.0005,
+        "loss": 1.1657,
+        "step": 11400
+      },
+      {
+        "epoch": 0.3638508881022992,
+        "grad_norm": 0.36290115118026733,
+        "learning_rate": 0.0005,
+        "loss": 1.1482,
+        "step": 11410
+      },
+      {
+        "epoch": 0.3641697758219331,
+        "grad_norm": 0.3666008412837982,
+        "learning_rate": 0.0005,
+        "loss": 1.1577,
+        "step": 11420
+      },
+      {
+        "epoch": 0.364488663541567,
+        "grad_norm": 0.36086878180503845,
+        "learning_rate": 0.0005,
+        "loss": 1.1377,
+        "step": 11430
+      },
+      {
+        "epoch": 0.36480755126120096,
+        "grad_norm": 0.36125829815864563,
+        "learning_rate": 0.0005,
+        "loss": 1.1607,
+        "step": 11440
+      },
+      {
+        "epoch": 0.36512643898083486,
+        "grad_norm": 0.35635900497436523,
+        "learning_rate": 0.0005,
+        "loss": 1.1602,
+        "step": 11450
+      },
+      {
+        "epoch": 0.36544532670046875,
+        "grad_norm": 0.3567068874835968,
+        "learning_rate": 0.0005,
+        "loss": 1.1432,
+        "step": 11460
+      },
+      {
+        "epoch": 0.3657642144201027,
+        "grad_norm": 0.37255966663360596,
+        "learning_rate": 0.0005,
+        "loss": 1.1605,
+        "step": 11470
+      },
+      {
+        "epoch": 0.3660831021397366,
+        "grad_norm": 0.36005812883377075,
+        "learning_rate": 0.0005,
+        "loss": 1.1489,
+        "step": 11480
+      },
+      {
+        "epoch": 0.3664019898593705,
+        "grad_norm": 0.36554327607154846,
+        "learning_rate": 0.0005,
+        "loss": 1.1594,
+        "step": 11490
+      },
+      {
+        "epoch": 0.36672087757900446,
+        "grad_norm": 0.3808325231075287,
+        "learning_rate": 0.0005,
+        "loss": 1.1783,
+        "step": 11500
+      },
+      {
+        "epoch": 0.36703976529863835,
+        "grad_norm": 0.35380932688713074,
+        "learning_rate": 0.0005,
+        "loss": 1.1485,
+        "step": 11510
+      },
+      {
+        "epoch": 0.36735865301827225,
+        "grad_norm": 0.3594566881656647,
+        "learning_rate": 0.0005,
+        "loss": 1.1728,
+        "step": 11520
+      },
+      {
+        "epoch": 0.3676775407379062,
+        "grad_norm": 0.36721861362457275,
+        "learning_rate": 0.0005,
+        "loss": 1.1462,
+        "step": 11530
+      },
+      {
+        "epoch": 0.3679964284575401,
+        "grad_norm": 0.3698878586292267,
+        "learning_rate": 0.0005,
+        "loss": 1.1349,
+        "step": 11540
+      },
+      {
+        "epoch": 0.368315316177174,
+        "grad_norm": 0.35534027218818665,
+        "learning_rate": 0.0005,
+        "loss": 1.1605,
+        "step": 11550
+      },
+      {
+        "epoch": 0.36863420389680795,
+        "grad_norm": 0.36262908577919006,
+        "learning_rate": 0.0005,
+        "loss": 1.1527,
+        "step": 11560
+      },
+      {
+        "epoch": 0.36895309161644185,
+        "grad_norm": 0.3791918158531189,
+        "learning_rate": 0.0005,
+        "loss": 1.1658,
+        "step": 11570
+      },
+      {
+        "epoch": 0.36927197933607575,
+        "grad_norm": 0.413955956697464,
+        "learning_rate": 0.0005,
+        "loss": 1.1337,
+        "step": 11580
+      },
+      {
+        "epoch": 0.3695908670557097,
+        "grad_norm": 0.3605935275554657,
+        "learning_rate": 0.0005,
+        "loss": 1.1422,
+        "step": 11590
+      },
+      {
+        "epoch": 0.3699097547753436,
+        "grad_norm": 0.36769089102745056,
+        "learning_rate": 0.0005,
+        "loss": 1.1756,
+        "step": 11600
+      },
+      {
+        "epoch": 0.3702286424949775,
+        "grad_norm": 0.35838136076927185,
+        "learning_rate": 0.0005,
+        "loss": 1.1598,
+        "step": 11610
+      },
+      {
+        "epoch": 0.37054753021461145,
+        "grad_norm": 0.3628617227077484,
+        "learning_rate": 0.0005,
+        "loss": 1.1712,
+        "step": 11620
+      },
+      {
+        "epoch": 0.37086641793424535,
+        "grad_norm": 0.35749825835227966,
+        "learning_rate": 0.0005,
+        "loss": 1.1641,
+        "step": 11630
+      },
+      {
+        "epoch": 0.37118530565387925,
+        "grad_norm": 0.3627100884914398,
+        "learning_rate": 0.0005,
+        "loss": 1.1821,
+        "step": 11640
+      },
+      {
+        "epoch": 0.3715041933735132,
+        "grad_norm": 0.3494734764099121,
+        "learning_rate": 0.0005,
+        "loss": 1.1492,
+        "step": 11650
+      },
+      {
+        "epoch": 0.3718230810931471,
+        "grad_norm": 0.3884241580963135,
+        "learning_rate": 0.0005,
+        "loss": 1.1744,
+        "step": 11660
+      },
+      {
+        "epoch": 0.372141968812781,
+        "grad_norm": 0.3681718111038208,
+        "learning_rate": 0.0005,
+        "loss": 1.1778,
+        "step": 11670
+      },
+      {
+        "epoch": 0.37246085653241495,
+        "grad_norm": 0.36322730779647827,
+        "learning_rate": 0.0005,
+        "loss": 1.1552,
+        "step": 11680
+      },
+      {
+        "epoch": 0.37277974425204885,
+        "grad_norm": 0.35371363162994385,
+        "learning_rate": 0.0005,
+        "loss": 1.1498,
+        "step": 11690
+      },
+      {
+        "epoch": 0.37309863197168275,
+        "grad_norm": 0.34495317935943604,
+        "learning_rate": 0.0005,
+        "loss": 1.1393,
+        "step": 11700
+      },
+      {
+        "epoch": 0.3734175196913167,
+        "grad_norm": 0.34707680344581604,
+        "learning_rate": 0.0005,
+        "loss": 1.1549,
+        "step": 11710
+      },
+      {
+        "epoch": 0.3737364074109506,
+        "grad_norm": 0.3565426766872406,
+        "learning_rate": 0.0005,
+        "loss": 1.1512,
+        "step": 11720
+      },
+      {
+        "epoch": 0.3740552951305845,
+        "grad_norm": 0.3783700466156006,
+        "learning_rate": 0.0005,
+        "loss": 1.1614,
+        "step": 11730
+      },
+      {
+        "epoch": 0.37437418285021845,
+        "grad_norm": 0.35486361384391785,
+        "learning_rate": 0.0005,
+        "loss": 1.1666,
+        "step": 11740
+      },
+      {
+        "epoch": 0.37469307056985235,
+        "grad_norm": 0.3659943640232086,
+        "learning_rate": 0.0005,
+        "loss": 1.1722,
+        "step": 11750
+      },
+      {
+        "epoch": 0.37501195828948625,
+        "grad_norm": 0.35426226258277893,
+        "learning_rate": 0.0005,
+        "loss": 1.1636,
+        "step": 11760
+      },
+      {
+        "epoch": 0.3753308460091202,
+        "grad_norm": 0.358793169260025,
+        "learning_rate": 0.0005,
+        "loss": 1.1649,
+        "step": 11770
+      },
+      {
+        "epoch": 0.3756497337287541,
+        "grad_norm": 0.35559338331222534,
+        "learning_rate": 0.0005,
+        "loss": 1.1591,
+        "step": 11780
+      },
+      {
+        "epoch": 0.375968621448388,
+        "grad_norm": 0.3519185781478882,
+        "learning_rate": 0.0005,
+        "loss": 1.1787,
+        "step": 11790
+      },
+      {
+        "epoch": 0.37628750916802195,
+        "grad_norm": 0.3474023640155792,
+        "learning_rate": 0.0005,
+        "loss": 1.1565,
+        "step": 11800
+      },
+      {
+        "epoch": 0.37660639688765585,
+        "grad_norm": 0.3500244915485382,
+        "learning_rate": 0.0005,
+        "loss": 1.1622,
+        "step": 11810
+      },
+      {
+        "epoch": 0.37692528460728975,
+        "grad_norm": 0.3565733730792999,
+        "learning_rate": 0.0005,
+        "loss": 1.164,
+        "step": 11820
+      },
+      {
+        "epoch": 0.3772441723269237,
+        "grad_norm": 0.36334407329559326,
+        "learning_rate": 0.0005,
+        "loss": 1.1387,
+        "step": 11830
+      },
+      {
+        "epoch": 0.3775630600465576,
+        "grad_norm": 0.3599485158920288,
+        "learning_rate": 0.0005,
+        "loss": 1.1615,
+        "step": 11840
+      },
+      {
+        "epoch": 0.37788194776619155,
+        "grad_norm": 0.3623838722705841,
+        "learning_rate": 0.0005,
+        "loss": 1.1589,
+        "step": 11850
+      },
+      {
+        "epoch": 0.37820083548582545,
+        "grad_norm": 0.3498895764350891,
+        "learning_rate": 0.0005,
+        "loss": 1.1644,
+        "step": 11860
+      },
+      {
+        "epoch": 0.37851972320545935,
+        "grad_norm": 0.36630934476852417,
+        "learning_rate": 0.0005,
+        "loss": 1.1632,
+        "step": 11870
+      },
+      {
+        "epoch": 0.3788386109250933,
+        "grad_norm": 0.3576071858406067,
+        "learning_rate": 0.0005,
+        "loss": 1.1638,
+        "step": 11880
+      },
+      {
+        "epoch": 0.3791574986447272,
+        "grad_norm": 0.34983667731285095,
+        "learning_rate": 0.0005,
+        "loss": 1.1516,
+        "step": 11890
+      },
+      {
+        "epoch": 0.3794763863643611,
+        "grad_norm": 0.35930803418159485,
+        "learning_rate": 0.0005,
+        "loss": 1.1447,
+        "step": 11900
+      },
+      {
+        "epoch": 0.37979527408399505,
+        "grad_norm": 0.3543728291988373,
+        "learning_rate": 0.0005,
+        "loss": 1.1642,
+        "step": 11910
+      },
+      {
+        "epoch": 0.38011416180362895,
+        "grad_norm": 0.35220077633857727,
+        "learning_rate": 0.0005,
+        "loss": 1.1647,
+        "step": 11920
+      },
+      {
+        "epoch": 0.38043304952326285,
+        "grad_norm": 0.3517967462539673,
+        "learning_rate": 0.0005,
+        "loss": 1.1344,
+        "step": 11930
+      },
+      {
+        "epoch": 0.3807519372428968,
+        "grad_norm": 0.34194129705429077,
+        "learning_rate": 0.0005,
+        "loss": 1.1277,
+        "step": 11940
+      },
+      {
+        "epoch": 0.3810708249625307,
+        "grad_norm": 0.35568177700042725,
+        "learning_rate": 0.0005,
+        "loss": 1.1492,
+        "step": 11950
+      },
+      {
+        "epoch": 0.3813897126821646,
+        "grad_norm": 0.36249402165412903,
+        "learning_rate": 0.0005,
+        "loss": 1.1629,
+        "step": 11960
+      },
+      {
+        "epoch": 0.38170860040179855,
+        "grad_norm": 0.3520861566066742,
+        "learning_rate": 0.0005,
+        "loss": 1.1556,
+        "step": 11970
+      },
+      {
+        "epoch": 0.38202748812143245,
+        "grad_norm": 0.3475063741207123,
+        "learning_rate": 0.0005,
+        "loss": 1.137,
+        "step": 11980
+      },
+      {
+        "epoch": 0.38234637584106634,
+        "grad_norm": 0.3599000573158264,
+        "learning_rate": 0.0005,
+        "loss": 1.147,
+        "step": 11990
+      },
+      {
+        "epoch": 0.3826652635607003,
+        "grad_norm": 0.38423648476600647,
+        "learning_rate": 0.0005,
+        "loss": 1.1504,
+        "step": 12000
+      },
+      {
+        "epoch": 0.3829841512803342,
+        "grad_norm": 0.3562023937702179,
+        "learning_rate": 0.0005,
+        "loss": 1.1617,
+        "step": 12010
+      },
+      {
+        "epoch": 0.3833030389999681,
+        "grad_norm": 0.3565387725830078,
+        "learning_rate": 0.0005,
+        "loss": 1.1744,
+        "step": 12020
+      },
+      {
+        "epoch": 0.38362192671960205,
+        "grad_norm": 0.34507426619529724,
+        "learning_rate": 0.0005,
+        "loss": 1.1502,
+        "step": 12030
+      },
+      {
+        "epoch": 0.38394081443923594,
+        "grad_norm": 0.36354437470436096,
+        "learning_rate": 0.0005,
+        "loss": 1.1442,
+        "step": 12040
+      },
+      {
+        "epoch": 0.38425970215886984,
+        "grad_norm": 0.34563297033309937,
+        "learning_rate": 0.0005,
+        "loss": 1.1528,
+        "step": 12050
+      },
+      {
+        "epoch": 0.3845785898785038,
+        "grad_norm": 0.3544037640094757,
+        "learning_rate": 0.0005,
+        "loss": 1.1394,
+        "step": 12060
+      },
+      {
+        "epoch": 0.3848974775981377,
+        "grad_norm": 0.34711146354675293,
+        "learning_rate": 0.0005,
+        "loss": 1.1584,
+        "step": 12070
+      },
+      {
+        "epoch": 0.3852163653177716,
+        "grad_norm": 0.353064626455307,
+        "learning_rate": 0.0005,
+        "loss": 1.1553,
+        "step": 12080
+      },
+      {
+        "epoch": 0.38553525303740555,
+        "grad_norm": 0.35694098472595215,
+        "learning_rate": 0.0005,
+        "loss": 1.1472,
+        "step": 12090
+      },
+      {
+        "epoch": 0.38585414075703944,
+        "grad_norm": 0.3677344024181366,
+        "learning_rate": 0.0005,
+        "loss": 1.1561,
+        "step": 12100
+      },
+      {
+        "epoch": 0.38617302847667334,
+        "grad_norm": 0.3596011698246002,
+        "learning_rate": 0.0005,
+        "loss": 1.1589,
+        "step": 12110
+      },
+      {
+        "epoch": 0.3864919161963073,
+        "grad_norm": 0.3607899248600006,
+        "learning_rate": 0.0005,
+        "loss": 1.1641,
+        "step": 12120
+      },
+      {
+        "epoch": 0.3868108039159412,
+        "grad_norm": 0.37189728021621704,
+        "learning_rate": 0.0005,
+        "loss": 1.1612,
+        "step": 12130
+      },
+      {
+        "epoch": 0.3871296916355751,
+        "grad_norm": 0.34136825799942017,
+        "learning_rate": 0.0005,
+        "loss": 1.1266,
+        "step": 12140
+      },
+      {
+        "epoch": 0.38744857935520904,
+        "grad_norm": 0.34620654582977295,
+        "learning_rate": 0.0005,
+        "loss": 1.1665,
+        "step": 12150
+      },
+      {
+        "epoch": 0.38776746707484294,
+        "grad_norm": 0.34414157271385193,
+        "learning_rate": 0.0005,
+        "loss": 1.1644,
+        "step": 12160
+      },
+      {
+        "epoch": 0.38808635479447684,
+        "grad_norm": 0.34810659289360046,
+        "learning_rate": 0.0005,
+        "loss": 1.1253,
+        "step": 12170
+      },
+      {
+        "epoch": 0.3884052425141108,
+        "grad_norm": 0.34679773449897766,
+        "learning_rate": 0.0005,
+        "loss": 1.1718,
+        "step": 12180
+      },
+      {
+        "epoch": 0.3887241302337447,
+        "grad_norm": 0.3541298806667328,
+        "learning_rate": 0.0005,
+        "loss": 1.1441,
+        "step": 12190
+      },
+      {
+        "epoch": 0.3890430179533786,
+        "grad_norm": 0.35697078704833984,
+        "learning_rate": 0.0005,
+        "loss": 1.1455,
+        "step": 12200
+      },
+      {
+        "epoch": 0.38936190567301254,
+        "grad_norm": 0.34624364972114563,
+        "learning_rate": 0.0005,
+        "loss": 1.1545,
+        "step": 12210
+      },
+      {
+        "epoch": 0.38968079339264644,
+        "grad_norm": 0.3443002998828888,
+        "learning_rate": 0.0005,
+        "loss": 1.1549,
+        "step": 12220
+      },
+      {
+        "epoch": 0.38999968111228034,
+        "grad_norm": 0.34411001205444336,
+        "learning_rate": 0.0005,
+        "loss": 1.1437,
+        "step": 12230
+      },
+      {
+        "epoch": 0.3903185688319143,
+        "grad_norm": 0.38664522767066956,
+        "learning_rate": 0.0005,
+        "loss": 1.1393,
+        "step": 12240
+      },
+      {
+        "epoch": 0.3906374565515482,
+        "grad_norm": 0.36069822311401367,
+        "learning_rate": 0.0005,
+        "loss": 1.1302,
+        "step": 12250
+      },
+      {
+        "epoch": 0.39095634427118214,
+        "grad_norm": 0.35054412484169006,
+        "learning_rate": 0.0005,
+        "loss": 1.1452,
+        "step": 12260
+      },
+      {
+        "epoch": 0.39127523199081604,
+        "grad_norm": 0.3593745827674866,
+        "learning_rate": 0.0005,
+        "loss": 1.1593,
+        "step": 12270
+      },
+      {
+        "epoch": 0.39159411971044994,
+        "grad_norm": 0.36931827664375305,
+        "learning_rate": 0.0005,
+        "loss": 1.1431,
+        "step": 12280
+      },
+      {
+        "epoch": 0.3919130074300839,
+        "grad_norm": 0.3705731928348541,
+        "learning_rate": 0.0005,
+        "loss": 1.1467,
+        "step": 12290
+      },
+      {
+        "epoch": 0.3922318951497178,
+        "grad_norm": 0.35201239585876465,
+        "learning_rate": 0.0005,
+        "loss": 1.1658,
+        "step": 12300
+      },
+      {
+        "epoch": 0.3925507828693517,
+        "grad_norm": 0.33797210454940796,
+        "learning_rate": 0.0005,
+        "loss": 1.1491,
+        "step": 12310
+      },
+      {
+        "epoch": 0.39286967058898564,
+        "grad_norm": 0.3512865900993347,
+        "learning_rate": 0.0005,
+        "loss": 1.1571,
+        "step": 12320
+      },
+      {
+        "epoch": 0.39318855830861954,
+        "grad_norm": 0.35392460227012634,
+        "learning_rate": 0.0005,
+        "loss": 1.1542,
+        "step": 12330
+      },
+      {
+        "epoch": 0.39350744602825344,
+        "grad_norm": 0.34811460971832275,
+        "learning_rate": 0.0005,
+        "loss": 1.1351,
+        "step": 12340
+      },
+      {
+        "epoch": 0.3938263337478874,
+        "grad_norm": 0.3421926200389862,
+        "learning_rate": 0.0005,
+        "loss": 1.1344,
+        "step": 12350
+      },
+      {
+        "epoch": 0.3941452214675213,
+        "grad_norm": 0.3451977074146271,
+        "learning_rate": 0.0005,
+        "loss": 1.176,
+        "step": 12360
+      },
+      {
+        "epoch": 0.3944641091871552,
+        "grad_norm": 0.33949071168899536,
+        "learning_rate": 0.0005,
+        "loss": 1.1455,
+        "step": 12370
+      },
+      {
+        "epoch": 0.39478299690678914,
+        "grad_norm": 0.35217025876045227,
+        "learning_rate": 0.0005,
+        "loss": 1.1395,
+        "step": 12380
+      },
+      {
+        "epoch": 0.39510188462642304,
+        "grad_norm": 0.35102054476737976,
+        "learning_rate": 0.0005,
+        "loss": 1.1419,
+        "step": 12390
+      },
+      {
+        "epoch": 0.39542077234605694,
+        "grad_norm": 0.3483518362045288,
+        "learning_rate": 0.0005,
+        "loss": 1.1571,
+        "step": 12400
+      },
+      {
+        "epoch": 0.3957396600656909,
+        "grad_norm": 0.35172149538993835,
+        "learning_rate": 0.0005,
+        "loss": 1.1394,
+        "step": 12410
+      },
+      {
+        "epoch": 0.3960585477853248,
+        "grad_norm": 0.3434382379055023,
+        "learning_rate": 0.0005,
+        "loss": 1.1504,
+        "step": 12420
+      },
+      {
+        "epoch": 0.3963774355049587,
+        "grad_norm": 0.3492888808250427,
+        "learning_rate": 0.0005,
+        "loss": 1.1534,
+        "step": 12430
+      },
+      {
+        "epoch": 0.39669632322459264,
+        "grad_norm": 0.34085673093795776,
+        "learning_rate": 0.0005,
+        "loss": 1.1288,
+        "step": 12440
+      },
+      {
+        "epoch": 0.39701521094422654,
+        "grad_norm": 0.35518237948417664,
+        "learning_rate": 0.0005,
+        "loss": 1.1519,
+        "step": 12450
+      },
+      {
+        "epoch": 0.39733409866386044,
+        "grad_norm": 0.3471197187900543,
+        "learning_rate": 0.0005,
+        "loss": 1.1358,
+        "step": 12460
+      },
+      {
+        "epoch": 0.3976529863834944,
+        "grad_norm": 0.36180561780929565,
+        "learning_rate": 0.0005,
+        "loss": 1.1609,
+        "step": 12470
+      },
+      {
+        "epoch": 0.3979718741031283,
+        "grad_norm": 0.3461403250694275,
+        "learning_rate": 0.0005,
+        "loss": 1.158,
+        "step": 12480
+      },
+      {
+        "epoch": 0.3982907618227622,
+        "grad_norm": 0.342271089553833,
+        "learning_rate": 0.0005,
+        "loss": 1.144,
+        "step": 12490
+      },
+      {
+        "epoch": 0.39860964954239614,
+        "grad_norm": 0.3423519730567932,
+        "learning_rate": 0.0005,
+        "loss": 1.1634,
+        "step": 12500
+      },
+      {
+        "epoch": 0.39892853726203004,
+        "grad_norm": 0.365598201751709,
+        "learning_rate": 0.0005,
+        "loss": 1.1384,
+        "step": 12510
+      },
+      {
+        "epoch": 0.39924742498166393,
+        "grad_norm": 0.3581569492816925,
+        "learning_rate": 0.0005,
+        "loss": 1.1419,
+        "step": 12520
+      },
+      {
+        "epoch": 0.3995663127012979,
+        "grad_norm": 0.3360135853290558,
+        "learning_rate": 0.0005,
+        "loss": 1.1578,
+        "step": 12530
+      },
+      {
+        "epoch": 0.3998852004209318,
+        "grad_norm": 0.3541604280471802,
+        "learning_rate": 0.0005,
+        "loss": 1.144,
+        "step": 12540
+      },
+      {
+        "epoch": 0.4002040881405657,
+        "grad_norm": 0.3437075912952423,
+        "learning_rate": 0.0005,
+        "loss": 1.1229,
+        "step": 12550
+      },
+      {
+        "epoch": 0.40052297586019964,
+        "grad_norm": 0.3445467948913574,
+        "learning_rate": 0.0005,
+        "loss": 1.1386,
+        "step": 12560
+      },
+      {
+        "epoch": 0.40084186357983353,
+        "grad_norm": 0.35498735308647156,
+        "learning_rate": 0.0005,
+        "loss": 1.1668,
+        "step": 12570
+      },
+      {
+        "epoch": 0.40116075129946743,
+        "grad_norm": 0.35733136534690857,
+        "learning_rate": 0.0005,
+        "loss": 1.1215,
+        "step": 12580
+      },
+      {
+        "epoch": 0.4014796390191014,
+        "grad_norm": 0.3459886312484741,
+        "learning_rate": 0.0005,
+        "loss": 1.1686,
+        "step": 12590
+      },
+      {
+        "epoch": 0.4017985267387353,
+        "grad_norm": 0.3461032509803772,
+        "learning_rate": 0.0005,
+        "loss": 1.1407,
+        "step": 12600
+      },
+      {
+        "epoch": 0.4021174144583692,
+        "grad_norm": 0.3554673492908478,
+        "learning_rate": 0.0005,
+        "loss": 1.1458,
+        "step": 12610
+      },
+      {
+        "epoch": 0.40243630217800314,
+        "grad_norm": 0.3475288450717926,
+        "learning_rate": 0.0005,
+        "loss": 1.1496,
+        "step": 12620
+      },
+      {
+        "epoch": 0.40275518989763703,
+        "grad_norm": 0.348635196685791,
+        "learning_rate": 0.0005,
+        "loss": 1.1268,
+        "step": 12630
+      },
+      {
+        "epoch": 0.40307407761727093,
+        "grad_norm": 0.3567737638950348,
+        "learning_rate": 0.0005,
+        "loss": 1.1291,
+        "step": 12640
+      },
+      {
+        "epoch": 0.4033929653369049,
+        "grad_norm": 0.3558914363384247,
+        "learning_rate": 0.0005,
+        "loss": 1.1519,
+        "step": 12650
+      },
+      {
+        "epoch": 0.4037118530565388,
+        "grad_norm": 0.35246145725250244,
+        "learning_rate": 0.0005,
+        "loss": 1.1425,
+        "step": 12660
+      },
+      {
+        "epoch": 0.40403074077617274,
+        "grad_norm": 0.3489568531513214,
+        "learning_rate": 0.0005,
+        "loss": 1.1396,
+        "step": 12670
+      },
+      {
+        "epoch": 0.40434962849580663,
+        "grad_norm": 0.361677348613739,
+        "learning_rate": 0.0005,
+        "loss": 1.1468,
+        "step": 12680
+      },
+      {
+        "epoch": 0.40466851621544053,
+        "grad_norm": 0.34218665957450867,
+        "learning_rate": 0.0005,
+        "loss": 1.1329,
+        "step": 12690
+      },
+      {
+        "epoch": 0.4049874039350745,
+        "grad_norm": 0.3400156795978546,
+        "learning_rate": 0.0005,
+        "loss": 1.1464,
+        "step": 12700
+      },
+      {
+        "epoch": 0.4053062916547084,
+        "grad_norm": 0.35337018966674805,
+        "learning_rate": 0.0005,
+        "loss": 1.131,
+        "step": 12710
+      },
+      {
+        "epoch": 0.4056251793743423,
+        "grad_norm": 0.3584569990634918,
+        "learning_rate": 0.0005,
+        "loss": 1.1508,
+        "step": 12720
+      },
+      {
+        "epoch": 0.40594406709397624,
+        "grad_norm": 0.3542274236679077,
+        "learning_rate": 0.0005,
+        "loss": 1.1348,
+        "step": 12730
+      },
+      {
+        "epoch": 0.40626295481361013,
+        "grad_norm": 0.3413817882537842,
+        "learning_rate": 0.0005,
+        "loss": 1.148,
+        "step": 12740
+      },
+      {
+        "epoch": 0.40658184253324403,
+        "grad_norm": 0.3511422872543335,
+        "learning_rate": 0.0005,
+        "loss": 1.1378,
+        "step": 12750
+      },
+      {
+        "epoch": 0.406900730252878,
+        "grad_norm": 0.3467022478580475,
+        "learning_rate": 0.0005,
+        "loss": 1.1405,
+        "step": 12760
+      },
+      {
+        "epoch": 0.4072196179725119,
+        "grad_norm": 0.35287347435951233,
+        "learning_rate": 0.0005,
+        "loss": 1.1562,
+        "step": 12770
+      },
+      {
+        "epoch": 0.4075385056921458,
+        "grad_norm": 0.35349148511886597,
+        "learning_rate": 0.0005,
+        "loss": 1.1361,
+        "step": 12780
+      },
+      {
+        "epoch": 0.40785739341177973,
+        "grad_norm": 0.35082828998565674,
+        "learning_rate": 0.0005,
+        "loss": 1.1388,
+        "step": 12790
+      },
+      {
+        "epoch": 0.40817628113141363,
+        "grad_norm": 0.34728944301605225,
+        "learning_rate": 0.0005,
+        "loss": 1.1577,
+        "step": 12800
+      },
+      {
+        "epoch": 0.40849516885104753,
+        "grad_norm": 0.35114023089408875,
+        "learning_rate": 0.0005,
+        "loss": 1.1352,
+        "step": 12810
+      },
+      {
+        "epoch": 0.4088140565706815,
+        "grad_norm": 0.34104374051094055,
+        "learning_rate": 0.0005,
+        "loss": 1.1343,
+        "step": 12820
+      },
+      {
+        "epoch": 0.4091329442903154,
+        "grad_norm": 0.33483564853668213,
+        "learning_rate": 0.0005,
+        "loss": 1.1331,
+        "step": 12830
+      },
+      {
+        "epoch": 0.4094518320099493,
+        "grad_norm": 0.3484364449977875,
+        "learning_rate": 0.0005,
+        "loss": 1.1079,
+        "step": 12840
+      },
+      {
+        "epoch": 0.40977071972958323,
+        "grad_norm": 0.3626136779785156,
+        "learning_rate": 0.0005,
+        "loss": 1.147,
+        "step": 12850
+      },
+      {
+        "epoch": 0.41008960744921713,
+        "grad_norm": 0.3401944935321808,
+        "learning_rate": 0.0005,
+        "loss": 1.1235,
+        "step": 12860
+      },
+      {
+        "epoch": 0.41040849516885103,
+        "grad_norm": 0.33367717266082764,
+        "learning_rate": 0.0005,
+        "loss": 1.1324,
+        "step": 12870
+      },
+      {
+        "epoch": 0.410727382888485,
+        "grad_norm": 0.3408673405647278,
+        "learning_rate": 0.0005,
+        "loss": 1.1421,
+        "step": 12880
+      },
+      {
+        "epoch": 0.4110462706081189,
+        "grad_norm": 0.37122371792793274,
+        "learning_rate": 0.0005,
+        "loss": 1.1486,
+        "step": 12890
+      },
+      {
+        "epoch": 0.4113651583277528,
+        "grad_norm": 0.3381376564502716,
+        "learning_rate": 0.0005,
+        "loss": 1.1311,
+        "step": 12900
+      },
+      {
+        "epoch": 0.41168404604738673,
+        "grad_norm": 0.3362289369106293,
+        "learning_rate": 0.0005,
+        "loss": 1.1246,
+        "step": 12910
+      },
+      {
+        "epoch": 0.41200293376702063,
+        "grad_norm": 0.355113685131073,
+        "learning_rate": 0.0005,
+        "loss": 1.1365,
+        "step": 12920
+      },
+      {
+        "epoch": 0.4123218214866545,
+        "grad_norm": 0.3402760624885559,
+        "learning_rate": 0.0005,
+        "loss": 1.1308,
+        "step": 12930
+      },
+      {
+        "epoch": 0.4126407092062885,
+        "grad_norm": 0.34513506293296814,
+        "learning_rate": 0.0005,
+        "loss": 1.1456,
+        "step": 12940
+      },
+      {
+        "epoch": 0.4129595969259224,
+        "grad_norm": 0.3438687026500702,
+        "learning_rate": 0.0005,
+        "loss": 1.1437,
+        "step": 12950
+      },
+      {
+        "epoch": 0.4132784846455563,
+        "grad_norm": 0.33485645055770874,
+        "learning_rate": 0.0005,
+        "loss": 1.1441,
+        "step": 12960
+      },
+      {
+        "epoch": 0.41359737236519023,
+        "grad_norm": 0.35648801922798157,
+        "learning_rate": 0.0005,
+        "loss": 1.1465,
+        "step": 12970
+      },
+      {
+        "epoch": 0.41391626008482413,
+        "grad_norm": 0.35223162174224854,
+        "learning_rate": 0.0005,
+        "loss": 1.1502,
+        "step": 12980
+      },
+      {
+        "epoch": 0.414235147804458,
+        "grad_norm": 0.3574160039424896,
+        "learning_rate": 0.0005,
+        "loss": 1.1337,
+        "step": 12990
+      },
+      {
+        "epoch": 0.414554035524092,
+        "grad_norm": 0.3410550057888031,
+        "learning_rate": 0.0005,
+        "loss": 1.1398,
+        "step": 13000
+      },
+      {
+        "epoch": 0.4148729232437259,
+        "grad_norm": 0.3610963523387909,
+        "learning_rate": 0.0005,
+        "loss": 1.1207,
+        "step": 13010
+      },
+      {
+        "epoch": 0.4151918109633598,
+        "grad_norm": 0.36963796615600586,
+        "learning_rate": 0.0005,
+        "loss": 1.1547,
+        "step": 13020
+      },
+      {
+        "epoch": 0.41551069868299373,
+        "grad_norm": 0.36384981870651245,
+        "learning_rate": 0.0005,
+        "loss": 1.1513,
+        "step": 13030
+      },
+      {
+        "epoch": 0.4158295864026276,
+        "grad_norm": 0.3315349817276001,
+        "learning_rate": 0.0005,
+        "loss": 1.1315,
+        "step": 13040
+      },
+      {
+        "epoch": 0.4161484741222615,
+        "grad_norm": 0.33881279826164246,
+        "learning_rate": 0.0005,
+        "loss": 1.1367,
+        "step": 13050
+      },
+      {
+        "epoch": 0.4164673618418955,
+        "grad_norm": 0.35474470257759094,
+        "learning_rate": 0.0005,
+        "loss": 1.1565,
+        "step": 13060
+      },
+      {
+        "epoch": 0.4167862495615294,
+        "grad_norm": 0.3389039635658264,
+        "learning_rate": 0.0005,
+        "loss": 1.1348,
+        "step": 13070
+      },
+      {
+        "epoch": 0.41710513728116333,
+        "grad_norm": 0.33887627720832825,
+        "learning_rate": 0.0005,
+        "loss": 1.1439,
+        "step": 13080
+      },
+      {
+        "epoch": 0.4174240250007972,
+        "grad_norm": 0.35486432909965515,
+        "learning_rate": 0.0005,
+        "loss": 1.1419,
+        "step": 13090
+      },
+      {
+        "epoch": 0.4177429127204311,
+        "grad_norm": 0.33916404843330383,
+        "learning_rate": 0.0005,
+        "loss": 1.1335,
+        "step": 13100
+      },
+      {
+        "epoch": 0.4180618004400651,
+        "grad_norm": 0.3448198437690735,
+        "learning_rate": 0.0005,
+        "loss": 1.1418,
+        "step": 13110
+      },
+      {
+        "epoch": 0.418380688159699,
+        "grad_norm": 0.364322692155838,
+        "learning_rate": 0.0005,
+        "loss": 1.1499,
+        "step": 13120
+      },
+      {
+        "epoch": 0.4186995758793329,
+        "grad_norm": 0.34698808193206787,
+        "learning_rate": 0.0005,
+        "loss": 1.1451,
+        "step": 13130
+      },
+      {
+        "epoch": 0.41901846359896683,
+        "grad_norm": 0.3550267517566681,
+        "learning_rate": 0.0005,
+        "loss": 1.1343,
+        "step": 13140
+      },
+      {
+        "epoch": 0.4193373513186007,
+        "grad_norm": 0.34121349453926086,
+        "learning_rate": 0.0005,
+        "loss": 1.1459,
+        "step": 13150
+      },
+      {
+        "epoch": 0.4196562390382346,
+        "grad_norm": 0.3349049687385559,
+        "learning_rate": 0.0005,
+        "loss": 1.1293,
+        "step": 13160
+      },
+      {
+        "epoch": 0.4199751267578686,
+        "grad_norm": 0.34862592816352844,
+        "learning_rate": 0.0005,
+        "loss": 1.137,
+        "step": 13170
+      },
+      {
+        "epoch": 0.4202940144775025,
+        "grad_norm": 0.3531326651573181,
+        "learning_rate": 0.0005,
+        "loss": 1.1218,
+        "step": 13180
+      },
+      {
+        "epoch": 0.4206129021971364,
+        "grad_norm": 0.34977883100509644,
+        "learning_rate": 0.0005,
+        "loss": 1.1375,
+        "step": 13190
+      },
+      {
+        "epoch": 0.4209317899167703,
+        "grad_norm": 0.35819658637046814,
+        "learning_rate": 0.0005,
+        "loss": 1.1542,
+        "step": 13200
+      },
+      {
+        "epoch": 0.4212506776364042,
+        "grad_norm": 0.3428193926811218,
+        "learning_rate": 0.0005,
+        "loss": 1.142,
+        "step": 13210
+      },
+      {
+        "epoch": 0.4215695653560381,
+        "grad_norm": 0.3502006530761719,
+        "learning_rate": 0.0005,
+        "loss": 1.1359,
+        "step": 13220
+      },
+      {
+        "epoch": 0.4218884530756721,
+        "grad_norm": 0.33796781301498413,
+        "learning_rate": 0.0005,
+        "loss": 1.144,
+        "step": 13230
+      },
+      {
+        "epoch": 0.422207340795306,
+        "grad_norm": 0.35478195548057556,
+        "learning_rate": 0.0005,
+        "loss": 1.1256,
+        "step": 13240
+      },
+      {
+        "epoch": 0.42252622851493987,
+        "grad_norm": 0.33419090509414673,
+        "learning_rate": 0.0005,
+        "loss": 1.1144,
+        "step": 13250
+      },
+      {
+        "epoch": 0.4228451162345738,
+        "grad_norm": 0.3418004810810089,
+        "learning_rate": 0.0005,
+        "loss": 1.1271,
+        "step": 13260
+      },
+      {
+        "epoch": 0.4231640039542077,
+        "grad_norm": 0.3498900532722473,
+        "learning_rate": 0.0005,
+        "loss": 1.1429,
+        "step": 13270
+      },
+      {
+        "epoch": 0.4234828916738416,
+        "grad_norm": 0.36065348982810974,
+        "learning_rate": 0.0005,
+        "loss": 1.1184,
+        "step": 13280
+      },
+      {
+        "epoch": 0.4238017793934756,
+        "grad_norm": 0.3300412595272064,
+        "learning_rate": 0.0005,
+        "loss": 1.1276,
+        "step": 13290
+      },
+      {
+        "epoch": 0.4241206671131095,
+        "grad_norm": 0.3515334725379944,
+        "learning_rate": 0.0005,
+        "loss": 1.151,
+        "step": 13300
+      },
+      {
+        "epoch": 0.42443955483274337,
+        "grad_norm": 0.33978763222694397,
+        "learning_rate": 0.0005,
+        "loss": 1.1498,
+        "step": 13310
+      },
+      {
+        "epoch": 0.4247584425523773,
+        "grad_norm": 0.32748571038246155,
+        "learning_rate": 0.0005,
+        "loss": 1.1273,
+        "step": 13320
+      },
+      {
+        "epoch": 0.4250773302720112,
+        "grad_norm": 0.34806540608406067,
+        "learning_rate": 0.0005,
+        "loss": 1.1401,
+        "step": 13330
+      },
+      {
+        "epoch": 0.4253962179916451,
+        "grad_norm": 0.3434291183948517,
+        "learning_rate": 0.0005,
+        "loss": 1.1291,
+        "step": 13340
+      },
+      {
+        "epoch": 0.4257151057112791,
+        "grad_norm": 0.3458356559276581,
+        "learning_rate": 0.0005,
+        "loss": 1.1373,
+        "step": 13350
+      },
+      {
+        "epoch": 0.42603399343091297,
+        "grad_norm": 0.3550995886325836,
+        "learning_rate": 0.0005,
+        "loss": 1.1265,
+        "step": 13360
+      },
+      {
+        "epoch": 0.42635288115054687,
+        "grad_norm": 0.3559800088405609,
+        "learning_rate": 0.0005,
+        "loss": 1.1417,
+        "step": 13370
+      },
+      {
+        "epoch": 0.4266717688701808,
+        "grad_norm": 0.3434266746044159,
+        "learning_rate": 0.0005,
+        "loss": 1.1189,
+        "step": 13380
+      },
+      {
+        "epoch": 0.4269906565898147,
+        "grad_norm": 0.35063812136650085,
+        "learning_rate": 0.0005,
+        "loss": 1.1267,
+        "step": 13390
+      },
+      {
+        "epoch": 0.4273095443094486,
+        "grad_norm": 0.34082141518592834,
+        "learning_rate": 0.0005,
+        "loss": 1.1221,
+        "step": 13400
+      },
+      {
+        "epoch": 0.4276284320290826,
+        "grad_norm": 0.3359578251838684,
+        "learning_rate": 0.0005,
+        "loss": 1.1136,
+        "step": 13410
+      },
+      {
+        "epoch": 0.42794731974871647,
+        "grad_norm": 0.3475954532623291,
+        "learning_rate": 0.0005,
+        "loss": 1.146,
+        "step": 13420
+      },
+      {
+        "epoch": 0.42826620746835037,
+        "grad_norm": 0.339847207069397,
+        "learning_rate": 0.0005,
+        "loss": 1.1373,
+        "step": 13430
+      },
+      {
+        "epoch": 0.4285850951879843,
+        "grad_norm": 0.3585573434829712,
+        "learning_rate": 0.0005,
+        "loss": 1.1432,
+        "step": 13440
+      },
+      {
+        "epoch": 0.4289039829076182,
+        "grad_norm": 0.3391568958759308,
+        "learning_rate": 0.0005,
+        "loss": 1.131,
+        "step": 13450
+      },
+      {
+        "epoch": 0.4292228706272521,
+        "grad_norm": 0.34588027000427246,
+        "learning_rate": 0.0005,
+        "loss": 1.1336,
+        "step": 13460
+      },
+      {
+        "epoch": 0.42954175834688607,
+        "grad_norm": 0.3400319516658783,
+        "learning_rate": 0.0005,
+        "loss": 1.1366,
+        "step": 13470
+      },
+      {
+        "epoch": 0.42986064606651997,
+        "grad_norm": 0.35178929567337036,
+        "learning_rate": 0.0005,
+        "loss": 1.1506,
+        "step": 13480
+      },
+      {
+        "epoch": 0.4301795337861539,
+        "grad_norm": 0.3204183876514435,
+        "learning_rate": 0.0005,
+        "loss": 1.1251,
+        "step": 13490
+      },
+      {
+        "epoch": 0.4304984215057878,
+        "grad_norm": 0.3399496376514435,
+        "learning_rate": 0.0005,
+        "loss": 1.1321,
+        "step": 13500
+      },
+      {
+        "epoch": 0.4308173092254217,
+        "grad_norm": 0.3532831072807312,
+        "learning_rate": 0.0005,
+        "loss": 1.1276,
+        "step": 13510
+      },
+      {
+        "epoch": 0.43113619694505567,
+        "grad_norm": 0.3372787833213806,
+        "learning_rate": 0.0005,
+        "loss": 1.1514,
+        "step": 13520
+      },
+      {
+        "epoch": 0.43145508466468957,
+        "grad_norm": 0.34379592537879944,
+        "learning_rate": 0.0005,
+        "loss": 1.1256,
+        "step": 13530
+      },
+      {
+        "epoch": 0.43177397238432347,
+        "grad_norm": 0.3416652977466583,
+        "learning_rate": 0.0005,
+        "loss": 1.1366,
+        "step": 13540
+      },
+      {
+        "epoch": 0.4320928601039574,
+        "grad_norm": 0.3551793098449707,
+        "learning_rate": 0.0005,
+        "loss": 1.1432,
+        "step": 13550
+      },
+      {
+        "epoch": 0.4324117478235913,
+        "grad_norm": 0.33919355273246765,
+        "learning_rate": 0.0005,
+        "loss": 1.1323,
+        "step": 13560
+      },
+      {
+        "epoch": 0.4327306355432252,
+        "grad_norm": 0.3380427360534668,
+        "learning_rate": 0.0005,
+        "loss": 1.1333,
+        "step": 13570
+      },
+      {
+        "epoch": 0.43304952326285917,
+        "grad_norm": 0.3556479215621948,
+        "learning_rate": 0.0005,
+        "loss": 1.1445,
+        "step": 13580
+      },
+      {
+        "epoch": 0.43336841098249307,
+        "grad_norm": 0.33824047446250916,
+        "learning_rate": 0.0005,
+        "loss": 1.1287,
+        "step": 13590
+      },
+      {
+        "epoch": 0.43368729870212697,
+        "grad_norm": 0.3541043996810913,
+        "learning_rate": 0.0005,
+        "loss": 1.1281,
+        "step": 13600
+      },
+      {
+        "epoch": 0.4340061864217609,
+        "grad_norm": 0.35882213711738586,
+        "learning_rate": 0.0005,
+        "loss": 1.1314,
+        "step": 13610
+      },
+      {
+        "epoch": 0.4343250741413948,
+        "grad_norm": 0.3262694478034973,
+        "learning_rate": 0.0005,
+        "loss": 1.1505,
+        "step": 13620
+      },
+      {
+        "epoch": 0.4346439618610287,
+        "grad_norm": 0.33700451254844666,
+        "learning_rate": 0.0005,
+        "loss": 1.1143,
+        "step": 13630
+      },
+      {
+        "epoch": 0.43496284958066267,
+        "grad_norm": 0.3422270119190216,
+        "learning_rate": 0.0005,
+        "loss": 1.1396,
+        "step": 13640
+      },
+      {
+        "epoch": 0.43528173730029657,
+        "grad_norm": 0.3300841450691223,
+        "learning_rate": 0.0005,
+        "loss": 1.1296,
+        "step": 13650
+      },
+      {
+        "epoch": 0.43560062501993047,
+        "grad_norm": 0.3336879014968872,
+        "learning_rate": 0.0005,
+        "loss": 1.1198,
+        "step": 13660
+      },
+      {
+        "epoch": 0.4359195127395644,
+        "grad_norm": 0.33946508169174194,
+        "learning_rate": 0.0005,
+        "loss": 1.1379,
+        "step": 13670
+      },
+      {
+        "epoch": 0.4362384004591983,
+        "grad_norm": 0.3411202132701874,
+        "learning_rate": 0.0005,
+        "loss": 1.1249,
+        "step": 13680
+      },
+      {
+        "epoch": 0.4365572881788322,
+        "grad_norm": 0.3332071304321289,
+        "learning_rate": 0.0005,
+        "loss": 1.1257,
+        "step": 13690
+      },
+      {
+        "epoch": 0.43687617589846617,
+        "grad_norm": 0.35336050391197205,
+        "learning_rate": 0.0005,
+        "loss": 1.1404,
+        "step": 13700
+      },
+      {
+        "epoch": 0.43719506361810007,
+        "grad_norm": 0.34526899456977844,
+        "learning_rate": 0.0005,
+        "loss": 1.126,
+        "step": 13710
+      },
+      {
+        "epoch": 0.43751395133773396,
+        "grad_norm": 0.34036850929260254,
+        "learning_rate": 0.0005,
+        "loss": 1.1086,
+        "step": 13720
+      },
+      {
+        "epoch": 0.4378328390573679,
+        "grad_norm": 0.34457531571388245,
+        "learning_rate": 0.0005,
+        "loss": 1.1478,
+        "step": 13730
+      },
+      {
+        "epoch": 0.4381517267770018,
+        "grad_norm": 0.34224215149879456,
+        "learning_rate": 0.0005,
+        "loss": 1.1355,
+        "step": 13740
+      },
+      {
+        "epoch": 0.4384706144966357,
+        "grad_norm": 0.34279027581214905,
+        "learning_rate": 0.0005,
+        "loss": 1.1249,
+        "step": 13750
+      },
+      {
+        "epoch": 0.43878950221626967,
+        "grad_norm": 0.3422126770019531,
+        "learning_rate": 0.0005,
+        "loss": 1.1215,
+        "step": 13760
+      },
+      {
+        "epoch": 0.43910838993590356,
+        "grad_norm": 0.3363499641418457,
+        "learning_rate": 0.0005,
+        "loss": 1.1344,
+        "step": 13770
+      },
+      {
+        "epoch": 0.43942727765553746,
+        "grad_norm": 0.3555680811405182,
+        "learning_rate": 0.0005,
+        "loss": 1.1377,
+        "step": 13780
+      },
+      {
+        "epoch": 0.4397461653751714,
+        "grad_norm": 0.3454040586948395,
+        "learning_rate": 0.0005,
+        "loss": 1.1365,
+        "step": 13790
+      },
+      {
+        "epoch": 0.4400650530948053,
+        "grad_norm": 0.34297218918800354,
+        "learning_rate": 0.0005,
+        "loss": 1.1291,
+        "step": 13800
+      },
+      {
+        "epoch": 0.4403839408144392,
+        "grad_norm": 0.34979772567749023,
+        "learning_rate": 0.0005,
+        "loss": 1.1187,
+        "step": 13810
+      },
+      {
+        "epoch": 0.44070282853407317,
+        "grad_norm": 0.3685065805912018,
+        "learning_rate": 0.0005,
+        "loss": 1.1194,
+        "step": 13820
+      },
+      {
+        "epoch": 0.44102171625370706,
+        "grad_norm": 0.33668696880340576,
+        "learning_rate": 0.0005,
+        "loss": 1.1245,
+        "step": 13830
+      },
+      {
+        "epoch": 0.44134060397334096,
+        "grad_norm": 0.3478093147277832,
+        "learning_rate": 0.0005,
+        "loss": 1.1124,
+        "step": 13840
+      },
+      {
+        "epoch": 0.4416594916929749,
+        "grad_norm": 0.3335479199886322,
+        "learning_rate": 0.0005,
+        "loss": 1.1241,
+        "step": 13850
+      },
+      {
+        "epoch": 0.4419783794126088,
+        "grad_norm": 0.3367275893688202,
+        "learning_rate": 0.0005,
+        "loss": 1.1258,
+        "step": 13860
+      },
+      {
+        "epoch": 0.4422972671322427,
+        "grad_norm": 0.3387754261493683,
+        "learning_rate": 0.0005,
+        "loss": 1.1147,
+        "step": 13870
+      },
+      {
+        "epoch": 0.44261615485187666,
+        "grad_norm": 0.33694881200790405,
+        "learning_rate": 0.0005,
+        "loss": 1.1038,
+        "step": 13880
+      },
+      {
+        "epoch": 0.44293504257151056,
+        "grad_norm": 0.326871782541275,
+        "learning_rate": 0.0005,
+        "loss": 1.1115,
+        "step": 13890
+      },
+      {
+        "epoch": 0.4432539302911445,
+        "grad_norm": 0.3470039367675781,
+        "learning_rate": 0.0005,
+        "loss": 1.1341,
+        "step": 13900
+      },
+      {
+        "epoch": 0.4435728180107784,
+        "grad_norm": 0.32792967557907104,
+        "learning_rate": 0.0005,
+        "loss": 1.1149,
+        "step": 13910
+      },
+      {
+        "epoch": 0.4438917057304123,
+        "grad_norm": 0.3368305265903473,
+        "learning_rate": 0.0005,
+        "loss": 1.1328,
+        "step": 13920
+      },
+      {
+        "epoch": 0.44421059345004626,
+        "grad_norm": 0.3338005840778351,
+        "learning_rate": 0.0005,
+        "loss": 1.1386,
+        "step": 13930
+      },
+      {
+        "epoch": 0.44452948116968016,
+        "grad_norm": 0.33262819051742554,
+        "learning_rate": 0.0005,
+        "loss": 1.1092,
+        "step": 13940
+      },
+      {
+        "epoch": 0.44484836888931406,
+        "grad_norm": 0.34088337421417236,
+        "learning_rate": 0.0005,
+        "loss": 1.1255,
+        "step": 13950
+      },
+      {
+        "epoch": 0.445167256608948,
+        "grad_norm": 0.33699917793273926,
+        "learning_rate": 0.0005,
+        "loss": 1.1453,
+        "step": 13960
+      },
+      {
+        "epoch": 0.4454861443285819,
+        "grad_norm": 0.3505221903324127,
+        "learning_rate": 0.0005,
+        "loss": 1.1192,
+        "step": 13970
+      },
+      {
+        "epoch": 0.4458050320482158,
+        "grad_norm": 0.3351021707057953,
+        "learning_rate": 0.0005,
+        "loss": 1.1242,
+        "step": 13980
+      },
+      {
+        "epoch": 0.44612391976784976,
+        "grad_norm": 0.3325427770614624,
+        "learning_rate": 0.0005,
+        "loss": 1.121,
+        "step": 13990
+      },
+      {
+        "epoch": 0.44644280748748366,
+        "grad_norm": 0.34713199734687805,
+        "learning_rate": 0.0005,
+        "loss": 1.1168,
+        "step": 14000
+      },
+      {
+        "epoch": 0.44676169520711756,
+        "grad_norm": 0.3391982614994049,
+        "learning_rate": 0.0005,
+        "loss": 1.1378,
+        "step": 14010
+      },
+      {
+        "epoch": 0.4470805829267515,
+        "grad_norm": 0.3458082675933838,
+        "learning_rate": 0.0005,
+        "loss": 1.1308,
+        "step": 14020
+      },
+      {
+        "epoch": 0.4473994706463854,
+        "grad_norm": 0.3400178849697113,
+        "learning_rate": 0.0005,
+        "loss": 1.1312,
+        "step": 14030
+      },
+      {
+        "epoch": 0.4477183583660193,
+        "grad_norm": 0.33684873580932617,
+        "learning_rate": 0.0005,
+        "loss": 1.116,
+        "step": 14040
+      },
+      {
+        "epoch": 0.44803724608565326,
+        "grad_norm": 0.3401698172092438,
+        "learning_rate": 0.0005,
+        "loss": 1.1092,
+        "step": 14050
+      },
+      {
+        "epoch": 0.44835613380528716,
+        "grad_norm": 0.34139543771743774,
+        "learning_rate": 0.0005,
+        "loss": 1.1152,
+        "step": 14060
+      },
+      {
+        "epoch": 0.44867502152492106,
+        "grad_norm": 0.3483545780181885,
+        "learning_rate": 0.0005,
+        "loss": 1.1096,
+        "step": 14070
+      },
+      {
+        "epoch": 0.448993909244555,
+        "grad_norm": 0.3576740026473999,
+        "learning_rate": 0.0005,
+        "loss": 1.1229,
+        "step": 14080
+      },
+      {
+        "epoch": 0.4493127969641889,
+        "grad_norm": 0.3445344865322113,
+        "learning_rate": 0.0005,
+        "loss": 1.1173,
+        "step": 14090
+      },
+      {
+        "epoch": 0.4496316846838228,
+        "grad_norm": 0.32825151085853577,
+        "learning_rate": 0.0005,
+        "loss": 1.1222,
+        "step": 14100
+      },
+      {
+        "epoch": 0.44995057240345676,
+        "grad_norm": 0.3463006317615509,
+        "learning_rate": 0.0005,
+        "loss": 1.1239,
+        "step": 14110
+      },
+      {
+        "epoch": 0.45026946012309066,
+        "grad_norm": 0.3307591676712036,
+        "learning_rate": 0.0005,
+        "loss": 1.112,
+        "step": 14120
+      },
+      {
+        "epoch": 0.45058834784272456,
+        "grad_norm": 0.33084917068481445,
+        "learning_rate": 0.0005,
+        "loss": 1.1167,
+        "step": 14130
+      },
+      {
+        "epoch": 0.4509072355623585,
+        "grad_norm": 0.3341514468193054,
+        "learning_rate": 0.0005,
+        "loss": 1.1146,
+        "step": 14140
+      },
+      {
+        "epoch": 0.4512261232819924,
+        "grad_norm": 0.3277684450149536,
+        "learning_rate": 0.0005,
+        "loss": 1.144,
+        "step": 14150
+      },
+      {
+        "epoch": 0.4515450110016263,
+        "grad_norm": 0.34516265988349915,
+        "learning_rate": 0.0005,
+        "loss": 1.1197,
+        "step": 14160
+      },
+      {
+        "epoch": 0.45186389872126026,
+        "grad_norm": 0.32295408844947815,
+        "learning_rate": 0.0005,
+        "loss": 1.1117,
+        "step": 14170
+      },
+      {
+        "epoch": 0.45218278644089416,
+        "grad_norm": 0.3291379511356354,
+        "learning_rate": 0.0005,
+        "loss": 1.1206,
+        "step": 14180
+      },
+      {
+        "epoch": 0.45250167416052806,
+        "grad_norm": 0.3296593129634857,
+        "learning_rate": 0.0005,
+        "loss": 1.1071,
+        "step": 14190
+      },
+      {
+        "epoch": 0.452820561880162,
+        "grad_norm": 0.3422677516937256,
+        "learning_rate": 0.0005,
+        "loss": 1.1355,
+        "step": 14200
+      },
+      {
+        "epoch": 0.4531394495997959,
+        "grad_norm": 0.3407352566719055,
+        "learning_rate": 0.0005,
+        "loss": 1.1322,
+        "step": 14210
+      },
+      {
+        "epoch": 0.4534583373194298,
+        "grad_norm": 0.3275332748889923,
+        "learning_rate": 0.0005,
+        "loss": 1.1203,
+        "step": 14220
+      },
+      {
+        "epoch": 0.45377722503906376,
+        "grad_norm": 0.33814772963523865,
+        "learning_rate": 0.0005,
+        "loss": 1.1084,
+        "step": 14230
+      },
+      {
+        "epoch": 0.45409611275869766,
+        "grad_norm": 0.3463352918624878,
+        "learning_rate": 0.0005,
+        "loss": 1.1066,
+        "step": 14240
+      },
+      {
+        "epoch": 0.45441500047833155,
+        "grad_norm": 0.3440384864807129,
+        "learning_rate": 0.0005,
+        "loss": 1.1264,
+        "step": 14250
+      },
+      {
+        "epoch": 0.4547338881979655,
+        "grad_norm": 0.3311120271682739,
+        "learning_rate": 0.0005,
+        "loss": 1.1377,
+        "step": 14260
+      },
+      {
+        "epoch": 0.4550527759175994,
+        "grad_norm": 0.3387615978717804,
+        "learning_rate": 0.0005,
+        "loss": 1.1166,
+        "step": 14270
+      },
+      {
+        "epoch": 0.4553716636372333,
+        "grad_norm": 0.33410459756851196,
+        "learning_rate": 0.0005,
+        "loss": 1.1278,
+        "step": 14280
+      },
+      {
+        "epoch": 0.45569055135686726,
+        "grad_norm": 0.34034672379493713,
+        "learning_rate": 0.0005,
+        "loss": 1.134,
+        "step": 14290
+      },
+      {
+        "epoch": 0.45600943907650116,
+        "grad_norm": 0.33677056431770325,
+        "learning_rate": 0.0005,
+        "loss": 1.1329,
+        "step": 14300
+      },
+      {
+        "epoch": 0.45632832679613505,
+        "grad_norm": 0.3402319550514221,
+        "learning_rate": 0.0005,
+        "loss": 1.1325,
+        "step": 14310
+      },
+      {
+        "epoch": 0.456647214515769,
+        "grad_norm": 0.3517909646034241,
+        "learning_rate": 0.0005,
+        "loss": 1.1216,
+        "step": 14320
+      },
+      {
+        "epoch": 0.4569661022354029,
+        "grad_norm": 0.3427649438381195,
+        "learning_rate": 0.0005,
+        "loss": 1.1195,
+        "step": 14330
+      },
+      {
+        "epoch": 0.45728498995503686,
+        "grad_norm": 0.339320570230484,
+        "learning_rate": 0.0005,
+        "loss": 1.1099,
+        "step": 14340
+      },
+      {
+        "epoch": 0.45760387767467076,
+        "grad_norm": 0.3295920193195343,
+        "learning_rate": 0.0005,
+        "loss": 1.12,
+        "step": 14350
+      },
+      {
+        "epoch": 0.45792276539430465,
+        "grad_norm": 0.331986665725708,
+        "learning_rate": 0.0005,
+        "loss": 1.1168,
+        "step": 14360
+      },
+      {
+        "epoch": 0.4582416531139386,
+        "grad_norm": 0.3381638526916504,
+        "learning_rate": 0.0005,
+        "loss": 1.1441,
+        "step": 14370
+      },
+      {
+        "epoch": 0.4585605408335725,
+        "grad_norm": 0.3370720446109772,
+        "learning_rate": 0.0005,
+        "loss": 1.1319,
+        "step": 14380
+      },
+      {
+        "epoch": 0.4588794285532064,
+        "grad_norm": 0.3332827091217041,
+        "learning_rate": 0.0005,
+        "loss": 1.1128,
+        "step": 14390
+      },
+      {
+        "epoch": 0.45919831627284036,
+        "grad_norm": 0.33136168122291565,
+        "learning_rate": 0.0005,
+        "loss": 1.138,
+        "step": 14400
+      },
+      {
+        "epoch": 0.45951720399247425,
+        "grad_norm": 0.32325100898742676,
+        "learning_rate": 0.0005,
+        "loss": 1.0948,
+        "step": 14410
+      },
+      {
+        "epoch": 0.45983609171210815,
+        "grad_norm": 0.3420909643173218,
+        "learning_rate": 0.0005,
+        "loss": 1.1158,
+        "step": 14420
+      },
+      {
+        "epoch": 0.4601549794317421,
+        "grad_norm": 0.35374853014945984,
+        "learning_rate": 0.0005,
+        "loss": 1.1158,
+        "step": 14430
+      },
+      {
+        "epoch": 0.460473867151376,
+        "grad_norm": 0.338075190782547,
+        "learning_rate": 0.0005,
+        "loss": 1.1328,
+        "step": 14440
+      },
+      {
+        "epoch": 0.4607927548710099,
+        "grad_norm": 0.3302494287490845,
+        "learning_rate": 0.0005,
+        "loss": 1.1088,
+        "step": 14450
+      },
+      {
+        "epoch": 0.46111164259064386,
+        "grad_norm": 0.33893561363220215,
+        "learning_rate": 0.0005,
+        "loss": 1.1255,
+        "step": 14460
+      },
+      {
+        "epoch": 0.46143053031027775,
+        "grad_norm": 0.3488835394382477,
+        "learning_rate": 0.0005,
+        "loss": 1.114,
+        "step": 14470
+      },
+      {
+        "epoch": 0.46174941802991165,
+        "grad_norm": 0.3383651077747345,
+        "learning_rate": 0.0005,
+        "loss": 1.1055,
+        "step": 14480
+      },
+      {
+        "epoch": 0.4620683057495456,
+        "grad_norm": 0.3396001160144806,
+        "learning_rate": 0.0005,
+        "loss": 1.1138,
+        "step": 14490
+      },
+      {
+        "epoch": 0.4623871934691795,
+        "grad_norm": 0.3394538164138794,
+        "learning_rate": 0.0005,
+        "loss": 1.1194,
+        "step": 14500
+      },
+      {
+        "epoch": 0.4627060811888134,
+        "grad_norm": 0.34485116600990295,
+        "learning_rate": 0.0005,
+        "loss": 1.1136,
+        "step": 14510
+      },
+      {
+        "epoch": 0.46302496890844735,
+        "grad_norm": 0.3296126425266266,
+        "learning_rate": 0.0005,
+        "loss": 1.1114,
+        "step": 14520
+      },
+      {
+        "epoch": 0.46334385662808125,
+        "grad_norm": 0.34639716148376465,
+        "learning_rate": 0.0005,
+        "loss": 1.138,
+        "step": 14530
+      },
+      {
+        "epoch": 0.46366274434771515,
+        "grad_norm": 0.32055237889289856,
+        "learning_rate": 0.0005,
+        "loss": 1.1053,
+        "step": 14540
+      },
+      {
+        "epoch": 0.4639816320673491,
+        "grad_norm": 0.3470534682273865,
+        "learning_rate": 0.0005,
+        "loss": 1.1337,
+        "step": 14550
+      },
+      {
+        "epoch": 0.464300519786983,
+        "grad_norm": 0.3348836898803711,
+        "learning_rate": 0.0005,
+        "loss": 1.1207,
+        "step": 14560
+      },
+      {
+        "epoch": 0.4646194075066169,
+        "grad_norm": 0.34529122710227966,
+        "learning_rate": 0.0005,
+        "loss": 1.1196,
+        "step": 14570
+      },
+      {
+        "epoch": 0.46493829522625085,
+        "grad_norm": 0.34491243958473206,
+        "learning_rate": 0.0005,
+        "loss": 1.1148,
+        "step": 14580
+      },
+      {
+        "epoch": 0.46525718294588475,
+        "grad_norm": 0.32686635851860046,
+        "learning_rate": 0.0005,
+        "loss": 1.1122,
+        "step": 14590
+      },
+      {
+        "epoch": 0.46557607066551865,
+        "grad_norm": 0.33966755867004395,
+        "learning_rate": 0.0005,
+        "loss": 1.1222,
+        "step": 14600
+      },
+      {
+        "epoch": 0.4658949583851526,
+        "grad_norm": 0.3253386914730072,
+        "learning_rate": 0.0005,
+        "loss": 1.1237,
+        "step": 14610
+      },
+      {
+        "epoch": 0.4662138461047865,
+        "grad_norm": 0.3355180621147156,
+        "learning_rate": 0.0005,
+        "loss": 1.1192,
+        "step": 14620
+      },
+      {
+        "epoch": 0.4665327338244204,
+        "grad_norm": 0.3447900414466858,
+        "learning_rate": 0.0005,
+        "loss": 1.1223,
+        "step": 14630
+      },
+      {
+        "epoch": 0.46685162154405435,
+        "grad_norm": 0.3398261070251465,
+        "learning_rate": 0.0005,
+        "loss": 1.0987,
+        "step": 14640
+      },
+      {
+        "epoch": 0.46717050926368825,
+        "grad_norm": 0.32602953910827637,
+        "learning_rate": 0.0005,
+        "loss": 1.1209,
+        "step": 14650
+      },
+      {
+        "epoch": 0.46748939698332215,
+        "grad_norm": 0.3350808322429657,
+        "learning_rate": 0.0005,
+        "loss": 1.1139,
+        "step": 14660
+      },
+      {
+        "epoch": 0.4678082847029561,
+        "grad_norm": 0.3368006944656372,
+        "learning_rate": 0.0005,
+        "loss": 1.1172,
+        "step": 14670
+      },
+      {
+        "epoch": 0.46812717242259,
+        "grad_norm": 0.33167609572410583,
+        "learning_rate": 0.0005,
+        "loss": 1.1054,
+        "step": 14680
+      },
+      {
+        "epoch": 0.4684460601422239,
+        "grad_norm": 0.350157767534256,
+        "learning_rate": 0.0005,
+        "loss": 1.1092,
+        "step": 14690
+      },
+      {
+        "epoch": 0.46876494786185785,
+        "grad_norm": 0.339697003364563,
+        "learning_rate": 0.0005,
+        "loss": 1.1128,
+        "step": 14700
+      },
+      {
+        "epoch": 0.46908383558149175,
+        "grad_norm": 0.3364179730415344,
+        "learning_rate": 0.0005,
+        "loss": 1.1288,
+        "step": 14710
+      },
+      {
+        "epoch": 0.46940272330112565,
+        "grad_norm": 0.33690929412841797,
+        "learning_rate": 0.0005,
+        "loss": 1.1194,
+        "step": 14720
+      },
+      {
+        "epoch": 0.4697216110207596,
+        "grad_norm": 0.334957480430603,
+        "learning_rate": 0.0005,
+        "loss": 1.1152,
+        "step": 14730
+      },
+      {
+        "epoch": 0.4700404987403935,
+        "grad_norm": 0.343614786863327,
+        "learning_rate": 0.0005,
+        "loss": 1.129,
+        "step": 14740
+      },
+      {
+        "epoch": 0.47035938646002745,
+        "grad_norm": 0.3393288254737854,
+        "learning_rate": 0.0005,
+        "loss": 1.1313,
+        "step": 14750
+      },
+      {
+        "epoch": 0.47067827417966135,
+        "grad_norm": 0.3249128460884094,
+        "learning_rate": 0.0005,
+        "loss": 1.1023,
+        "step": 14760
+      },
+      {
+        "epoch": 0.47099716189929525,
+        "grad_norm": 0.3428761065006256,
+        "learning_rate": 0.0005,
+        "loss": 1.0921,
+        "step": 14770
+      },
+      {
+        "epoch": 0.4713160496189292,
+        "grad_norm": 0.3276873826980591,
+        "learning_rate": 0.0005,
+        "loss": 1.1204,
+        "step": 14780
+      },
+      {
+        "epoch": 0.4716349373385631,
+        "grad_norm": 0.3779163062572479,
+        "learning_rate": 0.0005,
+        "loss": 1.1314,
+        "step": 14790
+      },
+      {
+        "epoch": 0.471953825058197,
+        "grad_norm": 0.334412157535553,
+        "learning_rate": 0.0005,
+        "loss": 1.1271,
+        "step": 14800
+      },
+      {
+        "epoch": 0.47227271277783095,
+        "grad_norm": 0.3291500508785248,
+        "learning_rate": 0.0005,
+        "loss": 1.115,
+        "step": 14810
+      },
+      {
+        "epoch": 0.47259160049746485,
+        "grad_norm": 0.34810540080070496,
+        "learning_rate": 0.0005,
+        "loss": 1.1239,
+        "step": 14820
+      },
+      {
+        "epoch": 0.47291048821709875,
+        "grad_norm": 0.3348764479160309,
+        "learning_rate": 0.0005,
+        "loss": 1.1267,
+        "step": 14830
+      },
+      {
+        "epoch": 0.4732293759367327,
+        "grad_norm": 0.3247257173061371,
+        "learning_rate": 0.0005,
+        "loss": 1.1244,
+        "step": 14840
+      },
+      {
+        "epoch": 0.4735482636563666,
+        "grad_norm": 0.32146748900413513,
+        "learning_rate": 0.0005,
+        "loss": 1.1196,
+        "step": 14850
+      },
+      {
+        "epoch": 0.4738671513760005,
+        "grad_norm": 0.32966938614845276,
+        "learning_rate": 0.0005,
+        "loss": 1.1019,
+        "step": 14860
+      },
+      {
+        "epoch": 0.47418603909563445,
+        "grad_norm": 0.3306904435157776,
+        "learning_rate": 0.0005,
+        "loss": 1.1064,
+        "step": 14870
+      },
+      {
+        "epoch": 0.47450492681526835,
+        "grad_norm": 0.3338717222213745,
+        "learning_rate": 0.0005,
+        "loss": 1.1099,
+        "step": 14880
+      },
+      {
+        "epoch": 0.47482381453490224,
+        "grad_norm": 0.3372955322265625,
+        "learning_rate": 0.0005,
+        "loss": 1.117,
+        "step": 14890
+      },
+      {
+        "epoch": 0.4751427022545362,
+        "grad_norm": 0.34013262391090393,
+        "learning_rate": 0.0005,
+        "loss": 1.119,
+        "step": 14900
+      },
+      {
+        "epoch": 0.4754615899741701,
+        "grad_norm": 0.33529746532440186,
+        "learning_rate": 0.0005,
+        "loss": 1.1284,
+        "step": 14910
+      },
+      {
+        "epoch": 0.475780477693804,
+        "grad_norm": 0.3290479779243469,
+        "learning_rate": 0.0005,
+        "loss": 1.1167,
+        "step": 14920
+      },
+      {
+        "epoch": 0.47609936541343795,
+        "grad_norm": 0.333850234746933,
+        "learning_rate": 0.0005,
+        "loss": 1.124,
+        "step": 14930
+      },
+      {
+        "epoch": 0.47641825313307185,
+        "grad_norm": 0.32731056213378906,
+        "learning_rate": 0.0005,
+        "loss": 1.1333,
+        "step": 14940
+      },
+      {
+        "epoch": 0.47673714085270574,
+        "grad_norm": 0.3414525091648102,
+        "learning_rate": 0.0005,
+        "loss": 1.1063,
+        "step": 14950
+      },
+      {
+        "epoch": 0.4770560285723397,
+        "grad_norm": 0.3164287507534027,
+        "learning_rate": 0.0005,
+        "loss": 1.1181,
+        "step": 14960
+      },
+      {
+        "epoch": 0.4773749162919736,
+        "grad_norm": 0.33766913414001465,
+        "learning_rate": 0.0005,
+        "loss": 1.1238,
+        "step": 14970
+      },
+      {
+        "epoch": 0.4776938040116075,
+        "grad_norm": 0.33459576964378357,
+        "learning_rate": 0.0005,
+        "loss": 1.1084,
+        "step": 14980
+      },
+      {
+        "epoch": 0.47801269173124145,
+        "grad_norm": 0.3171371519565582,
+        "learning_rate": 0.0005,
+        "loss": 1.0986,
+        "step": 14990
+      },
+      {
+        "epoch": 0.47833157945087534,
+        "grad_norm": 0.3457358479499817,
+        "learning_rate": 0.0005,
+        "loss": 1.1307,
+        "step": 15000
+      },
+      {
+        "epoch": 0.47865046717050924,
+        "grad_norm": 0.329092800617218,
+        "learning_rate": 0.0005,
+        "loss": 1.1265,
+        "step": 15010
+      },
+      {
+        "epoch": 0.4789693548901432,
+        "grad_norm": 0.34381380677223206,
+        "learning_rate": 0.0005,
+        "loss": 1.1096,
+        "step": 15020
+      },
+      {
+        "epoch": 0.4792882426097771,
+        "grad_norm": 0.3248065710067749,
+        "learning_rate": 0.0005,
+        "loss": 1.1278,
+        "step": 15030
+      },
+      {
+        "epoch": 0.479607130329411,
+        "grad_norm": 0.3197561800479889,
+        "learning_rate": 0.0005,
+        "loss": 1.1023,
+        "step": 15040
+      },
+      {
+        "epoch": 0.47992601804904494,
+        "grad_norm": 0.3313983082771301,
+        "learning_rate": 0.0005,
+        "loss": 1.1176,
+        "step": 15050
+      },
+      {
+        "epoch": 0.48024490576867884,
+        "grad_norm": 0.3262551724910736,
+        "learning_rate": 0.0005,
+        "loss": 1.109,
+        "step": 15060
+      },
+      {
+        "epoch": 0.48056379348831274,
+        "grad_norm": 0.3366710841655731,
+        "learning_rate": 0.0005,
+        "loss": 1.1223,
+        "step": 15070
+      },
+      {
+        "epoch": 0.4808826812079467,
+        "grad_norm": 0.32736074924468994,
+        "learning_rate": 0.0005,
+        "loss": 1.1187,
+        "step": 15080
+      },
+      {
+        "epoch": 0.4812015689275806,
+        "grad_norm": 0.33949264883995056,
+        "learning_rate": 0.0005,
+        "loss": 1.1115,
+        "step": 15090
+      },
+      {
+        "epoch": 0.4815204566472145,
+        "grad_norm": 0.3322353661060333,
+        "learning_rate": 0.0005,
+        "loss": 1.1256,
+        "step": 15100
+      },
+      {
+        "epoch": 0.48183934436684844,
+        "grad_norm": 0.3546300530433655,
+        "learning_rate": 0.0005,
+        "loss": 1.099,
+        "step": 15110
+      },
+      {
+        "epoch": 0.48215823208648234,
+        "grad_norm": 0.31901800632476807,
+        "learning_rate": 0.0005,
+        "loss": 1.0945,
+        "step": 15120
+      },
+      {
+        "epoch": 0.48247711980611624,
+        "grad_norm": 0.3382745087146759,
+        "learning_rate": 0.0005,
+        "loss": 1.1004,
+        "step": 15130
+      },
+      {
+        "epoch": 0.4827960075257502,
+        "grad_norm": 0.3378961980342865,
+        "learning_rate": 0.0005,
+        "loss": 1.1076,
+        "step": 15140
+      },
+      {
+        "epoch": 0.4831148952453841,
+        "grad_norm": 0.3255893290042877,
+        "learning_rate": 0.0005,
+        "loss": 1.0986,
+        "step": 15150
+      },
+      {
+        "epoch": 0.48343378296501804,
+        "grad_norm": 0.324762225151062,
+        "learning_rate": 0.0005,
+        "loss": 1.1073,
+        "step": 15160
+      },
+      {
+        "epoch": 0.48375267068465194,
+        "grad_norm": 0.3625001907348633,
+        "learning_rate": 0.0005,
+        "loss": 1.1125,
+        "step": 15170
+      },
+      {
+        "epoch": 0.48407155840428584,
+        "grad_norm": 0.35341137647628784,
+        "learning_rate": 0.0005,
+        "loss": 1.1037,
+        "step": 15180
+      },
+      {
+        "epoch": 0.4843904461239198,
+        "grad_norm": 0.3476621210575104,
+        "learning_rate": 0.0005,
+        "loss": 1.0979,
+        "step": 15190
+      },
+      {
+        "epoch": 0.4847093338435537,
+        "grad_norm": 0.339345246553421,
+        "learning_rate": 0.0005,
+        "loss": 1.1371,
+        "step": 15200
+      },
+      {
+        "epoch": 0.4850282215631876,
+        "grad_norm": 0.3191971182823181,
+        "learning_rate": 0.0005,
+        "loss": 1.0994,
+        "step": 15210
+      },
+      {
+        "epoch": 0.48534710928282154,
+        "grad_norm": 0.3185611665248871,
+        "learning_rate": 0.0005,
+        "loss": 1.111,
+        "step": 15220
+      },
+      {
+        "epoch": 0.48566599700245544,
+        "grad_norm": 0.33163318037986755,
+        "learning_rate": 0.0005,
+        "loss": 1.1218,
+        "step": 15230
+      },
+      {
+        "epoch": 0.48598488472208934,
+        "grad_norm": 0.3259880840778351,
+        "learning_rate": 0.0005,
+        "loss": 1.121,
+        "step": 15240
+      },
+      {
+        "epoch": 0.4863037724417233,
+        "grad_norm": 0.32738441228866577,
+        "learning_rate": 0.0005,
+        "loss": 1.1032,
+        "step": 15250
+      },
+      {
+        "epoch": 0.4866226601613572,
+        "grad_norm": 0.3310328423976898,
+        "learning_rate": 0.0005,
+        "loss": 1.1088,
+        "step": 15260
+      },
+      {
+        "epoch": 0.4869415478809911,
+        "grad_norm": 0.32167747616767883,
+        "learning_rate": 0.0005,
+        "loss": 1.106,
+        "step": 15270
+      },
+      {
+        "epoch": 0.48726043560062504,
+        "grad_norm": 0.32243403792381287,
+        "learning_rate": 0.0005,
+        "loss": 1.09,
+        "step": 15280
+      },
+      {
+        "epoch": 0.48757932332025894,
+        "grad_norm": 0.3249606788158417,
+        "learning_rate": 0.0005,
+        "loss": 1.1143,
+        "step": 15290
+      },
+      {
+        "epoch": 0.48789821103989284,
+        "grad_norm": 0.32354891300201416,
+        "learning_rate": 0.0005,
+        "loss": 1.1013,
+        "step": 15300
+      },
+      {
+        "epoch": 0.4882170987595268,
+        "grad_norm": 0.3281603157520294,
+        "learning_rate": 0.0005,
+        "loss": 1.0913,
+        "step": 15310
+      },
+      {
+        "epoch": 0.4885359864791607,
+        "grad_norm": 0.3279193639755249,
+        "learning_rate": 0.0005,
+        "loss": 1.1088,
+        "step": 15320
+      },
+      {
+        "epoch": 0.4888548741987946,
+        "grad_norm": 0.34850239753723145,
+        "learning_rate": 0.0005,
+        "loss": 1.1121,
+        "step": 15330
+      },
+      {
+        "epoch": 0.48917376191842854,
+        "grad_norm": 0.34309151768684387,
+        "learning_rate": 0.0005,
+        "loss": 1.0991,
+        "step": 15340
+      },
+      {
+        "epoch": 0.48949264963806244,
+        "grad_norm": 0.3212507665157318,
+        "learning_rate": 0.0005,
+        "loss": 1.1025,
+        "step": 15350
+      },
+      {
+        "epoch": 0.48981153735769634,
+        "grad_norm": 0.329383909702301,
+        "learning_rate": 0.0005,
+        "loss": 1.0944,
+        "step": 15360
+      },
+      {
+        "epoch": 0.4901304250773303,
+        "grad_norm": 0.32786908745765686,
+        "learning_rate": 0.0005,
+        "loss": 1.1184,
+        "step": 15370
+      },
+      {
+        "epoch": 0.4904493127969642,
+        "grad_norm": 0.3314971327781677,
+        "learning_rate": 0.0005,
+        "loss": 1.1109,
+        "step": 15380
+      },
+      {
+        "epoch": 0.4907682005165981,
+        "grad_norm": 0.3312240540981293,
+        "learning_rate": 0.0005,
+        "loss": 1.1164,
+        "step": 15390
+      },
+      {
+        "epoch": 0.49108708823623204,
+        "grad_norm": 0.33236539363861084,
+        "learning_rate": 0.0005,
+        "loss": 1.0962,
+        "step": 15400
+      },
+      {
+        "epoch": 0.49140597595586594,
+        "grad_norm": 0.33051598072052,
+        "learning_rate": 0.0005,
+        "loss": 1.1167,
+        "step": 15410
+      },
+      {
+        "epoch": 0.49172486367549983,
+        "grad_norm": 0.33629119396209717,
+        "learning_rate": 0.0005,
+        "loss": 1.1157,
+        "step": 15420
+      },
+      {
+        "epoch": 0.4920437513951338,
+        "grad_norm": 0.33263641595840454,
+        "learning_rate": 0.0005,
+        "loss": 1.1083,
+        "step": 15430
+      },
+      {
+        "epoch": 0.4923626391147677,
+        "grad_norm": 0.32079949975013733,
+        "learning_rate": 0.0005,
+        "loss": 1.1283,
+        "step": 15440
+      },
+      {
+        "epoch": 0.4926815268344016,
+        "grad_norm": 0.34400150179862976,
+        "learning_rate": 0.0005,
+        "loss": 1.1332,
+        "step": 15450
+      },
+      {
+        "epoch": 0.49300041455403554,
+        "grad_norm": 0.32528164982795715,
+        "learning_rate": 0.0005,
+        "loss": 1.1287,
+        "step": 15460
+      },
+      {
+        "epoch": 0.49331930227366944,
+        "grad_norm": 0.34158089756965637,
+        "learning_rate": 0.0005,
+        "loss": 1.1292,
+        "step": 15470
+      },
+      {
+        "epoch": 0.49363818999330333,
+        "grad_norm": 0.3264855444431305,
+        "learning_rate": 0.0005,
+        "loss": 1.0951,
+        "step": 15480
+      },
+      {
+        "epoch": 0.4939570777129373,
+        "grad_norm": 0.3351936638355255,
+        "learning_rate": 0.0005,
+        "loss": 1.1266,
+        "step": 15490
+      },
+      {
+        "epoch": 0.4942759654325712,
+        "grad_norm": 0.3334040939807892,
+        "learning_rate": 0.0005,
+        "loss": 1.102,
+        "step": 15500
+      },
+      {
+        "epoch": 0.4945948531522051,
+        "grad_norm": 0.34181714057922363,
+        "learning_rate": 0.0005,
+        "loss": 1.1126,
+        "step": 15510
+      },
+      {
+        "epoch": 0.49491374087183904,
+        "grad_norm": 0.32649973034858704,
+        "learning_rate": 0.0005,
+        "loss": 1.1222,
+        "step": 15520
+      },
+      {
+        "epoch": 0.49523262859147293,
+        "grad_norm": 0.33181843161582947,
+        "learning_rate": 0.0005,
+        "loss": 1.1101,
+        "step": 15530
+      },
+      {
+        "epoch": 0.49555151631110683,
+        "grad_norm": 0.3195984661579132,
+        "learning_rate": 0.0005,
+        "loss": 1.11,
+        "step": 15540
+      },
+      {
+        "epoch": 0.4958704040307408,
+        "grad_norm": 0.3320752680301666,
+        "learning_rate": 0.0005,
+        "loss": 1.1157,
+        "step": 15550
+      },
+      {
+        "epoch": 0.4961892917503747,
+        "grad_norm": 0.3402622640132904,
+        "learning_rate": 0.0005,
+        "loss": 1.1054,
+        "step": 15560
+      },
+      {
+        "epoch": 0.49650817947000864,
+        "grad_norm": 0.32147037982940674,
+        "learning_rate": 0.0005,
+        "loss": 1.1161,
+        "step": 15570
+      },
+      {
+        "epoch": 0.49682706718964254,
+        "grad_norm": 0.32301703095436096,
+        "learning_rate": 0.0005,
+        "loss": 1.1137,
+        "step": 15580
+      },
+      {
+        "epoch": 0.49714595490927643,
+        "grad_norm": 0.3312668800354004,
+        "learning_rate": 0.0005,
+        "loss": 1.1193,
+        "step": 15590
+      },
+      {
+        "epoch": 0.4974648426289104,
+        "grad_norm": 0.3265191316604614,
+        "learning_rate": 0.0005,
+        "loss": 1.1028,
+        "step": 15600
+      },
+      {
+        "epoch": 0.4977837303485443,
+        "grad_norm": 0.3307774066925049,
+        "learning_rate": 0.0005,
+        "loss": 1.1019,
+        "step": 15610
+      },
+      {
+        "epoch": 0.4981026180681782,
+        "grad_norm": 0.3307434320449829,
+        "learning_rate": 0.0005,
+        "loss": 1.1119,
+        "step": 15620
+      },
+      {
+        "epoch": 0.49842150578781214,
+        "grad_norm": 0.3392607867717743,
+        "learning_rate": 0.0005,
+        "loss": 1.0973,
+        "step": 15630
+      },
+      {
+        "epoch": 0.49874039350744603,
+        "grad_norm": 0.33843058347702026,
+        "learning_rate": 0.0005,
+        "loss": 1.1075,
+        "step": 15640
+      },
+      {
+        "epoch": 0.49905928122707993,
+        "grad_norm": 0.3343130946159363,
+        "learning_rate": 0.0005,
+        "loss": 1.1091,
+        "step": 15650
+      },
+      {
+        "epoch": 0.4993781689467139,
+        "grad_norm": 0.3436983823776245,
+        "learning_rate": 0.0005,
+        "loss": 1.1301,
+        "step": 15660
+      },
+      {
+        "epoch": 0.4996970566663478,
+        "grad_norm": 0.33115842938423157,
+        "learning_rate": 0.0005,
+        "loss": 1.1082,
+        "step": 15670
+      },
+      {
+        "epoch": 0.5000159443859817,
+        "grad_norm": 0.3339487612247467,
+        "learning_rate": 0.0005,
+        "loss": 1.1121,
+        "step": 15680
+      },
+      {
+        "epoch": 0.5003348321056156,
+        "grad_norm": 0.32088544964790344,
+        "learning_rate": 0.0005,
+        "loss": 1.1045,
+        "step": 15690
+      },
+      {
+        "epoch": 0.5006537198252495,
+        "grad_norm": 0.32219377160072327,
+        "learning_rate": 0.0005,
+        "loss": 1.1053,
+        "step": 15700
+      },
+      {
+        "epoch": 0.5009726075448835,
+        "grad_norm": 0.32998543977737427,
+        "learning_rate": 0.0005,
+        "loss": 1.1107,
+        "step": 15710
+      },
+      {
+        "epoch": 0.5012914952645173,
+        "grad_norm": 0.32973790168762207,
+        "learning_rate": 0.0005,
+        "loss": 1.1204,
+        "step": 15720
+      },
+      {
+        "epoch": 0.5016103829841513,
+        "grad_norm": 0.32395172119140625,
+        "learning_rate": 0.0005,
+        "loss": 1.1005,
+        "step": 15730
+      },
+      {
+        "epoch": 0.5019292707037852,
+        "grad_norm": 0.328885942697525,
+        "learning_rate": 0.0005,
+        "loss": 1.1309,
+        "step": 15740
+      },
+      {
+        "epoch": 0.5022481584234191,
+        "grad_norm": 0.33086928725242615,
+        "learning_rate": 0.0005,
+        "loss": 1.1112,
+        "step": 15750
+      },
+      {
+        "epoch": 0.502567046143053,
+        "grad_norm": 0.3351631760597229,
+        "learning_rate": 0.0005,
+        "loss": 1.0884,
+        "step": 15760
+      },
+      {
+        "epoch": 0.502885933862687,
+        "grad_norm": 0.33762550354003906,
+        "learning_rate": 0.0005,
+        "loss": 1.1138,
+        "step": 15770
+      },
+      {
+        "epoch": 0.5032048215823208,
+        "grad_norm": 0.35110947489738464,
+        "learning_rate": 0.0005,
+        "loss": 1.1083,
+        "step": 15780
+      },
+      {
+        "epoch": 0.5035237093019548,
+        "grad_norm": 0.3249277174472809,
+        "learning_rate": 0.0005,
+        "loss": 1.1036,
+        "step": 15790
+      },
+      {
+        "epoch": 0.5038425970215887,
+        "grad_norm": 0.3232463300228119,
+        "learning_rate": 0.0005,
+        "loss": 1.1012,
+        "step": 15800
+      },
+      {
+        "epoch": 0.5041614847412226,
+        "grad_norm": 0.33399292826652527,
+        "learning_rate": 0.0005,
+        "loss": 1.1157,
+        "step": 15810
+      },
+      {
+        "epoch": 0.5044803724608565,
+        "grad_norm": 0.3410165309906006,
+        "learning_rate": 0.0005,
+        "loss": 1.1029,
+        "step": 15820
+      },
+      {
+        "epoch": 0.5047992601804905,
+        "grad_norm": 0.32605990767478943,
+        "learning_rate": 0.0005,
+        "loss": 1.0959,
+        "step": 15830
+      },
+      {
+        "epoch": 0.5051181479001243,
+        "grad_norm": 0.3320138156414032,
+        "learning_rate": 0.0005,
+        "loss": 1.1122,
+        "step": 15840
+      },
+      {
+        "epoch": 0.5054370356197583,
+        "grad_norm": 0.3410665690898895,
+        "learning_rate": 0.0005,
+        "loss": 1.1164,
+        "step": 15850
+      },
+      {
+        "epoch": 0.5057559233393922,
+        "grad_norm": 0.34686121344566345,
+        "learning_rate": 0.0005,
+        "loss": 1.1192,
+        "step": 15860
+      },
+      {
+        "epoch": 0.5060748110590261,
+        "grad_norm": 0.34597936272621155,
+        "learning_rate": 0.0005,
+        "loss": 1.0986,
+        "step": 15870
+      },
+      {
+        "epoch": 0.50639369877866,
+        "grad_norm": 0.3277563750743866,
+        "learning_rate": 0.0005,
+        "loss": 1.1092,
+        "step": 15880
+      },
+      {
+        "epoch": 0.506712586498294,
+        "grad_norm": 0.31833529472351074,
+        "learning_rate": 0.0005,
+        "loss": 1.0957,
+        "step": 15890
+      },
+      {
+        "epoch": 0.5070314742179278,
+        "grad_norm": 0.33636802434921265,
+        "learning_rate": 0.0005,
+        "loss": 1.1011,
+        "step": 15900
+      },
+      {
+        "epoch": 0.5073503619375618,
+        "grad_norm": 0.32934021949768066,
+        "learning_rate": 0.0005,
+        "loss": 1.0999,
+        "step": 15910
+      },
+      {
+        "epoch": 0.5076692496571957,
+        "grad_norm": 0.33739903569221497,
+        "learning_rate": 0.0005,
+        "loss": 1.099,
+        "step": 15920
+      },
+      {
+        "epoch": 0.5079881373768296,
+        "grad_norm": 0.32094883918762207,
+        "learning_rate": 0.0005,
+        "loss": 1.1056,
+        "step": 15930
+      },
+      {
+        "epoch": 0.5083070250964635,
+        "grad_norm": 0.31300342082977295,
+        "learning_rate": 0.0005,
+        "loss": 1.0974,
+        "step": 15940
+      },
+      {
+        "epoch": 0.5086259128160975,
+        "grad_norm": 0.3224886655807495,
+        "learning_rate": 0.0005,
+        "loss": 1.1031,
+        "step": 15950
+      },
+      {
+        "epoch": 0.5089448005357313,
+        "grad_norm": 0.33514395356178284,
+        "learning_rate": 0.0005,
+        "loss": 1.11,
+        "step": 15960
+      },
+      {
+        "epoch": 0.5092636882553653,
+        "grad_norm": 0.320422887802124,
+        "learning_rate": 0.0005,
+        "loss": 1.111,
+        "step": 15970
+      },
+      {
+        "epoch": 0.5095825759749992,
+        "grad_norm": 0.3200617730617523,
+        "learning_rate": 0.0005,
+        "loss": 1.0999,
+        "step": 15980
+      },
+      {
+        "epoch": 0.5099014636946331,
+        "grad_norm": 0.32363903522491455,
+        "learning_rate": 0.0005,
+        "loss": 1.0888,
+        "step": 15990
+      },
+      {
+        "epoch": 0.510220351414267,
+        "grad_norm": 0.32799628376960754,
+        "learning_rate": 0.0005,
+        "loss": 1.094,
+        "step": 16000
+      },
+      {
+        "epoch": 0.510539239133901,
+        "grad_norm": 0.33881473541259766,
+        "learning_rate": 0.0005,
+        "loss": 1.1014,
+        "step": 16010
+      },
+      {
+        "epoch": 0.5108581268535348,
+        "grad_norm": 0.32602325081825256,
+        "learning_rate": 0.0005,
+        "loss": 1.1111,
+        "step": 16020
+      },
+      {
+        "epoch": 0.5111770145731688,
+        "grad_norm": 0.33234670758247375,
+        "learning_rate": 0.0005,
+        "loss": 1.0863,
+        "step": 16030
+      },
+      {
+        "epoch": 0.5114959022928027,
+        "grad_norm": 0.3310084640979767,
+        "learning_rate": 0.0005,
+        "loss": 1.1222,
+        "step": 16040
+      },
+      {
+        "epoch": 0.5118147900124366,
+        "grad_norm": 0.3287167549133301,
+        "learning_rate": 0.0005,
+        "loss": 1.0986,
+        "step": 16050
+      },
+      {
+        "epoch": 0.5121336777320705,
+        "grad_norm": 0.33373862504959106,
+        "learning_rate": 0.0005,
+        "loss": 1.1214,
+        "step": 16060
+      },
+      {
+        "epoch": 0.5124525654517045,
+        "grad_norm": 0.3199145793914795,
+        "learning_rate": 0.0005,
+        "loss": 1.0752,
+        "step": 16070
+      },
+      {
+        "epoch": 0.5127714531713383,
+        "grad_norm": 0.33313313126564026,
+        "learning_rate": 0.0005,
+        "loss": 1.0977,
+        "step": 16080
+      },
+      {
+        "epoch": 0.5130903408909723,
+        "grad_norm": 0.31770849227905273,
+        "learning_rate": 0.0005,
+        "loss": 1.1104,
+        "step": 16090
+      },
+      {
+        "epoch": 0.5134092286106062,
+        "grad_norm": 0.32164034247398376,
+        "learning_rate": 0.0005,
+        "loss": 1.1103,
+        "step": 16100
+      },
+      {
+        "epoch": 0.5137281163302401,
+        "grad_norm": 0.3254896104335785,
+        "learning_rate": 0.0005,
+        "loss": 1.1007,
+        "step": 16110
+      },
+      {
+        "epoch": 0.514047004049874,
+        "grad_norm": 0.3215978443622589,
+        "learning_rate": 0.0005,
+        "loss": 1.0996,
+        "step": 16120
+      },
+      {
+        "epoch": 0.514365891769508,
+        "grad_norm": 0.31912076473236084,
+        "learning_rate": 0.0005,
+        "loss": 1.1018,
+        "step": 16130
+      },
+      {
+        "epoch": 0.5146847794891418,
+        "grad_norm": 0.31868767738342285,
+        "learning_rate": 0.0005,
+        "loss": 1.1013,
+        "step": 16140
+      },
+      {
+        "epoch": 0.5150036672087758,
+        "grad_norm": 0.3318631649017334,
+        "learning_rate": 0.0005,
+        "loss": 1.1092,
+        "step": 16150
+      },
+      {
+        "epoch": 0.5153225549284097,
+        "grad_norm": 0.3318668603897095,
+        "learning_rate": 0.0005,
+        "loss": 1.1043,
+        "step": 16160
+      },
+      {
+        "epoch": 0.5156414426480436,
+        "grad_norm": 0.3503141403198242,
+        "learning_rate": 0.0005,
+        "loss": 1.0892,
+        "step": 16170
+      },
+      {
+        "epoch": 0.5159603303676775,
+        "grad_norm": 0.33148205280303955,
+        "learning_rate": 0.0005,
+        "loss": 1.0893,
+        "step": 16180
+      },
+      {
+        "epoch": 0.5162792180873115,
+        "grad_norm": 0.320092111825943,
+        "learning_rate": 0.0005,
+        "loss": 1.0984,
+        "step": 16190
+      },
+      {
+        "epoch": 0.5165981058069454,
+        "grad_norm": 0.3361297845840454,
+        "learning_rate": 0.0005,
+        "loss": 1.109,
+        "step": 16200
+      },
+      {
+        "epoch": 0.5169169935265793,
+        "grad_norm": 0.3223029673099518,
+        "learning_rate": 0.0005,
+        "loss": 1.1007,
+        "step": 16210
+      },
+      {
+        "epoch": 0.5172358812462132,
+        "grad_norm": 0.3160358667373657,
+        "learning_rate": 0.0005,
+        "loss": 1.1045,
+        "step": 16220
+      },
+      {
+        "epoch": 0.5175547689658472,
+        "grad_norm": 0.3212055563926697,
+        "learning_rate": 0.0005,
+        "loss": 1.0979,
+        "step": 16230
+      },
+      {
+        "epoch": 0.517873656685481,
+        "grad_norm": 0.33216482400894165,
+        "learning_rate": 0.0005,
+        "loss": 1.1005,
+        "step": 16240
+      },
+      {
+        "epoch": 0.518192544405115,
+        "grad_norm": 0.32473310828208923,
+        "learning_rate": 0.0005,
+        "loss": 1.1105,
+        "step": 16250
+      },
+      {
+        "epoch": 0.5185114321247489,
+        "grad_norm": 0.3209489583969116,
+        "learning_rate": 0.0005,
+        "loss": 1.1025,
+        "step": 16260
+      },
+      {
+        "epoch": 0.5188303198443828,
+        "grad_norm": 0.31811580061912537,
+        "learning_rate": 0.0005,
+        "loss": 1.0915,
+        "step": 16270
+      },
+      {
+        "epoch": 0.5191492075640167,
+        "grad_norm": 0.32100510597229004,
+        "learning_rate": 0.0005,
+        "loss": 1.0897,
+        "step": 16280
+      },
+      {
+        "epoch": 0.5194680952836507,
+        "grad_norm": 0.3171219229698181,
+        "learning_rate": 0.0005,
+        "loss": 1.0954,
+        "step": 16290
+      },
+      {
+        "epoch": 0.5197869830032845,
+        "grad_norm": 0.3136873245239258,
+        "learning_rate": 0.0005,
+        "loss": 1.1141,
+        "step": 16300
+      },
+      {
+        "epoch": 0.5201058707229185,
+        "grad_norm": 0.3253128230571747,
+        "learning_rate": 0.0005,
+        "loss": 1.1143,
+        "step": 16310
+      },
+      {
+        "epoch": 0.5204247584425524,
+        "grad_norm": 0.3093385696411133,
+        "learning_rate": 0.0005,
+        "loss": 1.0969,
+        "step": 16320
+      },
+      {
+        "epoch": 0.5207436461621863,
+        "grad_norm": 0.33357712626457214,
+        "learning_rate": 0.0005,
+        "loss": 1.1024,
+        "step": 16330
+      },
+      {
+        "epoch": 0.5210625338818202,
+        "grad_norm": 0.32705700397491455,
+        "learning_rate": 0.0005,
+        "loss": 1.1166,
+        "step": 16340
+      },
+      {
+        "epoch": 0.5213814216014542,
+        "grad_norm": 0.3455062508583069,
+        "learning_rate": 0.0005,
+        "loss": 1.0951,
+        "step": 16350
+      },
+      {
+        "epoch": 0.521700309321088,
+        "grad_norm": 0.3206673264503479,
+        "learning_rate": 0.0005,
+        "loss": 1.08,
+        "step": 16360
+      },
+      {
+        "epoch": 0.522019197040722,
+        "grad_norm": 0.32268789410591125,
+        "learning_rate": 0.0005,
+        "loss": 1.0939,
+        "step": 16370
+      },
+      {
+        "epoch": 0.5223380847603559,
+        "grad_norm": 0.3400406241416931,
+        "learning_rate": 0.0005,
+        "loss": 1.0899,
+        "step": 16380
+      },
+      {
+        "epoch": 0.5226569724799898,
+        "grad_norm": 0.32846105098724365,
+        "learning_rate": 0.0005,
+        "loss": 1.1309,
+        "step": 16390
+      },
+      {
+        "epoch": 0.5229758601996237,
+        "grad_norm": 0.3352746069431305,
+        "learning_rate": 0.0005,
+        "loss": 1.0927,
+        "step": 16400
+      },
+      {
+        "epoch": 0.5232947479192577,
+        "grad_norm": 0.335052490234375,
+        "learning_rate": 0.0005,
+        "loss": 1.0957,
+        "step": 16410
+      },
+      {
+        "epoch": 0.5236136356388915,
+        "grad_norm": 0.3198140859603882,
+        "learning_rate": 0.0005,
+        "loss": 1.0907,
+        "step": 16420
+      },
+      {
+        "epoch": 0.5239325233585255,
+        "grad_norm": 0.33067774772644043,
+        "learning_rate": 0.0005,
+        "loss": 1.1192,
+        "step": 16430
+      },
+      {
+        "epoch": 0.5242514110781594,
+        "grad_norm": 0.3276583254337311,
+        "learning_rate": 0.0005,
+        "loss": 1.1113,
+        "step": 16440
+      },
+      {
+        "epoch": 0.5245702987977933,
+        "grad_norm": 0.3308953046798706,
+        "learning_rate": 0.0005,
+        "loss": 1.113,
+        "step": 16450
+      },
+      {
+        "epoch": 0.5248891865174272,
+        "grad_norm": 0.32457342743873596,
+        "learning_rate": 0.0005,
+        "loss": 1.0961,
+        "step": 16460
+      },
+      {
+        "epoch": 0.5252080742370612,
+        "grad_norm": 0.3347114324569702,
+        "learning_rate": 0.0005,
+        "loss": 1.1081,
+        "step": 16470
+      },
+      {
+        "epoch": 0.525526961956695,
+        "grad_norm": 0.3195747435092926,
+        "learning_rate": 0.0005,
+        "loss": 1.1221,
+        "step": 16480
+      },
+      {
+        "epoch": 0.525845849676329,
+        "grad_norm": 0.3215080499649048,
+        "learning_rate": 0.0005,
+        "loss": 1.1107,
+        "step": 16490
+      },
+      {
+        "epoch": 0.5261647373959629,
+        "grad_norm": 0.32074692845344543,
+        "learning_rate": 0.0005,
+        "loss": 1.0845,
+        "step": 16500
+      },
+      {
+        "epoch": 0.5264836251155968,
+        "grad_norm": 0.3176482617855072,
+        "learning_rate": 0.0005,
+        "loss": 1.0918,
+        "step": 16510
+      },
+      {
+        "epoch": 0.5268025128352307,
+        "grad_norm": 0.3347056806087494,
+        "learning_rate": 0.0005,
+        "loss": 1.1051,
+        "step": 16520
+      },
+      {
+        "epoch": 0.5271214005548647,
+        "grad_norm": 0.31140273809432983,
+        "learning_rate": 0.0005,
+        "loss": 1.0929,
+        "step": 16530
+      },
+      {
+        "epoch": 0.5274402882744985,
+        "grad_norm": 0.33070501685142517,
+        "learning_rate": 0.0005,
+        "loss": 1.1311,
+        "step": 16540
+      },
+      {
+        "epoch": 0.5277591759941325,
+        "grad_norm": 0.31990689039230347,
+        "learning_rate": 0.0005,
+        "loss": 1.0851,
+        "step": 16550
+      },
+      {
+        "epoch": 0.5280780637137664,
+        "grad_norm": 0.3268377482891083,
+        "learning_rate": 0.0005,
+        "loss": 1.1039,
+        "step": 16560
+      },
+      {
+        "epoch": 0.5283969514334003,
+        "grad_norm": 0.33449193835258484,
+        "learning_rate": 0.0005,
+        "loss": 1.0925,
+        "step": 16570
+      },
+      {
+        "epoch": 0.5287158391530342,
+        "grad_norm": 0.3253713548183441,
+        "learning_rate": 0.0005,
+        "loss": 1.0776,
+        "step": 16580
+      },
+      {
+        "epoch": 0.5290347268726682,
+        "grad_norm": 0.32997438311576843,
+        "learning_rate": 0.0005,
+        "loss": 1.0932,
+        "step": 16590
+      },
+      {
+        "epoch": 0.529353614592302,
+        "grad_norm": 0.30654647946357727,
+        "learning_rate": 0.0005,
+        "loss": 1.0962,
+        "step": 16600
+      },
+      {
+        "epoch": 0.529672502311936,
+        "grad_norm": 0.3221551775932312,
+        "learning_rate": 0.0005,
+        "loss": 1.0841,
+        "step": 16610
+      },
+      {
+        "epoch": 0.5299913900315699,
+        "grad_norm": 0.3215653896331787,
+        "learning_rate": 0.0005,
+        "loss": 1.1215,
+        "step": 16620
+      },
+      {
+        "epoch": 0.5303102777512038,
+        "grad_norm": 0.3248690962791443,
+        "learning_rate": 0.0005,
+        "loss": 1.1035,
+        "step": 16630
+      },
+      {
+        "epoch": 0.5306291654708377,
+        "grad_norm": 0.3199111223220825,
+        "learning_rate": 0.0005,
+        "loss": 1.1017,
+        "step": 16640
+      },
+      {
+        "epoch": 0.5309480531904717,
+        "grad_norm": 0.3270610570907593,
+        "learning_rate": 0.0005,
+        "loss": 1.1033,
+        "step": 16650
+      },
+      {
+        "epoch": 0.5312669409101055,
+        "grad_norm": 0.31784605979919434,
+        "learning_rate": 0.0005,
+        "loss": 1.1037,
+        "step": 16660
+      },
+      {
+        "epoch": 0.5315858286297395,
+        "grad_norm": 0.31250184774398804,
+        "learning_rate": 0.0005,
+        "loss": 1.0831,
+        "step": 16670
+      },
+      {
+        "epoch": 0.5319047163493734,
+        "grad_norm": 0.31594663858413696,
+        "learning_rate": 0.0005,
+        "loss": 1.1303,
+        "step": 16680
+      },
+      {
+        "epoch": 0.5322236040690073,
+        "grad_norm": 0.31342098116874695,
+        "learning_rate": 0.0005,
+        "loss": 1.0912,
+        "step": 16690
+      },
+      {
+        "epoch": 0.5325424917886412,
+        "grad_norm": 0.33595171570777893,
+        "learning_rate": 0.0005,
+        "loss": 1.0904,
+        "step": 16700
+      },
+      {
+        "epoch": 0.5328613795082752,
+        "grad_norm": 0.32776495814323425,
+        "learning_rate": 0.0005,
+        "loss": 1.0886,
+        "step": 16710
+      },
+      {
+        "epoch": 0.533180267227909,
+        "grad_norm": 0.324022114276886,
+        "learning_rate": 0.0005,
+        "loss": 1.1122,
+        "step": 16720
+      },
+      {
+        "epoch": 0.533499154947543,
+        "grad_norm": 0.32876861095428467,
+        "learning_rate": 0.0005,
+        "loss": 1.1067,
+        "step": 16730
+      },
+      {
+        "epoch": 0.5338180426671769,
+        "grad_norm": 0.3247472643852234,
+        "learning_rate": 0.0005,
+        "loss": 1.0947,
+        "step": 16740
+      },
+      {
+        "epoch": 0.5341369303868108,
+        "grad_norm": 0.3151411712169647,
+        "learning_rate": 0.0005,
+        "loss": 1.0858,
+        "step": 16750
+      },
+      {
+        "epoch": 0.5344558181064447,
+        "grad_norm": 0.3146439790725708,
+        "learning_rate": 0.0005,
+        "loss": 1.1091,
+        "step": 16760
+      },
+      {
+        "epoch": 0.5347747058260787,
+        "grad_norm": 0.3136953115463257,
+        "learning_rate": 0.0005,
+        "loss": 1.0998,
+        "step": 16770
+      },
+      {
+        "epoch": 0.5350935935457125,
+        "grad_norm": 0.3351533114910126,
+        "learning_rate": 0.0005,
+        "loss": 1.1047,
+        "step": 16780
+      },
+      {
+        "epoch": 0.5354124812653465,
+        "grad_norm": 0.3329567611217499,
+        "learning_rate": 0.0005,
+        "loss": 1.1035,
+        "step": 16790
+      },
+      {
+        "epoch": 0.5357313689849804,
+        "grad_norm": 0.3350735306739807,
+        "learning_rate": 0.0005,
+        "loss": 1.0927,
+        "step": 16800
+      },
+      {
+        "epoch": 0.5360502567046143,
+        "grad_norm": 0.3343573212623596,
+        "learning_rate": 0.0005,
+        "loss": 1.1199,
+        "step": 16810
+      },
+      {
+        "epoch": 0.5363691444242482,
+        "grad_norm": 0.3330814242362976,
+        "learning_rate": 0.0005,
+        "loss": 1.0983,
+        "step": 16820
+      },
+      {
+        "epoch": 0.5366880321438822,
+        "grad_norm": 0.3203672766685486,
+        "learning_rate": 0.0005,
+        "loss": 1.0973,
+        "step": 16830
+      },
+      {
+        "epoch": 0.537006919863516,
+        "grad_norm": 0.3277007043361664,
+        "learning_rate": 0.0005,
+        "loss": 1.1185,
+        "step": 16840
+      },
+      {
+        "epoch": 0.53732580758315,
+        "grad_norm": 0.3185708820819855,
+        "learning_rate": 0.0005,
+        "loss": 1.1009,
+        "step": 16850
+      },
+      {
+        "epoch": 0.5376446953027839,
+        "grad_norm": 0.3192184567451477,
+        "learning_rate": 0.0005,
+        "loss": 1.1048,
+        "step": 16860
+      },
+      {
+        "epoch": 0.5379635830224178,
+        "grad_norm": 0.33029595017433167,
+        "learning_rate": 0.0005,
+        "loss": 1.0923,
+        "step": 16870
+      },
+      {
+        "epoch": 0.5382824707420517,
+        "grad_norm": 0.3171393871307373,
+        "learning_rate": 0.0005,
+        "loss": 1.0793,
+        "step": 16880
+      },
+      {
+        "epoch": 0.5386013584616857,
+        "grad_norm": 0.30802884697914124,
+        "learning_rate": 0.0005,
+        "loss": 1.1076,
+        "step": 16890
+      },
+      {
+        "epoch": 0.5389202461813195,
+        "grad_norm": 0.31417855620384216,
+        "learning_rate": 0.0005,
+        "loss": 1.1079,
+        "step": 16900
+      },
+      {
+        "epoch": 0.5392391339009535,
+        "grad_norm": 0.3120649456977844,
+        "learning_rate": 0.0005,
+        "loss": 1.0814,
+        "step": 16910
+      },
+      {
+        "epoch": 0.5395580216205874,
+        "grad_norm": 0.3179928958415985,
+        "learning_rate": 0.0005,
+        "loss": 1.1141,
+        "step": 16920
+      },
+      {
+        "epoch": 0.5398769093402213,
+        "grad_norm": 0.31776371598243713,
+        "learning_rate": 0.0005,
+        "loss": 1.0932,
+        "step": 16930
+      },
+      {
+        "epoch": 0.5401957970598552,
+        "grad_norm": 0.31800293922424316,
+        "learning_rate": 0.0005,
+        "loss": 1.0993,
+        "step": 16940
+      },
+      {
+        "epoch": 0.5405146847794892,
+        "grad_norm": 0.3233039081096649,
+        "learning_rate": 0.0005,
+        "loss": 1.0912,
+        "step": 16950
+      },
+      {
+        "epoch": 0.540833572499123,
+        "grad_norm": 0.34119749069213867,
+        "learning_rate": 0.0005,
+        "loss": 1.0959,
+        "step": 16960
+      },
+      {
+        "epoch": 0.541152460218757,
+        "grad_norm": 0.3114102780818939,
+        "learning_rate": 0.0005,
+        "loss": 1.0944,
+        "step": 16970
+      },
+      {
+        "epoch": 0.5414713479383909,
+        "grad_norm": 0.32388511300086975,
+        "learning_rate": 0.0005,
+        "loss": 1.0894,
+        "step": 16980
+      },
+      {
+        "epoch": 0.5417902356580248,
+        "grad_norm": 0.3139544129371643,
+        "learning_rate": 0.0005,
+        "loss": 1.1136,
+        "step": 16990
+      },
+      {
+        "epoch": 0.5421091233776587,
+        "grad_norm": 0.31297817826271057,
+        "learning_rate": 0.0005,
+        "loss": 1.0754,
+        "step": 17000
+      },
+      {
+        "epoch": 0.5424280110972927,
+        "grad_norm": 0.3118950128555298,
+        "learning_rate": 0.0005,
+        "loss": 1.0771,
+        "step": 17010
+      },
+      {
+        "epoch": 0.5427468988169266,
+        "grad_norm": 0.3142285943031311,
+        "learning_rate": 0.0005,
+        "loss": 1.0684,
+        "step": 17020
+      },
+      {
+        "epoch": 0.5430657865365605,
+        "grad_norm": 0.320333868265152,
+        "learning_rate": 0.0005,
+        "loss": 1.0987,
+        "step": 17030
+      },
+      {
+        "epoch": 0.5433846742561944,
+        "grad_norm": 0.3135775327682495,
+        "learning_rate": 0.0005,
+        "loss": 1.0792,
+        "step": 17040
+      },
+      {
+        "epoch": 0.5437035619758284,
+        "grad_norm": 0.3219400644302368,
+        "learning_rate": 0.0005,
+        "loss": 1.0789,
+        "step": 17050
+      },
+      {
+        "epoch": 0.5440224496954622,
+        "grad_norm": 0.3215932250022888,
+        "learning_rate": 0.0005,
+        "loss": 1.1065,
+        "step": 17060
+      },
+      {
+        "epoch": 0.5443413374150962,
+        "grad_norm": 0.3155306279659271,
+        "learning_rate": 0.0005,
+        "loss": 1.1091,
+        "step": 17070
+      },
+      {
+        "epoch": 0.5446602251347301,
+        "grad_norm": 0.3236227333545685,
+        "learning_rate": 0.0005,
+        "loss": 1.111,
+        "step": 17080
+      },
+      {
+        "epoch": 0.544979112854364,
+        "grad_norm": 0.333713561296463,
+        "learning_rate": 0.0005,
+        "loss": 1.1022,
+        "step": 17090
+      },
+      {
+        "epoch": 0.5452980005739979,
+        "grad_norm": 0.3219466805458069,
+        "learning_rate": 0.0005,
+        "loss": 1.0927,
+        "step": 17100
+      },
+      {
+        "epoch": 0.5456168882936319,
+        "grad_norm": 0.3192996680736542,
+        "learning_rate": 0.0005,
+        "loss": 1.1043,
+        "step": 17110
+      },
+      {
+        "epoch": 0.5459357760132657,
+        "grad_norm": 0.32369479537010193,
+        "learning_rate": 0.0005,
+        "loss": 1.1067,
+        "step": 17120
+      },
+      {
+        "epoch": 0.5462546637328997,
+        "grad_norm": 0.32851865887641907,
+        "learning_rate": 0.0005,
+        "loss": 1.0983,
+        "step": 17130
+      },
+      {
+        "epoch": 0.5465735514525336,
+        "grad_norm": 0.3223077952861786,
+        "learning_rate": 0.0005,
+        "loss": 1.0899,
+        "step": 17140
+      },
+      {
+        "epoch": 0.5468924391721675,
+        "grad_norm": 0.3217220902442932,
+        "learning_rate": 0.0005,
+        "loss": 1.0957,
+        "step": 17150
+      },
+      {
+        "epoch": 0.5472113268918014,
+        "grad_norm": 0.31770792603492737,
+        "learning_rate": 0.0005,
+        "loss": 1.0947,
+        "step": 17160
+      },
+      {
+        "epoch": 0.5475302146114354,
+        "grad_norm": 0.3142428398132324,
+        "learning_rate": 0.0005,
+        "loss": 1.0879,
+        "step": 17170
+      },
+      {
+        "epoch": 0.5478491023310692,
+        "grad_norm": 0.3312866687774658,
+        "learning_rate": 0.0005,
+        "loss": 1.107,
+        "step": 17180
+      },
+      {
+        "epoch": 0.5481679900507032,
+        "grad_norm": 0.32655006647109985,
+        "learning_rate": 0.0005,
+        "loss": 1.0836,
+        "step": 17190
+      },
+      {
+        "epoch": 0.5484868777703371,
+        "grad_norm": 0.3185037076473236,
+        "learning_rate": 0.0005,
+        "loss": 1.1043,
+        "step": 17200
+      },
+      {
+        "epoch": 0.548805765489971,
+        "grad_norm": 0.3207053542137146,
+        "learning_rate": 0.0005,
+        "loss": 1.0953,
+        "step": 17210
+      },
+      {
+        "epoch": 0.5491246532096049,
+        "grad_norm": 0.31583043932914734,
+        "learning_rate": 0.0005,
+        "loss": 1.0864,
+        "step": 17220
+      },
+      {
+        "epoch": 0.5494435409292389,
+        "grad_norm": 0.3145662248134613,
+        "learning_rate": 0.0005,
+        "loss": 1.0854,
+        "step": 17230
+      },
+      {
+        "epoch": 0.5497624286488727,
+        "grad_norm": 0.3155752718448639,
+        "learning_rate": 0.0005,
+        "loss": 1.1009,
+        "step": 17240
+      },
+      {
+        "epoch": 0.5500813163685067,
+        "grad_norm": 0.32154303789138794,
+        "learning_rate": 0.0005,
+        "loss": 1.0996,
+        "step": 17250
+      },
+      {
+        "epoch": 0.5504002040881406,
+        "grad_norm": 0.3479996621608734,
+        "learning_rate": 0.0005,
+        "loss": 1.0944,
+        "step": 17260
+      },
+      {
+        "epoch": 0.5507190918077745,
+        "grad_norm": 0.33058908581733704,
+        "learning_rate": 0.0005,
+        "loss": 1.1195,
+        "step": 17270
+      },
+      {
+        "epoch": 0.5510379795274084,
+        "grad_norm": 0.31440943479537964,
+        "learning_rate": 0.0005,
+        "loss": 1.0773,
+        "step": 17280
+      },
+      {
+        "epoch": 0.5513568672470424,
+        "grad_norm": 0.32479768991470337,
+        "learning_rate": 0.0005,
+        "loss": 1.1164,
+        "step": 17290
+      },
+      {
+        "epoch": 0.5516757549666762,
+        "grad_norm": 0.3268716037273407,
+        "learning_rate": 0.0005,
+        "loss": 1.0876,
+        "step": 17300
+      },
+      {
+        "epoch": 0.5519946426863102,
+        "grad_norm": 0.3237885534763336,
+        "learning_rate": 0.0005,
+        "loss": 1.0971,
+        "step": 17310
+      },
+      {
+        "epoch": 0.5523135304059441,
+        "grad_norm": 0.31847426295280457,
+        "learning_rate": 0.0005,
+        "loss": 1.0974,
+        "step": 17320
+      },
+      {
+        "epoch": 0.552632418125578,
+        "grad_norm": 0.3174494802951813,
+        "learning_rate": 0.0005,
+        "loss": 1.0852,
+        "step": 17330
+      },
+      {
+        "epoch": 0.5529513058452119,
+        "grad_norm": 0.3171641528606415,
+        "learning_rate": 0.0005,
+        "loss": 1.0966,
+        "step": 17340
+      },
+      {
+        "epoch": 0.5532701935648459,
+        "grad_norm": 0.31176289916038513,
+        "learning_rate": 0.0005,
+        "loss": 1.0982,
+        "step": 17350
+      },
+      {
+        "epoch": 0.5535890812844797,
+        "grad_norm": 0.31715038418769836,
+        "learning_rate": 0.0005,
+        "loss": 1.0745,
+        "step": 17360
+      },
+      {
+        "epoch": 0.5539079690041137,
+        "grad_norm": 0.3252870440483093,
+        "learning_rate": 0.0005,
+        "loss": 1.0895,
+        "step": 17370
+      },
+      {
+        "epoch": 0.5542268567237476,
+        "grad_norm": 0.3222629725933075,
+        "learning_rate": 0.0005,
+        "loss": 1.1029,
+        "step": 17380
+      },
+      {
+        "epoch": 0.5545457444433814,
+        "grad_norm": 0.3321414291858673,
+        "learning_rate": 0.0005,
+        "loss": 1.0894,
+        "step": 17390
+      },
+      {
+        "epoch": 0.5548646321630154,
+        "grad_norm": 0.31622350215911865,
+        "learning_rate": 0.0005,
+        "loss": 1.1194,
+        "step": 17400
+      },
+      {
+        "epoch": 0.5551835198826494,
+        "grad_norm": 0.33685582876205444,
+        "learning_rate": 0.0005,
+        "loss": 1.0993,
+        "step": 17410
+      },
+      {
+        "epoch": 0.5555024076022832,
+        "grad_norm": 0.3243965208530426,
+        "learning_rate": 0.0005,
+        "loss": 1.0831,
+        "step": 17420
+      },
+      {
+        "epoch": 0.5558212953219172,
+        "grad_norm": 0.331458181142807,
+        "learning_rate": 0.0005,
+        "loss": 1.0939,
+        "step": 17430
+      },
+      {
+        "epoch": 0.5561401830415511,
+        "grad_norm": 0.32177624106407166,
+        "learning_rate": 0.0005,
+        "loss": 1.0987,
+        "step": 17440
+      },
+      {
+        "epoch": 0.556459070761185,
+        "grad_norm": 0.31778931617736816,
+        "learning_rate": 0.0005,
+        "loss": 1.0989,
+        "step": 17450
+      },
+      {
+        "epoch": 0.5567779584808189,
+        "grad_norm": 0.3121318519115448,
+        "learning_rate": 0.0005,
+        "loss": 1.0955,
+        "step": 17460
+      },
+      {
+        "epoch": 0.5570968462004529,
+        "grad_norm": 0.3124259114265442,
+        "learning_rate": 0.0005,
+        "loss": 1.1019,
+        "step": 17470
+      },
+      {
+        "epoch": 0.5574157339200867,
+        "grad_norm": 0.3164832890033722,
+        "learning_rate": 0.0005,
+        "loss": 1.0751,
+        "step": 17480
+      },
+      {
+        "epoch": 0.5577346216397207,
+        "grad_norm": 0.31521815061569214,
+        "learning_rate": 0.0005,
+        "loss": 1.1034,
+        "step": 17490
+      },
+      {
+        "epoch": 0.5580535093593546,
+        "grad_norm": 0.3184082508087158,
+        "learning_rate": 0.0005,
+        "loss": 1.094,
+        "step": 17500
+      },
+      {
+        "epoch": 0.5583723970789884,
+        "grad_norm": 0.33521050214767456,
+        "learning_rate": 0.0005,
+        "loss": 1.0955,
+        "step": 17510
+      },
+      {
+        "epoch": 0.5586912847986224,
+        "grad_norm": 0.3602006733417511,
+        "learning_rate": 0.0005,
+        "loss": 1.0959,
+        "step": 17520
+      },
+      {
+        "epoch": 0.5590101725182564,
+        "grad_norm": 0.33869659900665283,
+        "learning_rate": 0.0005,
+        "loss": 1.0808,
+        "step": 17530
+      },
+      {
+        "epoch": 0.5593290602378902,
+        "grad_norm": 0.30730921030044556,
+        "learning_rate": 0.0005,
+        "loss": 1.0845,
+        "step": 17540
+      },
+      {
+        "epoch": 0.5596479479575242,
+        "grad_norm": 0.32410943508148193,
+        "learning_rate": 0.0005,
+        "loss": 1.1208,
+        "step": 17550
+      },
+      {
+        "epoch": 0.5599668356771581,
+        "grad_norm": 0.323948472738266,
+        "learning_rate": 0.0005,
+        "loss": 1.0782,
+        "step": 17560
+      },
+      {
+        "epoch": 0.560285723396792,
+        "grad_norm": 0.35217639803886414,
+        "learning_rate": 0.0005,
+        "loss": 1.0914,
+        "step": 17570
+      },
+      {
+        "epoch": 0.5606046111164259,
+        "grad_norm": 0.32672759890556335,
+        "learning_rate": 0.0005,
+        "loss": 1.0978,
+        "step": 17580
+      },
+      {
+        "epoch": 0.5609234988360599,
+        "grad_norm": 0.3202705979347229,
+        "learning_rate": 0.0005,
+        "loss": 1.0918,
+        "step": 17590
+      },
+      {
+        "epoch": 0.5612423865556937,
+        "grad_norm": 0.31115081906318665,
+        "learning_rate": 0.0005,
+        "loss": 1.0894,
+        "step": 17600
+      },
+      {
+        "epoch": 0.5615612742753276,
+        "grad_norm": 0.3230699598789215,
+        "learning_rate": 0.0005,
+        "loss": 1.0941,
+        "step": 17610
+      },
+      {
+        "epoch": 0.5618801619949616,
+        "grad_norm": 0.31278496980667114,
+        "learning_rate": 0.0005,
+        "loss": 1.093,
+        "step": 17620
+      },
+      {
+        "epoch": 0.5621990497145954,
+        "grad_norm": 0.32693833112716675,
+        "learning_rate": 0.0005,
+        "loss": 1.098,
+        "step": 17630
+      },
+      {
+        "epoch": 0.5625179374342294,
+        "grad_norm": 0.33379238843917847,
+        "learning_rate": 0.0005,
+        "loss": 1.0821,
+        "step": 17640
+      },
+      {
+        "epoch": 0.5628368251538634,
+        "grad_norm": 0.3170439600944519,
+        "learning_rate": 0.0005,
+        "loss": 1.0699,
+        "step": 17650
+      },
+      {
+        "epoch": 0.5631557128734972,
+        "grad_norm": 0.3304067850112915,
+        "learning_rate": 0.0005,
+        "loss": 1.0921,
+        "step": 17660
+      },
+      {
+        "epoch": 0.5634746005931311,
+        "grad_norm": 0.3296552300453186,
+        "learning_rate": 0.0005,
+        "loss": 1.0899,
+        "step": 17670
+      },
+      {
+        "epoch": 0.5637934883127651,
+        "grad_norm": 0.3277881145477295,
+        "learning_rate": 0.0005,
+        "loss": 1.0861,
+        "step": 17680
+      },
+      {
+        "epoch": 0.5641123760323989,
+        "grad_norm": 0.3206186592578888,
+        "learning_rate": 0.0005,
+        "loss": 1.0976,
+        "step": 17690
+      },
+      {
+        "epoch": 0.5644312637520329,
+        "grad_norm": 0.3092396557331085,
+        "learning_rate": 0.0005,
+        "loss": 1.0925,
+        "step": 17700
+      },
+      {
+        "epoch": 0.5647501514716669,
+        "grad_norm": 0.32194575667381287,
+        "learning_rate": 0.0005,
+        "loss": 1.0954,
+        "step": 17710
+      },
+      {
+        "epoch": 0.5650690391913007,
+        "grad_norm": 0.32105258107185364,
+        "learning_rate": 0.0005,
+        "loss": 1.0902,
+        "step": 17720
+      },
+      {
+        "epoch": 0.5653879269109346,
+        "grad_norm": 0.30944812297821045,
+        "learning_rate": 0.0005,
+        "loss": 1.0711,
+        "step": 17730
+      },
+      {
+        "epoch": 0.5657068146305686,
+        "grad_norm": 0.31964802742004395,
+        "learning_rate": 0.0005,
+        "loss": 1.0906,
+        "step": 17740
+      },
+      {
+        "epoch": 0.5660257023502024,
+        "grad_norm": 0.317140132188797,
+        "learning_rate": 0.0005,
+        "loss": 1.0794,
+        "step": 17750
+      },
+      {
+        "epoch": 0.5663445900698364,
+        "grad_norm": 0.31247347593307495,
+        "learning_rate": 0.0005,
+        "loss": 1.0903,
+        "step": 17760
+      },
+      {
+        "epoch": 0.5666634777894703,
+        "grad_norm": 0.317495733499527,
+        "learning_rate": 0.0005,
+        "loss": 1.0792,
+        "step": 17770
+      },
+      {
+        "epoch": 0.5669823655091042,
+        "grad_norm": 0.324897438287735,
+        "learning_rate": 0.0005,
+        "loss": 1.0885,
+        "step": 17780
+      },
+      {
+        "epoch": 0.5673012532287381,
+        "grad_norm": 0.3336643874645233,
+        "learning_rate": 0.0005,
+        "loss": 1.1004,
+        "step": 17790
+      },
+      {
+        "epoch": 0.5676201409483721,
+        "grad_norm": 0.3118143081665039,
+        "learning_rate": 0.0005,
+        "loss": 1.0736,
+        "step": 17800
+      },
+      {
+        "epoch": 0.5679390286680059,
+        "grad_norm": 0.3250415623188019,
+        "learning_rate": 0.0005,
+        "loss": 1.0936,
+        "step": 17810
+      },
+      {
+        "epoch": 0.5682579163876399,
+        "grad_norm": 0.32718950510025024,
+        "learning_rate": 0.0005,
+        "loss": 1.0887,
+        "step": 17820
+      },
+      {
+        "epoch": 0.5685768041072738,
+        "grad_norm": 0.3129884898662567,
+        "learning_rate": 0.0005,
+        "loss": 1.0821,
+        "step": 17830
+      },
+      {
+        "epoch": 0.5688956918269078,
+        "grad_norm": 0.3273843228816986,
+        "learning_rate": 0.0005,
+        "loss": 1.0775,
+        "step": 17840
+      },
+      {
+        "epoch": 0.5692145795465416,
+        "grad_norm": 0.3269795775413513,
+        "learning_rate": 0.0005,
+        "loss": 1.0889,
+        "step": 17850
+      },
+      {
+        "epoch": 0.5695334672661756,
+        "grad_norm": 0.3237743079662323,
+        "learning_rate": 0.0005,
+        "loss": 1.0871,
+        "step": 17860
+      },
+      {
+        "epoch": 0.5698523549858096,
+        "grad_norm": 0.31734833121299744,
+        "learning_rate": 0.0005,
+        "loss": 1.0841,
+        "step": 17870
+      },
+      {
+        "epoch": 0.5701712427054434,
+        "grad_norm": 0.33148670196533203,
+        "learning_rate": 0.0005,
+        "loss": 1.0911,
+        "step": 17880
+      },
+      {
+        "epoch": 0.5704901304250773,
+        "grad_norm": 0.31050851941108704,
+        "learning_rate": 0.0005,
+        "loss": 1.0832,
+        "step": 17890
+      },
+      {
+        "epoch": 0.5708090181447113,
+        "grad_norm": 0.327975869178772,
+        "learning_rate": 0.0005,
+        "loss": 1.0871,
+        "step": 17900
+      },
+      {
+        "epoch": 0.5711279058643451,
+        "grad_norm": 0.30302131175994873,
+        "learning_rate": 0.0005,
+        "loss": 1.0877,
+        "step": 17910
+      },
+      {
+        "epoch": 0.5714467935839791,
+        "grad_norm": 0.30800721049308777,
+        "learning_rate": 0.0005,
+        "loss": 1.0849,
+        "step": 17920
+      },
+      {
+        "epoch": 0.571765681303613,
+        "grad_norm": 0.3125106394290924,
+        "learning_rate": 0.0005,
+        "loss": 1.087,
+        "step": 17930
+      },
+      {
+        "epoch": 0.5720845690232469,
+        "grad_norm": 0.3220685124397278,
+        "learning_rate": 0.0005,
+        "loss": 1.1014,
+        "step": 17940
+      },
+      {
+        "epoch": 0.5724034567428808,
+        "grad_norm": 0.3280767798423767,
+        "learning_rate": 0.0005,
+        "loss": 1.0834,
+        "step": 17950
+      },
+      {
+        "epoch": 0.5727223444625148,
+        "grad_norm": 0.3060448467731476,
+        "learning_rate": 0.0005,
+        "loss": 1.0684,
+        "step": 17960
+      },
+      {
+        "epoch": 0.5730412321821486,
+        "grad_norm": 0.30415078997612,
+        "learning_rate": 0.0005,
+        "loss": 1.0799,
+        "step": 17970
+      },
+      {
+        "epoch": 0.5733601199017826,
+        "grad_norm": 0.32798802852630615,
+        "learning_rate": 0.0005,
+        "loss": 1.0938,
+        "step": 17980
+      },
+      {
+        "epoch": 0.5736790076214165,
+        "grad_norm": 0.31149980425834656,
+        "learning_rate": 0.0005,
+        "loss": 1.0994,
+        "step": 17990
+      },
+      {
+        "epoch": 0.5739978953410504,
+        "grad_norm": 0.3258190155029297,
+        "learning_rate": 0.0005,
+        "loss": 1.0796,
+        "step": 18000
+      },
+      {
+        "epoch": 0.5743167830606843,
+        "grad_norm": 0.3247692883014679,
+        "learning_rate": 0.0005,
+        "loss": 1.0778,
+        "step": 18010
+      },
+      {
+        "epoch": 0.5746356707803183,
+        "grad_norm": 0.3152218759059906,
+        "learning_rate": 0.0005,
+        "loss": 1.0898,
+        "step": 18020
+      },
+      {
+        "epoch": 0.5749545584999521,
+        "grad_norm": 0.32625725865364075,
+        "learning_rate": 0.0005,
+        "loss": 1.0934,
+        "step": 18030
+      },
+      {
+        "epoch": 0.5752734462195861,
+        "grad_norm": 0.3017418682575226,
+        "learning_rate": 0.0005,
+        "loss": 1.0719,
+        "step": 18040
+      },
+      {
+        "epoch": 0.57559233393922,
+        "grad_norm": 0.3239911198616028,
+        "learning_rate": 0.0005,
+        "loss": 1.0818,
+        "step": 18050
+      },
+      {
+        "epoch": 0.5759112216588539,
+        "grad_norm": 0.3368164002895355,
+        "learning_rate": 0.0005,
+        "loss": 1.078,
+        "step": 18060
+      },
+      {
+        "epoch": 0.5762301093784878,
+        "grad_norm": 0.3111822307109833,
+        "learning_rate": 0.0005,
+        "loss": 1.0862,
+        "step": 18070
+      },
+      {
+        "epoch": 0.5765489970981218,
+        "grad_norm": 0.3116544485092163,
+        "learning_rate": 0.0005,
+        "loss": 1.0813,
+        "step": 18080
+      },
+      {
+        "epoch": 0.5768678848177556,
+        "grad_norm": 0.3096770942211151,
+        "learning_rate": 0.0005,
+        "loss": 1.1032,
+        "step": 18090
+      },
+      {
+        "epoch": 0.5771867725373896,
+        "grad_norm": 0.3190501630306244,
+        "learning_rate": 0.0005,
+        "loss": 1.0994,
+        "step": 18100
+      },
+      {
+        "epoch": 0.5775056602570235,
+        "grad_norm": 0.31052494049072266,
+        "learning_rate": 0.0005,
+        "loss": 1.0836,
+        "step": 18110
+      },
+      {
+        "epoch": 0.5778245479766574,
+        "grad_norm": 0.33043143153190613,
+        "learning_rate": 0.0005,
+        "loss": 1.0933,
+        "step": 18120
+      },
+      {
+        "epoch": 0.5781434356962913,
+        "grad_norm": 0.30940407514572144,
+        "learning_rate": 0.0005,
+        "loss": 1.0694,
+        "step": 18130
+      },
+      {
+        "epoch": 0.5784623234159253,
+        "grad_norm": 0.3174685537815094,
+        "learning_rate": 0.0005,
+        "loss": 1.0847,
+        "step": 18140
+      },
+      {
+        "epoch": 0.5787812111355591,
+        "grad_norm": 0.30869245529174805,
+        "learning_rate": 0.0005,
+        "loss": 1.0561,
+        "step": 18150
+      },
+      {
+        "epoch": 0.5791000988551931,
+        "grad_norm": 0.32156649231910706,
+        "learning_rate": 0.0005,
+        "loss": 1.0796,
+        "step": 18160
+      },
+      {
+        "epoch": 0.579418986574827,
+        "grad_norm": 0.3175433278083801,
+        "learning_rate": 0.0005,
+        "loss": 1.0694,
+        "step": 18170
+      },
+      {
+        "epoch": 0.5797378742944609,
+        "grad_norm": 0.3061865270137787,
+        "learning_rate": 0.0005,
+        "loss": 1.0898,
+        "step": 18180
+      },
+      {
+        "epoch": 0.5800567620140948,
+        "grad_norm": 0.31781259179115295,
+        "learning_rate": 0.0005,
+        "loss": 1.0893,
+        "step": 18190
+      },
+      {
+        "epoch": 0.5803756497337288,
+        "grad_norm": 0.32358020544052124,
+        "learning_rate": 0.0005,
+        "loss": 1.0951,
+        "step": 18200
+      },
+      {
+        "epoch": 0.5806945374533626,
+        "grad_norm": 0.3164525330066681,
+        "learning_rate": 0.0005,
+        "loss": 1.0843,
+        "step": 18210
+      },
+      {
+        "epoch": 0.5810134251729966,
+        "grad_norm": 0.3080329895019531,
+        "learning_rate": 0.0005,
+        "loss": 1.0892,
+        "step": 18220
+      },
+      {
+        "epoch": 0.5813323128926305,
+        "grad_norm": 0.30992263555526733,
+        "learning_rate": 0.0005,
+        "loss": 1.1065,
+        "step": 18230
+      },
+      {
+        "epoch": 0.5816512006122644,
+        "grad_norm": 0.31690654158592224,
+        "learning_rate": 0.0005,
+        "loss": 1.0786,
+        "step": 18240
+      },
+      {
+        "epoch": 0.5819700883318983,
+        "grad_norm": 0.3134094774723053,
+        "learning_rate": 0.0005,
+        "loss": 1.0891,
+        "step": 18250
+      },
+      {
+        "epoch": 0.5822889760515323,
+        "grad_norm": 0.3203316032886505,
+        "learning_rate": 0.0005,
+        "loss": 1.0813,
+        "step": 18260
+      },
+      {
+        "epoch": 0.5826078637711661,
+        "grad_norm": 0.3246511220932007,
+        "learning_rate": 0.0005,
+        "loss": 1.0917,
+        "step": 18270
+      },
+      {
+        "epoch": 0.5829267514908001,
+        "grad_norm": 0.3299003541469574,
+        "learning_rate": 0.0005,
+        "loss": 1.0738,
+        "step": 18280
+      },
+      {
+        "epoch": 0.583245639210434,
+        "grad_norm": 0.3138013184070587,
+        "learning_rate": 0.0005,
+        "loss": 1.0879,
+        "step": 18290
+      },
+      {
+        "epoch": 0.5835645269300679,
+        "grad_norm": 0.3196541666984558,
+        "learning_rate": 0.0005,
+        "loss": 1.0802,
+        "step": 18300
+      },
+      {
+        "epoch": 0.5838834146497018,
+        "grad_norm": 0.31367239356040955,
+        "learning_rate": 0.0005,
+        "loss": 1.0718,
+        "step": 18310
+      },
+      {
+        "epoch": 0.5842023023693358,
+        "grad_norm": 0.31627488136291504,
+        "learning_rate": 0.0005,
+        "loss": 1.0858,
+        "step": 18320
+      },
+      {
+        "epoch": 0.5845211900889696,
+        "grad_norm": 0.32422131299972534,
+        "learning_rate": 0.0005,
+        "loss": 1.0874,
+        "step": 18330
+      },
+      {
+        "epoch": 0.5848400778086036,
+        "grad_norm": 0.3215131163597107,
+        "learning_rate": 0.0005,
+        "loss": 1.079,
+        "step": 18340
+      },
+      {
+        "epoch": 0.5851589655282375,
+        "grad_norm": 0.29539674520492554,
+        "learning_rate": 0.0005,
+        "loss": 1.0892,
+        "step": 18350
+      },
+      {
+        "epoch": 0.5854778532478714,
+        "grad_norm": 0.3192177712917328,
+        "learning_rate": 0.0005,
+        "loss": 1.1016,
+        "step": 18360
+      },
+      {
+        "epoch": 0.5857967409675053,
+        "grad_norm": 0.32366615533828735,
+        "learning_rate": 0.0005,
+        "loss": 1.1019,
+        "step": 18370
+      },
+      {
+        "epoch": 0.5861156286871393,
+        "grad_norm": 0.34069007635116577,
+        "learning_rate": 0.0005,
+        "loss": 1.1053,
+        "step": 18380
+      },
+      {
+        "epoch": 0.5864345164067731,
+        "grad_norm": 0.3245450556278229,
+        "learning_rate": 0.0005,
+        "loss": 1.0755,
+        "step": 18390
+      },
+      {
+        "epoch": 0.5867534041264071,
+        "grad_norm": 0.31849008798599243,
+        "learning_rate": 0.0005,
+        "loss": 1.0748,
+        "step": 18400
+      },
+      {
+        "epoch": 0.587072291846041,
+        "grad_norm": 0.3136410713195801,
+        "learning_rate": 0.0005,
+        "loss": 1.0805,
+        "step": 18410
+      },
+      {
+        "epoch": 0.5873911795656749,
+        "grad_norm": 0.32406342029571533,
+        "learning_rate": 0.0005,
+        "loss": 1.0971,
+        "step": 18420
+      },
+      {
+        "epoch": 0.5877100672853088,
+        "grad_norm": 0.3156557083129883,
+        "learning_rate": 0.0005,
+        "loss": 1.0971,
+        "step": 18430
+      },
+      {
+        "epoch": 0.5880289550049428,
+        "grad_norm": 0.31240981817245483,
+        "learning_rate": 0.0005,
+        "loss": 1.0888,
+        "step": 18440
+      },
+      {
+        "epoch": 0.5883478427245766,
+        "grad_norm": 0.3174673914909363,
+        "learning_rate": 0.0005,
+        "loss": 1.0876,
+        "step": 18450
+      },
+      {
+        "epoch": 0.5886667304442106,
+        "grad_norm": 0.30896031856536865,
+        "learning_rate": 0.0005,
+        "loss": 1.0677,
+        "step": 18460
+      },
+      {
+        "epoch": 0.5889856181638445,
+        "grad_norm": 0.3110184967517853,
+        "learning_rate": 0.0005,
+        "loss": 1.0785,
+        "step": 18470
+      },
+      {
+        "epoch": 0.5893045058834784,
+        "grad_norm": 0.3145858645439148,
+        "learning_rate": 0.0005,
+        "loss": 1.0896,
+        "step": 18480
+      },
+      {
+        "epoch": 0.5896233936031123,
+        "grad_norm": 0.30939409136772156,
+        "learning_rate": 0.0005,
+        "loss": 1.0787,
+        "step": 18490
+      },
+      {
+        "epoch": 0.5899422813227463,
+        "grad_norm": 0.30431243777275085,
+        "learning_rate": 0.0005,
+        "loss": 1.087,
+        "step": 18500
+      },
+      {
+        "epoch": 0.5902611690423801,
+        "grad_norm": 0.31970348954200745,
+        "learning_rate": 0.0005,
+        "loss": 1.0854,
+        "step": 18510
+      },
+      {
+        "epoch": 0.5905800567620141,
+        "grad_norm": 0.3159744441509247,
+        "learning_rate": 0.0005,
+        "loss": 1.0932,
+        "step": 18520
+      },
+      {
+        "epoch": 0.590898944481648,
+        "grad_norm": 0.32231128215789795,
+        "learning_rate": 0.0005,
+        "loss": 1.0746,
+        "step": 18530
+      },
+      {
+        "epoch": 0.5912178322012819,
+        "grad_norm": 0.31338217854499817,
+        "learning_rate": 0.0005,
+        "loss": 1.0857,
+        "step": 18540
+      },
+      {
+        "epoch": 0.5915367199209158,
+        "grad_norm": 0.31467384099960327,
+        "learning_rate": 0.0005,
+        "loss": 1.0889,
+        "step": 18550
+      },
+      {
+        "epoch": 0.5918556076405498,
+        "grad_norm": 0.3193754255771637,
+        "learning_rate": 0.0005,
+        "loss": 1.1047,
+        "step": 18560
+      },
+      {
+        "epoch": 0.5921744953601836,
+        "grad_norm": 0.3180255591869354,
+        "learning_rate": 0.0005,
+        "loss": 1.0657,
+        "step": 18570
+      },
+      {
+        "epoch": 0.5924933830798176,
+        "grad_norm": 0.33617210388183594,
+        "learning_rate": 0.0005,
+        "loss": 1.0757,
+        "step": 18580
+      },
+      {
+        "epoch": 0.5928122707994515,
+        "grad_norm": 0.31531816720962524,
+        "learning_rate": 0.0005,
+        "loss": 1.0864,
+        "step": 18590
+      },
+      {
+        "epoch": 0.5931311585190854,
+        "grad_norm": 0.31398406624794006,
+        "learning_rate": 0.0005,
+        "loss": 1.0818,
+        "step": 18600
+      },
+      {
+        "epoch": 0.5934500462387193,
+        "grad_norm": 0.3130438029766083,
+        "learning_rate": 0.0005,
+        "loss": 1.0916,
+        "step": 18610
+      },
+      {
+        "epoch": 0.5937689339583533,
+        "grad_norm": 0.31848466396331787,
+        "learning_rate": 0.0005,
+        "loss": 1.0892,
+        "step": 18620
+      },
+      {
+        "epoch": 0.5940878216779871,
+        "grad_norm": 0.3218449652194977,
+        "learning_rate": 0.0005,
+        "loss": 1.0859,
+        "step": 18630
+      },
+      {
+        "epoch": 0.5944067093976211,
+        "grad_norm": 0.3119860589504242,
+        "learning_rate": 0.0005,
+        "loss": 1.083,
+        "step": 18640
+      },
+      {
+        "epoch": 0.594725597117255,
+        "grad_norm": 0.3109588325023651,
+        "learning_rate": 0.0005,
+        "loss": 1.0874,
+        "step": 18650
+      },
+      {
+        "epoch": 0.5950444848368889,
+        "grad_norm": 0.320926696062088,
+        "learning_rate": 0.0005,
+        "loss": 1.0579,
+        "step": 18660
+      },
+      {
+        "epoch": 0.5953633725565228,
+        "grad_norm": 0.3240651786327362,
+        "learning_rate": 0.0005,
+        "loss": 1.0754,
+        "step": 18670
+      },
+      {
+        "epoch": 0.5956822602761568,
+        "grad_norm": 0.31916317343711853,
+        "learning_rate": 0.0005,
+        "loss": 1.0845,
+        "step": 18680
+      },
+      {
+        "epoch": 0.5960011479957907,
+        "grad_norm": 0.3060884177684784,
+        "learning_rate": 0.0005,
+        "loss": 1.0777,
+        "step": 18690
+      },
+      {
+        "epoch": 0.5963200357154246,
+        "grad_norm": 0.3231722414493561,
+        "learning_rate": 0.0005,
+        "loss": 1.0662,
+        "step": 18700
+      },
+      {
+        "epoch": 0.5966389234350585,
+        "grad_norm": 0.31437253952026367,
+        "learning_rate": 0.0005,
+        "loss": 1.0897,
+        "step": 18710
+      },
+      {
+        "epoch": 0.5969578111546925,
+        "grad_norm": 0.3079948425292969,
+        "learning_rate": 0.0005,
+        "loss": 1.0843,
+        "step": 18720
+      },
+      {
+        "epoch": 0.5972766988743263,
+        "grad_norm": 0.3151026964187622,
+        "learning_rate": 0.0005,
+        "loss": 1.0825,
+        "step": 18730
+      },
+      {
+        "epoch": 0.5975955865939603,
+        "grad_norm": 0.30802813172340393,
+        "learning_rate": 0.0005,
+        "loss": 1.0981,
+        "step": 18740
+      },
+      {
+        "epoch": 0.5979144743135942,
+        "grad_norm": 0.305775910615921,
+        "learning_rate": 0.0005,
+        "loss": 1.0685,
+        "step": 18750
+      },
+      {
+        "epoch": 0.5982333620332281,
+        "grad_norm": 0.325735479593277,
+        "learning_rate": 0.0005,
+        "loss": 1.0901,
+        "step": 18760
+      },
+      {
+        "epoch": 0.598552249752862,
+        "grad_norm": 0.3160100281238556,
+        "learning_rate": 0.0005,
+        "loss": 1.081,
+        "step": 18770
+      },
+      {
+        "epoch": 0.598871137472496,
+        "grad_norm": 0.31380367279052734,
+        "learning_rate": 0.0005,
+        "loss": 1.0778,
+        "step": 18780
+      },
+      {
+        "epoch": 0.5991900251921298,
+        "grad_norm": 0.3192298114299774,
+        "learning_rate": 0.0005,
+        "loss": 1.1119,
+        "step": 18790
+      },
+      {
+        "epoch": 0.5995089129117638,
+        "grad_norm": 0.3143872618675232,
+        "learning_rate": 0.0005,
+        "loss": 1.0516,
+        "step": 18800
+      },
+      {
+        "epoch": 0.5998278006313977,
+        "grad_norm": 0.31682097911834717,
+        "learning_rate": 0.0005,
+        "loss": 1.0708,
+        "step": 18810
+      },
+      {
+        "epoch": 0.6001466883510316,
+        "grad_norm": 0.31039610505104065,
+        "learning_rate": 0.0005,
+        "loss": 1.0906,
+        "step": 18820
+      },
+      {
+        "epoch": 0.6004655760706655,
+        "grad_norm": 0.32007476687431335,
+        "learning_rate": 0.0005,
+        "loss": 1.0852,
+        "step": 18830
+      },
+      {
+        "epoch": 0.6007844637902995,
+        "grad_norm": 0.3164651393890381,
+        "learning_rate": 0.0005,
+        "loss": 1.0755,
+        "step": 18840
+      },
+      {
+        "epoch": 0.6011033515099333,
+        "grad_norm": 0.32360073924064636,
+        "learning_rate": 0.0005,
+        "loss": 1.1139,
+        "step": 18850
+      },
+      {
+        "epoch": 0.6014222392295673,
+        "grad_norm": 0.3149234354496002,
+        "learning_rate": 0.0005,
+        "loss": 1.0762,
+        "step": 18860
+      },
+      {
+        "epoch": 0.6017411269492012,
+        "grad_norm": 0.3178025484085083,
+        "learning_rate": 0.0005,
+        "loss": 1.0824,
+        "step": 18870
+      },
+      {
+        "epoch": 0.6020600146688351,
+        "grad_norm": 0.324581116437912,
+        "learning_rate": 0.0005,
+        "loss": 1.068,
+        "step": 18880
+      },
+      {
+        "epoch": 0.602378902388469,
+        "grad_norm": 0.3058132231235504,
+        "learning_rate": 0.0005,
+        "loss": 1.0791,
+        "step": 18890
+      },
+      {
+        "epoch": 0.602697790108103,
+        "grad_norm": 0.30854079127311707,
+        "learning_rate": 0.0005,
+        "loss": 1.0819,
+        "step": 18900
+      },
+      {
+        "epoch": 0.6030166778277368,
+        "grad_norm": 0.31456077098846436,
+        "learning_rate": 0.0005,
+        "loss": 1.0844,
+        "step": 18910
+      },
+      {
+        "epoch": 0.6033355655473708,
+        "grad_norm": 0.31017518043518066,
+        "learning_rate": 0.0005,
+        "loss": 1.089,
+        "step": 18920
+      },
+      {
+        "epoch": 0.6036544532670047,
+        "grad_norm": 0.3030060827732086,
+        "learning_rate": 0.0005,
+        "loss": 1.0601,
+        "step": 18930
+      },
+      {
+        "epoch": 0.6039733409866386,
+        "grad_norm": 0.34125179052352905,
+        "learning_rate": 0.0005,
+        "loss": 1.059,
+        "step": 18940
+      },
+      {
+        "epoch": 0.6042922287062725,
+        "grad_norm": 0.3149068355560303,
+        "learning_rate": 0.0005,
+        "loss": 1.0878,
+        "step": 18950
+      },
+      {
+        "epoch": 0.6046111164259065,
+        "grad_norm": 0.32189345359802246,
+        "learning_rate": 0.0005,
+        "loss": 1.0925,
+        "step": 18960
+      },
+      {
+        "epoch": 0.6049300041455403,
+        "grad_norm": 0.32223328948020935,
+        "learning_rate": 0.0005,
+        "loss": 1.0847,
+        "step": 18970
+      },
+      {
+        "epoch": 0.6052488918651743,
+        "grad_norm": 0.3165595829486847,
+        "learning_rate": 0.0005,
+        "loss": 1.0891,
+        "step": 18980
+      },
+      {
+        "epoch": 0.6055677795848082,
+        "grad_norm": 0.32246798276901245,
+        "learning_rate": 0.0005,
+        "loss": 1.0869,
+        "step": 18990
+      },
+      {
+        "epoch": 0.6058866673044421,
+        "grad_norm": 0.3145750164985657,
+        "learning_rate": 0.0005,
+        "loss": 1.0796,
+        "step": 19000
+      },
+      {
+        "epoch": 0.606205555024076,
+        "grad_norm": 0.3327442705631256,
+        "learning_rate": 0.0005,
+        "loss": 1.0733,
+        "step": 19010
+      },
+      {
+        "epoch": 0.60652444274371,
+        "grad_norm": 0.3172445595264435,
+        "learning_rate": 0.0005,
+        "loss": 1.0942,
+        "step": 19020
+      },
+      {
+        "epoch": 0.6068433304633438,
+        "grad_norm": 0.30171164870262146,
+        "learning_rate": 0.0005,
+        "loss": 1.0749,
+        "step": 19030
+      },
+      {
+        "epoch": 0.6071622181829778,
+        "grad_norm": 0.3355121314525604,
+        "learning_rate": 0.0005,
+        "loss": 1.0649,
+        "step": 19040
+      },
+      {
+        "epoch": 0.6074811059026117,
+        "grad_norm": 0.3107321262359619,
+        "learning_rate": 0.0005,
+        "loss": 1.0701,
+        "step": 19050
+      },
+      {
+        "epoch": 0.6077999936222456,
+        "grad_norm": 0.2994323968887329,
+        "learning_rate": 0.0005,
+        "loss": 1.0784,
+        "step": 19060
+      },
+      {
+        "epoch": 0.6081188813418795,
+        "grad_norm": 0.3183419108390808,
+        "learning_rate": 0.0005,
+        "loss": 1.0828,
+        "step": 19070
+      },
+      {
+        "epoch": 0.6084377690615135,
+        "grad_norm": 0.30589038133621216,
+        "learning_rate": 0.0005,
+        "loss": 1.0728,
+        "step": 19080
+      },
+      {
+        "epoch": 0.6087566567811473,
+        "grad_norm": 0.3241669833660126,
+        "learning_rate": 0.0005,
+        "loss": 1.0593,
+        "step": 19090
+      },
+      {
+        "epoch": 0.6090755445007813,
+        "grad_norm": 0.32239434123039246,
+        "learning_rate": 0.0005,
+        "loss": 1.0817,
+        "step": 19100
+      },
+      {
+        "epoch": 0.6093944322204152,
+        "grad_norm": 0.30899012088775635,
+        "learning_rate": 0.0005,
+        "loss": 1.0617,
+        "step": 19110
+      },
+      {
+        "epoch": 0.6097133199400491,
+        "grad_norm": 0.32763203978538513,
+        "learning_rate": 0.0005,
+        "loss": 1.0819,
+        "step": 19120
+      },
+      {
+        "epoch": 0.610032207659683,
+        "grad_norm": 0.3291943371295929,
+        "learning_rate": 0.0005,
+        "loss": 1.0924,
+        "step": 19130
+      },
+      {
+        "epoch": 0.610351095379317,
+        "grad_norm": 0.3210895359516144,
+        "learning_rate": 0.0005,
+        "loss": 1.0946,
+        "step": 19140
+      },
+      {
+        "epoch": 0.6106699830989508,
+        "grad_norm": 0.3067147433757782,
+        "learning_rate": 0.0005,
+        "loss": 1.0917,
+        "step": 19150
+      },
+      {
+        "epoch": 0.6109888708185848,
+        "grad_norm": 0.31053534150123596,
+        "learning_rate": 0.0005,
+        "loss": 1.0649,
+        "step": 19160
+      },
+      {
+        "epoch": 0.6113077585382187,
+        "grad_norm": 0.3062504827976227,
+        "learning_rate": 0.0005,
+        "loss": 1.0925,
+        "step": 19170
+      },
+      {
+        "epoch": 0.6116266462578526,
+        "grad_norm": 0.3294965326786041,
+        "learning_rate": 0.0005,
+        "loss": 1.09,
+        "step": 19180
+      },
+      {
+        "epoch": 0.6119455339774865,
+        "grad_norm": 0.3160441219806671,
+        "learning_rate": 0.0005,
+        "loss": 1.0911,
+        "step": 19190
+      },
+      {
+        "epoch": 0.6122644216971205,
+        "grad_norm": 0.31811705231666565,
+        "learning_rate": 0.0005,
+        "loss": 1.0721,
+        "step": 19200
+      },
+      {
+        "epoch": 0.6125833094167543,
+        "grad_norm": 0.3058662414550781,
+        "learning_rate": 0.0005,
+        "loss": 1.0591,
+        "step": 19210
+      },
+      {
+        "epoch": 0.6129021971363883,
+        "grad_norm": 0.31138697266578674,
+        "learning_rate": 0.0005,
+        "loss": 1.0823,
+        "step": 19220
+      },
+      {
+        "epoch": 0.6132210848560222,
+        "grad_norm": 0.30498701333999634,
+        "learning_rate": 0.0005,
+        "loss": 1.0655,
+        "step": 19230
+      },
+      {
+        "epoch": 0.6135399725756561,
+        "grad_norm": 0.3057716488838196,
+        "learning_rate": 0.0005,
+        "loss": 1.0745,
+        "step": 19240
+      },
+      {
+        "epoch": 0.61385886029529,
+        "grad_norm": 0.31259825825691223,
+        "learning_rate": 0.0005,
+        "loss": 1.0592,
+        "step": 19250
+      },
+      {
+        "epoch": 0.614177748014924,
+        "grad_norm": 0.32048723101615906,
+        "learning_rate": 0.0005,
+        "loss": 1.0882,
+        "step": 19260
+      },
+      {
+        "epoch": 0.6144966357345578,
+        "grad_norm": 0.3062502443790436,
+        "learning_rate": 0.0005,
+        "loss": 1.0894,
+        "step": 19270
+      },
+      {
+        "epoch": 0.6148155234541918,
+        "grad_norm": 0.30581751465797424,
+        "learning_rate": 0.0005,
+        "loss": 1.0821,
+        "step": 19280
+      },
+      {
+        "epoch": 0.6151344111738257,
+        "grad_norm": 0.30549851059913635,
+        "learning_rate": 0.0005,
+        "loss": 1.0616,
+        "step": 19290
+      },
+      {
+        "epoch": 0.6154532988934596,
+        "grad_norm": 0.3184187710285187,
+        "learning_rate": 0.0005,
+        "loss": 1.0855,
+        "step": 19300
+      },
+      {
+        "epoch": 0.6157721866130935,
+        "grad_norm": 0.3128301799297333,
+        "learning_rate": 0.0005,
+        "loss": 1.0877,
+        "step": 19310
+      },
+      {
+        "epoch": 0.6160910743327275,
+        "grad_norm": 0.3164614737033844,
+        "learning_rate": 0.0005,
+        "loss": 1.071,
+        "step": 19320
+      },
+      {
+        "epoch": 0.6164099620523613,
+        "grad_norm": 0.30828580260276794,
+        "learning_rate": 0.0005,
+        "loss": 1.0978,
+        "step": 19330
+      },
+      {
+        "epoch": 0.6167288497719953,
+        "grad_norm": 0.30957871675491333,
+        "learning_rate": 0.0005,
+        "loss": 1.0753,
+        "step": 19340
+      },
+      {
+        "epoch": 0.6170477374916292,
+        "grad_norm": 0.31249192357063293,
+        "learning_rate": 0.0005,
+        "loss": 1.0781,
+        "step": 19350
+      },
+      {
+        "epoch": 0.6173666252112631,
+        "grad_norm": 0.32036566734313965,
+        "learning_rate": 0.0005,
+        "loss": 1.0794,
+        "step": 19360
+      },
+      {
+        "epoch": 0.617685512930897,
+        "grad_norm": 0.3068339228630066,
+        "learning_rate": 0.0005,
+        "loss": 1.0816,
+        "step": 19370
+      },
+      {
+        "epoch": 0.618004400650531,
+        "grad_norm": 0.3214913010597229,
+        "learning_rate": 0.0005,
+        "loss": 1.0882,
+        "step": 19380
+      },
+      {
+        "epoch": 0.6183232883701648,
+        "grad_norm": 0.3247632682323456,
+        "learning_rate": 0.0005,
+        "loss": 1.095,
+        "step": 19390
+      },
+      {
+        "epoch": 0.6186421760897988,
+        "grad_norm": 0.3041056990623474,
+        "learning_rate": 0.0005,
+        "loss": 1.0697,
+        "step": 19400
+      },
+      {
+        "epoch": 0.6189610638094327,
+        "grad_norm": 0.3186788856983185,
+        "learning_rate": 0.0005,
+        "loss": 1.0743,
+        "step": 19410
+      },
+      {
+        "epoch": 0.6192799515290666,
+        "grad_norm": 0.3210011124610901,
+        "learning_rate": 0.0005,
+        "loss": 1.0952,
+        "step": 19420
+      },
+      {
+        "epoch": 0.6195988392487005,
+        "grad_norm": 0.32189202308654785,
+        "learning_rate": 0.0005,
+        "loss": 1.076,
+        "step": 19430
+      },
+      {
+        "epoch": 0.6199177269683345,
+        "grad_norm": 0.30417484045028687,
+        "learning_rate": 0.0005,
+        "loss": 1.0834,
+        "step": 19440
+      },
+      {
+        "epoch": 0.6202366146879683,
+        "grad_norm": 0.31225481629371643,
+        "learning_rate": 0.0005,
+        "loss": 1.0592,
+        "step": 19450
+      },
+      {
+        "epoch": 0.6205555024076023,
+        "grad_norm": 0.30717405676841736,
+        "learning_rate": 0.0005,
+        "loss": 1.0985,
+        "step": 19460
+      },
+      {
+        "epoch": 0.6208743901272362,
+        "grad_norm": 0.31118044257164,
+        "learning_rate": 0.0005,
+        "loss": 1.0651,
+        "step": 19470
+      },
+      {
+        "epoch": 0.6211932778468701,
+        "grad_norm": 0.31739023327827454,
+        "learning_rate": 0.0005,
+        "loss": 1.075,
+        "step": 19480
+      },
+      {
+        "epoch": 0.621512165566504,
+        "grad_norm": 0.3207088112831116,
+        "learning_rate": 0.0005,
+        "loss": 1.0782,
+        "step": 19490
+      },
+      {
+        "epoch": 0.621831053286138,
+        "grad_norm": 0.3196359872817993,
+        "learning_rate": 0.0005,
+        "loss": 1.0844,
+        "step": 19500
+      },
+      {
+        "epoch": 0.6221499410057719,
+        "grad_norm": 0.30430689454078674,
+        "learning_rate": 0.0005,
+        "loss": 1.0849,
+        "step": 19510
+      },
+      {
+        "epoch": 0.6224688287254058,
+        "grad_norm": 0.3219655752182007,
+        "learning_rate": 0.0005,
+        "loss": 1.077,
+        "step": 19520
+      },
+      {
+        "epoch": 0.6227877164450397,
+        "grad_norm": 0.31854191422462463,
+        "learning_rate": 0.0005,
+        "loss": 1.0903,
+        "step": 19530
+      },
+      {
+        "epoch": 0.6231066041646737,
+        "grad_norm": 0.3191505968570709,
+        "learning_rate": 0.0005,
+        "loss": 1.071,
+        "step": 19540
+      },
+      {
+        "epoch": 0.6234254918843075,
+        "grad_norm": 0.3143983781337738,
+        "learning_rate": 0.0005,
+        "loss": 1.076,
+        "step": 19550
+      },
+      {
+        "epoch": 0.6237443796039415,
+        "grad_norm": 0.31715285778045654,
+        "learning_rate": 0.0005,
+        "loss": 1.0982,
+        "step": 19560
+      },
+      {
+        "epoch": 0.6240632673235754,
+        "grad_norm": 0.31628987193107605,
+        "learning_rate": 0.0005,
+        "loss": 1.0877,
+        "step": 19570
+      },
+      {
+        "epoch": 0.6243821550432093,
+        "grad_norm": 0.35582295060157776,
+        "learning_rate": 0.0005,
+        "loss": 1.0843,
+        "step": 19580
+      },
+      {
+        "epoch": 0.6247010427628432,
+        "grad_norm": 0.30786216259002686,
+        "learning_rate": 0.0005,
+        "loss": 1.0763,
+        "step": 19590
+      },
+      {
+        "epoch": 0.6250199304824772,
+        "grad_norm": 0.31700748205184937,
+        "learning_rate": 0.0005,
+        "loss": 1.0728,
+        "step": 19600
+      },
+      {
+        "epoch": 0.625338818202111,
+        "grad_norm": 0.31084170937538147,
+        "learning_rate": 0.0005,
+        "loss": 1.0858,
+        "step": 19610
+      },
+      {
+        "epoch": 0.625657705921745,
+        "grad_norm": 0.3071608245372772,
+        "learning_rate": 0.0005,
+        "loss": 1.0952,
+        "step": 19620
+      },
+      {
+        "epoch": 0.6259765936413789,
+        "grad_norm": 0.3199949562549591,
+        "learning_rate": 0.0005,
+        "loss": 1.07,
+        "step": 19630
+      },
+      {
+        "epoch": 0.6262954813610128,
+        "grad_norm": 0.3328166902065277,
+        "learning_rate": 0.0005,
+        "loss": 1.081,
+        "step": 19640
+      },
+      {
+        "epoch": 0.6266143690806467,
+        "grad_norm": 0.31577205657958984,
+        "learning_rate": 0.0005,
+        "loss": 1.082,
+        "step": 19650
+      },
+      {
+        "epoch": 0.6269332568002807,
+        "grad_norm": 0.3245375454425812,
+        "learning_rate": 0.0005,
+        "loss": 1.0764,
+        "step": 19660
+      },
+      {
+        "epoch": 0.6272521445199145,
+        "grad_norm": 0.3172302544116974,
+        "learning_rate": 0.0005,
+        "loss": 1.0731,
+        "step": 19670
+      },
+      {
+        "epoch": 0.6275710322395485,
+        "grad_norm": 0.31262245774269104,
+        "learning_rate": 0.0005,
+        "loss": 1.0866,
+        "step": 19680
+      },
+      {
+        "epoch": 0.6278899199591824,
+        "grad_norm": 0.30418139696121216,
+        "learning_rate": 0.0005,
+        "loss": 1.0758,
+        "step": 19690
+      },
+      {
+        "epoch": 0.6282088076788163,
+        "grad_norm": 0.31656044721603394,
+        "learning_rate": 0.0005,
+        "loss": 1.074,
+        "step": 19700
+      },
+      {
+        "epoch": 0.6285276953984502,
+        "grad_norm": 0.3122449517250061,
+        "learning_rate": 0.0005,
+        "loss": 1.0621,
+        "step": 19710
+      },
+      {
+        "epoch": 0.6288465831180842,
+        "grad_norm": 0.3014175593852997,
+        "learning_rate": 0.0005,
+        "loss": 1.0772,
+        "step": 19720
+      },
+      {
+        "epoch": 0.629165470837718,
+        "grad_norm": 0.3063112199306488,
+        "learning_rate": 0.0005,
+        "loss": 1.0704,
+        "step": 19730
+      },
+      {
+        "epoch": 0.629484358557352,
+        "grad_norm": 0.3160080313682556,
+        "learning_rate": 0.0005,
+        "loss": 1.0758,
+        "step": 19740
+      },
+      {
+        "epoch": 0.6298032462769859,
+        "grad_norm": 0.3040753901004791,
+        "learning_rate": 0.0005,
+        "loss": 1.0648,
+        "step": 19750
+      },
+      {
+        "epoch": 0.6301221339966198,
+        "grad_norm": 0.31027182936668396,
+        "learning_rate": 0.0005,
+        "loss": 1.0645,
+        "step": 19760
+      },
+      {
+        "epoch": 0.6304410217162537,
+        "grad_norm": 0.32170340418815613,
+        "learning_rate": 0.0005,
+        "loss": 1.0654,
+        "step": 19770
+      },
+      {
+        "epoch": 0.6307599094358877,
+        "grad_norm": 0.3243429660797119,
+        "learning_rate": 0.0005,
+        "loss": 1.0693,
+        "step": 19780
+      },
+      {
+        "epoch": 0.6310787971555215,
+        "grad_norm": 0.3192366361618042,
+        "learning_rate": 0.0005,
+        "loss": 1.0811,
+        "step": 19790
+      },
+      {
+        "epoch": 0.6313976848751555,
+        "grad_norm": 0.32173165678977966,
+        "learning_rate": 0.0005,
+        "loss": 1.072,
+        "step": 19800
+      },
+      {
+        "epoch": 0.6317165725947894,
+        "grad_norm": 0.3074743449687958,
+        "learning_rate": 0.0005,
+        "loss": 1.077,
+        "step": 19810
+      },
+      {
+        "epoch": 0.6320354603144233,
+        "grad_norm": 0.3193652927875519,
+        "learning_rate": 0.0005,
+        "loss": 1.0719,
+        "step": 19820
+      },
+      {
+        "epoch": 0.6323543480340572,
+        "grad_norm": 0.3196190297603607,
+        "learning_rate": 0.0005,
+        "loss": 1.0755,
+        "step": 19830
+      },
+      {
+        "epoch": 0.6326732357536912,
+        "grad_norm": 0.3098837435245514,
+        "learning_rate": 0.0005,
+        "loss": 1.0872,
+        "step": 19840
+      },
+      {
+        "epoch": 0.632992123473325,
+        "grad_norm": 0.31178563833236694,
+        "learning_rate": 0.0005,
+        "loss": 1.0707,
+        "step": 19850
+      },
+      {
+        "epoch": 0.633311011192959,
+        "grad_norm": 0.31588777899742126,
+        "learning_rate": 0.0005,
+        "loss": 1.076,
+        "step": 19860
+      },
+      {
+        "epoch": 0.6336298989125929,
+        "grad_norm": 0.3056369125843048,
+        "learning_rate": 0.0005,
+        "loss": 1.0688,
+        "step": 19870
+      },
+      {
+        "epoch": 0.6339487866322268,
+        "grad_norm": 0.30744126439094543,
+        "learning_rate": 0.0005,
+        "loss": 1.0759,
+        "step": 19880
+      },
+      {
+        "epoch": 0.6342676743518607,
+        "grad_norm": 0.3057192265987396,
+        "learning_rate": 0.0005,
+        "loss": 1.081,
+        "step": 19890
+      },
+      {
+        "epoch": 0.6345865620714947,
+        "grad_norm": 0.3178814947605133,
+        "learning_rate": 0.0005,
+        "loss": 1.0742,
+        "step": 19900
+      },
+      {
+        "epoch": 0.6349054497911285,
+        "grad_norm": 0.328322172164917,
+        "learning_rate": 0.0005,
+        "loss": 1.0863,
+        "step": 19910
+      },
+      {
+        "epoch": 0.6352243375107625,
+        "grad_norm": 0.32027745246887207,
+        "learning_rate": 0.0005,
+        "loss": 1.0823,
+        "step": 19920
+      },
+      {
+        "epoch": 0.6355432252303964,
+        "grad_norm": 0.3261570334434509,
+        "learning_rate": 0.0005,
+        "loss": 1.0639,
+        "step": 19930
+      },
+      {
+        "epoch": 0.6358621129500303,
+        "grad_norm": 0.321636825799942,
+        "learning_rate": 0.0005,
+        "loss": 1.0653,
+        "step": 19940
+      },
+      {
+        "epoch": 0.6361810006696642,
+        "grad_norm": 0.2975068986415863,
+        "learning_rate": 0.0005,
+        "loss": 1.0604,
+        "step": 19950
+      },
+      {
+        "epoch": 0.6364998883892982,
+        "grad_norm": 0.30960360169410706,
+        "learning_rate": 0.0005,
+        "loss": 1.0758,
+        "step": 19960
+      },
+      {
+        "epoch": 0.636818776108932,
+        "grad_norm": 0.31270766258239746,
+        "learning_rate": 0.0005,
+        "loss": 1.0889,
+        "step": 19970
+      },
+      {
+        "epoch": 0.637137663828566,
+        "grad_norm": 0.3046696186065674,
+        "learning_rate": 0.0005,
+        "loss": 1.089,
+        "step": 19980
+      },
+      {
+        "epoch": 0.6374565515481999,
+        "grad_norm": 0.30438730120658875,
+        "learning_rate": 0.0005,
+        "loss": 1.0811,
+        "step": 19990
+      },
+      {
+        "epoch": 0.6377754392678338,
+        "grad_norm": 0.3202705383300781,
+        "learning_rate": 0.0005,
+        "loss": 1.0734,
+        "step": 20000
+      },
+      {
+        "epoch": 0.6380943269874677,
+        "grad_norm": 0.30801746249198914,
+        "learning_rate": 0.0005,
+        "loss": 1.0845,
+        "step": 20010
+      },
+      {
+        "epoch": 0.6384132147071017,
+        "grad_norm": 0.3115306794643402,
+        "learning_rate": 0.0005,
+        "loss": 1.0866,
+        "step": 20020
+      },
+      {
+        "epoch": 0.6387321024267355,
+        "grad_norm": 0.316930890083313,
+        "learning_rate": 0.0005,
+        "loss": 1.0956,
+        "step": 20030
+      },
+      {
+        "epoch": 0.6390509901463695,
+        "grad_norm": 0.2994682192802429,
+        "learning_rate": 0.0005,
+        "loss": 1.0702,
+        "step": 20040
+      },
+      {
+        "epoch": 0.6393698778660034,
+        "grad_norm": 0.29843565821647644,
+        "learning_rate": 0.0005,
+        "loss": 1.0666,
+        "step": 20050
+      },
+      {
+        "epoch": 0.6396887655856373,
+        "grad_norm": 0.3006264567375183,
+        "learning_rate": 0.0005,
+        "loss": 1.0761,
+        "step": 20060
+      },
+      {
+        "epoch": 0.6400076533052712,
+        "grad_norm": 0.30730554461479187,
+        "learning_rate": 0.0005,
+        "loss": 1.0755,
+        "step": 20070
+      },
+      {
+        "epoch": 0.6403265410249052,
+        "grad_norm": 0.3036216199398041,
+        "learning_rate": 0.0005,
+        "loss": 1.0657,
+        "step": 20080
+      },
+      {
+        "epoch": 0.640645428744539,
+        "grad_norm": 0.29510772228240967,
+        "learning_rate": 0.0005,
+        "loss": 1.0851,
+        "step": 20090
+      },
+      {
+        "epoch": 0.640964316464173,
+        "grad_norm": 0.3131539821624756,
+        "learning_rate": 0.0005,
+        "loss": 1.0887,
+        "step": 20100
+      },
+      {
+        "epoch": 0.6412832041838069,
+        "grad_norm": 0.3116014301776886,
+        "learning_rate": 0.0005,
+        "loss": 1.0644,
+        "step": 20110
+      },
+      {
+        "epoch": 0.6416020919034408,
+        "grad_norm": 0.30821144580841064,
+        "learning_rate": 0.0005,
+        "loss": 1.081,
+        "step": 20120
+      },
+      {
+        "epoch": 0.6419209796230747,
+        "grad_norm": 0.3133336007595062,
+        "learning_rate": 0.0005,
+        "loss": 1.0814,
+        "step": 20130
+      },
+      {
+        "epoch": 0.6422398673427087,
+        "grad_norm": 0.30954423546791077,
+        "learning_rate": 0.0005,
+        "loss": 1.0738,
+        "step": 20140
+      },
+      {
+        "epoch": 0.6425587550623425,
+        "grad_norm": 0.31236380338668823,
+        "learning_rate": 0.0005,
+        "loss": 1.08,
+        "step": 20150
+      },
+      {
+        "epoch": 0.6428776427819765,
+        "grad_norm": 0.3041278123855591,
+        "learning_rate": 0.0005,
+        "loss": 1.083,
+        "step": 20160
+      },
+      {
+        "epoch": 0.6431965305016104,
+        "grad_norm": 0.31309154629707336,
+        "learning_rate": 0.0005,
+        "loss": 1.0839,
+        "step": 20170
+      },
+      {
+        "epoch": 0.6435154182212443,
+        "grad_norm": 0.30441009998321533,
+        "learning_rate": 0.0005,
+        "loss": 1.0804,
+        "step": 20180
+      },
+      {
+        "epoch": 0.6438343059408782,
+        "grad_norm": 0.31017962098121643,
+        "learning_rate": 0.0005,
+        "loss": 1.0484,
+        "step": 20190
+      },
+      {
+        "epoch": 0.6441531936605122,
+        "grad_norm": 0.3073233664035797,
+        "learning_rate": 0.0005,
+        "loss": 1.0729,
+        "step": 20200
+      },
+      {
+        "epoch": 0.644472081380146,
+        "grad_norm": 0.3151942491531372,
+        "learning_rate": 0.0005,
+        "loss": 1.0896,
+        "step": 20210
+      },
+      {
+        "epoch": 0.64479096909978,
+        "grad_norm": 0.31230977177619934,
+        "learning_rate": 0.0005,
+        "loss": 1.0653,
+        "step": 20220
+      },
+      {
+        "epoch": 0.6451098568194139,
+        "grad_norm": 0.3134220838546753,
+        "learning_rate": 0.0005,
+        "loss": 1.0781,
+        "step": 20230
+      },
+      {
+        "epoch": 0.6454287445390477,
+        "grad_norm": 0.31314539909362793,
+        "learning_rate": 0.0005,
+        "loss": 1.0615,
+        "step": 20240
+      },
+      {
+        "epoch": 0.6457476322586817,
+        "grad_norm": 0.3104364573955536,
+        "learning_rate": 0.0005,
+        "loss": 1.0555,
+        "step": 20250
+      },
+      {
+        "epoch": 0.6460665199783157,
+        "grad_norm": 0.3331508934497833,
+        "learning_rate": 0.0005,
+        "loss": 1.0692,
+        "step": 20260
+      },
+      {
+        "epoch": 0.6463854076979495,
+        "grad_norm": 0.30866751074790955,
+        "learning_rate": 0.0005,
+        "loss": 1.0753,
+        "step": 20270
+      },
+      {
+        "epoch": 0.6467042954175835,
+        "grad_norm": 0.30994173884391785,
+        "learning_rate": 0.0005,
+        "loss": 1.0934,
+        "step": 20280
+      },
+      {
+        "epoch": 0.6470231831372174,
+        "grad_norm": 0.3175150454044342,
+        "learning_rate": 0.0005,
+        "loss": 1.0595,
+        "step": 20290
+      },
+      {
+        "epoch": 0.6473420708568512,
+        "grad_norm": 0.32161569595336914,
+        "learning_rate": 0.0005,
+        "loss": 1.0798,
+        "step": 20300
+      },
+      {
+        "epoch": 0.6476609585764852,
+        "grad_norm": 0.31296443939208984,
+        "learning_rate": 0.0005,
+        "loss": 1.0628,
+        "step": 20310
+      },
+      {
+        "epoch": 0.6479798462961192,
+        "grad_norm": 0.3004898428916931,
+        "learning_rate": 0.0005,
+        "loss": 1.0651,
+        "step": 20320
+      },
+      {
+        "epoch": 0.6482987340157531,
+        "grad_norm": 0.30378371477127075,
+        "learning_rate": 0.0005,
+        "loss": 1.0462,
+        "step": 20330
+      },
+      {
+        "epoch": 0.648617621735387,
+        "grad_norm": 0.3166843056678772,
+        "learning_rate": 0.0005,
+        "loss": 1.0789,
+        "step": 20340
+      },
+      {
+        "epoch": 0.6489365094550209,
+        "grad_norm": 0.31301385164260864,
+        "learning_rate": 0.0005,
+        "loss": 1.0887,
+        "step": 20350
+      },
+      {
+        "epoch": 0.6492553971746549,
+        "grad_norm": 0.30465880036354065,
+        "learning_rate": 0.0005,
+        "loss": 1.0783,
+        "step": 20360
+      },
+      {
+        "epoch": 0.6495742848942887,
+        "grad_norm": 0.3161429166793823,
+        "learning_rate": 0.0005,
+        "loss": 1.0658,
+        "step": 20370
+      },
+      {
+        "epoch": 0.6498931726139227,
+        "grad_norm": 0.30736565589904785,
+        "learning_rate": 0.0005,
+        "loss": 1.0837,
+        "step": 20380
+      },
+      {
+        "epoch": 0.6502120603335566,
+        "grad_norm": 0.3056853413581848,
+        "learning_rate": 0.0004987293446490035,
+        "loss": 1.047,
+        "step": 20390
+      },
+      {
+        "epoch": 0.6505309480531904,
+        "grad_norm": 0.32306766510009766,
+        "learning_rate": 0.000496824661048599,
+        "loss": 1.0778,
+        "step": 20400
+      },
+      {
+        "epoch": 0.6508498357728244,
+        "grad_norm": 0.3094659745693207,
+        "learning_rate": 0.0004949272515732417,
+        "loss": 1.0593,
+        "step": 20410
+      },
+      {
+        "epoch": 0.6511687234924584,
+        "grad_norm": 0.3049302101135254,
+        "learning_rate": 0.0004930370884425194,
+        "loss": 1.0757,
+        "step": 20420
+      },
+      {
+        "epoch": 0.6514876112120922,
+        "grad_norm": 0.3081049919128418,
+        "learning_rate": 0.0004911541439821156,
+        "loss": 1.0774,
+        "step": 20430
+      },
+      {
+        "epoch": 0.6518064989317262,
+        "grad_norm": 0.31899651885032654,
+        "learning_rate": 0.000489278390623404,
+        "loss": 1.0995,
+        "step": 20440
+      },
+      {
+        "epoch": 0.6521253866513601,
+        "grad_norm": 0.30184099078178406,
+        "learning_rate": 0.0004874098009030445,
+        "loss": 1.0715,
+        "step": 20450
+      },
+      {
+        "epoch": 0.652444274370994,
+        "grad_norm": 0.31541329622268677,
+        "learning_rate": 0.0004855483474625822,
+        "loss": 1.0667,
+        "step": 20460
+      },
+      {
+        "epoch": 0.6527631620906279,
+        "grad_norm": 0.3150116801261902,
+        "learning_rate": 0.0004836940030480454,
+        "loss": 1.0679,
+        "step": 20470
+      },
+      {
+        "epoch": 0.6530820498102619,
+        "grad_norm": 0.3065341114997864,
+        "learning_rate": 0.0004818467405095477,
+        "loss": 1.04,
+        "step": 20480
+      },
+      {
+        "epoch": 0.6534009375298957,
+        "grad_norm": 0.3076440691947937,
+        "learning_rate": 0.00048000653280089,
+        "loss": 1.0669,
+        "step": 20490
+      },
+      {
+        "epoch": 0.6537198252495297,
+        "grad_norm": 0.31127315759658813,
+        "learning_rate": 0.00047817335297916465,
+        "loss": 1.0612,
+        "step": 20500
+      },
+      {
+        "epoch": 0.6540387129691636,
+        "grad_norm": 0.30329588055610657,
+        "learning_rate": 0.0004763471742043608,
+        "loss": 1.0636,
+        "step": 20510
+      },
+      {
+        "epoch": 0.6543576006887974,
+        "grad_norm": 0.31193721294403076,
+        "learning_rate": 0.00047452796973897166,
+        "loss": 1.0676,
+        "step": 20520
+      },
+      {
+        "epoch": 0.6546764884084314,
+        "grad_norm": 0.30703896284103394,
+        "learning_rate": 0.0004727157129476028,
+        "loss": 1.0552,
+        "step": 20530
+      },
+      {
+        "epoch": 0.6549953761280654,
+        "grad_norm": 0.31341981887817383,
+        "learning_rate": 0.0004709103772965826,
+        "loss": 1.074,
+        "step": 20540
+      },
+      {
+        "epoch": 0.6553142638476992,
+        "grad_norm": 0.3001408874988556,
+        "learning_rate": 0.0004691119363535731,
+        "loss": 1.0641,
+        "step": 20550
+      },
+      {
+        "epoch": 0.6556331515673332,
+        "grad_norm": 0.2930058240890503,
+        "learning_rate": 0.0004673203637871834,
+        "loss": 1.0539,
+        "step": 20560
+      },
+      {
+        "epoch": 0.6559520392869671,
+        "grad_norm": 0.309577614068985,
+        "learning_rate": 0.0004655356333665843,
+        "loss": 1.0751,
+        "step": 20570
+      },
+      {
+        "epoch": 0.656270927006601,
+        "grad_norm": 0.3096447288990021,
+        "learning_rate": 0.000463757718961124,
+        "loss": 1.0723,
+        "step": 20580
+      },
+      {
+        "epoch": 0.6565898147262349,
+        "grad_norm": 0.308167427778244,
+        "learning_rate": 0.00046198659453994544,
+        "loss": 1.0593,
+        "step": 20590
+      },
+      {
+        "epoch": 0.6569087024458689,
+        "grad_norm": 0.3029198944568634,
+        "learning_rate": 0.0004602222341716053,
+        "loss": 1.0637,
+        "step": 20600
+      },
+      {
+        "epoch": 0.6572275901655027,
+        "grad_norm": 0.3003123104572296,
+        "learning_rate": 0.0004584646120236944,
+        "loss": 1.0652,
+        "step": 20610
+      },
+      {
+        "epoch": 0.6575464778851366,
+        "grad_norm": 0.30862095952033997,
+        "learning_rate": 0.0004567137023624594,
+        "loss": 1.0593,
+        "step": 20620
+      },
+      {
+        "epoch": 0.6578653656047706,
+        "grad_norm": 0.29509639739990234,
+        "learning_rate": 0.0004549694795524261,
+        "loss": 1.0419,
+        "step": 20630
+      },
+      {
+        "epoch": 0.6581842533244044,
+        "grad_norm": 0.3133094608783722,
+        "learning_rate": 0.0004532319180560237,
+        "loss": 1.1013,
+        "step": 20640
+      },
+      {
+        "epoch": 0.6585031410440384,
+        "grad_norm": 0.3254597783088684,
+        "learning_rate": 0.00045150099243321175,
+        "loss": 1.0659,
+        "step": 20650
+      },
+      {
+        "epoch": 0.6588220287636724,
+        "grad_norm": 0.3049736022949219,
+        "learning_rate": 0.00044977667734110685,
+        "loss": 1.044,
+        "step": 20660
+      },
+      {
+        "epoch": 0.6591409164833062,
+        "grad_norm": 0.30839237570762634,
+        "learning_rate": 0.00044805894753361183,
+        "loss": 1.061,
+        "step": 20670
+      },
+      {
+        "epoch": 0.6594598042029401,
+        "grad_norm": 0.2982349991798401,
+        "learning_rate": 0.0004463477778610465,
+        "loss": 1.0611,
+        "step": 20680
+      },
+      {
+        "epoch": 0.6597786919225741,
+        "grad_norm": 0.30832356214523315,
+        "learning_rate": 0.0004446431432697789,
+        "loss": 1.055,
+        "step": 20690
+      },
+      {
+        "epoch": 0.6600975796422079,
+        "grad_norm": 0.2984870970249176,
+        "learning_rate": 0.0004429450188018587,
+        "loss": 1.0611,
+        "step": 20700
+      },
+      {
+        "epoch": 0.6604164673618419,
+        "grad_norm": 0.3130272924900055,
+        "learning_rate": 0.00044125337959465206,
+        "loss": 1.0632,
+        "step": 20710
+      },
+      {
+        "epoch": 0.6607353550814759,
+        "grad_norm": 0.3037204146385193,
+        "learning_rate": 0.00043956820088047696,
+        "loss": 1.0297,
+        "step": 20720
+      },
+      {
+        "epoch": 0.6610542428011097,
+        "grad_norm": 0.30242738127708435,
+        "learning_rate": 0.0004378894579862412,
+        "loss": 1.0539,
+        "step": 20730
+      },
+      {
+        "epoch": 0.6613731305207436,
+        "grad_norm": 0.3087151050567627,
+        "learning_rate": 0.00043621712633308096,
+        "loss": 1.0575,
+        "step": 20740
+      },
+      {
+        "epoch": 0.6616920182403776,
+        "grad_norm": 0.3058840036392212,
+        "learning_rate": 0.00043455118143600075,
+        "loss": 1.0649,
+        "step": 20750
+      },
+      {
+        "epoch": 0.6620109059600114,
+        "grad_norm": 0.2988451421260834,
+        "learning_rate": 0.00043289159890351497,
+        "loss": 1.0449,
+        "step": 20760
+      },
+      {
+        "epoch": 0.6623297936796454,
+        "grad_norm": 0.31583669781684875,
+        "learning_rate": 0.00043123835443729117,
+        "loss": 1.0668,
+        "step": 20770
+      },
+      {
+        "epoch": 0.6626486813992793,
+        "grad_norm": 0.3084590435028076,
+        "learning_rate": 0.0004295914238317937,
+        "loss": 1.0578,
+        "step": 20780
+      },
+      {
+        "epoch": 0.6629675691189132,
+        "grad_norm": 0.3043973445892334,
+        "learning_rate": 0.00042795078297392985,
+        "loss": 1.0539,
+        "step": 20790
+      },
+      {
+        "epoch": 0.6632864568385471,
+        "grad_norm": 0.3038099706172943,
+        "learning_rate": 0.0004263164078426964,
+        "loss": 1.0719,
+        "step": 20800
+      },
+      {
+        "epoch": 0.6636053445581811,
+        "grad_norm": 0.3076210021972656,
+        "learning_rate": 0.0004246882745088281,
+        "loss": 1.0376,
+        "step": 20810
+      },
+      {
+        "epoch": 0.6639242322778149,
+        "grad_norm": 0.30958908796310425,
+        "learning_rate": 0.0004230663591344474,
+        "loss": 1.0645,
+        "step": 20820
+      },
+      {
+        "epoch": 0.6642431199974489,
+        "grad_norm": 0.3039079010486603,
+        "learning_rate": 0.00042145063797271515,
+        "loss": 1.0444,
+        "step": 20830
+      },
+      {
+        "epoch": 0.6645620077170828,
+        "grad_norm": 0.30851301550865173,
+        "learning_rate": 0.00041984108736748337,
+        "loss": 1.0453,
+        "step": 20840
+      },
+      {
+        "epoch": 0.6648808954367167,
+        "grad_norm": 0.300443172454834,
+        "learning_rate": 0.00041823768375294853,
+        "loss": 1.0507,
+        "step": 20850
+      },
+      {
+        "epoch": 0.6651997831563506,
+        "grad_norm": 0.30188149213790894,
+        "learning_rate": 0.0004166404036533064,
+        "loss": 1.0443,
+        "step": 20860
+      },
+      {
+        "epoch": 0.6655186708759846,
+        "grad_norm": 0.30435100197792053,
+        "learning_rate": 0.0004150492236824088,
+        "loss": 1.0586,
+        "step": 20870
+      },
+      {
+        "epoch": 0.6658375585956184,
+        "grad_norm": 0.31833893060684204,
+        "learning_rate": 0.000413464120543421,
+        "loss": 1.0498,
+        "step": 20880
+      },
+      {
+        "epoch": 0.6661564463152524,
+        "grad_norm": 0.3018447756767273,
+        "learning_rate": 0.00041188507102848026,
+        "loss": 1.0474,
+        "step": 20890
+      },
+      {
+        "epoch": 0.6664753340348863,
+        "grad_norm": 0.30486419796943665,
+        "learning_rate": 0.00041031205201835666,
+        "loss": 1.0525,
+        "step": 20900
+      },
+      {
+        "epoch": 0.6667942217545202,
+        "grad_norm": 0.3087126910686493,
+        "learning_rate": 0.0004087450404821141,
+        "loss": 1.0435,
+        "step": 20910
+      },
+      {
+        "epoch": 0.6671131094741541,
+        "grad_norm": 0.2946838140487671,
+        "learning_rate": 0.0004071840134767734,
+        "loss": 1.0272,
+        "step": 20920
+      },
+      {
+        "epoch": 0.6674319971937881,
+        "grad_norm": 0.28987836837768555,
+        "learning_rate": 0.0004056289481469763,
+        "loss": 1.0473,
+        "step": 20930
+      },
+      {
+        "epoch": 0.6677508849134219,
+        "grad_norm": 0.3040097951889038,
+        "learning_rate": 0.00040407982172465073,
+        "loss": 1.0351,
+        "step": 20940
+      },
+      {
+        "epoch": 0.6680697726330559,
+        "grad_norm": 0.31401920318603516,
+        "learning_rate": 0.00040253661152867774,
+        "loss": 1.0547,
+        "step": 20950
+      },
+      {
+        "epoch": 0.6683886603526898,
+        "grad_norm": 0.30797773599624634,
+        "learning_rate": 0.00040099929496455883,
+        "loss": 1.0548,
+        "step": 20960
+      },
+      {
+        "epoch": 0.6687075480723237,
+        "grad_norm": 0.2939365804195404,
+        "learning_rate": 0.0003994678495240859,
+        "loss": 1.0335,
+        "step": 20970
+      },
+      {
+        "epoch": 0.6690264357919576,
+        "grad_norm": 0.29866331815719604,
+        "learning_rate": 0.0003979422527850112,
+        "loss": 1.0369,
+        "step": 20980
+      },
+      {
+        "epoch": 0.6693453235115916,
+        "grad_norm": 0.30503347516059875,
+        "learning_rate": 0.00039642248241071923,
+        "loss": 1.0408,
+        "step": 20990
+      },
+      {
+        "epoch": 0.6696642112312254,
+        "grad_norm": 0.29835569858551025,
+        "learning_rate": 0.0003949085161498995,
+        "loss": 1.0557,
+        "step": 21000
+      },
+      {
+        "epoch": 0.6699830989508594,
+        "grad_norm": 0.2976466417312622,
+        "learning_rate": 0.00039340033183622096,
+        "loss": 1.0394,
+        "step": 21010
+      },
+      {
+        "epoch": 0.6703019866704933,
+        "grad_norm": 0.3077746629714966,
+        "learning_rate": 0.00039189790738800755,
+        "loss": 1.0403,
+        "step": 21020
+      },
+      {
+        "epoch": 0.6706208743901272,
+        "grad_norm": 0.32718950510025024,
+        "learning_rate": 0.00039040122080791445,
+        "loss": 1.0463,
+        "step": 21030
+      },
+      {
+        "epoch": 0.6709397621097611,
+        "grad_norm": 0.30107590556144714,
+        "learning_rate": 0.0003889102501826066,
+        "loss": 1.043,
+        "step": 21040
+      },
+      {
+        "epoch": 0.6712586498293951,
+        "grad_norm": 0.2878934144973755,
+        "learning_rate": 0.00038742497368243723,
+        "loss": 1.0468,
+        "step": 21050
+      },
+      {
+        "epoch": 0.6715775375490289,
+        "grad_norm": 0.3092693090438843,
+        "learning_rate": 0.00038594536956112894,
+        "loss": 1.0431,
+        "step": 21060
+      },
+      {
+        "epoch": 0.6718964252686629,
+        "grad_norm": 0.3057478964328766,
+        "learning_rate": 0.00038447141615545455,
+        "loss": 1.0429,
+        "step": 21070
+      },
+      {
+        "epoch": 0.6722153129882968,
+        "grad_norm": 0.3098829388618469,
+        "learning_rate": 0.00038300309188492066,
+        "loss": 1.0655,
+        "step": 21080
+      },
+      {
+        "epoch": 0.6725342007079307,
+        "grad_norm": 0.30573147535324097,
+        "learning_rate": 0.00038154037525145126,
+        "loss": 1.0372,
+        "step": 21090
+      },
+      {
+        "epoch": 0.6728530884275646,
+        "grad_norm": 0.29434239864349365,
+        "learning_rate": 0.0003800832448390728,
+        "loss": 1.0116,
+        "step": 21100
+      },
+      {
+        "epoch": 0.6731719761471986,
+        "grad_norm": 0.3012748658657074,
+        "learning_rate": 0.00037863167931360134,
+        "loss": 1.06,
+        "step": 21110
+      },
+      {
+        "epoch": 0.6734908638668324,
+        "grad_norm": 0.3058761656284332,
+        "learning_rate": 0.00037718565742232924,
+        "loss": 1.0392,
+        "step": 21120
+      },
+      {
+        "epoch": 0.6738097515864664,
+        "grad_norm": 0.30524468421936035,
+        "learning_rate": 0.0003757451579937149,
+        "loss": 1.024,
+        "step": 21130
+      },
+      {
+        "epoch": 0.6741286393061003,
+        "grad_norm": 0.3032868206501007,
+        "learning_rate": 0.00037431015993707214,
+        "loss": 1.0362,
+        "step": 21140
+      },
+      {
+        "epoch": 0.6744475270257342,
+        "grad_norm": 0.30639830231666565,
+        "learning_rate": 0.00037288064224226184,
+        "loss": 1.0327,
+        "step": 21150
+      },
+      {
+        "epoch": 0.6747664147453681,
+        "grad_norm": 0.29574814438819885,
+        "learning_rate": 0.00037145658397938383,
+        "loss": 1.0589,
+        "step": 21160
+      },
+      {
+        "epoch": 0.6750853024650021,
+        "grad_norm": 0.3002104163169861,
+        "learning_rate": 0.0003700379642984711,
+        "loss": 1.0235,
+        "step": 21170
+      },
+      {
+        "epoch": 0.675404190184636,
+        "grad_norm": 0.303792268037796,
+        "learning_rate": 0.0003686247624291839,
+        "loss": 1.0553,
+        "step": 21180
+      },
+      {
+        "epoch": 0.6757230779042699,
+        "grad_norm": 0.30552342534065247,
+        "learning_rate": 0.000367216957680506,
+        "loss": 1.0126,
+        "step": 21190
+      },
+      {
+        "epoch": 0.6760419656239038,
+        "grad_norm": 0.29988184571266174,
+        "learning_rate": 0.00036581452944044174,
+        "loss": 1.0194,
+        "step": 21200
+      },
+      {
+        "epoch": 0.6763608533435378,
+        "grad_norm": 0.29784929752349854,
+        "learning_rate": 0.0003644174571757142,
+        "loss": 1.0048,
+        "step": 21210
+      },
+      {
+        "epoch": 0.6766797410631716,
+        "grad_norm": 0.291316956281662,
+        "learning_rate": 0.00036302572043146427,
+        "loss": 1.0252,
+        "step": 21220
+      },
+      {
+        "epoch": 0.6769986287828056,
+        "grad_norm": 0.2924594283103943,
+        "learning_rate": 0.00036163929883095176,
+        "loss": 1.0144,
+        "step": 21230
+      },
+      {
+        "epoch": 0.6773175165024395,
+        "grad_norm": 0.30861392617225647,
+        "learning_rate": 0.0003602581720752565,
+        "loss": 1.0258,
+        "step": 21240
+      },
+      {
+        "epoch": 0.6776364042220734,
+        "grad_norm": 0.29646366834640503,
+        "learning_rate": 0.00035888231994298154,
+        "loss": 1.0436,
+        "step": 21250
+      },
+      {
+        "epoch": 0.6779552919417073,
+        "grad_norm": 0.299898236989975,
+        "learning_rate": 0.00035751172228995677,
+        "loss": 1.0235,
+        "step": 21260
+      },
+      {
+        "epoch": 0.6782741796613413,
+        "grad_norm": 0.3147020936012268,
+        "learning_rate": 0.00035614635904894417,
+        "loss": 1.0335,
+        "step": 21270
+      },
+      {
+        "epoch": 0.6785930673809751,
+        "grad_norm": 0.299283504486084,
+        "learning_rate": 0.000354786210229344,
+        "loss": 1.044,
+        "step": 21280
+      },
+      {
+        "epoch": 0.6789119551006091,
+        "grad_norm": 0.2966497242450714,
+        "learning_rate": 0.0003534312559169021,
+        "loss": 1.0162,
+        "step": 21290
+      },
+      {
+        "epoch": 0.679230842820243,
+        "grad_norm": 0.29665714502334595,
+        "learning_rate": 0.00035208147627341824,
+        "loss": 1.0322,
+        "step": 21300
+      },
+      {
+        "epoch": 0.6795497305398769,
+        "grad_norm": 0.29577651619911194,
+        "learning_rate": 0.0003507368515364557,
+        "loss": 1.0338,
+        "step": 21310
+      },
+      {
+        "epoch": 0.6798686182595108,
+        "grad_norm": 0.3065530061721802,
+        "learning_rate": 0.00034939736201905193,
+        "loss": 1.0233,
+        "step": 21320
+      },
+      {
+        "epoch": 0.6801875059791448,
+        "grad_norm": 0.29793888330459595,
+        "learning_rate": 0.0003480629881094304,
+        "loss": 1.0468,
+        "step": 21330
+      },
+      {
+        "epoch": 0.6805063936987786,
+        "grad_norm": 0.3028135895729065,
+        "learning_rate": 0.0003467337102707133,
+        "loss": 1.0133,
+        "step": 21340
+      },
+      {
+        "epoch": 0.6808252814184126,
+        "grad_norm": 0.3011240065097809,
+        "learning_rate": 0.0003454095090406356,
+        "loss": 1.0428,
+        "step": 21350
+      },
+      {
+        "epoch": 0.6811441691380465,
+        "grad_norm": 0.3000783622264862,
+        "learning_rate": 0.00034409036503126023,
+        "loss": 1.0193,
+        "step": 21360
+      },
+      {
+        "epoch": 0.6814630568576804,
+        "grad_norm": 0.3108273446559906,
+        "learning_rate": 0.00034277625892869374,
+        "loss": 1.0302,
+        "step": 21370
+      },
+      {
+        "epoch": 0.6817819445773143,
+        "grad_norm": 0.2920081615447998,
+        "learning_rate": 0.0003414671714928041,
+        "loss": 1.0516,
+        "step": 21380
+      },
+      {
+        "epoch": 0.6821008322969483,
+        "grad_norm": 0.30891862511634827,
+        "learning_rate": 0.00034016308355693865,
+        "loss": 1.0285,
+        "step": 21390
+      },
+      {
+        "epoch": 0.6824197200165821,
+        "grad_norm": 0.308363676071167,
+        "learning_rate": 0.0003388639760276436,
+        "loss": 1.03,
+        "step": 21400
+      },
+      {
+        "epoch": 0.6827386077362161,
+        "grad_norm": 0.2917180359363556,
+        "learning_rate": 0.0003375698298843843,
+        "loss": 1.0295,
+        "step": 21410
+      },
+      {
+        "epoch": 0.68305749545585,
+        "grad_norm": 0.2991756796836853,
+        "learning_rate": 0.00033628062617926724,
+        "loss": 1.0186,
+        "step": 21420
+      },
+      {
+        "epoch": 0.6833763831754839,
+        "grad_norm": 0.3052626848220825,
+        "learning_rate": 0.0003349963460367619,
+        "loss": 1.02,
+        "step": 21430
+      },
+      {
+        "epoch": 0.6836952708951178,
+        "grad_norm": 0.29897376894950867,
+        "learning_rate": 0.0003337169706534251,
+        "loss": 1.0281,
+        "step": 21440
+      },
+      {
+        "epoch": 0.6840141586147518,
+        "grad_norm": 0.29893383383750916,
+        "learning_rate": 0.00033244248129762514,
+        "loss": 1.0148,
+        "step": 21450
+      },
+      {
+        "epoch": 0.6843330463343856,
+        "grad_norm": 0.2953765392303467,
+        "learning_rate": 0.00033117285930926786,
+        "loss": 1.0206,
+        "step": 21460
+      },
+      {
+        "epoch": 0.6846519340540196,
+        "grad_norm": 0.3016309142112732,
+        "learning_rate": 0.0003299080860995236,
+        "loss": 1.0222,
+        "step": 21470
+      },
+      {
+        "epoch": 0.6849708217736535,
+        "grad_norm": 0.28592649102211,
+        "learning_rate": 0.00032864814315055425,
+        "loss": 1.0199,
+        "step": 21480
+      },
+      {
+        "epoch": 0.6852897094932874,
+        "grad_norm": 0.3022935688495636,
+        "learning_rate": 0.00032739301201524317,
+        "loss": 1.0286,
+        "step": 21490
+      },
+      {
+        "epoch": 0.6856085972129213,
+        "grad_norm": 0.30078959465026855,
+        "learning_rate": 0.0003261426743169244,
+        "loss": 1.0419,
+        "step": 21500
+      },
+      {
+        "epoch": 0.6859274849325553,
+        "grad_norm": 0.2940540611743927,
+        "learning_rate": 0.0003248971117491138,
+        "loss": 1.0184,
+        "step": 21510
+      },
+      {
+        "epoch": 0.6862463726521891,
+        "grad_norm": 0.2969309389591217,
+        "learning_rate": 0.00032365630607524107,
+        "loss": 1.0269,
+        "step": 21520
+      },
+      {
+        "epoch": 0.6865652603718231,
+        "grad_norm": 0.28925058245658875,
+        "learning_rate": 0.00032242023912838264,
+        "loss": 1.0162,
+        "step": 21530
+      },
+      {
+        "epoch": 0.686884148091457,
+        "grad_norm": 0.2998692989349365,
+        "learning_rate": 0.00032118889281099574,
+        "loss": 1.0226,
+        "step": 21540
+      },
+      {
+        "epoch": 0.6872030358110909,
+        "grad_norm": 0.2965996265411377,
+        "learning_rate": 0.0003199622490946535,
+        "loss": 1.0165,
+        "step": 21550
+      },
+      {
+        "epoch": 0.6875219235307248,
+        "grad_norm": 0.3009476363658905,
+        "learning_rate": 0.0003187402900197809,
+        "loss": 1.0253,
+        "step": 21560
+      },
+      {
+        "epoch": 0.6878408112503588,
+        "grad_norm": 0.30146294832229614,
+        "learning_rate": 0.0003175229976953918,
+        "loss": 1.0075,
+        "step": 21570
+      },
+      {
+        "epoch": 0.6881596989699926,
+        "grad_norm": 0.3076365292072296,
+        "learning_rate": 0.000316310354298827,
+        "loss": 0.9962,
+        "step": 21580
+      },
+      {
+        "epoch": 0.6884785866896266,
+        "grad_norm": 0.3112267255783081,
+        "learning_rate": 0.00031510234207549347,
+        "loss": 1.0198,
+        "step": 21590
+      },
+      {
+        "epoch": 0.6887974744092605,
+        "grad_norm": 0.29509472846984863,
+        "learning_rate": 0.0003138989433386042,
+        "loss": 1.0115,
+        "step": 21600
+      },
+      {
+        "epoch": 0.6891163621288944,
+        "grad_norm": 0.2932247221469879,
+        "learning_rate": 0.0003127001404689193,
+        "loss": 1.0168,
+        "step": 21610
+      },
+      {
+        "epoch": 0.6894352498485283,
+        "grad_norm": 0.29951944947242737,
+        "learning_rate": 0.0003115059159144878,
+        "loss": 1.0189,
+        "step": 21620
+      },
+      {
+        "epoch": 0.6897541375681623,
+        "grad_norm": 0.29388749599456787,
+        "learning_rate": 0.0003103162521903914,
+        "loss": 1.0033,
+        "step": 21630
+      },
+      {
+        "epoch": 0.6900730252877961,
+        "grad_norm": 0.3006175458431244,
+        "learning_rate": 0.00030913113187848755,
+        "loss": 1.0147,
+        "step": 21640
+      },
+      {
+        "epoch": 0.6903919130074301,
+        "grad_norm": 0.29276975989341736,
+        "learning_rate": 0.00030795053762715507,
+        "loss": 1.0185,
+        "step": 21650
+      },
+      {
+        "epoch": 0.690710800727064,
+        "grad_norm": 0.2958676517009735,
+        "learning_rate": 0.00030677445215103975,
+        "loss": 1.0211,
+        "step": 21660
+      },
+      {
+        "epoch": 0.6910296884466979,
+        "grad_norm": 0.30803918838500977,
+        "learning_rate": 0.00030560285823080156,
+        "loss": 1.008,
+        "step": 21670
+      },
+      {
+        "epoch": 0.6913485761663318,
+        "grad_norm": 0.30229854583740234,
+        "learning_rate": 0.00030443573871286224,
+        "loss": 1.0236,
+        "step": 21680
+      },
+      {
+        "epoch": 0.6916674638859658,
+        "grad_norm": 0.30095160007476807,
+        "learning_rate": 0.0003032730765091543,
+        "loss": 1.0158,
+        "step": 21690
+      },
+      {
+        "epoch": 0.6919863516055996,
+        "grad_norm": 0.295236736536026,
+        "learning_rate": 0.000302114854596871,
+        "loss": 1.0213,
+        "step": 21700
+      },
+      {
+        "epoch": 0.6923052393252336,
+        "grad_norm": 0.2985754609107971,
+        "learning_rate": 0.00030096105601821663,
+        "loss": 1.0131,
+        "step": 21710
+      },
+      {
+        "epoch": 0.6926241270448675,
+        "grad_norm": 0.30802398920059204,
+        "learning_rate": 0.0002998116638801587,
+        "loss": 1.009,
+        "step": 21720
+      },
+      {
+        "epoch": 0.6929430147645014,
+        "grad_norm": 0.3082250952720642,
+        "learning_rate": 0.0002986666613541806,
+        "loss": 1.0245,
+        "step": 21730
+      },
+      {
+        "epoch": 0.6932619024841353,
+        "grad_norm": 0.29745855927467346,
+        "learning_rate": 0.00029752603167603484,
+        "loss": 1.0063,
+        "step": 21740
+      },
+      {
+        "epoch": 0.6935807902037693,
+        "grad_norm": 0.29345378279685974,
+        "learning_rate": 0.00029638975814549775,
+        "loss": 1.0124,
+        "step": 21750
+      },
+      {
+        "epoch": 0.6938996779234031,
+        "grad_norm": 0.2981776297092438,
+        "learning_rate": 0.0002952578241261252,
+        "loss": 1.0184,
+        "step": 21760
+      },
+      {
+        "epoch": 0.6942185656430371,
+        "grad_norm": 0.30407872796058655,
+        "learning_rate": 0.00029413021304500875,
+        "loss": 1.0117,
+        "step": 21770
+      },
+      {
+        "epoch": 0.694537453362671,
+        "grad_norm": 0.29665490984916687,
+        "learning_rate": 0.0002930069083925332,
+        "loss": 1.0199,
+        "step": 21780
+      },
+      {
+        "epoch": 0.6948563410823049,
+        "grad_norm": 0.29667824506759644,
+        "learning_rate": 0.0002918878937221347,
+        "loss": 1.0138,
+        "step": 21790
+      },
+      {
+        "epoch": 0.6951752288019388,
+        "grad_norm": 0.2990490794181824,
+        "learning_rate": 0.00029077315265005994,
+        "loss": 1.0146,
+        "step": 21800
+      },
+      {
+        "epoch": 0.6954941165215728,
+        "grad_norm": 0.3064717948436737,
+        "learning_rate": 0.00028966266885512655,
+        "loss": 1.0067,
+        "step": 21810
+      },
+      {
+        "epoch": 0.6958130042412066,
+        "grad_norm": 0.3038541376590729,
+        "learning_rate": 0.0002885564260784838,
+        "loss": 1.0198,
+        "step": 21820
+      },
+      {
+        "epoch": 0.6961318919608406,
+        "grad_norm": 0.30450543761253357,
+        "learning_rate": 0.00028745440812337464,
+        "loss": 0.9972,
+        "step": 21830
+      },
+      {
+        "epoch": 0.6964507796804745,
+        "grad_norm": 0.2951158285140991,
+        "learning_rate": 0.00028635659885489876,
+        "loss": 0.9993,
+        "step": 21840
+      },
+      {
+        "epoch": 0.6967696674001084,
+        "grad_norm": 0.30965596437454224,
+        "learning_rate": 0.00028526298219977615,
+        "loss": 1.0153,
+        "step": 21850
+      },
+      {
+        "epoch": 0.6970885551197423,
+        "grad_norm": 0.3073263168334961,
+        "learning_rate": 0.0002841735421461118,
+        "loss": 1.0039,
+        "step": 21860
+      },
+      {
+        "epoch": 0.6974074428393763,
+        "grad_norm": 0.2883033752441406,
+        "learning_rate": 0.00028308826274316135,
+        "loss": 1.0199,
+        "step": 21870
+      },
+      {
+        "epoch": 0.6977263305590101,
+        "grad_norm": 0.3025035262107849,
+        "learning_rate": 0.00028200712810109736,
+        "loss": 1.0074,
+        "step": 21880
+      },
+      {
+        "epoch": 0.6980452182786441,
+        "grad_norm": 0.29780346155166626,
+        "learning_rate": 0.00028093012239077697,
+        "loss": 1.0152,
+        "step": 21890
+      },
+      {
+        "epoch": 0.698364105998278,
+        "grad_norm": 0.3000519871711731,
+        "learning_rate": 0.0002798572298435098,
+        "loss": 1.0103,
+        "step": 21900
+      },
+      {
+        "epoch": 0.6986829937179119,
+        "grad_norm": 0.28480494022369385,
+        "learning_rate": 0.00027878843475082743,
+        "loss": 0.9929,
+        "step": 21910
+      },
+      {
+        "epoch": 0.6990018814375458,
+        "grad_norm": 0.29734286665916443,
+        "learning_rate": 0.00027772372146425304,
+        "loss": 0.9953,
+        "step": 21920
+      },
+      {
+        "epoch": 0.6993207691571798,
+        "grad_norm": 0.3029066026210785,
+        "learning_rate": 0.00027666307439507253,
+        "loss": 1.008,
+        "step": 21930
+      },
+      {
+        "epoch": 0.6996396568768136,
+        "grad_norm": 0.3026646673679352,
+        "learning_rate": 0.0002756064780141064,
+        "loss": 0.9897,
+        "step": 21940
+      },
+      {
+        "epoch": 0.6999585445964476,
+        "grad_norm": 0.2930282652378082,
+        "learning_rate": 0.00027455391685148186,
+        "loss": 0.9982,
+        "step": 21950
+      },
+      {
+        "epoch": 0.7002774323160815,
+        "grad_norm": 0.30149659514427185,
+        "learning_rate": 0.0002735053754964071,
+        "loss": 1.0087,
+        "step": 21960
+      },
+      {
+        "epoch": 0.7005963200357154,
+        "grad_norm": 0.2959401309490204,
+        "learning_rate": 0.00027246083859694485,
+        "loss": 0.9837,
+        "step": 21970
+      },
+      {
+        "epoch": 0.7009152077553493,
+        "grad_norm": 0.29587122797966003,
+        "learning_rate": 0.0002714202908597884,
+        "loss": 1.0099,
+        "step": 21980
+      },
+      {
+        "epoch": 0.7012340954749833,
+        "grad_norm": 0.2973346412181854,
+        "learning_rate": 0.00027038371705003693,
+        "loss": 1.0117,
+        "step": 21990
+      },
+      {
+        "epoch": 0.7015529831946172,
+        "grad_norm": 0.30666384100914,
+        "learning_rate": 0.0002693511019909731,
+        "loss": 1.0027,
+        "step": 22000
+      },
+      {
+        "epoch": 0.7018718709142511,
+        "grad_norm": 0.2986239194869995,
+        "learning_rate": 0.0002683224305638403,
+        "loss": 1.0127,
+        "step": 22010
+      },
+      {
+        "epoch": 0.702190758633885,
+        "grad_norm": 0.30097588896751404,
+        "learning_rate": 0.0002672976877076218,
+        "loss": 0.9949,
+        "step": 22020
+      },
+      {
+        "epoch": 0.702509646353519,
+        "grad_norm": 0.3022991120815277,
+        "learning_rate": 0.0002662768584188198,
+        "loss": 1.009,
+        "step": 22030
+      },
+      {
+        "epoch": 0.7028285340731528,
+        "grad_norm": 0.29063522815704346,
+        "learning_rate": 0.0002652599277512359,
+        "loss": 0.9988,
+        "step": 22040
+      },
+      {
+        "epoch": 0.7031474217927868,
+        "grad_norm": 0.3014470636844635,
+        "learning_rate": 0.00026424688081575263,
+        "loss": 1.0068,
+        "step": 22050
+      },
+      {
+        "epoch": 0.7034663095124207,
+        "grad_norm": 0.2925295829772949,
+        "learning_rate": 0.0002632377027801149,
+        "loss": 0.9807,
+        "step": 22060
+      },
+      {
+        "epoch": 0.7037851972320546,
+        "grad_norm": 0.3046445846557617,
+        "learning_rate": 0.000262232378868713,
+        "loss": 1.0161,
+        "step": 22070
+      },
+      {
+        "epoch": 0.7041040849516885,
+        "grad_norm": 0.30122342705726624,
+        "learning_rate": 0.00026123089436236655,
+        "loss": 1.0032,
+        "step": 22080
+      },
+      {
+        "epoch": 0.7044229726713225,
+        "grad_norm": 0.29070591926574707,
+        "learning_rate": 0.0002602332345981087,
+        "loss": 0.9921,
+        "step": 22090
+      },
+      {
+        "epoch": 0.7047418603909563,
+        "grad_norm": 0.29186105728149414,
+        "learning_rate": 0.00025923938496897155,
+        "loss": 1.0109,
+        "step": 22100
+      },
+      {
+        "epoch": 0.7050607481105903,
+        "grad_norm": 0.3077830970287323,
+        "learning_rate": 0.0002582493309237722,
+        "loss": 0.9994,
+        "step": 22110
+      },
+      {
+        "epoch": 0.7053796358302242,
+        "grad_norm": 0.2919236719608307,
+        "learning_rate": 0.0002572630579669,
+        "loss": 0.9873,
+        "step": 22120
+      },
+      {
+        "epoch": 0.7056985235498581,
+        "grad_norm": 0.2887571454048157,
+        "learning_rate": 0.0002562805516581037,
+        "loss": 1.0074,
+        "step": 22130
+      },
+      {
+        "epoch": 0.706017411269492,
+        "grad_norm": 0.28118544816970825,
+        "learning_rate": 0.00025530179761228084,
+        "loss": 0.9953,
+        "step": 22140
+      },
+      {
+        "epoch": 0.706336298989126,
+        "grad_norm": 0.2915787100791931,
+        "learning_rate": 0.00025432678149926643,
+        "loss": 0.9959,
+        "step": 22150
+      },
+      {
+        "epoch": 0.7066551867087598,
+        "grad_norm": 0.2891466021537781,
+        "learning_rate": 0.00025335548904362355,
+        "loss": 1.01,
+        "step": 22160
+      },
+      {
+        "epoch": 0.7069740744283938,
+        "grad_norm": 0.29659250378608704,
+        "learning_rate": 0.0002523879060244341,
+        "loss": 0.9755,
+        "step": 22170
+      },
+      {
+        "epoch": 0.7072929621480277,
+        "grad_norm": 0.30222389101982117,
+        "learning_rate": 0.00025142401827509084,
+        "loss": 0.9955,
+        "step": 22180
+      },
+      {
+        "epoch": 0.7076118498676616,
+        "grad_norm": 0.2955167293548584,
+        "learning_rate": 0.00025046381168308975,
+        "loss": 1.005,
+        "step": 22190
+      },
+      {
+        "epoch": 0.7079307375872955,
+        "grad_norm": 0.2983476221561432,
+        "learning_rate": 0.00024950727218982346,
+        "loss": 0.9933,
+        "step": 22200
+      },
+      {
+        "epoch": 0.7082496253069295,
+        "grad_norm": 0.29820889234542847,
+        "learning_rate": 0.00024855438579037556,
+        "loss": 0.9969,
+        "step": 22210
+      },
+      {
+        "epoch": 0.7085685130265633,
+        "grad_norm": 0.2924354076385498,
+        "learning_rate": 0.0002476051385333152,
+        "loss": 0.9908,
+        "step": 22220
+      },
+      {
+        "epoch": 0.7088874007461973,
+        "grad_norm": 0.28841105103492737,
+        "learning_rate": 0.0002466595165204934,
+        "loss": 1.0096,
+        "step": 22230
+      },
+      {
+        "epoch": 0.7092062884658312,
+        "grad_norm": 0.30202123522758484,
+        "learning_rate": 0.000245717505906839,
+        "loss": 1.0128,
+        "step": 22240
+      },
+      {
+        "epoch": 0.7095251761854651,
+        "grad_norm": 0.28892645239830017,
+        "learning_rate": 0.00024477909290015614,
+        "loss": 0.9837,
+        "step": 22250
+      },
+      {
+        "epoch": 0.709844063905099,
+        "grad_norm": 0.2854585349559784,
+        "learning_rate": 0.00024384426376092245,
+        "loss": 0.987,
+        "step": 22260
+      },
+      {
+        "epoch": 0.710162951624733,
+        "grad_norm": 0.2966926395893097,
+        "learning_rate": 0.00024291300480208775,
+        "loss": 0.9888,
+        "step": 22270
+      },
+      {
+        "epoch": 0.7104818393443668,
+        "grad_norm": 0.2984680235385895,
+        "learning_rate": 0.00024198530238887377,
+        "loss": 0.9936,
+        "step": 22280
+      },
+      {
+        "epoch": 0.7108007270640008,
+        "grad_norm": 0.29596757888793945,
+        "learning_rate": 0.00024106114293857438,
+        "loss": 1.0019,
+        "step": 22290
+      },
+      {
+        "epoch": 0.7111196147836347,
+        "grad_norm": 0.2984013557434082,
+        "learning_rate": 0.00024014051292035676,
+        "loss": 1.0001,
+        "step": 22300
+      },
+      {
+        "epoch": 0.7114385025032686,
+        "grad_norm": 0.2908388376235962,
+        "learning_rate": 0.00023922339885506342,
+        "loss": 0.996,
+        "step": 22310
+      },
+      {
+        "epoch": 0.7117573902229025,
+        "grad_norm": 0.2907799780368805,
+        "learning_rate": 0.00023830978731501472,
+        "loss": 0.997,
+        "step": 22320
+      },
+      {
+        "epoch": 0.7120762779425365,
+        "grad_norm": 0.30171695351600647,
+        "learning_rate": 0.00023739966492381225,
+        "loss": 0.9943,
+        "step": 22330
+      },
+      {
+        "epoch": 0.7123951656621703,
+        "grad_norm": 0.2960367500782013,
+        "learning_rate": 0.0002364930183561431,
+        "loss": 0.9891,
+        "step": 22340
+      },
+      {
+        "epoch": 0.7127140533818043,
+        "grad_norm": 0.2947923243045807,
+        "learning_rate": 0.00023558983433758467,
+        "loss": 0.9766,
+        "step": 22350
+      },
+      {
+        "epoch": 0.7130329411014382,
+        "grad_norm": 0.28949424624443054,
+        "learning_rate": 0.0002346900996444103,
+        "loss": 0.9916,
+        "step": 22360
+      },
+      {
+        "epoch": 0.7133518288210721,
+        "grad_norm": 0.29011088609695435,
+        "learning_rate": 0.00023379380110339585,
+        "loss": 0.9879,
+        "step": 22370
+      },
+      {
+        "epoch": 0.713670716540706,
+        "grad_norm": 0.28879514336586,
+        "learning_rate": 0.00023290092559162653,
+        "loss": 0.9758,
+        "step": 22380
+      },
+      {
+        "epoch": 0.71398960426034,
+        "grad_norm": 0.288805216550827,
+        "learning_rate": 0.0002320114600363049,
+        "loss": 0.9858,
+        "step": 22390
+      },
+      {
+        "epoch": 0.7143084919799738,
+        "grad_norm": 0.2958844006061554,
+        "learning_rate": 0.0002311253914145597,
+        "loss": 0.9706,
+        "step": 22400
+      },
+      {
+        "epoch": 0.7146273796996078,
+        "grad_norm": 0.3001909852027893,
+        "learning_rate": 0.00023024270675325468,
+        "loss": 1.0048,
+        "step": 22410
+      },
+      {
+        "epoch": 0.7149462674192417,
+        "grad_norm": 0.29309847950935364,
+        "learning_rate": 0.0002293633931287991,
+        "loss": 0.9978,
+        "step": 22420
+      },
+      {
+        "epoch": 0.7152651551388756,
+        "grad_norm": 0.28979870676994324,
+        "learning_rate": 0.00022848743766695835,
+        "loss": 0.9768,
+        "step": 22430
+      },
+      {
+        "epoch": 0.7155840428585095,
+        "grad_norm": 0.2978128492832184,
+        "learning_rate": 0.00022761482754266545,
+        "loss": 0.981,
+        "step": 22440
+      },
+      {
+        "epoch": 0.7159029305781435,
+        "grad_norm": 0.29963478446006775,
+        "learning_rate": 0.00022674554997983317,
+        "loss": 0.9862,
+        "step": 22450
+      },
+      {
+        "epoch": 0.7162218182977773,
+        "grad_norm": 0.2936873733997345,
+        "learning_rate": 0.0002258795922511673,
+        "loss": 0.9954,
+        "step": 22460
+      },
+      {
+        "epoch": 0.7165407060174113,
+        "grad_norm": 0.2975631356239319,
+        "learning_rate": 0.00022501694167797987,
+        "loss": 0.983,
+        "step": 22470
+      },
+      {
+        "epoch": 0.7168595937370452,
+        "grad_norm": 0.2997070550918579,
+        "learning_rate": 0.00022415758563000388,
+        "loss": 0.9788,
+        "step": 22480
+      },
+      {
+        "epoch": 0.7171784814566791,
+        "grad_norm": 0.2969716787338257,
+        "learning_rate": 0.0002233015115252082,
+        "loss": 0.9866,
+        "step": 22490
+      },
+      {
+        "epoch": 0.717497369176313,
+        "grad_norm": 0.29699286818504333,
+        "learning_rate": 0.00022244870682961338,
+        "loss": 0.99,
+        "step": 22500
+      },
+      {
+        "epoch": 0.717816256895947,
+        "grad_norm": 0.3004414737224579,
+        "learning_rate": 0.00022159915905710817,
+        "loss": 0.9883,
+        "step": 22510
+      },
+      {
+        "epoch": 0.7181351446155808,
+        "grad_norm": 0.29157665371894836,
+        "learning_rate": 0.00022075285576926662,
+        "loss": 0.9781,
+        "step": 22520
+      },
+      {
+        "epoch": 0.7184540323352148,
+        "grad_norm": 0.3094356656074524,
+        "learning_rate": 0.00021990978457516616,
+        "loss": 0.9807,
+        "step": 22530
+      },
+      {
+        "epoch": 0.7187729200548487,
+        "grad_norm": 0.29095223546028137,
+        "learning_rate": 0.00021906993313120594,
+        "loss": 0.9825,
+        "step": 22540
+      },
+      {
+        "epoch": 0.7190918077744826,
+        "grad_norm": 0.29207363724708557,
+        "learning_rate": 0.00021823328914092629,
+        "loss": 0.9792,
+        "step": 22550
+      },
+      {
+        "epoch": 0.7194106954941165,
+        "grad_norm": 0.2970081865787506,
+        "learning_rate": 0.00021739984035482846,
+        "loss": 0.9807,
+        "step": 22560
+      },
+      {
+        "epoch": 0.7197295832137505,
+        "grad_norm": 0.2927705943584442,
+        "learning_rate": 0.00021656957457019568,
+        "loss": 0.9855,
+        "step": 22570
+      },
+      {
+        "epoch": 0.7200484709333843,
+        "grad_norm": 0.29955190420150757,
+        "learning_rate": 0.00021574247963091408,
+        "loss": 0.9804,
+        "step": 22580
+      },
+      {
+        "epoch": 0.7203673586530183,
+        "grad_norm": 0.3011538088321686,
+        "learning_rate": 0.00021491854342729487,
+        "loss": 0.9766,
+        "step": 22590
+      },
+      {
+        "epoch": 0.7206862463726522,
+        "grad_norm": 0.2975776195526123,
+        "learning_rate": 0.00021409775389589705,
+        "loss": 0.9849,
+        "step": 22600
+      },
+      {
+        "epoch": 0.7210051340922861,
+        "grad_norm": 0.29041531682014465,
+        "learning_rate": 0.00021328009901935087,
+        "loss": 0.9671,
+        "step": 22610
+      },
+      {
+        "epoch": 0.72132402181192,
+        "grad_norm": 0.2931303083896637,
+        "learning_rate": 0.00021246556682618162,
+        "loss": 0.9847,
+        "step": 22620
+      },
+      {
+        "epoch": 0.721642909531554,
+        "grad_norm": 0.2981482744216919,
+        "learning_rate": 0.0002116541453906347,
+        "loss": 0.9809,
+        "step": 22630
+      },
+      {
+        "epoch": 0.7219617972511878,
+        "grad_norm": 0.28368112444877625,
+        "learning_rate": 0.0002108458228325007,
+        "loss": 0.9686,
+        "step": 22640
+      },
+      {
+        "epoch": 0.7222806849708218,
+        "grad_norm": 0.2948022782802582,
+        "learning_rate": 0.0002100405873169417,
+        "loss": 0.9818,
+        "step": 22650
+      },
+      {
+        "epoch": 0.7225995726904557,
+        "grad_norm": 0.3083368241786957,
+        "learning_rate": 0.0002092384270543178,
+        "loss": 0.9774,
+        "step": 22660
+      },
+      {
+        "epoch": 0.7229184604100896,
+        "grad_norm": 0.2966402769088745,
+        "learning_rate": 0.00020843933030001454,
+        "loss": 0.9715,
+        "step": 22670
+      },
+      {
+        "epoch": 0.7232373481297235,
+        "grad_norm": 0.2903192937374115,
+        "learning_rate": 0.00020764328535427122,
+        "loss": 0.977,
+        "step": 22680
+      },
+      {
+        "epoch": 0.7235562358493575,
+        "grad_norm": 0.2952418327331543,
+        "learning_rate": 0.0002068502805620091,
+        "loss": 0.9908,
+        "step": 22690
+      },
+      {
+        "epoch": 0.7238751235689913,
+        "grad_norm": 0.29269060492515564,
+        "learning_rate": 0.0002060603043126613,
+        "loss": 0.9872,
+        "step": 22700
+      },
+      {
+        "epoch": 0.7241940112886253,
+        "grad_norm": 0.30094051361083984,
+        "learning_rate": 0.00020527334504000228,
+        "loss": 0.9802,
+        "step": 22710
+      },
+      {
+        "epoch": 0.7245128990082592,
+        "grad_norm": 0.2994566261768341,
+        "learning_rate": 0.00020448939122197894,
+        "loss": 0.9669,
+        "step": 22720
+      },
+      {
+        "epoch": 0.7248317867278931,
+        "grad_norm": 0.2904956042766571,
+        "learning_rate": 0.00020370843138054165,
+        "loss": 0.9748,
+        "step": 22730
+      },
+      {
+        "epoch": 0.725150674447527,
+        "grad_norm": 0.308773010969162,
+        "learning_rate": 0.00020293045408147637,
+        "loss": 0.9635,
+        "step": 22740
+      },
+      {
+        "epoch": 0.725469562167161,
+        "grad_norm": 0.3033744990825653,
+        "learning_rate": 0.00020215544793423702,
+        "loss": 0.9724,
+        "step": 22750
+      },
+      {
+        "epoch": 0.7257884498867948,
+        "grad_norm": 0.30875164270401,
+        "learning_rate": 0.00020138340159177898,
+        "loss": 0.9818,
+        "step": 22760
+      },
+      {
+        "epoch": 0.7261073376064288,
+        "grad_norm": 0.3124573230743408,
+        "learning_rate": 0.00020061430375039268,
+        "loss": 0.9786,
+        "step": 22770
+      },
+      {
+        "epoch": 0.7264262253260627,
+        "grad_norm": 0.29787442088127136,
+        "learning_rate": 0.00019984814314953845,
+        "loss": 0.9688,
+        "step": 22780
+      },
+      {
+        "epoch": 0.7267451130456966,
+        "grad_norm": 0.29551681876182556,
+        "learning_rate": 0.0001990849085716812,
+        "loss": 0.9926,
+        "step": 22790
+      },
+      {
+        "epoch": 0.7270640007653305,
+        "grad_norm": 0.2911129593849182,
+        "learning_rate": 0.0001983245888421265,
+        "loss": 0.972,
+        "step": 22800
+      },
+      {
+        "epoch": 0.7273828884849645,
+        "grad_norm": 0.2950027287006378,
+        "learning_rate": 0.000197567172828857,
+        "loss": 0.9871,
+        "step": 22810
+      },
+      {
+        "epoch": 0.7277017762045984,
+        "grad_norm": 0.30667251348495483,
+        "learning_rate": 0.0001968126494423692,
+        "loss": 0.9906,
+        "step": 22820
+      },
+      {
+        "epoch": 0.7280206639242323,
+        "grad_norm": 0.29852592945098877,
+        "learning_rate": 0.0001960610076355113,
+        "loss": 0.9819,
+        "step": 22830
+      },
+      {
+        "epoch": 0.7283395516438662,
+        "grad_norm": 0.3074007034301758,
+        "learning_rate": 0.00019531223640332132,
+        "loss": 0.9905,
+        "step": 22840
+      },
+      {
+        "epoch": 0.7286584393635002,
+        "grad_norm": 0.2953300178050995,
+        "learning_rate": 0.00019456632478286607,
+        "loss": 0.9926,
+        "step": 22850
+      },
+      {
+        "epoch": 0.728977327083134,
+        "grad_norm": 0.29797452688217163,
+        "learning_rate": 0.00019382326185308054,
+        "loss": 0.9698,
+        "step": 22860
+      },
+      {
+        "epoch": 0.729296214802768,
+        "grad_norm": 0.2951212525367737,
+        "learning_rate": 0.00019308303673460815,
+        "loss": 0.9786,
+        "step": 22870
+      },
+      {
+        "epoch": 0.7296151025224019,
+        "grad_norm": 0.3035532236099243,
+        "learning_rate": 0.0001923456385896413,
+        "loss": 0.9626,
+        "step": 22880
+      },
+      {
+        "epoch": 0.7299339902420358,
+        "grad_norm": 0.3000469207763672,
+        "learning_rate": 0.0001916110566217628,
+        "loss": 0.971,
+        "step": 22890
+      },
+      {
+        "epoch": 0.7302528779616697,
+        "grad_norm": 0.29106569290161133,
+        "learning_rate": 0.00019087928007578776,
+        "loss": 0.987,
+        "step": 22900
+      },
+      {
+        "epoch": 0.7305717656813037,
+        "grad_norm": 0.28790828585624695,
+        "learning_rate": 0.0001901502982376061,
+        "loss": 0.9702,
+        "step": 22910
+      },
+      {
+        "epoch": 0.7308906534009375,
+        "grad_norm": 0.2918194532394409,
+        "learning_rate": 0.00018942410043402573,
+        "loss": 0.9686,
+        "step": 22920
+      },
+      {
+        "epoch": 0.7312095411205715,
+        "grad_norm": 0.299245148897171,
+        "learning_rate": 0.0001887006760326162,
+        "loss": 0.984,
+        "step": 22930
+      },
+      {
+        "epoch": 0.7315284288402054,
+        "grad_norm": 0.29804402589797974,
+        "learning_rate": 0.00018798001444155318,
+        "loss": 0.9651,
+        "step": 22940
+      },
+      {
+        "epoch": 0.7318473165598393,
+        "grad_norm": 0.29632264375686646,
+        "learning_rate": 0.00018726210510946323,
+        "loss": 0.967,
+        "step": 22950
+      },
+      {
+        "epoch": 0.7321662042794732,
+        "grad_norm": 0.29573965072631836,
+        "learning_rate": 0.00018654693752526939,
+        "loss": 0.9698,
+        "step": 22960
+      },
+      {
+        "epoch": 0.7324850919991072,
+        "grad_norm": 0.29792520403862,
+        "learning_rate": 0.00018583450121803722,
+        "loss": 0.9609,
+        "step": 22970
+      },
+      {
+        "epoch": 0.732803979718741,
+        "grad_norm": 0.2980731427669525,
+        "learning_rate": 0.00018512478575682153,
+        "loss": 0.9676,
+        "step": 22980
+      },
+      {
+        "epoch": 0.733122867438375,
+        "grad_norm": 0.2982822060585022,
+        "learning_rate": 0.00018441778075051386,
+        "loss": 0.9586,
+        "step": 22990
+      },
+      {
+        "epoch": 0.7334417551580089,
+        "grad_norm": 0.31218892335891724,
+        "learning_rate": 0.00018371347584768985,
+        "loss": 0.9756,
+        "step": 23000
+      },
+      {
+        "epoch": 0.7337606428776428,
+        "grad_norm": 0.2934929430484772,
+        "learning_rate": 0.0001830118607364582,
+        "loss": 0.9658,
+        "step": 23010
+      },
+      {
+        "epoch": 0.7340795305972767,
+        "grad_norm": 0.3041274845600128,
+        "learning_rate": 0.00018231292514430937,
+        "loss": 0.9774,
+        "step": 23020
+      },
+      {
+        "epoch": 0.7343984183169107,
+        "grad_norm": 0.29541394114494324,
+        "learning_rate": 0.00018161665883796536,
+        "loss": 0.9876,
+        "step": 23030
+      },
+      {
+        "epoch": 0.7347173060365445,
+        "grad_norm": 0.28990721702575684,
+        "learning_rate": 0.0001809230516232297,
+        "loss": 0.9596,
+        "step": 23040
+      },
+      {
+        "epoch": 0.7350361937561785,
+        "grad_norm": 0.29748308658599854,
+        "learning_rate": 0.0001802320933448384,
+        "loss": 0.9863,
+        "step": 23050
+      },
+      {
+        "epoch": 0.7353550814758124,
+        "grad_norm": 0.2978643476963043,
+        "learning_rate": 0.0001795437738863111,
+        "loss": 0.9724,
+        "step": 23060
+      },
+      {
+        "epoch": 0.7356739691954463,
+        "grad_norm": 0.3006666898727417,
+        "learning_rate": 0.000178858083169803,
+        "loss": 0.9686,
+        "step": 23070
+      },
+      {
+        "epoch": 0.7359928569150802,
+        "grad_norm": 0.2904208302497864,
+        "learning_rate": 0.00017817501115595735,
+        "loss": 0.9763,
+        "step": 23080
+      },
+      {
+        "epoch": 0.7363117446347142,
+        "grad_norm": 0.2943444848060608,
+        "learning_rate": 0.0001774945478437585,
+        "loss": 0.9506,
+        "step": 23090
+      },
+      {
+        "epoch": 0.736630632354348,
+        "grad_norm": 0.2955239415168762,
+        "learning_rate": 0.00017681668327038525,
+        "loss": 0.9692,
+        "step": 23100
+      },
+      {
+        "epoch": 0.736949520073982,
+        "grad_norm": 0.29572662711143494,
+        "learning_rate": 0.00017614140751106525,
+        "loss": 0.954,
+        "step": 23110
+      },
+      {
+        "epoch": 0.7372684077936159,
+        "grad_norm": 0.298018217086792,
+        "learning_rate": 0.00017546871067892955,
+        "loss": 0.9701,
+        "step": 23120
+      },
+      {
+        "epoch": 0.7375872955132498,
+        "grad_norm": 0.30332425236701965,
+        "learning_rate": 0.00017479858292486792,
+        "loss": 0.9623,
+        "step": 23130
+      },
+      {
+        "epoch": 0.7379061832328837,
+        "grad_norm": 0.3037513196468353,
+        "learning_rate": 0.0001741310144373845,
+        "loss": 0.9921,
+        "step": 23140
+      },
+      {
+        "epoch": 0.7382250709525177,
+        "grad_norm": 0.2905147969722748,
+        "learning_rate": 0.00017346599544245428,
+        "loss": 0.983,
+        "step": 23150
+      },
+      {
+        "epoch": 0.7385439586721515,
+        "grad_norm": 0.2991820275783539,
+        "learning_rate": 0.00017280351620338,
+        "loss": 0.9676,
+        "step": 23160
+      },
+      {
+        "epoch": 0.7388628463917855,
+        "grad_norm": 0.29384827613830566,
+        "learning_rate": 0.00017214356702064952,
+        "loss": 0.9778,
+        "step": 23170
+      },
+      {
+        "epoch": 0.7391817341114194,
+        "grad_norm": 0.29975441098213196,
+        "learning_rate": 0.00017148613823179387,
+        "loss": 0.9632,
+        "step": 23180
+      },
+      {
+        "epoch": 0.7395006218310533,
+        "grad_norm": 0.2996976375579834,
+        "learning_rate": 0.00017083122021124575,
+        "loss": 0.9638,
+        "step": 23190
+      },
+      {
+        "epoch": 0.7398195095506872,
+        "grad_norm": 0.29453301429748535,
+        "learning_rate": 0.00017017880337019854,
+        "loss": 0.9618,
+        "step": 23200
+      },
+      {
+        "epoch": 0.7401383972703212,
+        "grad_norm": 0.29755935072898865,
+        "learning_rate": 0.00016952887815646604,
+        "loss": 0.9716,
+        "step": 23210
+      },
+      {
+        "epoch": 0.740457284989955,
+        "grad_norm": 0.2969668209552765,
+        "learning_rate": 0.0001688814350543425,
+        "loss": 0.9548,
+        "step": 23220
+      },
+      {
+        "epoch": 0.740776172709589,
+        "grad_norm": 0.2998278737068176,
+        "learning_rate": 0.00016823646458446337,
+        "loss": 0.9516,
+        "step": 23230
+      },
+      {
+        "epoch": 0.7410950604292229,
+        "grad_norm": 0.30308350920677185,
+        "learning_rate": 0.00016759395730366647,
+        "loss": 0.9581,
+        "step": 23240
+      },
+      {
+        "epoch": 0.7414139481488567,
+        "grad_norm": 0.30527162551879883,
+        "learning_rate": 0.00016695390380485372,
+        "loss": 0.9601,
+        "step": 23250
+      },
+      {
+        "epoch": 0.7417328358684907,
+        "grad_norm": 0.2971056401729584,
+        "learning_rate": 0.00016631629471685345,
+        "loss": 0.961,
+        "step": 23260
+      },
+      {
+        "epoch": 0.7420517235881247,
+        "grad_norm": 0.29665809869766235,
+        "learning_rate": 0.0001656811207042832,
+        "loss": 0.9591,
+        "step": 23270
+      },
+      {
+        "epoch": 0.7423706113077585,
+        "grad_norm": 0.30231061577796936,
+        "learning_rate": 0.00016504837246741303,
+        "loss": 0.9541,
+        "step": 23280
+      },
+      {
+        "epoch": 0.7426894990273925,
+        "grad_norm": 0.29970023036003113,
+        "learning_rate": 0.00016441804074202924,
+        "loss": 0.9554,
+        "step": 23290
+      },
+      {
+        "epoch": 0.7430083867470264,
+        "grad_norm": 0.3048689663410187,
+        "learning_rate": 0.00016379011629929894,
+        "loss": 0.9787,
+        "step": 23300
+      },
+      {
+        "epoch": 0.7433272744666602,
+        "grad_norm": 0.2947291433811188,
+        "learning_rate": 0.00016316458994563475,
+        "loss": 0.9644,
+        "step": 23310
+      },
+      {
+        "epoch": 0.7436461621862942,
+        "grad_norm": 0.30350422859191895,
+        "learning_rate": 0.00016254145252256035,
+        "loss": 0.9737,
+        "step": 23320
+      },
+      {
+        "epoch": 0.7439650499059282,
+        "grad_norm": 0.29172733426094055,
+        "learning_rate": 0.0001619206949065762,
+        "loss": 0.9717,
+        "step": 23330
+      },
+      {
+        "epoch": 0.744283937625562,
+        "grad_norm": 0.29668304324150085,
+        "learning_rate": 0.00016130230800902616,
+        "loss": 0.9663,
+        "step": 23340
+      },
+      {
+        "epoch": 0.744602825345196,
+        "grad_norm": 0.2947177588939667,
+        "learning_rate": 0.00016068628277596432,
+        "loss": 0.9654,
+        "step": 23350
+      },
+      {
+        "epoch": 0.7449217130648299,
+        "grad_norm": 0.29447776079177856,
+        "learning_rate": 0.00016007261018802238,
+        "loss": 0.9474,
+        "step": 23360
+      },
+      {
+        "epoch": 0.7452406007844637,
+        "grad_norm": 0.28898507356643677,
+        "learning_rate": 0.00015946128126027772,
+        "loss": 0.9569,
+        "step": 23370
+      },
+      {
+        "epoch": 0.7455594885040977,
+        "grad_norm": 0.30178964138031006,
+        "learning_rate": 0.0001588522870421218,
+        "loss": 0.9586,
+        "step": 23380
+      },
+      {
+        "epoch": 0.7458783762237317,
+        "grad_norm": 0.313784122467041,
+        "learning_rate": 0.00015824561861712907,
+        "loss": 0.9641,
+        "step": 23390
+      },
+      {
+        "epoch": 0.7461972639433655,
+        "grad_norm": 0.3013688325881958,
+        "learning_rate": 0.00015764126710292645,
+        "loss": 0.9634,
+        "step": 23400
+      },
+      {
+        "epoch": 0.7465161516629994,
+        "grad_norm": 0.29923033714294434,
+        "learning_rate": 0.0001570392236510633,
+        "loss": 0.975,
+        "step": 23410
+      },
+      {
+        "epoch": 0.7468350393826334,
+        "grad_norm": 0.3004629611968994,
+        "learning_rate": 0.00015643947944688188,
+        "loss": 0.9622,
+        "step": 23420
+      },
+      {
+        "epoch": 0.7471539271022672,
+        "grad_norm": 0.2845831513404846,
+        "learning_rate": 0.00015584202570938826,
+        "loss": 0.954,
+        "step": 23430
+      },
+      {
+        "epoch": 0.7474728148219012,
+        "grad_norm": 0.29445093870162964,
+        "learning_rate": 0.00015524685369112376,
+        "loss": 0.9471,
+        "step": 23440
+      },
+      {
+        "epoch": 0.7477917025415352,
+        "grad_norm": 0.2833983898162842,
+        "learning_rate": 0.00015465395467803683,
+        "loss": 0.9486,
+        "step": 23450
+      },
+      {
+        "epoch": 0.748110590261169,
+        "grad_norm": 0.30625349283218384,
+        "learning_rate": 0.00015406331998935564,
+        "loss": 0.9747,
+        "step": 23460
+      },
+      {
+        "epoch": 0.748429477980803,
+        "grad_norm": 0.29102441668510437,
+        "learning_rate": 0.00015347494097746068,
+        "loss": 0.9649,
+        "step": 23470
+      },
+      {
+        "epoch": 0.7487483657004369,
+        "grad_norm": 0.3019132614135742,
+        "learning_rate": 0.0001528888090277585,
+        "loss": 0.9533,
+        "step": 23480
+      },
+      {
+        "epoch": 0.7490672534200707,
+        "grad_norm": 0.30258646607398987,
+        "learning_rate": 0.0001523049155585553,
+        "loss": 0.9544,
+        "step": 23490
+      },
+      {
+        "epoch": 0.7493861411397047,
+        "grad_norm": 0.3062969744205475,
+        "learning_rate": 0.00015172325202093142,
+        "loss": 0.9646,
+        "step": 23500
+      },
+      {
+        "epoch": 0.7497050288593387,
+        "grad_norm": 0.28206557035446167,
+        "learning_rate": 0.00015114380989861618,
+        "loss": 0.9502,
+        "step": 23510
+      },
+      {
+        "epoch": 0.7500239165789725,
+        "grad_norm": 0.29351454973220825,
+        "learning_rate": 0.0001505665807078631,
+        "loss": 0.9459,
+        "step": 23520
+      },
+      {
+        "epoch": 0.7503428042986064,
+        "grad_norm": 0.30770137906074524,
+        "learning_rate": 0.00014999155599732582,
+        "loss": 0.9596,
+        "step": 23530
+      },
+      {
+        "epoch": 0.7506616920182404,
+        "grad_norm": 0.28911292552948,
+        "learning_rate": 0.00014941872734793423,
+        "loss": 0.9553,
+        "step": 23540
+      },
+      {
+        "epoch": 0.7509805797378742,
+        "grad_norm": 0.29347413778305054,
+        "learning_rate": 0.0001488480863727712,
+        "loss": 0.9649,
+        "step": 23550
+      },
+      {
+        "epoch": 0.7512994674575082,
+        "grad_norm": 0.2877577245235443,
+        "learning_rate": 0.00014827962471694999,
+        "loss": 0.9516,
+        "step": 23560
+      },
+      {
+        "epoch": 0.7516183551771422,
+        "grad_norm": 0.2972458302974701,
+        "learning_rate": 0.00014771333405749166,
+        "loss": 0.9637,
+        "step": 23570
+      },
+      {
+        "epoch": 0.751937242896776,
+        "grad_norm": 0.2873701751232147,
+        "learning_rate": 0.00014714920610320336,
+        "loss": 0.9532,
+        "step": 23580
+      },
+      {
+        "epoch": 0.75225613061641,
+        "grad_norm": 0.2910812795162201,
+        "learning_rate": 0.00014658723259455698,
+        "loss": 0.9506,
+        "step": 23590
+      },
+      {
+        "epoch": 0.7525750183360439,
+        "grad_norm": 0.29596349596977234,
+        "learning_rate": 0.00014602740530356807,
+        "loss": 0.9468,
+        "step": 23600
+      },
+      {
+        "epoch": 0.7528939060556777,
+        "grad_norm": 0.298368364572525,
+        "learning_rate": 0.00014546971603367545,
+        "loss": 0.9631,
+        "step": 23610
+      },
+      {
+        "epoch": 0.7532127937753117,
+        "grad_norm": 0.29450052976608276,
+        "learning_rate": 0.00014491415661962125,
+        "loss": 0.9641,
+        "step": 23620
+      },
+      {
+        "epoch": 0.7535316814949456,
+        "grad_norm": 0.2979263663291931,
+        "learning_rate": 0.00014436071892733128,
+        "loss": 0.9673,
+        "step": 23630
+      },
+      {
+        "epoch": 0.7538505692145795,
+        "grad_norm": 0.2885546088218689,
+        "learning_rate": 0.00014380939485379596,
+        "loss": 0.9397,
+        "step": 23640
+      },
+      {
+        "epoch": 0.7541694569342134,
+        "grad_norm": 0.30781450867652893,
+        "learning_rate": 0.00014326017632695172,
+        "loss": 0.9604,
+        "step": 23650
+      },
+      {
+        "epoch": 0.7544883446538474,
+        "grad_norm": 0.30378997325897217,
+        "learning_rate": 0.00014271305530556277,
+        "loss": 0.9551,
+        "step": 23660
+      },
+      {
+        "epoch": 0.7548072323734814,
+        "grad_norm": 0.29330143332481384,
+        "learning_rate": 0.00014216802377910344,
+        "loss": 0.9503,
+        "step": 23670
+      },
+      {
+        "epoch": 0.7551261200931152,
+        "grad_norm": 0.2918219268321991,
+        "learning_rate": 0.00014162507376764066,
+        "loss": 0.9498,
+        "step": 23680
+      },
+      {
+        "epoch": 0.7554450078127491,
+        "grad_norm": 0.30905428528785706,
+        "learning_rate": 0.0001410841973217176,
+        "loss": 0.9724,
+        "step": 23690
+      },
+      {
+        "epoch": 0.7557638955323831,
+        "grad_norm": 0.29354947805404663,
+        "learning_rate": 0.0001405453865222367,
+        "loss": 0.9546,
+        "step": 23700
+      },
+      {
+        "epoch": 0.7560827832520169,
+        "grad_norm": 0.2877028286457062,
+        "learning_rate": 0.0001400086334803442,
+        "loss": 0.9587,
+        "step": 23710
+      },
+      {
+        "epoch": 0.7564016709716509,
+        "grad_norm": 0.3082736134529114,
+        "learning_rate": 0.0001394739303373143,
+        "loss": 0.9457,
+        "step": 23720
+      },
+      {
+        "epoch": 0.7567205586912849,
+        "grad_norm": 0.3034181594848633,
+        "learning_rate": 0.0001389412692644344,
+        "loss": 0.9521,
+        "step": 23730
+      },
+      {
+        "epoch": 0.7570394464109187,
+        "grad_norm": 0.2895594835281372,
+        "learning_rate": 0.00013841064246289018,
+        "loss": 0.9519,
+        "step": 23740
+      },
+      {
+        "epoch": 0.7573583341305526,
+        "grad_norm": 0.2918117642402649,
+        "learning_rate": 0.00013788204216365162,
+        "loss": 0.946,
+        "step": 23750
+      },
+      {
+        "epoch": 0.7576772218501866,
+        "grad_norm": 0.30359402298927307,
+        "learning_rate": 0.00013735546062735915,
+        "loss": 0.938,
+        "step": 23760
+      },
+      {
+        "epoch": 0.7579961095698204,
+        "grad_norm": 0.30913326144218445,
+        "learning_rate": 0.00013683089014421051,
+        "loss": 0.9617,
+        "step": 23770
+      },
+      {
+        "epoch": 0.7583149972894544,
+        "grad_norm": 0.2882547676563263,
+        "learning_rate": 0.0001363083230338476,
+        "loss": 0.9395,
+        "step": 23780
+      },
+      {
+        "epoch": 0.7586338850090883,
+        "grad_norm": 0.29031381011009216,
+        "learning_rate": 0.00013578775164524418,
+        "loss": 0.9587,
+        "step": 23790
+      },
+      {
+        "epoch": 0.7589527727287222,
+        "grad_norm": 0.30304664373397827,
+        "learning_rate": 0.00013526916835659387,
+        "loss": 0.9743,
+        "step": 23800
+      },
+      {
+        "epoch": 0.7592716604483561,
+        "grad_norm": 0.3043603003025055,
+        "learning_rate": 0.00013475256557519852,
+        "loss": 0.9396,
+        "step": 23810
+      },
+      {
+        "epoch": 0.7595905481679901,
+        "grad_norm": 0.30265742540359497,
+        "learning_rate": 0.00013423793573735703,
+        "loss": 0.9545,
+        "step": 23820
+      },
+      {
+        "epoch": 0.7599094358876239,
+        "grad_norm": 0.29440733790397644,
+        "learning_rate": 0.00013372527130825463,
+        "loss": 0.9535,
+        "step": 23830
+      },
+      {
+        "epoch": 0.7602283236072579,
+        "grad_norm": 0.2963486313819885,
+        "learning_rate": 0.00013321456478185254,
+        "loss": 0.9536,
+        "step": 23840
+      },
+      {
+        "epoch": 0.7605472113268918,
+        "grad_norm": 0.2968265116214752,
+        "learning_rate": 0.00013270580868077813,
+        "loss": 0.9552,
+        "step": 23850
+      },
+      {
+        "epoch": 0.7608660990465257,
+        "grad_norm": 0.303342342376709,
+        "learning_rate": 0.00013219899555621538,
+        "loss": 0.9403,
+        "step": 23860
+      },
+      {
+        "epoch": 0.7611849867661596,
+        "grad_norm": 0.29479485750198364,
+        "learning_rate": 0.00013169411798779585,
+        "loss": 0.9686,
+        "step": 23870
+      },
+      {
+        "epoch": 0.7615038744857936,
+        "grad_norm": 0.2946151793003082,
+        "learning_rate": 0.00013119116858348995,
+        "loss": 0.9551,
+        "step": 23880
+      },
+      {
+        "epoch": 0.7618227622054274,
+        "grad_norm": 0.29338327050209045,
+        "learning_rate": 0.0001306901399794989,
+        "loss": 0.9594,
+        "step": 23890
+      },
+      {
+        "epoch": 0.7621416499250614,
+        "grad_norm": 0.3011675179004669,
+        "learning_rate": 0.0001301910248401467,
+        "loss": 0.942,
+        "step": 23900
+      },
+      {
+        "epoch": 0.7624605376446953,
+        "grad_norm": 0.30116263031959534,
+        "learning_rate": 0.0001296938158577729,
+        "loss": 0.9561,
+        "step": 23910
+      },
+      {
+        "epoch": 0.7627794253643292,
+        "grad_norm": 0.2906716465950012,
+        "learning_rate": 0.0001291985057526255,
+        "loss": 0.9438,
+        "step": 23920
+      },
+      {
+        "epoch": 0.7630983130839631,
+        "grad_norm": 0.29096871614456177,
+        "learning_rate": 0.00012870508727275446,
+        "loss": 0.9429,
+        "step": 23930
+      },
+      {
+        "epoch": 0.7634172008035971,
+        "grad_norm": 0.30688127875328064,
+        "learning_rate": 0.0001282135531939054,
+        "loss": 0.9586,
+        "step": 23940
+      },
+      {
+        "epoch": 0.7637360885232309,
+        "grad_norm": 0.29994523525238037,
+        "learning_rate": 0.00012772389631941392,
+        "loss": 0.9515,
+        "step": 23950
+      },
+      {
+        "epoch": 0.7640549762428649,
+        "grad_norm": 0.302988737821579,
+        "learning_rate": 0.00012723610948010017,
+        "loss": 0.9508,
+        "step": 23960
+      },
+      {
+        "epoch": 0.7643738639624988,
+        "grad_norm": 0.29745492339134216,
+        "learning_rate": 0.00012675018553416398,
+        "loss": 0.9381,
+        "step": 23970
+      },
+      {
+        "epoch": 0.7646927516821327,
+        "grad_norm": 0.294744074344635,
+        "learning_rate": 0.00012626611736708018,
+        "loss": 0.9523,
+        "step": 23980
+      },
+      {
+        "epoch": 0.7650116394017666,
+        "grad_norm": 0.3051568269729614,
+        "learning_rate": 0.00012578389789149453,
+        "loss": 0.9519,
+        "step": 23990
+      },
+      {
+        "epoch": 0.7653305271214006,
+        "grad_norm": 0.2979709208011627,
+        "learning_rate": 0.00012530352004711987,
+        "loss": 0.9629,
+        "step": 24000
+      },
+      {
+        "epoch": 0.7656494148410344,
+        "grad_norm": 0.2889893054962158,
+        "learning_rate": 0.00012482497680063275,
+        "loss": 0.9319,
+        "step": 24010
+      },
+      {
+        "epoch": 0.7659683025606684,
+        "grad_norm": 0.28444361686706543,
+        "learning_rate": 0.00012434826114557067,
+        "loss": 0.9473,
+        "step": 24020
+      },
+      {
+        "epoch": 0.7662871902803023,
+        "grad_norm": 0.3054760992527008,
+        "learning_rate": 0.00012387336610222914,
+        "loss": 0.9636,
+        "step": 24030
+      },
+      {
+        "epoch": 0.7666060779999362,
+        "grad_norm": 0.29499685764312744,
+        "learning_rate": 0.00012340028471755974,
+        "loss": 0.954,
+        "step": 24040
+      },
+      {
+        "epoch": 0.7669249657195701,
+        "grad_norm": 0.294760525226593,
+        "learning_rate": 0.00012292901006506822,
+        "loss": 0.9515,
+        "step": 24050
+      },
+      {
+        "epoch": 0.7672438534392041,
+        "grad_norm": 0.2936845123767853,
+        "learning_rate": 0.00012245953524471316,
+        "loss": 0.957,
+        "step": 24060
+      },
+      {
+        "epoch": 0.7675627411588379,
+        "grad_norm": 0.292159765958786,
+        "learning_rate": 0.00012199185338280484,
+        "loss": 0.9705,
+        "step": 24070
+      },
+      {
+        "epoch": 0.7678816288784719,
+        "grad_norm": 0.2962867319583893,
+        "learning_rate": 0.00012152595763190471,
+        "loss": 0.9582,
+        "step": 24080
+      },
+      {
+        "epoch": 0.7682005165981058,
+        "grad_norm": 0.295032799243927,
+        "learning_rate": 0.00012106184117072507,
+        "loss": 0.9469,
+        "step": 24090
+      },
+      {
+        "epoch": 0.7685194043177397,
+        "grad_norm": 0.2952403426170349,
+        "learning_rate": 0.00012059949720402918,
+        "loss": 0.9507,
+        "step": 24100
+      },
+      {
+        "epoch": 0.7688382920373736,
+        "grad_norm": 0.30328744649887085,
+        "learning_rate": 0.00012013891896253189,
+        "loss": 0.9545,
+        "step": 24110
+      },
+      {
+        "epoch": 0.7691571797570076,
+        "grad_norm": 0.292563796043396,
+        "learning_rate": 0.00011968009970280033,
+        "loss": 0.945,
+        "step": 24120
+      },
+      {
+        "epoch": 0.7694760674766414,
+        "grad_norm": 0.30118969082832336,
+        "learning_rate": 0.00011922303270715539,
+        "loss": 0.9483,
+        "step": 24130
+      },
+      {
+        "epoch": 0.7697949551962754,
+        "grad_norm": 0.2996894419193268,
+        "learning_rate": 0.00011876771128357317,
+        "loss": 0.9486,
+        "step": 24140
+      },
+      {
+        "epoch": 0.7701138429159093,
+        "grad_norm": 0.29649966955184937,
+        "learning_rate": 0.00011831412876558719,
+        "loss": 0.9512,
+        "step": 24150
+      },
+      {
+        "epoch": 0.7704327306355432,
+        "grad_norm": 0.29212257266044617,
+        "learning_rate": 0.00011786227851219062,
+        "loss": 0.935,
+        "step": 24160
+      },
+      {
+        "epoch": 0.7707516183551771,
+        "grad_norm": 0.29390260577201843,
+        "learning_rate": 0.00011741215390773915,
+        "loss": 0.9384,
+        "step": 24170
+      },
+      {
+        "epoch": 0.7710705060748111,
+        "grad_norm": 0.29803451895713806,
+        "learning_rate": 0.00011696374836185405,
+        "loss": 0.9535,
+        "step": 24180
+      },
+      {
+        "epoch": 0.7713893937944449,
+        "grad_norm": 0.30166253447532654,
+        "learning_rate": 0.00011651705530932584,
+        "loss": 0.9464,
+        "step": 24190
+      },
+      {
+        "epoch": 0.7717082815140789,
+        "grad_norm": 0.29554587602615356,
+        "learning_rate": 0.00011607206821001794,
+        "loss": 0.9386,
+        "step": 24200
+      },
+      {
+        "epoch": 0.7720271692337128,
+        "grad_norm": 0.29633378982543945,
+        "learning_rate": 0.00011562878054877104,
+        "loss": 0.9422,
+        "step": 24210
+      },
+      {
+        "epoch": 0.7723460569533467,
+        "grad_norm": 0.29746270179748535,
+        "learning_rate": 0.00011518718583530775,
+        "loss": 0.9347,
+        "step": 24220
+      },
+      {
+        "epoch": 0.7726649446729806,
+        "grad_norm": 0.30118975043296814,
+        "learning_rate": 0.00011474727760413749,
+        "loss": 0.9196,
+        "step": 24230
+      },
+      {
+        "epoch": 0.7729838323926146,
+        "grad_norm": 0.29916834831237793,
+        "learning_rate": 0.00011430904941446183,
+        "loss": 0.9439,
+        "step": 24240
+      },
+      {
+        "epoch": 0.7733027201122484,
+        "grad_norm": 0.2930695414543152,
+        "learning_rate": 0.00011387249485008026,
+        "loss": 0.9367,
+        "step": 24250
+      },
+      {
+        "epoch": 0.7736216078318824,
+        "grad_norm": 0.3068337142467499,
+        "learning_rate": 0.0001134376075192962,
+        "loss": 0.9356,
+        "step": 24260
+      },
+      {
+        "epoch": 0.7739404955515163,
+        "grad_norm": 0.29748132824897766,
+        "learning_rate": 0.00011300438105482337,
+        "loss": 0.9471,
+        "step": 24270
+      },
+      {
+        "epoch": 0.7742593832711502,
+        "grad_norm": 0.30024072527885437,
+        "learning_rate": 0.00011257280911369273,
+        "loss": 0.9564,
+        "step": 24280
+      },
+      {
+        "epoch": 0.7745782709907841,
+        "grad_norm": 0.3009824752807617,
+        "learning_rate": 0.00011214288537715942,
+        "loss": 0.9544,
+        "step": 24290
+      },
+      {
+        "epoch": 0.7748971587104181,
+        "grad_norm": 0.28870582580566406,
+        "learning_rate": 0.00011171460355061031,
+        "loss": 0.9459,
+        "step": 24300
+      },
+      {
+        "epoch": 0.7752160464300519,
+        "grad_norm": 0.29776766896247864,
+        "learning_rate": 0.0001112879573634719,
+        "loss": 0.9492,
+        "step": 24310
+      },
+      {
+        "epoch": 0.7755349341496859,
+        "grad_norm": 0.3070945143699646,
+        "learning_rate": 0.00011086294056911845,
+        "loss": 0.9562,
+        "step": 24320
+      },
+      {
+        "epoch": 0.7758538218693198,
+        "grad_norm": 0.2858162522315979,
+        "learning_rate": 0.00011043954694478053,
+        "loss": 0.9385,
+        "step": 24330
+      },
+      {
+        "epoch": 0.7761727095889537,
+        "grad_norm": 0.30373987555503845,
+        "learning_rate": 0.00011001777029145393,
+        "loss": 0.9516,
+        "step": 24340
+      },
+      {
+        "epoch": 0.7764915973085876,
+        "grad_norm": 0.2940308153629303,
+        "learning_rate": 0.00010959760443380887,
+        "loss": 0.9504,
+        "step": 24350
+      },
+      {
+        "epoch": 0.7768104850282216,
+        "grad_norm": 0.2989041805267334,
+        "learning_rate": 0.00010917904322009965,
+        "loss": 0.9447,
+        "step": 24360
+      },
+      {
+        "epoch": 0.7771293727478554,
+        "grad_norm": 0.294559508562088,
+        "learning_rate": 0.00010876208052207448,
+        "loss": 0.9328,
+        "step": 24370
+      },
+      {
+        "epoch": 0.7774482604674894,
+        "grad_norm": 0.29151490330696106,
+        "learning_rate": 0.00010834671023488582,
+        "loss": 0.925,
+        "step": 24380
+      },
+      {
+        "epoch": 0.7777671481871233,
+        "grad_norm": 0.29237502813339233,
+        "learning_rate": 0.000107932926277001,
+        "loss": 0.9441,
+        "step": 24390
+      },
+      {
+        "epoch": 0.7780860359067572,
+        "grad_norm": 0.3009505271911621,
+        "learning_rate": 0.00010752072259011318,
+        "loss": 0.9462,
+        "step": 24400
+      },
+      {
+        "epoch": 0.7784049236263911,
+        "grad_norm": 0.2872629463672638,
+        "learning_rate": 0.00010711009313905259,
+        "loss": 0.9472,
+        "step": 24410
+      },
+      {
+        "epoch": 0.7787238113460251,
+        "grad_norm": 0.2956571578979492,
+        "learning_rate": 0.00010670103191169825,
+        "loss": 0.9524,
+        "step": 24420
+      },
+      {
+        "epoch": 0.7790426990656589,
+        "grad_norm": 0.2941124439239502,
+        "learning_rate": 0.00010629353291888986,
+        "loss": 0.9457,
+        "step": 24430
+      },
+      {
+        "epoch": 0.7793615867852929,
+        "grad_norm": 0.29650887846946716,
+        "learning_rate": 0.00010588759019434024,
+        "loss": 0.9447,
+        "step": 24440
+      },
+      {
+        "epoch": 0.7796804745049268,
+        "grad_norm": 0.28828340768814087,
+        "learning_rate": 0.00010548319779454782,
+        "loss": 0.933,
+        "step": 24450
+      },
+      {
+        "epoch": 0.7799993622245607,
+        "grad_norm": 0.29822880029678345,
+        "learning_rate": 0.00010508034979870972,
+        "loss": 0.9441,
+        "step": 24460
+      },
+      {
+        "epoch": 0.7803182499441946,
+        "grad_norm": 0.2889900505542755,
+        "learning_rate": 0.00010467904030863507,
+        "loss": 0.938,
+        "step": 24470
+      },
+      {
+        "epoch": 0.7806371376638286,
+        "grad_norm": 0.2908421754837036,
+        "learning_rate": 0.00010427926344865853,
+        "loss": 0.9256,
+        "step": 24480
+      },
+      {
+        "epoch": 0.7809560253834625,
+        "grad_norm": 0.29898807406425476,
+        "learning_rate": 0.00010388101336555442,
+        "loss": 0.9449,
+        "step": 24490
+      },
+      {
+        "epoch": 0.7812749131030964,
+        "grad_norm": 0.30551448464393616,
+        "learning_rate": 0.00010348428422845097,
+        "loss": 0.947,
+        "step": 24500
+      },
+      {
+        "epoch": 0.7815938008227303,
+        "grad_norm": 0.3013573884963989,
+        "learning_rate": 0.00010308907022874489,
+        "loss": 0.9266,
+        "step": 24510
+      },
+      {
+        "epoch": 0.7819126885423643,
+        "grad_norm": 0.29684606194496155,
+        "learning_rate": 0.00010269536558001635,
+        "loss": 0.9466,
+        "step": 24520
+      },
+      {
+        "epoch": 0.7822315762619981,
+        "grad_norm": 0.29465001821517944,
+        "learning_rate": 0.00010230316451794439,
+        "loss": 0.9542,
+        "step": 24530
+      },
+      {
+        "epoch": 0.7825504639816321,
+        "grad_norm": 0.2919856905937195,
+        "learning_rate": 0.00010191246130022226,
+        "loss": 0.9328,
+        "step": 24540
+      },
+      {
+        "epoch": 0.782869351701266,
+        "grad_norm": 0.29406240582466125,
+        "learning_rate": 0.00010152325020647362,
+        "loss": 0.9406,
+        "step": 24550
+      },
+      {
+        "epoch": 0.7831882394208999,
+        "grad_norm": 0.3012252151966095,
+        "learning_rate": 0.00010113552553816855,
+        "loss": 0.9365,
+        "step": 24560
+      },
+      {
+        "epoch": 0.7835071271405338,
+        "grad_norm": 0.3084176182746887,
+        "learning_rate": 0.00010074928161854036,
+        "loss": 0.9421,
+        "step": 24570
+      },
+      {
+        "epoch": 0.7838260148601678,
+        "grad_norm": 0.2966445982456207,
+        "learning_rate": 0.00010036451279250222,
+        "loss": 0.9372,
+        "step": 24580
+      },
+      {
+        "epoch": 0.7841449025798016,
+        "grad_norm": 0.28686439990997314,
+        "learning_rate": 9.998121342656457e-05,
+        "loss": 0.9429,
+        "step": 24590
+      },
+      {
+        "epoch": 0.7844637902994356,
+        "grad_norm": 0.2919677197933197,
+        "learning_rate": 9.959937790875249e-05,
+        "loss": 0.9405,
+        "step": 24600
+      },
+      {
+        "epoch": 0.7847826780190695,
+        "grad_norm": 0.2990611791610718,
+        "learning_rate": 9.92190006485237e-05,
+        "loss": 0.9352,
+        "step": 24610
+      },
+      {
+        "epoch": 0.7851015657387034,
+        "grad_norm": 0.29399555921554565,
+        "learning_rate": 9.884007607668652e-05,
+        "loss": 0.9236,
+        "step": 24620
+      },
+      {
+        "epoch": 0.7854204534583373,
+        "grad_norm": 0.3035459518432617,
+        "learning_rate": 9.846259864531842e-05,
+        "loss": 0.9397,
+        "step": 24630
+      },
+      {
+        "epoch": 0.7857393411779713,
+        "grad_norm": 0.30628782510757446,
+        "learning_rate": 9.808656282768486e-05,
+        "loss": 0.9341,
+        "step": 24640
+      },
+      {
+        "epoch": 0.7860582288976051,
+        "grad_norm": 0.3014283776283264,
+        "learning_rate": 9.77119631181582e-05,
+        "loss": 0.9338,
+        "step": 24650
+      },
+      {
+        "epoch": 0.7863771166172391,
+        "grad_norm": 0.29893139004707336,
+        "learning_rate": 9.733879403213728e-05,
+        "loss": 0.9387,
+        "step": 24660
+      },
+      {
+        "epoch": 0.786696004336873,
+        "grad_norm": 0.287754625082016,
+        "learning_rate": 9.696705010596698e-05,
+        "loss": 0.9269,
+        "step": 24670
+      },
+      {
+        "epoch": 0.7870148920565069,
+        "grad_norm": 0.2931799590587616,
+        "learning_rate": 9.65967258968583e-05,
+        "loss": 0.9466,
+        "step": 24680
+      },
+      {
+        "epoch": 0.7873337797761408,
+        "grad_norm": 0.3002002537250519,
+        "learning_rate": 9.622781598280861e-05,
+        "loss": 0.9433,
+        "step": 24690
+      },
+      {
+        "epoch": 0.7876526674957748,
+        "grad_norm": 0.2931079566478729,
+        "learning_rate": 9.586031496252237e-05,
+        "loss": 0.9258,
+        "step": 24700
+      },
+      {
+        "epoch": 0.7879715552154086,
+        "grad_norm": 0.3031649887561798,
+        "learning_rate": 9.549421745533192e-05,
+        "loss": 0.9424,
+        "step": 24710
+      },
+      {
+        "epoch": 0.7882904429350426,
+        "grad_norm": 0.29497942328453064,
+        "learning_rate": 9.512951810111879e-05,
+        "loss": 0.9243,
+        "step": 24720
+      },
+      {
+        "epoch": 0.7886093306546765,
+        "grad_norm": 0.29368606209754944,
+        "learning_rate": 9.476621156023519e-05,
+        "loss": 0.9426,
+        "step": 24730
+      },
+      {
+        "epoch": 0.7889282183743104,
+        "grad_norm": 0.3123050630092621,
+        "learning_rate": 9.44042925134258e-05,
+        "loss": 0.9408,
+        "step": 24740
+      },
+      {
+        "epoch": 0.7892471060939443,
+        "grad_norm": 0.28893402218818665,
+        "learning_rate": 9.404375566174994e-05,
+        "loss": 0.9569,
+        "step": 24750
+      },
+      {
+        "epoch": 0.7895659938135783,
+        "grad_norm": 0.2901870608329773,
+        "learning_rate": 9.368459572650401e-05,
+        "loss": 0.9323,
+        "step": 24760
+      },
+      {
+        "epoch": 0.7898848815332121,
+        "grad_norm": 0.2980130612850189,
+        "learning_rate": 9.33268074491441e-05,
+        "loss": 0.9224,
+        "step": 24770
+      },
+      {
+        "epoch": 0.7902037692528461,
+        "grad_norm": 0.2842077314853668,
+        "learning_rate": 9.297038559120912e-05,
+        "loss": 0.9443,
+        "step": 24780
+      },
+      {
+        "epoch": 0.79052265697248,
+        "grad_norm": 0.2993188798427582,
+        "learning_rate": 9.2615324934244e-05,
+        "loss": 0.9566,
+        "step": 24790
+      },
+      {
+        "epoch": 0.7908415446921139,
+        "grad_norm": 0.2908771336078644,
+        "learning_rate": 9.226162027972337e-05,
+        "loss": 0.9255,
+        "step": 24800
+      },
+      {
+        "epoch": 0.7911604324117478,
+        "grad_norm": 0.28995591402053833,
+        "learning_rate": 9.190926644897531e-05,
+        "loss": 0.9434,
+        "step": 24810
+      },
+      {
+        "epoch": 0.7914793201313818,
+        "grad_norm": 0.30371615290641785,
+        "learning_rate": 9.155825828310578e-05,
+        "loss": 0.9343,
+        "step": 24820
+      },
+      {
+        "epoch": 0.7917982078510156,
+        "grad_norm": 0.30678293108940125,
+        "learning_rate": 9.120859064292278e-05,
+        "loss": 0.9372,
+        "step": 24830
+      },
+      {
+        "epoch": 0.7921170955706496,
+        "grad_norm": 0.2907579839229584,
+        "learning_rate": 9.086025840886135e-05,
+        "loss": 0.93,
+        "step": 24840
+      },
+      {
+        "epoch": 0.7924359832902835,
+        "grad_norm": 0.30206260085105896,
+        "learning_rate": 9.05132564809085e-05,
+        "loss": 0.94,
+        "step": 24850
+      },
+      {
+        "epoch": 0.7927548710099174,
+        "grad_norm": 0.3013043999671936,
+        "learning_rate": 9.016757977852851e-05,
+        "loss": 0.931,
+        "step": 24860
+      },
+      {
+        "epoch": 0.7930737587295513,
+        "grad_norm": 0.29542598128318787,
+        "learning_rate": 8.982322324058868e-05,
+        "loss": 0.9251,
+        "step": 24870
+      },
+      {
+        "epoch": 0.7933926464491853,
+        "grad_norm": 0.2961618900299072,
+        "learning_rate": 8.948018182528511e-05,
+        "loss": 0.9327,
+        "step": 24880
+      },
+      {
+        "epoch": 0.7937115341688191,
+        "grad_norm": 0.2946419417858124,
+        "learning_rate": 8.913845051006889e-05,
+        "loss": 0.9261,
+        "step": 24890
+      },
+      {
+        "epoch": 0.7940304218884531,
+        "grad_norm": 0.2816670835018158,
+        "learning_rate": 8.879802429157259e-05,
+        "loss": 0.9261,
+        "step": 24900
+      },
+      {
+        "epoch": 0.794349309608087,
+        "grad_norm": 0.30013415217399597,
+        "learning_rate": 8.845889818553704e-05,
+        "loss": 0.9395,
+        "step": 24910
+      },
+      {
+        "epoch": 0.7946681973277209,
+        "grad_norm": 0.2884807288646698,
+        "learning_rate": 8.81210672267383e-05,
+        "loss": 0.9348,
+        "step": 24920
+      },
+      {
+        "epoch": 0.7949870850473548,
+        "grad_norm": 0.296625554561615,
+        "learning_rate": 8.778452646891497e-05,
+        "loss": 0.9279,
+        "step": 24930
+      },
+      {
+        "epoch": 0.7953059727669888,
+        "grad_norm": 0.2994830906391144,
+        "learning_rate": 8.744927098469577e-05,
+        "loss": 0.932,
+        "step": 24940
+      },
+      {
+        "epoch": 0.7956248604866226,
+        "grad_norm": 0.29800304770469666,
+        "learning_rate": 8.711529586552748e-05,
+        "loss": 0.9428,
+        "step": 24950
+      },
+      {
+        "epoch": 0.7959437482062566,
+        "grad_norm": 0.2980853319168091,
+        "learning_rate": 8.678259622160293e-05,
+        "loss": 0.9293,
+        "step": 24960
+      },
+      {
+        "epoch": 0.7962626359258905,
+        "grad_norm": 0.29429447650909424,
+        "learning_rate": 8.645116718178947e-05,
+        "loss": 0.9274,
+        "step": 24970
+      },
+      {
+        "epoch": 0.7965815236455244,
+        "grad_norm": 0.29675599932670593,
+        "learning_rate": 8.612100389355772e-05,
+        "loss": 0.9342,
+        "step": 24980
+      },
+      {
+        "epoch": 0.7969004113651583,
+        "grad_norm": 0.29180872440338135,
+        "learning_rate": 8.579210152291042e-05,
+        "loss": 0.9107,
+        "step": 24990
+      },
+      {
+        "epoch": 0.7972192990847923,
+        "grad_norm": 0.29538294672966003,
+        "learning_rate": 8.546445525431173e-05,
+        "loss": 0.9268,
+        "step": 25000
+      },
+      {
+        "epoch": 0.7975381868044261,
+        "grad_norm": 0.30392956733703613,
+        "learning_rate": 8.513806029061663e-05,
+        "loss": 0.9478,
+        "step": 25010
+      },
+      {
+        "epoch": 0.7978570745240601,
+        "grad_norm": 0.30718958377838135,
+        "learning_rate": 8.481291185300078e-05,
+        "loss": 0.9272,
+        "step": 25020
+      },
+      {
+        "epoch": 0.798175962243694,
+        "grad_norm": 0.31373822689056396,
+        "learning_rate": 8.448900518089058e-05,
+        "loss": 0.9256,
+        "step": 25030
+      },
+      {
+        "epoch": 0.7984948499633279,
+        "grad_norm": 0.31905749440193176,
+        "learning_rate": 8.416633553189332e-05,
+        "loss": 0.9426,
+        "step": 25040
+      },
+      {
+        "epoch": 0.7988137376829618,
+        "grad_norm": 0.288647323846817,
+        "learning_rate": 8.384489818172784e-05,
+        "loss": 0.9257,
+        "step": 25050
+      },
+      {
+        "epoch": 0.7991326254025958,
+        "grad_norm": 0.2976011037826538,
+        "learning_rate": 8.35246884241554e-05,
+        "loss": 0.9251,
+        "step": 25060
+      },
+      {
+        "epoch": 0.7994515131222296,
+        "grad_norm": 0.298535019159317,
+        "learning_rate": 8.32057015709107e-05,
+        "loss": 0.9444,
+        "step": 25070
+      },
+      {
+        "epoch": 0.7997704008418636,
+        "grad_norm": 0.3076132833957672,
+        "learning_rate": 8.288793295163325e-05,
+        "loss": 0.9376,
+        "step": 25080
+      },
+      {
+        "epoch": 0.8000892885614975,
+        "grad_norm": 0.29532676935195923,
+        "learning_rate": 8.257137791379903e-05,
+        "loss": 0.9432,
+        "step": 25090
+      },
+      {
+        "epoch": 0.8004081762811314,
+        "grad_norm": 0.29447340965270996,
+        "learning_rate": 8.225603182265234e-05,
+        "loss": 0.9187,
+        "step": 25100
+      },
+      {
+        "epoch": 0.8007270640007653,
+        "grad_norm": 0.29353076219558716,
+        "learning_rate": 8.194189006113793e-05,
+        "loss": 0.9317,
+        "step": 25110
+      },
+      {
+        "epoch": 0.8010459517203993,
+        "grad_norm": 0.2983883023262024,
+        "learning_rate": 8.162894802983348e-05,
+        "loss": 0.9314,
+        "step": 25120
+      },
+      {
+        "epoch": 0.8013648394400331,
+        "grad_norm": 0.3010222613811493,
+        "learning_rate": 8.131720114688214e-05,
+        "loss": 0.9255,
+        "step": 25130
+      },
+      {
+        "epoch": 0.8016837271596671,
+        "grad_norm": 0.3008621335029602,
+        "learning_rate": 8.100664484792551e-05,
+        "loss": 0.9321,
+        "step": 25140
+      },
+      {
+        "epoch": 0.802002614879301,
+        "grad_norm": 0.3051190972328186,
+        "learning_rate": 8.069727458603683e-05,
+        "loss": 0.9206,
+        "step": 25150
+      },
+      {
+        "epoch": 0.8023215025989349,
+        "grad_norm": 0.2958706021308899,
+        "learning_rate": 8.038908583165436e-05,
+        "loss": 0.9458,
+        "step": 25160
+      },
+      {
+        "epoch": 0.8026403903185688,
+        "grad_norm": 0.30285048484802246,
+        "learning_rate": 8.008207407251511e-05,
+        "loss": 0.9097,
+        "step": 25170
+      },
+      {
+        "epoch": 0.8029592780382028,
+        "grad_norm": 0.29344502091407776,
+        "learning_rate": 7.977623481358872e-05,
+        "loss": 0.9226,
+        "step": 25180
+      },
+      {
+        "epoch": 0.8032781657578366,
+        "grad_norm": 0.2978428304195404,
+        "learning_rate": 7.947156357701168e-05,
+        "loss": 0.9196,
+        "step": 25190
+      },
+      {
+        "epoch": 0.8035970534774706,
+        "grad_norm": 0.29757606983184814,
+        "learning_rate": 7.916805590202182e-05,
+        "loss": 0.9299,
+        "step": 25200
+      },
+      {
+        "epoch": 0.8039159411971045,
+        "grad_norm": 0.2951332926750183,
+        "learning_rate": 7.886570734489289e-05,
+        "loss": 0.9392,
+        "step": 25210
+      },
+      {
+        "epoch": 0.8042348289167384,
+        "grad_norm": 0.29570409655570984,
+        "learning_rate": 7.856451347886955e-05,
+        "loss": 0.9252,
+        "step": 25220
+      },
+      {
+        "epoch": 0.8045537166363723,
+        "grad_norm": 0.2930358350276947,
+        "learning_rate": 7.82644698941026e-05,
+        "loss": 0.9182,
+        "step": 25230
+      },
+      {
+        "epoch": 0.8048726043560063,
+        "grad_norm": 0.30780941247940063,
+        "learning_rate": 7.796557219758433e-05,
+        "loss": 0.9442,
+        "step": 25240
+      },
+      {
+        "epoch": 0.8051914920756401,
+        "grad_norm": 0.28759080171585083,
+        "learning_rate": 7.766781601308432e-05,
+        "loss": 0.9377,
+        "step": 25250
+      },
+      {
+        "epoch": 0.8055103797952741,
+        "grad_norm": 0.3078532814979553,
+        "learning_rate": 7.737119698108516e-05,
+        "loss": 0.9119,
+        "step": 25260
+      },
+      {
+        "epoch": 0.805829267514908,
+        "grad_norm": 0.2813483774662018,
+        "learning_rate": 7.70757107587189e-05,
+        "loss": 0.9083,
+        "step": 25270
+      },
+      {
+        "epoch": 0.8061481552345419,
+        "grad_norm": 0.29622021317481995,
+        "learning_rate": 7.678135301970321e-05,
+        "loss": 0.9323,
+        "step": 25280
+      },
+      {
+        "epoch": 0.8064670429541758,
+        "grad_norm": 0.294413298368454,
+        "learning_rate": 7.648811945427822e-05,
+        "loss": 0.9312,
+        "step": 25290
+      },
+      {
+        "epoch": 0.8067859306738098,
+        "grad_norm": 0.2925923466682434,
+        "learning_rate": 7.61960057691433e-05,
+        "loss": 0.9284,
+        "step": 25300
+      },
+      {
+        "epoch": 0.8071048183934437,
+        "grad_norm": 0.302916020154953,
+        "learning_rate": 7.590500768739427e-05,
+        "loss": 0.9162,
+        "step": 25310
+      },
+      {
+        "epoch": 0.8074237061130776,
+        "grad_norm": 0.2981068789958954,
+        "learning_rate": 7.561512094846075e-05,
+        "loss": 0.9496,
+        "step": 25320
+      },
+      {
+        "epoch": 0.8077425938327115,
+        "grad_norm": 0.2955712080001831,
+        "learning_rate": 7.532634130804376e-05,
+        "loss": 0.9274,
+        "step": 25330
+      },
+      {
+        "epoch": 0.8080614815523455,
+        "grad_norm": 0.29238399863243103,
+        "learning_rate": 7.503866453805368e-05,
+        "loss": 0.9292,
+        "step": 25340
+      },
+      {
+        "epoch": 0.8083803692719793,
+        "grad_norm": 0.2969522476196289,
+        "learning_rate": 7.475208642654819e-05,
+        "loss": 0.9279,
+        "step": 25350
+      },
+      {
+        "epoch": 0.8086992569916133,
+        "grad_norm": 0.29493582248687744,
+        "learning_rate": 7.44666027776707e-05,
+        "loss": 0.9236,
+        "step": 25360
+      },
+      {
+        "epoch": 0.8090181447112472,
+        "grad_norm": 0.29591089487075806,
+        "learning_rate": 7.418220941158897e-05,
+        "loss": 0.9205,
+        "step": 25370
+      },
+      {
+        "epoch": 0.8093370324308811,
+        "grad_norm": 0.2962036430835724,
+        "learning_rate": 7.389890216443378e-05,
+        "loss": 0.9343,
+        "step": 25380
+      },
+      {
+        "epoch": 0.809655920150515,
+        "grad_norm": 0.29654377698898315,
+        "learning_rate": 7.361667688823803e-05,
+        "loss": 0.9211,
+        "step": 25390
+      },
+      {
+        "epoch": 0.809974807870149,
+        "grad_norm": 0.29060810804367065,
+        "learning_rate": 7.333552945087602e-05,
+        "loss": 0.9202,
+        "step": 25400
+      },
+      {
+        "epoch": 0.8102936955897828,
+        "grad_norm": 0.2932351231575012,
+        "learning_rate": 7.305545573600293e-05,
+        "loss": 0.9223,
+        "step": 25410
+      },
+      {
+        "epoch": 0.8106125833094168,
+        "grad_norm": 0.30790457129478455,
+        "learning_rate": 7.277645164299459e-05,
+        "loss": 0.9295,
+        "step": 25420
+      },
+      {
+        "epoch": 0.8109314710290507,
+        "grad_norm": 0.30155086517333984,
+        "learning_rate": 7.249851308688739e-05,
+        "loss": 0.9283,
+        "step": 25430
+      },
+      {
+        "epoch": 0.8112503587486846,
+        "grad_norm": 0.3012137711048126,
+        "learning_rate": 7.222163599831843e-05,
+        "loss": 0.9385,
+        "step": 25440
+      },
+      {
+        "epoch": 0.8115692464683185,
+        "grad_norm": 0.30086320638656616,
+        "learning_rate": 7.194581632346617e-05,
+        "loss": 0.9277,
+        "step": 25450
+      },
+      {
+        "epoch": 0.8118881341879525,
+        "grad_norm": 0.29009512066841125,
+        "learning_rate": 7.167105002399073e-05,
+        "loss": 0.9119,
+        "step": 25460
+      },
+      {
+        "epoch": 0.8122070219075863,
+        "grad_norm": 0.30384042859077454,
+        "learning_rate": 7.139733307697503e-05,
+        "loss": 0.9378,
+        "step": 25470
+      },
+      {
+        "epoch": 0.8125259096272203,
+        "grad_norm": 0.28765690326690674,
+        "learning_rate": 7.112466147486579e-05,
+        "loss": 0.9416,
+        "step": 25480
+      },
+      {
+        "epoch": 0.8128447973468542,
+        "grad_norm": 0.3002883791923523,
+        "learning_rate": 7.08530312254149e-05,
+        "loss": 0.9368,
+        "step": 25490
+      },
+      {
+        "epoch": 0.8131636850664881,
+        "grad_norm": 0.3096667230129242,
+        "learning_rate": 7.058243835162084e-05,
+        "loss": 0.9346,
+        "step": 25500
+      },
+      {
+        "epoch": 0.813482572786122,
+        "grad_norm": 0.29634806513786316,
+        "learning_rate": 7.031287889167066e-05,
+        "loss": 0.9294,
+        "step": 25510
+      },
+      {
+        "epoch": 0.813801460505756,
+        "grad_norm": 0.3080945909023285,
+        "learning_rate": 7.004434889888181e-05,
+        "loss": 0.9251,
+        "step": 25520
+      },
+      {
+        "epoch": 0.8141203482253898,
+        "grad_norm": 0.2975804805755615,
+        "learning_rate": 6.977684444164437e-05,
+        "loss": 0.9313,
+        "step": 25530
+      },
+      {
+        "epoch": 0.8144392359450238,
+        "grad_norm": 0.2953619062900543,
+        "learning_rate": 6.951036160336359e-05,
+        "loss": 0.9318,
+        "step": 25540
+      },
+      {
+        "epoch": 0.8147581236646577,
+        "grad_norm": 0.30130645632743835,
+        "learning_rate": 6.924489648240246e-05,
+        "loss": 0.9289,
+        "step": 25550
+      },
+      {
+        "epoch": 0.8150770113842916,
+        "grad_norm": 0.29083284735679626,
+        "learning_rate": 6.89804451920246e-05,
+        "loss": 0.9212,
+        "step": 25560
+      },
+      {
+        "epoch": 0.8153958991039255,
+        "grad_norm": 0.2993018329143524,
+        "learning_rate": 6.871700386033734e-05,
+        "loss": 0.9303,
+        "step": 25570
+      },
+      {
+        "epoch": 0.8157147868235595,
+        "grad_norm": 0.29969847202301025,
+        "learning_rate": 6.845456863023507e-05,
+        "loss": 0.927,
+        "step": 25580
+      },
+      {
+        "epoch": 0.8160336745431933,
+        "grad_norm": 0.3024032711982727,
+        "learning_rate": 6.819313565934276e-05,
+        "loss": 0.9385,
+        "step": 25590
+      },
+      {
+        "epoch": 0.8163525622628273,
+        "grad_norm": 0.2944593131542206,
+        "learning_rate": 6.793270111995963e-05,
+        "loss": 0.9268,
+        "step": 25600
+      },
+      {
+        "epoch": 0.8166714499824612,
+        "grad_norm": 0.29249638319015503,
+        "learning_rate": 6.767326119900324e-05,
+        "loss": 0.9226,
+        "step": 25610
+      },
+      {
+        "epoch": 0.8169903377020951,
+        "grad_norm": 0.2940239906311035,
+        "learning_rate": 6.741481209795356e-05,
+        "loss": 0.9279,
+        "step": 25620
+      },
+      {
+        "epoch": 0.817309225421729,
+        "grad_norm": 0.30038586258888245,
+        "learning_rate": 6.715735003279736e-05,
+        "loss": 0.935,
+        "step": 25630
+      },
+      {
+        "epoch": 0.817628113141363,
+        "grad_norm": 0.294821172952652,
+        "learning_rate": 6.69008712339729e-05,
+        "loss": 0.9313,
+        "step": 25640
+      },
+      {
+        "epoch": 0.8179470008609968,
+        "grad_norm": 0.2999110519886017,
+        "learning_rate": 6.664537194631458e-05,
+        "loss": 0.9118,
+        "step": 25650
+      },
+      {
+        "epoch": 0.8182658885806308,
+        "grad_norm": 0.30320724844932556,
+        "learning_rate": 6.639084842899812e-05,
+        "loss": 0.9299,
+        "step": 25660
+      },
+      {
+        "epoch": 0.8185847763002647,
+        "grad_norm": 0.3023329973220825,
+        "learning_rate": 6.613729695548567e-05,
+        "loss": 0.9234,
+        "step": 25670
+      },
+      {
+        "epoch": 0.8189036640198986,
+        "grad_norm": 0.3076138198375702,
+        "learning_rate": 6.588471381347133e-05,
+        "loss": 0.9323,
+        "step": 25680
+      },
+      {
+        "epoch": 0.8192225517395325,
+        "grad_norm": 0.3051240146160126,
+        "learning_rate": 6.563309530482677e-05,
+        "loss": 0.9331,
+        "step": 25690
+      },
+      {
+        "epoch": 0.8195414394591665,
+        "grad_norm": 0.31167781352996826,
+        "learning_rate": 6.538243774554706e-05,
+        "loss": 0.937,
+        "step": 25700
+      },
+      {
+        "epoch": 0.8198603271788003,
+        "grad_norm": 0.30333802103996277,
+        "learning_rate": 6.513273746569676e-05,
+        "loss": 0.9235,
+        "step": 25710
+      },
+      {
+        "epoch": 0.8201792148984343,
+        "grad_norm": 0.2919188439846039,
+        "learning_rate": 6.48839908093562e-05,
+        "loss": 0.9078,
+        "step": 25720
+      },
+      {
+        "epoch": 0.8204981026180682,
+        "grad_norm": 0.30451634526252747,
+        "learning_rate": 6.463619413456787e-05,
+        "loss": 0.9274,
+        "step": 25730
+      },
+      {
+        "epoch": 0.8208169903377021,
+        "grad_norm": 0.30046597123146057,
+        "learning_rate": 6.438934381328326e-05,
+        "loss": 0.9228,
+        "step": 25740
+      },
+      {
+        "epoch": 0.821135878057336,
+        "grad_norm": 0.29581958055496216,
+        "learning_rate": 6.414343623130956e-05,
+        "loss": 0.9238,
+        "step": 25750
+      },
+      {
+        "epoch": 0.82145476577697,
+        "grad_norm": 0.29217979311943054,
+        "learning_rate": 6.389846778825685e-05,
+        "loss": 0.9004,
+        "step": 25760
+      },
+      {
+        "epoch": 0.8217736534966038,
+        "grad_norm": 0.29643428325653076,
+        "learning_rate": 6.365443489748536e-05,
+        "loss": 0.9239,
+        "step": 25770
+      },
+      {
+        "epoch": 0.8220925412162378,
+        "grad_norm": 0.30537211894989014,
+        "learning_rate": 6.341133398605295e-05,
+        "loss": 0.9208,
+        "step": 25780
+      },
+      {
+        "epoch": 0.8224114289358717,
+        "grad_norm": 0.30653655529022217,
+        "learning_rate": 6.316916149466283e-05,
+        "loss": 0.9282,
+        "step": 25790
+      },
+      {
+        "epoch": 0.8227303166555056,
+        "grad_norm": 0.3039204180240631,
+        "learning_rate": 6.292791387761143e-05,
+        "loss": 0.905,
+        "step": 25800
+      },
+      {
+        "epoch": 0.8230492043751395,
+        "grad_norm": 0.2952761650085449,
+        "learning_rate": 6.268758760273642e-05,
+        "loss": 0.9261,
+        "step": 25810
+      },
+      {
+        "epoch": 0.8233680920947735,
+        "grad_norm": 0.29730528593063354,
+        "learning_rate": 6.24481791513651e-05,
+        "loss": 0.9288,
+        "step": 25820
+      },
+      {
+        "epoch": 0.8236869798144073,
+        "grad_norm": 0.3040090799331665,
+        "learning_rate": 6.220968501826283e-05,
+        "loss": 0.9172,
+        "step": 25830
+      },
+      {
+        "epoch": 0.8240058675340413,
+        "grad_norm": 0.2928590476512909,
+        "learning_rate": 6.197210171158173e-05,
+        "loss": 0.9149,
+        "step": 25840
+      },
+      {
+        "epoch": 0.8243247552536752,
+        "grad_norm": 0.3051118850708008,
+        "learning_rate": 6.173542575280949e-05,
+        "loss": 0.9294,
+        "step": 25850
+      },
+      {
+        "epoch": 0.824643642973309,
+        "grad_norm": 0.3019244074821472,
+        "learning_rate": 6.149965367671856e-05,
+        "loss": 0.9181,
+        "step": 25860
+      },
+      {
+        "epoch": 0.824962530692943,
+        "grad_norm": 0.2956985831260681,
+        "learning_rate": 6.126478203131529e-05,
+        "loss": 0.906,
+        "step": 25870
+      },
+      {
+        "epoch": 0.825281418412577,
+        "grad_norm": 0.29620710015296936,
+        "learning_rate": 6.1030807377789486e-05,
+        "loss": 0.9138,
+        "step": 25880
+      },
+      {
+        "epoch": 0.8256003061322108,
+        "grad_norm": 0.30234384536743164,
+        "learning_rate": 6.0797726290463996e-05,
+        "loss": 0.927,
+        "step": 25890
+      },
+      {
+        "epoch": 0.8259191938518448,
+        "grad_norm": 0.30620619654655457,
+        "learning_rate": 6.056553535674458e-05,
+        "loss": 0.9098,
+        "step": 25900
+      },
+      {
+        "epoch": 0.8262380815714787,
+        "grad_norm": 0.3129229247570038,
+        "learning_rate": 6.033423117706994e-05,
+        "loss": 0.9256,
+        "step": 25910
+      },
+      {
+        "epoch": 0.8265569692911126,
+        "grad_norm": 0.29301196336746216,
+        "learning_rate": 6.0103810364861955e-05,
+        "loss": 0.9254,
+        "step": 25920
+      },
+      {
+        "epoch": 0.8268758570107465,
+        "grad_norm": 0.30466556549072266,
+        "learning_rate": 5.9874269546476105e-05,
+        "loss": 0.9175,
+        "step": 25930
+      },
+      {
+        "epoch": 0.8271947447303805,
+        "grad_norm": 0.2885163724422455,
+        "learning_rate": 5.964560536115204e-05,
+        "loss": 0.9272,
+        "step": 25940
+      },
+      {
+        "epoch": 0.8275136324500143,
+        "grad_norm": 0.2929341495037079,
+        "learning_rate": 5.941781446096441e-05,
+        "loss": 0.8968,
+        "step": 25950
+      },
+      {
+        "epoch": 0.8278325201696483,
+        "grad_norm": 0.3047708570957184,
+        "learning_rate": 5.9190893510773834e-05,
+        "loss": 0.9265,
+        "step": 25960
+      },
+      {
+        "epoch": 0.8281514078892822,
+        "grad_norm": 0.3061845600605011,
+        "learning_rate": 5.896483918817807e-05,
+        "loss": 0.9216,
+        "step": 25970
+      },
+      {
+        "epoch": 0.828470295608916,
+        "grad_norm": 0.31115177273750305,
+        "learning_rate": 5.873964818346338e-05,
+        "loss": 0.9316,
+        "step": 25980
+      },
+      {
+        "epoch": 0.82878918332855,
+        "grad_norm": 0.31330665946006775,
+        "learning_rate": 5.8515317199556014e-05,
+        "loss": 0.9172,
+        "step": 25990
+      },
+      {
+        "epoch": 0.829108071048184,
+        "grad_norm": 0.3075343072414398,
+        "learning_rate": 5.829184295197409e-05,
+        "loss": 0.9142,
+        "step": 26000
+      },
+      {
+        "epoch": 0.8294269587678178,
+        "grad_norm": 0.2841055691242218,
+        "learning_rate": 5.806922216877932e-05,
+        "loss": 0.912,
+        "step": 26010
+      },
+      {
+        "epoch": 0.8297458464874518,
+        "grad_norm": 0.29682737588882446,
+        "learning_rate": 5.784745159052919e-05,
+        "loss": 0.9113,
+        "step": 26020
+      },
+      {
+        "epoch": 0.8300647342070857,
+        "grad_norm": 0.29654332995414734,
+        "learning_rate": 5.7626527970229256e-05,
+        "loss": 0.9276,
+        "step": 26030
+      },
+      {
+        "epoch": 0.8303836219267196,
+        "grad_norm": 0.3100140690803528,
+        "learning_rate": 5.7406448073285566e-05,
+        "loss": 0.9193,
+        "step": 26040
+      },
+      {
+        "epoch": 0.8307025096463535,
+        "grad_norm": 0.3016508221626282,
+        "learning_rate": 5.718720867745734e-05,
+        "loss": 0.9318,
+        "step": 26050
+      },
+      {
+        "epoch": 0.8310213973659875,
+        "grad_norm": 0.2981314957141876,
+        "learning_rate": 5.6968806572809736e-05,
+        "loss": 0.9169,
+        "step": 26060
+      },
+      {
+        "epoch": 0.8313402850856213,
+        "grad_norm": 0.29724085330963135,
+        "learning_rate": 5.675123856166692e-05,
+        "loss": 0.9076,
+        "step": 26070
+      },
+      {
+        "epoch": 0.8316591728052553,
+        "grad_norm": 0.30090001225471497,
+        "learning_rate": 5.653450145856519e-05,
+        "loss": 0.9193,
+        "step": 26080
+      },
+      {
+        "epoch": 0.8319780605248892,
+        "grad_norm": 0.3024527430534363,
+        "learning_rate": 5.631859209020637e-05,
+        "loss": 0.9148,
+        "step": 26090
+      },
+      {
+        "epoch": 0.832296948244523,
+        "grad_norm": 0.3030632436275482,
+        "learning_rate": 5.6103507295411355e-05,
+        "loss": 0.9211,
+        "step": 26100
+      },
+      {
+        "epoch": 0.832615835964157,
+        "grad_norm": 0.29895177483558655,
+        "learning_rate": 5.5889243925073783e-05,
+        "loss": 0.952,
+        "step": 26110
+      },
+      {
+        "epoch": 0.832934723683791,
+        "grad_norm": 0.3039778172969818,
+        "learning_rate": 5.5675798842113984e-05,
+        "loss": 0.9227,
+        "step": 26120
+      },
+      {
+        "epoch": 0.8332536114034248,
+        "grad_norm": 0.29388052225112915,
+        "learning_rate": 5.546316892143301e-05,
+        "loss": 0.9155,
+        "step": 26130
+      },
+      {
+        "epoch": 0.8335724991230588,
+        "grad_norm": 0.3008153736591339,
+        "learning_rate": 5.525135104986689e-05,
+        "loss": 0.91,
+        "step": 26140
+      },
+      {
+        "epoch": 0.8338913868426927,
+        "grad_norm": 0.2974579930305481,
+        "learning_rate": 5.5040342126141065e-05,
+        "loss": 0.92,
+        "step": 26150
+      },
+      {
+        "epoch": 0.8342102745623267,
+        "grad_norm": 0.3058636486530304,
+        "learning_rate": 5.483013906082494e-05,
+        "loss": 0.9288,
+        "step": 26160
+      },
+      {
+        "epoch": 0.8345291622819605,
+        "grad_norm": 0.29824039340019226,
+        "learning_rate": 5.462073877628674e-05,
+        "loss": 0.8927,
+        "step": 26170
+      },
+      {
+        "epoch": 0.8348480500015945,
+        "grad_norm": 0.31605926156044006,
+        "learning_rate": 5.4412138206648335e-05,
+        "loss": 0.9409,
+        "step": 26180
+      },
+      {
+        "epoch": 0.8351669377212284,
+        "grad_norm": 0.3000921308994293,
+        "learning_rate": 5.420433429774042e-05,
+        "loss": 0.9149,
+        "step": 26190
+      },
+      {
+        "epoch": 0.8354858254408623,
+        "grad_norm": 0.300586462020874,
+        "learning_rate": 5.3997324007057795e-05,
+        "loss": 0.9032,
+        "step": 26200
+      },
+      {
+        "epoch": 0.8358047131604962,
+        "grad_norm": 0.3025199770927429,
+        "learning_rate": 5.379110430371478e-05,
+        "loss": 0.9097,
+        "step": 26210
+      },
+      {
+        "epoch": 0.8361236008801302,
+        "grad_norm": 0.29137352108955383,
+        "learning_rate": 5.358567216840091e-05,
+        "loss": 0.9149,
+        "step": 26220
+      },
+      {
+        "epoch": 0.836442488599764,
+        "grad_norm": 0.30015018582344055,
+        "learning_rate": 5.338102459333666e-05,
+        "loss": 0.9139,
+        "step": 26230
+      },
+      {
+        "epoch": 0.836761376319398,
+        "grad_norm": 0.2905537188053131,
+        "learning_rate": 5.317715858222943e-05,
+        "loss": 0.9161,
+        "step": 26240
+      },
+      {
+        "epoch": 0.8370802640390319,
+        "grad_norm": 0.2965790927410126,
+        "learning_rate": 5.297407115022969e-05,
+        "loss": 0.9137,
+        "step": 26250
+      },
+      {
+        "epoch": 0.8373991517586657,
+        "grad_norm": 0.29615485668182373,
+        "learning_rate": 5.277175932388726e-05,
+        "loss": 0.8942,
+        "step": 26260
+      },
+      {
+        "epoch": 0.8377180394782997,
+        "grad_norm": 0.30239808559417725,
+        "learning_rate": 5.2570220141107796e-05,
+        "loss": 0.9362,
+        "step": 26270
+      },
+      {
+        "epoch": 0.8380369271979337,
+        "grad_norm": 0.30178868770599365,
+        "learning_rate": 5.236945065110938e-05,
+        "loss": 0.9336,
+        "step": 26280
+      },
+      {
+        "epoch": 0.8383558149175675,
+        "grad_norm": 0.2944106161594391,
+        "learning_rate": 5.216944791437939e-05,
+        "loss": 0.9131,
+        "step": 26290
+      },
+      {
+        "epoch": 0.8386747026372015,
+        "grad_norm": 0.2962855100631714,
+        "learning_rate": 5.1970209002631356e-05,
+        "loss": 0.9228,
+        "step": 26300
+      },
+      {
+        "epoch": 0.8389935903568354,
+        "grad_norm": 0.30278846621513367,
+        "learning_rate": 5.177173099876221e-05,
+        "loss": 0.9166,
+        "step": 26310
+      },
+      {
+        "epoch": 0.8393124780764692,
+        "grad_norm": 0.2937107980251312,
+        "learning_rate": 5.157401099680946e-05,
+        "loss": 0.9147,
+        "step": 26320
+      },
+      {
+        "epoch": 0.8396313657961032,
+        "grad_norm": 0.303463876247406,
+        "learning_rate": 5.13770461019087e-05,
+        "loss": 0.9206,
+        "step": 26330
+      },
+      {
+        "epoch": 0.8399502535157372,
+        "grad_norm": 0.2950541377067566,
+        "learning_rate": 5.118083343025129e-05,
+        "loss": 0.9133,
+        "step": 26340
+      },
+      {
+        "epoch": 0.840269141235371,
+        "grad_norm": 0.2916299104690552,
+        "learning_rate": 5.0985370109041986e-05,
+        "loss": 0.9022,
+        "step": 26350
+      },
+      {
+        "epoch": 0.840588028955005,
+        "grad_norm": 0.29524877667427063,
+        "learning_rate": 5.079065327645699e-05,
+        "loss": 0.9277,
+        "step": 26360
+      },
+      {
+        "epoch": 0.8409069166746389,
+        "grad_norm": 0.30931761860847473,
+        "learning_rate": 5.059668008160201e-05,
+        "loss": 0.9224,
+        "step": 26370
+      },
+      {
+        "epoch": 0.8412258043942727,
+        "grad_norm": 0.3061833679676056,
+        "learning_rate": 5.0403447684470534e-05,
+        "loss": 0.9148,
+        "step": 26380
+      },
+      {
+        "epoch": 0.8415446921139067,
+        "grad_norm": 0.2954571843147278,
+        "learning_rate": 5.021095325590223e-05,
+        "loss": 0.9069,
+        "step": 26390
+      },
+      {
+        "epoch": 0.8418635798335407,
+        "grad_norm": 0.31159186363220215,
+        "learning_rate": 5.001919397754154e-05,
+        "loss": 0.9337,
+        "step": 26400
+      },
+      {
+        "epoch": 0.8421824675531745,
+        "grad_norm": 0.29583415389060974,
+        "learning_rate": 4.9828167041796415e-05,
+        "loss": 0.9238,
+        "step": 26410
+      },
+      {
+        "epoch": 0.8425013552728084,
+        "grad_norm": 0.3016620874404907,
+        "learning_rate": 4.9637869651797196e-05,
+        "loss": 0.8986,
+        "step": 26420
+      },
+      {
+        "epoch": 0.8428202429924424,
+        "grad_norm": 0.303551584482193,
+        "learning_rate": 4.944829902135568e-05,
+        "loss": 0.918,
+        "step": 26430
+      },
+      {
+        "epoch": 0.8431391307120762,
+        "grad_norm": 0.29466649889945984,
+        "learning_rate": 4.925945237492433e-05,
+        "loss": 0.9219,
+        "step": 26440
+      },
+      {
+        "epoch": 0.8434580184317102,
+        "grad_norm": 0.30812641978263855,
+        "learning_rate": 4.9071326947555595e-05,
+        "loss": 0.9345,
+        "step": 26450
+      },
+      {
+        "epoch": 0.8437769061513442,
+        "grad_norm": 0.29291924834251404,
+        "learning_rate": 4.8883919984861484e-05,
+        "loss": 0.9256,
+        "step": 26460
+      },
+      {
+        "epoch": 0.844095793870978,
+        "grad_norm": 0.3033911883831024,
+        "learning_rate": 4.86972287429732e-05,
+        "loss": 0.9032,
+        "step": 26470
+      },
+      {
+        "epoch": 0.844414681590612,
+        "grad_norm": 0.2836741805076599,
+        "learning_rate": 4.8511250488501005e-05,
+        "loss": 0.9112,
+        "step": 26480
+      },
+      {
+        "epoch": 0.8447335693102459,
+        "grad_norm": 0.3043772876262665,
+        "learning_rate": 4.832598249849414e-05,
+        "loss": 0.9109,
+        "step": 26490
+      },
+      {
+        "epoch": 0.8450524570298797,
+        "grad_norm": 0.2880835235118866,
+        "learning_rate": 4.814142206040099e-05,
+        "loss": 0.9149,
+        "step": 26500
+      },
+      {
+        "epoch": 0.8453713447495137,
+        "grad_norm": 0.2935115098953247,
+        "learning_rate": 4.795756647202945e-05,
+        "loss": 0.9118,
+        "step": 26510
+      },
+      {
+        "epoch": 0.8456902324691477,
+        "grad_norm": 0.2969006597995758,
+        "learning_rate": 4.777441304150719e-05,
+        "loss": 0.906,
+        "step": 26520
+      },
+      {
+        "epoch": 0.8460091201887815,
+        "grad_norm": 0.30467066168785095,
+        "learning_rate": 4.759195908724238e-05,
+        "loss": 0.9099,
+        "step": 26530
+      },
+      {
+        "epoch": 0.8463280079084154,
+        "grad_norm": 0.29597529768943787,
+        "learning_rate": 4.741020193788437e-05,
+        "loss": 0.906,
+        "step": 26540
+      },
+      {
+        "epoch": 0.8466468956280494,
+        "grad_norm": 0.3062756359577179,
+        "learning_rate": 4.72291389322846e-05,
+        "loss": 0.9162,
+        "step": 26550
+      },
+      {
+        "epoch": 0.8469657833476832,
+        "grad_norm": 0.2979463040828705,
+        "learning_rate": 4.7048767419457626e-05,
+        "loss": 0.91,
+        "step": 26560
+      },
+      {
+        "epoch": 0.8472846710673172,
+        "grad_norm": 0.2843947112560272,
+        "learning_rate": 4.686908475854231e-05,
+        "loss": 0.9273,
+        "step": 26570
+      },
+      {
+        "epoch": 0.8476035587869512,
+        "grad_norm": 0.2960202395915985,
+        "learning_rate": 4.669008831876315e-05,
+        "loss": 0.909,
+        "step": 26580
+      },
+      {
+        "epoch": 0.847922446506585,
+        "grad_norm": 0.2985907196998596,
+        "learning_rate": 4.651177547939179e-05,
+        "loss": 0.9071,
+        "step": 26590
+      },
+      {
+        "epoch": 0.848241334226219,
+        "grad_norm": 0.3096676170825958,
+        "learning_rate": 4.633414362970859e-05,
+        "loss": 0.9192,
+        "step": 26600
+      },
+      {
+        "epoch": 0.8485602219458529,
+        "grad_norm": 0.31174537539482117,
+        "learning_rate": 4.6157190168964464e-05,
+        "loss": 0.9073,
+        "step": 26610
+      },
+      {
+        "epoch": 0.8488791096654867,
+        "grad_norm": 0.3036581575870514,
+        "learning_rate": 4.598091250634277e-05,
+        "loss": 0.9155,
+        "step": 26620
+      },
+      {
+        "epoch": 0.8491979973851207,
+        "grad_norm": 0.30269771814346313,
+        "learning_rate": 4.580530806092137e-05,
+        "loss": 0.9152,
+        "step": 26630
+      },
+      {
+        "epoch": 0.8495168851047546,
+        "grad_norm": 0.29733002185821533,
+        "learning_rate": 4.563037426163488e-05,
+        "loss": 0.9315,
+        "step": 26640
+      },
+      {
+        "epoch": 0.8498357728243885,
+        "grad_norm": 0.2943744659423828,
+        "learning_rate": 4.545610854723698e-05,
+        "loss": 0.9136,
+        "step": 26650
+      },
+      {
+        "epoch": 0.8501546605440224,
+        "grad_norm": 0.2963501513004303,
+        "learning_rate": 4.528250836626295e-05,
+        "loss": 0.9026,
+        "step": 26660
+      },
+      {
+        "epoch": 0.8504735482636564,
+        "grad_norm": 0.30102425813674927,
+        "learning_rate": 4.5109571176992264e-05,
+        "loss": 0.916,
+        "step": 26670
+      },
+      {
+        "epoch": 0.8507924359832902,
+        "grad_norm": 0.2941626012325287,
+        "learning_rate": 4.493729444741149e-05,
+        "loss": 0.9177,
+        "step": 26680
+      },
+      {
+        "epoch": 0.8511113237029242,
+        "grad_norm": 0.2860785126686096,
+        "learning_rate": 4.476567565517706e-05,
+        "loss": 0.9068,
+        "step": 26690
+      },
+      {
+        "epoch": 0.8514302114225581,
+        "grad_norm": 0.2924879789352417,
+        "learning_rate": 4.459471228757844e-05,
+        "loss": 0.9065,
+        "step": 26700
+      },
+      {
+        "epoch": 0.851749099142192,
+        "grad_norm": 0.2924700677394867,
+        "learning_rate": 4.442440184150135e-05,
+        "loss": 0.9184,
+        "step": 26710
+      },
+      {
+        "epoch": 0.8520679868618259,
+        "grad_norm": 0.2944236695766449,
+        "learning_rate": 4.425474182339106e-05,
+        "loss": 0.9126,
+        "step": 26720
+      },
+      {
+        "epoch": 0.8523868745814599,
+        "grad_norm": 0.2980422079563141,
+        "learning_rate": 4.40857297492159e-05,
+        "loss": 0.9211,
+        "step": 26730
+      },
+      {
+        "epoch": 0.8527057623010937,
+        "grad_norm": 0.3008560836315155,
+        "learning_rate": 4.391736314443091e-05,
+        "loss": 0.9262,
+        "step": 26740
+      },
+      {
+        "epoch": 0.8530246500207277,
+        "grad_norm": 0.2976393699645996,
+        "learning_rate": 4.37496395439416e-05,
+        "loss": 0.9227,
+        "step": 26750
+      },
+      {
+        "epoch": 0.8533435377403616,
+        "grad_norm": 0.31023627519607544,
+        "learning_rate": 4.3582556492067844e-05,
+        "loss": 0.9081,
+        "step": 26760
+      },
+      {
+        "epoch": 0.8536624254599955,
+        "grad_norm": 0.3027059733867645,
+        "learning_rate": 4.341611154250795e-05,
+        "loss": 0.9044,
+        "step": 26770
+      },
+      {
+        "epoch": 0.8539813131796294,
+        "grad_norm": 0.3005632758140564,
+        "learning_rate": 4.325030225830281e-05,
+        "loss": 0.9241,
+        "step": 26780
+      },
+      {
+        "epoch": 0.8543002008992634,
+        "grad_norm": 0.3049420118331909,
+        "learning_rate": 4.308512621180027e-05,
+        "loss": 0.9203,
+        "step": 26790
+      },
+      {
+        "epoch": 0.8546190886188972,
+        "grad_norm": 0.29897886514663696,
+        "learning_rate": 4.2920580984619533e-05,
+        "loss": 0.8969,
+        "step": 26800
+      },
+      {
+        "epoch": 0.8549379763385312,
+        "grad_norm": 0.2905649244785309,
+        "learning_rate": 4.275666416761577e-05,
+        "loss": 0.9149,
+        "step": 26810
+      },
+      {
+        "epoch": 0.8552568640581651,
+        "grad_norm": 0.30739012360572815,
+        "learning_rate": 4.259337336084486e-05,
+        "loss": 0.9262,
+        "step": 26820
+      },
+      {
+        "epoch": 0.855575751777799,
+        "grad_norm": 0.30363893508911133,
+        "learning_rate": 4.243070617352825e-05,
+        "loss": 0.9272,
+        "step": 26830
+      },
+      {
+        "epoch": 0.8558946394974329,
+        "grad_norm": 0.3002462089061737,
+        "learning_rate": 4.226866022401794e-05,
+        "loss": 0.9178,
+        "step": 26840
+      },
+      {
+        "epoch": 0.8562135272170669,
+        "grad_norm": 0.31431305408477783,
+        "learning_rate": 4.2107233139761615e-05,
+        "loss": 0.9074,
+        "step": 26850
+      },
+      {
+        "epoch": 0.8565324149367007,
+        "grad_norm": 0.3078189492225647,
+        "learning_rate": 4.194642255726791e-05,
+        "loss": 0.9078,
+        "step": 26860
+      },
+      {
+        "epoch": 0.8568513026563347,
+        "grad_norm": 0.30262264609336853,
+        "learning_rate": 4.1786226122071794e-05,
+        "loss": 0.9189,
+        "step": 26870
+      },
+      {
+        "epoch": 0.8571701903759686,
+        "grad_norm": 0.30763888359069824,
+        "learning_rate": 4.162664148870013e-05,
+        "loss": 0.9119,
+        "step": 26880
+      },
+      {
+        "epoch": 0.8574890780956025,
+        "grad_norm": 0.3012956976890564,
+        "learning_rate": 4.146766632063729e-05,
+        "loss": 0.9212,
+        "step": 26890
+      },
+      {
+        "epoch": 0.8578079658152364,
+        "grad_norm": 0.3108784556388855,
+        "learning_rate": 4.1309298290290994e-05,
+        "loss": 0.9221,
+        "step": 26900
+      },
+      {
+        "epoch": 0.8581268535348704,
+        "grad_norm": 0.3018873631954193,
+        "learning_rate": 4.1151535078958185e-05,
+        "loss": 0.9244,
+        "step": 26910
+      },
+      {
+        "epoch": 0.8584457412545042,
+        "grad_norm": 0.3096584975719452,
+        "learning_rate": 4.099437437679111e-05,
+        "loss": 0.9098,
+        "step": 26920
+      },
+      {
+        "epoch": 0.8587646289741382,
+        "grad_norm": 0.2895481288433075,
+        "learning_rate": 4.08378138827635e-05,
+        "loss": 0.9285,
+        "step": 26930
+      },
+      {
+        "epoch": 0.8590835166937721,
+        "grad_norm": 0.30135300755500793,
+        "learning_rate": 4.0681851304636857e-05,
+        "loss": 0.9091,
+        "step": 26940
+      },
+      {
+        "epoch": 0.859402404413406,
+        "grad_norm": 0.30391889810562134,
+        "learning_rate": 4.052648435892692e-05,
+        "loss": 0.9058,
+        "step": 26950
+      },
+      {
+        "epoch": 0.8597212921330399,
+        "grad_norm": 0.29739269614219666,
+        "learning_rate": 4.037171077087022e-05,
+        "loss": 0.9178,
+        "step": 26960
+      },
+      {
+        "epoch": 0.8600401798526739,
+        "grad_norm": 0.29668211936950684,
+        "learning_rate": 4.021752827439075e-05,
+        "loss": 0.8903,
+        "step": 26970
+      },
+      {
+        "epoch": 0.8603590675723078,
+        "grad_norm": 0.30262407660484314,
+        "learning_rate": 4.0063934612066855e-05,
+        "loss": 0.9003,
+        "step": 26980
+      },
+      {
+        "epoch": 0.8606779552919417,
+        "grad_norm": 0.3077700436115265,
+        "learning_rate": 3.991092753509812e-05,
+        "loss": 0.9104,
+        "step": 26990
+      },
+      {
+        "epoch": 0.8609968430115756,
+        "grad_norm": 0.29334044456481934,
+        "learning_rate": 3.975850480327241e-05,
+        "loss": 0.8973,
+        "step": 27000
+      },
+      {
+        "epoch": 0.8613157307312096,
+        "grad_norm": 0.2982562780380249,
+        "learning_rate": 3.960666418493324e-05,
+        "loss": 0.8958,
+        "step": 27010
+      },
+      {
+        "epoch": 0.8616346184508434,
+        "grad_norm": 0.3008376955986023,
+        "learning_rate": 3.9455403456946875e-05,
+        "loss": 0.9145,
+        "step": 27020
+      },
+      {
+        "epoch": 0.8619535061704774,
+        "grad_norm": 0.2966604232788086,
+        "learning_rate": 3.930472040466995e-05,
+        "loss": 0.9166,
+        "step": 27030
+      },
+      {
+        "epoch": 0.8622723938901113,
+        "grad_norm": 0.290862113237381,
+        "learning_rate": 3.915461282191693e-05,
+        "loss": 0.9264,
+        "step": 27040
+      },
+      {
+        "epoch": 0.8625912816097452,
+        "grad_norm": 0.3028498888015747,
+        "learning_rate": 3.900507851092791e-05,
+        "loss": 0.9287,
+        "step": 27050
+      },
+      {
+        "epoch": 0.8629101693293791,
+        "grad_norm": 0.3000180125236511,
+        "learning_rate": 3.885611528233638e-05,
+        "loss": 0.9223,
+        "step": 27060
+      },
+      {
+        "epoch": 0.8632290570490131,
+        "grad_norm": 0.3000931143760681,
+        "learning_rate": 3.870772095513717e-05,
+        "loss": 0.9042,
+        "step": 27070
+      },
+      {
+        "epoch": 0.8635479447686469,
+        "grad_norm": 0.2879917323589325,
+        "learning_rate": 3.855989335665453e-05,
+        "loss": 0.9207,
+        "step": 27080
+      },
+      {
+        "epoch": 0.8638668324882809,
+        "grad_norm": 0.30853670835494995,
+        "learning_rate": 3.841263032251032e-05,
+        "loss": 0.9243,
+        "step": 27090
+      },
+      {
+        "epoch": 0.8641857202079148,
+        "grad_norm": 0.2945482134819031,
+        "learning_rate": 3.8265929696592315e-05,
+        "loss": 0.9161,
+        "step": 27100
+      },
+      {
+        "epoch": 0.8645046079275487,
+        "grad_norm": 0.29502424597740173,
+        "learning_rate": 3.811978933102264e-05,
+        "loss": 0.9068,
+        "step": 27110
+      },
+      {
+        "epoch": 0.8648234956471826,
+        "grad_norm": 0.29816558957099915,
+        "learning_rate": 3.797420708612632e-05,
+        "loss": 0.9172,
+        "step": 27120
+      },
+      {
+        "epoch": 0.8651423833668166,
+        "grad_norm": 0.29900944232940674,
+        "learning_rate": 3.7829180830399963e-05,
+        "loss": 0.9034,
+        "step": 27130
+      },
+      {
+        "epoch": 0.8654612710864504,
+        "grad_norm": 0.29612571001052856,
+        "learning_rate": 3.768470844048052e-05,
+        "loss": 0.8965,
+        "step": 27140
+      },
+      {
+        "epoch": 0.8657801588060844,
+        "grad_norm": 0.29050344228744507,
+        "learning_rate": 3.7540787801114243e-05,
+        "loss": 0.9144,
+        "step": 27150
+      },
+      {
+        "epoch": 0.8660990465257183,
+        "grad_norm": 0.2979932725429535,
+        "learning_rate": 3.739741680512569e-05,
+        "loss": 0.9189,
+        "step": 27160
+      },
+      {
+        "epoch": 0.8664179342453522,
+        "grad_norm": 0.3041857182979584,
+        "learning_rate": 3.725459335338685e-05,
+        "loss": 0.9034,
+        "step": 27170
+      },
+      {
+        "epoch": 0.8667368219649861,
+        "grad_norm": 0.2961612343788147,
+        "learning_rate": 3.711231535478648e-05,
+        "loss": 0.9151,
+        "step": 27180
+      },
+      {
+        "epoch": 0.8670557096846201,
+        "grad_norm": 0.2977130711078644,
+        "learning_rate": 3.697058072619941e-05,
+        "loss": 0.912,
+        "step": 27190
+      },
+      {
+        "epoch": 0.8673745974042539,
+        "grad_norm": 0.2953384518623352,
+        "learning_rate": 3.6829387392456075e-05,
+        "loss": 0.9095,
+        "step": 27200
+      },
+      {
+        "epoch": 0.8676934851238879,
+        "grad_norm": 0.30030712485313416,
+        "learning_rate": 3.668873328631214e-05,
+        "loss": 0.908,
+        "step": 27210
+      },
+      {
+        "epoch": 0.8680123728435218,
+        "grad_norm": 0.29483762383461,
+        "learning_rate": 3.6548616348418236e-05,
+        "loss": 0.9091,
+        "step": 27220
+      },
+      {
+        "epoch": 0.8683312605631557,
+        "grad_norm": 0.2982817590236664,
+        "learning_rate": 3.640903452728978e-05,
+        "loss": 0.9046,
+        "step": 27230
+      },
+      {
+        "epoch": 0.8686501482827896,
+        "grad_norm": 0.3042912483215332,
+        "learning_rate": 3.626998577927698e-05,
+        "loss": 0.9053,
+        "step": 27240
+      },
+      {
+        "epoch": 0.8689690360024236,
+        "grad_norm": 0.30695784091949463,
+        "learning_rate": 3.6131468068534876e-05,
+        "loss": 0.9161,
+        "step": 27250
+      },
+      {
+        "epoch": 0.8692879237220574,
+        "grad_norm": 0.29794302582740784,
+        "learning_rate": 3.599347936699354e-05,
+        "loss": 0.9065,
+        "step": 27260
+      },
+      {
+        "epoch": 0.8696068114416914,
+        "grad_norm": 0.29503190517425537,
+        "learning_rate": 3.585601765432841e-05,
+        "loss": 0.9098,
+        "step": 27270
+      },
+      {
+        "epoch": 0.8699256991613253,
+        "grad_norm": 0.30191126465797424,
+        "learning_rate": 3.571908091793068e-05,
+        "loss": 0.9056,
+        "step": 27280
+      },
+      {
+        "epoch": 0.8702445868809592,
+        "grad_norm": 0.29695814847946167,
+        "learning_rate": 3.558266715287785e-05,
+        "loss": 0.9103,
+        "step": 27290
+      },
+      {
+        "epoch": 0.8705634746005931,
+        "grad_norm": 0.3085416555404663,
+        "learning_rate": 3.544677436190435e-05,
+        "loss": 0.9198,
+        "step": 27300
+      },
+      {
+        "epoch": 0.8708823623202271,
+        "grad_norm": 0.30874142050743103,
+        "learning_rate": 3.5311400555372326e-05,
+        "loss": 0.929,
+        "step": 27310
+      },
+      {
+        "epoch": 0.8712012500398609,
+        "grad_norm": 0.2925911545753479,
+        "learning_rate": 3.517654375124249e-05,
+        "loss": 0.9045,
+        "step": 27320
+      },
+      {
+        "epoch": 0.8715201377594949,
+        "grad_norm": 0.2910749018192291,
+        "learning_rate": 3.5042201975045115e-05,
+        "loss": 0.9206,
+        "step": 27330
+      },
+      {
+        "epoch": 0.8718390254791288,
+        "grad_norm": 0.29372018575668335,
+        "learning_rate": 3.490837325985108e-05,
+        "loss": 0.9189,
+        "step": 27340
+      },
+      {
+        "epoch": 0.8721579131987627,
+        "grad_norm": 0.29507362842559814,
+        "learning_rate": 3.4775055646243186e-05,
+        "loss": 0.9068,
+        "step": 27350
+      },
+      {
+        "epoch": 0.8724768009183966,
+        "grad_norm": 0.3026776611804962,
+        "learning_rate": 3.464224718228731e-05,
+        "loss": 0.9182,
+        "step": 27360
+      },
+      {
+        "epoch": 0.8727956886380306,
+        "grad_norm": 0.2964503765106201,
+        "learning_rate": 3.450994592350395e-05,
+        "loss": 0.9027,
+        "step": 27370
+      },
+      {
+        "epoch": 0.8731145763576644,
+        "grad_norm": 0.29183635115623474,
+        "learning_rate": 3.437814993283972e-05,
+        "loss": 0.9101,
+        "step": 27380
+      },
+      {
+        "epoch": 0.8734334640772984,
+        "grad_norm": 0.3055545687675476,
+        "learning_rate": 3.424685728063894e-05,
+        "loss": 0.9126,
+        "step": 27390
+      },
+      {
+        "epoch": 0.8737523517969323,
+        "grad_norm": 0.29662030935287476,
+        "learning_rate": 3.411606604461545e-05,
+        "loss": 0.9079,
+        "step": 27400
+      },
+      {
+        "epoch": 0.8740712395165662,
+        "grad_norm": 0.2959885895252228,
+        "learning_rate": 3.398577430982446e-05,
+        "loss": 0.8983,
+        "step": 27410
+      },
+      {
+        "epoch": 0.8743901272362001,
+        "grad_norm": 0.298469603061676,
+        "learning_rate": 3.385598016863445e-05,
+        "loss": 0.9029,
+        "step": 27420
+      },
+      {
+        "epoch": 0.8747090149558341,
+        "grad_norm": 0.2938048839569092,
+        "learning_rate": 3.372668172069933e-05,
+        "loss": 0.9121,
+        "step": 27430
+      },
+      {
+        "epoch": 0.8750279026754679,
+        "grad_norm": 0.3020491898059845,
+        "learning_rate": 3.3597877072930536e-05,
+        "loss": 0.9042,
+        "step": 27440
+      },
+      {
+        "epoch": 0.8753467903951019,
+        "grad_norm": 0.30387693643569946,
+        "learning_rate": 3.3469564339469355e-05,
+        "loss": 0.9113,
+        "step": 27450
+      },
+      {
+        "epoch": 0.8756656781147358,
+        "grad_norm": 0.2981345057487488,
+        "learning_rate": 3.334174164165931e-05,
+        "loss": 0.906,
+        "step": 27460
+      },
+      {
+        "epoch": 0.8759845658343697,
+        "grad_norm": 0.2940691113471985,
+        "learning_rate": 3.321440710801865e-05,
+        "loss": 0.9156,
+        "step": 27470
+      },
+      {
+        "epoch": 0.8763034535540036,
+        "grad_norm": 0.3027077317237854,
+        "learning_rate": 3.308755887421296e-05,
+        "loss": 0.9064,
+        "step": 27480
+      },
+      {
+        "epoch": 0.8766223412736376,
+        "grad_norm": 0.2986266613006592,
+        "learning_rate": 3.296119508302781e-05,
+        "loss": 0.8992,
+        "step": 27490
+      },
+      {
+        "epoch": 0.8769412289932714,
+        "grad_norm": 0.3028486669063568,
+        "learning_rate": 3.2835313884341655e-05,
+        "loss": 0.9183,
+        "step": 27500
+      },
+      {
+        "epoch": 0.8772601167129054,
+        "grad_norm": 0.3042612075805664,
+        "learning_rate": 3.2709913435098666e-05,
+        "loss": 0.9037,
+        "step": 27510
+      },
+      {
+        "epoch": 0.8775790044325393,
+        "grad_norm": 0.3072284758090973,
+        "learning_rate": 3.2584991899281825e-05,
+        "loss": 0.9171,
+        "step": 27520
+      },
+      {
+        "epoch": 0.8778978921521732,
+        "grad_norm": 0.29810744524002075,
+        "learning_rate": 3.246054744788594e-05,
+        "loss": 0.8848,
+        "step": 27530
+      },
+      {
+        "epoch": 0.8782167798718071,
+        "grad_norm": 0.2982387840747833,
+        "learning_rate": 3.233657825889095e-05,
+        "loss": 0.9106,
+        "step": 27540
+      },
+      {
+        "epoch": 0.8785356675914411,
+        "grad_norm": 0.3004058003425598,
+        "learning_rate": 3.221308251723522e-05,
+        "loss": 0.9152,
+        "step": 27550
+      },
+      {
+        "epoch": 0.8788545553110749,
+        "grad_norm": 0.2940619885921478,
+        "learning_rate": 3.2090058414788956e-05,
+        "loss": 0.9059,
+        "step": 27560
+      },
+      {
+        "epoch": 0.8791734430307089,
+        "grad_norm": 0.29345858097076416,
+        "learning_rate": 3.196750415032777e-05,
+        "loss": 0.9055,
+        "step": 27570
+      },
+      {
+        "epoch": 0.8794923307503428,
+        "grad_norm": 0.3052271902561188,
+        "learning_rate": 3.1845417929506246e-05,
+        "loss": 0.9127,
+        "step": 27580
+      },
+      {
+        "epoch": 0.8798112184699767,
+        "grad_norm": 0.30265307426452637,
+        "learning_rate": 3.1723797964831725e-05,
+        "loss": 0.9087,
+        "step": 27590
+      },
+      {
+        "epoch": 0.8801301061896106,
+        "grad_norm": 0.3081713318824768,
+        "learning_rate": 3.160264247563812e-05,
+        "loss": 0.908,
+        "step": 27600
+      },
+      {
+        "epoch": 0.8804489939092446,
+        "grad_norm": 0.302143931388855,
+        "learning_rate": 3.1481949688059806e-05,
+        "loss": 0.9051,
+        "step": 27610
+      },
+      {
+        "epoch": 0.8807678816288784,
+        "grad_norm": 0.3035591244697571,
+        "learning_rate": 3.1361717835005704e-05,
+        "loss": 0.9119,
+        "step": 27620
+      },
+      {
+        "epoch": 0.8810867693485124,
+        "grad_norm": 0.30098700523376465,
+        "learning_rate": 3.1241945156133386e-05,
+        "loss": 0.9125,
+        "step": 27630
+      },
+      {
+        "epoch": 0.8814056570681463,
+        "grad_norm": 0.3099716305732727,
+        "learning_rate": 3.1122629897823284e-05,
+        "loss": 0.906,
+        "step": 27640
+      },
+      {
+        "epoch": 0.8817245447877802,
+        "grad_norm": 0.2987038791179657,
+        "learning_rate": 3.100377031315304e-05,
+        "loss": 0.9013,
+        "step": 27650
+      },
+      {
+        "epoch": 0.8820434325074141,
+        "grad_norm": 0.30601197481155396,
+        "learning_rate": 3.088536466187193e-05,
+        "loss": 0.9189,
+        "step": 27660
+      },
+      {
+        "epoch": 0.8823623202270481,
+        "grad_norm": 0.29450488090515137,
+        "learning_rate": 3.076741121037534e-05,
+        "loss": 0.8981,
+        "step": 27670
+      },
+      {
+        "epoch": 0.8826812079466819,
+        "grad_norm": 0.3021726608276367,
+        "learning_rate": 3.064990823167945e-05,
+        "loss": 0.91,
+        "step": 27680
+      },
+      {
+        "epoch": 0.8830000956663159,
+        "grad_norm": 0.30047470331192017,
+        "learning_rate": 3.053285400539591e-05,
+        "loss": 0.9283,
+        "step": 27690
+      },
+      {
+        "epoch": 0.8833189833859498,
+        "grad_norm": 0.3061404526233673,
+        "learning_rate": 3.041624681770667e-05,
+        "loss": 0.9088,
+        "step": 27700
+      },
+      {
+        "epoch": 0.8836378711055837,
+        "grad_norm": 0.3043230175971985,
+        "learning_rate": 3.030008496133884e-05,
+        "loss": 0.8921,
+        "step": 27710
+      },
+      {
+        "epoch": 0.8839567588252176,
+        "grad_norm": 0.2976597845554352,
+        "learning_rate": 3.0184366735539748e-05,
+        "loss": 0.9011,
+        "step": 27720
+      },
+      {
+        "epoch": 0.8842756465448516,
+        "grad_norm": 0.29181239008903503,
+        "learning_rate": 3.006909044605203e-05,
+        "loss": 0.9071,
+        "step": 27730
+      },
+      {
+        "epoch": 0.8845945342644854,
+        "grad_norm": 0.30034399032592773,
+        "learning_rate": 2.995425440508881e-05,
+        "loss": 0.8866,
+        "step": 27740
+      },
+      {
+        "epoch": 0.8849134219841194,
+        "grad_norm": 0.29232257604599,
+        "learning_rate": 2.983985693130898e-05,
+        "loss": 0.9205,
+        "step": 27750
+      },
+      {
+        "epoch": 0.8852323097037533,
+        "grad_norm": 0.3077869713306427,
+        "learning_rate": 2.9725896349792608e-05,
+        "loss": 0.9076,
+        "step": 27760
+      },
+      {
+        "epoch": 0.8855511974233872,
+        "grad_norm": 0.2921328544616699,
+        "learning_rate": 2.9612370992016398e-05,
+        "loss": 0.908,
+        "step": 27770
+      },
+      {
+        "epoch": 0.8858700851430211,
+        "grad_norm": 0.3000868260860443,
+        "learning_rate": 2.9499279195829267e-05,
+        "loss": 0.8937,
+        "step": 27780
+      },
+      {
+        "epoch": 0.8861889728626551,
+        "grad_norm": 0.29569077491760254,
+        "learning_rate": 2.9386619305428005e-05,
+        "loss": 0.9016,
+        "step": 27790
+      },
+      {
+        "epoch": 0.886507860582289,
+        "grad_norm": 0.3077255189418793,
+        "learning_rate": 2.9274389671333044e-05,
+        "loss": 0.9156,
+        "step": 27800
+      },
+      {
+        "epoch": 0.8868267483019229,
+        "grad_norm": 0.3058718144893646,
+        "learning_rate": 2.9162588650364285e-05,
+        "loss": 0.9102,
+        "step": 27810
+      },
+      {
+        "epoch": 0.8871456360215568,
+        "grad_norm": 0.2974928021430969,
+        "learning_rate": 2.9051214605617067e-05,
+        "loss": 0.8916,
+        "step": 27820
+      },
+      {
+        "epoch": 0.8874645237411908,
+        "grad_norm": 0.2975204885005951,
+        "learning_rate": 2.8940265906438172e-05,
+        "loss": 0.9004,
+        "step": 27830
+      },
+      {
+        "epoch": 0.8877834114608246,
+        "grad_norm": 0.301376074552536,
+        "learning_rate": 2.882974092840196e-05,
+        "loss": 0.9053,
+        "step": 27840
+      },
+      {
+        "epoch": 0.8881022991804586,
+        "grad_norm": 0.29913774132728577,
+        "learning_rate": 2.871963805328664e-05,
+        "loss": 0.9204,
+        "step": 27850
+      },
+      {
+        "epoch": 0.8884211869000925,
+        "grad_norm": 0.298863023519516,
+        "learning_rate": 2.860995566905046e-05,
+        "loss": 0.9073,
+        "step": 27860
+      },
+      {
+        "epoch": 0.8887400746197264,
+        "grad_norm": 0.2911647856235504,
+        "learning_rate": 2.850069216980822e-05,
+        "loss": 0.9052,
+        "step": 27870
+      },
+      {
+        "epoch": 0.8890589623393603,
+        "grad_norm": 0.30974212288856506,
+        "learning_rate": 2.8391845955807693e-05,
+        "loss": 0.9118,
+        "step": 27880
+      },
+      {
+        "epoch": 0.8893778500589943,
+        "grad_norm": 0.3009479343891144,
+        "learning_rate": 2.8283415433406215e-05,
+        "loss": 0.8883,
+        "step": 27890
+      },
+      {
+        "epoch": 0.8896967377786281,
+        "grad_norm": 0.29356539249420166,
+        "learning_rate": 2.8175399015047376e-05,
+        "loss": 0.914,
+        "step": 27900
+      },
+      {
+        "epoch": 0.8900156254982621,
+        "grad_norm": 0.29100754857063293,
+        "learning_rate": 2.8067795119237755e-05,
+        "loss": 0.9132,
+        "step": 27910
+      },
+      {
+        "epoch": 0.890334513217896,
+        "grad_norm": 0.29773107171058655,
+        "learning_rate": 2.796060217052376e-05,
+        "loss": 0.9217,
+        "step": 27920
+      },
+      {
+        "epoch": 0.8906534009375299,
+        "grad_norm": 0.3045399487018585,
+        "learning_rate": 2.785381859946858e-05,
+        "loss": 0.9037,
+        "step": 27930
+      },
+      {
+        "epoch": 0.8909722886571638,
+        "grad_norm": 0.29568803310394287,
+        "learning_rate": 2.7747442842629192e-05,
+        "loss": 0.9042,
+        "step": 27940
+      },
+      {
+        "epoch": 0.8912911763767978,
+        "grad_norm": 0.2984471023082733,
+        "learning_rate": 2.7641473342533486e-05,
+        "loss": 0.9092,
+        "step": 27950
+      },
+      {
+        "epoch": 0.8916100640964316,
+        "grad_norm": 0.31042030453681946,
+        "learning_rate": 2.753590854765743e-05,
+        "loss": 0.9193,
+        "step": 27960
+      },
+      {
+        "epoch": 0.8919289518160656,
+        "grad_norm": 0.30285584926605225,
+        "learning_rate": 2.7430746912402397e-05,
+        "loss": 0.9192,
+        "step": 27970
+      },
+      {
+        "epoch": 0.8922478395356995,
+        "grad_norm": 0.30536022782325745,
+        "learning_rate": 2.73259868970725e-05,
+        "loss": 0.9267,
+        "step": 27980
+      },
+      {
+        "epoch": 0.8925667272553334,
+        "grad_norm": 0.3031696677207947,
+        "learning_rate": 2.722162696785207e-05,
+        "loss": 0.9038,
+        "step": 27990
+      },
+      {
+        "epoch": 0.8928856149749673,
+        "grad_norm": 0.2872668504714966,
+        "learning_rate": 2.711766559678319e-05,
+        "loss": 0.9067,
+        "step": 28000
+      },
+      {
+        "epoch": 0.8932045026946013,
+        "grad_norm": 0.30583876371383667,
+        "learning_rate": 2.7014101261743303e-05,
+        "loss": 0.9048,
+        "step": 28010
+      },
+      {
+        "epoch": 0.8935233904142351,
+        "grad_norm": 0.2939108610153198,
+        "learning_rate": 2.6910932446423003e-05,
+        "loss": 0.9092,
+        "step": 28020
+      },
+      {
+        "epoch": 0.8938422781338691,
+        "grad_norm": 0.29195529222488403,
+        "learning_rate": 2.6808157640303726e-05,
+        "loss": 0.9137,
+        "step": 28030
+      },
+      {
+        "epoch": 0.894161165853503,
+        "grad_norm": 0.29214227199554443,
+        "learning_rate": 2.670577533863571e-05,
+        "loss": 0.9005,
+        "step": 28040
+      },
+      {
+        "epoch": 0.8944800535731369,
+        "grad_norm": 0.3046804666519165,
+        "learning_rate": 2.6603784042415948e-05,
+        "loss": 0.8878,
+        "step": 28050
+      },
+      {
+        "epoch": 0.8947989412927708,
+        "grad_norm": 0.3029037117958069,
+        "learning_rate": 2.6502182258366217e-05,
+        "loss": 0.9043,
+        "step": 28060
+      },
+      {
+        "epoch": 0.8951178290124048,
+        "grad_norm": 0.3057042062282562,
+        "learning_rate": 2.640096849891124e-05,
+        "loss": 0.9032,
+        "step": 28070
+      },
+      {
+        "epoch": 0.8954367167320386,
+        "grad_norm": 0.2865537106990814,
+        "learning_rate": 2.630014128215691e-05,
+        "loss": 0.9048,
+        "step": 28080
+      },
+      {
+        "epoch": 0.8957556044516726,
+        "grad_norm": 0.3084664046764374,
+        "learning_rate": 2.6199699131868562e-05,
+        "loss": 0.9025,
+        "step": 28090
+      },
+      {
+        "epoch": 0.8960744921713065,
+        "grad_norm": 0.3067476749420166,
+        "learning_rate": 2.6099640577449387e-05,
+        "loss": 0.9095,
+        "step": 28100
+      },
+      {
+        "epoch": 0.8963933798909404,
+        "grad_norm": 0.2988032400608063,
+        "learning_rate": 2.5999964153918897e-05,
+        "loss": 0.9005,
+        "step": 28110
+      },
+      {
+        "epoch": 0.8967122676105743,
+        "grad_norm": 0.30241847038269043,
+        "learning_rate": 2.5900668401891458e-05,
+        "loss": 0.9179,
+        "step": 28120
+      },
+      {
+        "epoch": 0.8970311553302083,
+        "grad_norm": 0.2994360029697418,
+        "learning_rate": 2.5801751867554953e-05,
+        "loss": 0.8999,
+        "step": 28130
+      },
+      {
+        "epoch": 0.8973500430498421,
+        "grad_norm": 0.3065774440765381,
+        "learning_rate": 2.570321310264946e-05,
+        "loss": 0.9246,
+        "step": 28140
+      },
+      {
+        "epoch": 0.8976689307694761,
+        "grad_norm": 0.30343982577323914,
+        "learning_rate": 2.5605050664446088e-05,
+        "loss": 0.9103,
+        "step": 28150
+      },
+      {
+        "epoch": 0.89798781848911,
+        "grad_norm": 0.2930835485458374,
+        "learning_rate": 2.550726311572581e-05,
+        "loss": 0.9086,
+        "step": 28160
+      },
+      {
+        "epoch": 0.8983067062087439,
+        "grad_norm": 0.30507856607437134,
+        "learning_rate": 2.540984902475846e-05,
+        "loss": 0.8979,
+        "step": 28170
+      },
+      {
+        "epoch": 0.8986255939283778,
+        "grad_norm": 0.2992618680000305,
+        "learning_rate": 2.5312806965281734e-05,
+        "loss": 0.9004,
+        "step": 28180
+      },
+      {
+        "epoch": 0.8989444816480118,
+        "grad_norm": 0.29800593852996826,
+        "learning_rate": 2.5216135516480365e-05,
+        "loss": 0.9016,
+        "step": 28190
+      },
+      {
+        "epoch": 0.8992633693676456,
+        "grad_norm": 0.2914922833442688,
+        "learning_rate": 2.5119833262965248e-05,
+        "loss": 0.9049,
+        "step": 28200
+      },
+      {
+        "epoch": 0.8995822570872796,
+        "grad_norm": 0.303462952375412,
+        "learning_rate": 2.502389879475276e-05,
+        "loss": 0.91,
+        "step": 28210
+      },
+      {
+        "epoch": 0.8999011448069135,
+        "grad_norm": 0.3019823431968689,
+        "learning_rate": 2.4928330707244114e-05,
+        "loss": 0.9172,
+        "step": 28220
+      },
+      {
+        "epoch": 0.9002200325265474,
+        "grad_norm": 0.2922445833683014,
+        "learning_rate": 2.4833127601204777e-05,
+        "loss": 0.8916,
+        "step": 28230
+      },
+      {
+        "epoch": 0.9005389202461813,
+        "grad_norm": 0.30021733045578003,
+        "learning_rate": 2.4738288082744006e-05,
+        "loss": 0.9124,
+        "step": 28240
+      },
+      {
+        "epoch": 0.9008578079658153,
+        "grad_norm": 0.2954046130180359,
+        "learning_rate": 2.4643810763294427e-05,
+        "loss": 0.9114,
+        "step": 28250
+      },
+      {
+        "epoch": 0.9011766956854491,
+        "grad_norm": 0.29407021403312683,
+        "learning_rate": 2.4549694259591695e-05,
+        "loss": 0.8875,
+        "step": 28260
+      },
+      {
+        "epoch": 0.9014955834050831,
+        "grad_norm": 0.30686795711517334,
+        "learning_rate": 2.4455937193654268e-05,
+        "loss": 0.9125,
+        "step": 28270
+      },
+      {
+        "epoch": 0.901814471124717,
+        "grad_norm": 0.29066452383995056,
+        "learning_rate": 2.4362538192763198e-05,
+        "loss": 0.8989,
+        "step": 28280
+      },
+      {
+        "epoch": 0.9021333588443509,
+        "grad_norm": 0.2918499708175659,
+        "learning_rate": 2.4269495889442062e-05,
+        "loss": 0.9129,
+        "step": 28290
+      },
+      {
+        "epoch": 0.9024522465639848,
+        "grad_norm": 0.30791276693344116,
+        "learning_rate": 2.417680892143693e-05,
+        "loss": 0.9288,
+        "step": 28300
+      },
+      {
+        "epoch": 0.9027711342836188,
+        "grad_norm": 0.2873190939426422,
+        "learning_rate": 2.4084475931696406e-05,
+        "loss": 0.9119,
+        "step": 28310
+      },
+      {
+        "epoch": 0.9030900220032526,
+        "grad_norm": 0.3017880916595459,
+        "learning_rate": 2.399249556835179e-05,
+        "loss": 0.8997,
+        "step": 28320
+      },
+      {
+        "epoch": 0.9034089097228866,
+        "grad_norm": 0.2940405309200287,
+        "learning_rate": 2.390086648469725e-05,
+        "loss": 0.9052,
+        "step": 28330
+      },
+      {
+        "epoch": 0.9037277974425205,
+        "grad_norm": 0.2958362400531769,
+        "learning_rate": 2.3809587339170133e-05,
+        "loss": 0.8872,
+        "step": 28340
+      },
+      {
+        "epoch": 0.9040466851621544,
+        "grad_norm": 0.304292768239975,
+        "learning_rate": 2.3718656795331296e-05,
+        "loss": 0.9111,
+        "step": 28350
+      },
+      {
+        "epoch": 0.9043655728817883,
+        "grad_norm": 0.2933245599269867,
+        "learning_rate": 2.362807352184559e-05,
+        "loss": 0.9089,
+        "step": 28360
+      },
+      {
+        "epoch": 0.9046844606014223,
+        "grad_norm": 0.29516878724098206,
+        "learning_rate": 2.3537836192462286e-05,
+        "loss": 0.91,
+        "step": 28370
+      },
+      {
+        "epoch": 0.9050033483210561,
+        "grad_norm": 0.2948715388774872,
+        "learning_rate": 2.344794348599573e-05,
+        "loss": 0.8964,
+        "step": 28380
+      },
+      {
+        "epoch": 0.9053222360406901,
+        "grad_norm": 0.2943824231624603,
+        "learning_rate": 2.3358394086305966e-05,
+        "loss": 0.8966,
+        "step": 28390
+      },
+      {
+        "epoch": 0.905641123760324,
+        "grad_norm": 0.30269181728363037,
+        "learning_rate": 2.3269186682279475e-05,
+        "loss": 0.9035,
+        "step": 28400
+      },
+      {
+        "epoch": 0.9059600114799579,
+        "grad_norm": 0.30818644165992737,
+        "learning_rate": 2.3180319967809967e-05,
+        "loss": 0.9079,
+        "step": 28410
+      },
+      {
+        "epoch": 0.9062788991995918,
+        "grad_norm": 0.29804039001464844,
+        "learning_rate": 2.3091792641779272e-05,
+        "loss": 0.9004,
+        "step": 28420
+      },
+      {
+        "epoch": 0.9065977869192258,
+        "grad_norm": 0.2957077622413635,
+        "learning_rate": 2.300360340803829e-05,
+        "loss": 0.9246,
+        "step": 28430
+      },
+      {
+        "epoch": 0.9069166746388596,
+        "grad_norm": 0.30506861209869385,
+        "learning_rate": 2.2915750975388005e-05,
+        "loss": 0.9161,
+        "step": 28440
+      },
+      {
+        "epoch": 0.9072355623584936,
+        "grad_norm": 0.29112300276756287,
+        "learning_rate": 2.2828234057560574e-05,
+        "loss": 0.9136,
+        "step": 28450
+      },
+      {
+        "epoch": 0.9075544500781275,
+        "grad_norm": 0.30500057339668274,
+        "learning_rate": 2.2741051373200522e-05,
+        "loss": 0.8983,
+        "step": 28460
+      },
+      {
+        "epoch": 0.9078733377977614,
+        "grad_norm": 0.2991545796394348,
+        "learning_rate": 2.265420164584595e-05,
+        "loss": 0.9081,
+        "step": 28470
+      },
+      {
+        "epoch": 0.9081922255173953,
+        "grad_norm": 0.2967453896999359,
+        "learning_rate": 2.2567683603909864e-05,
+        "loss": 0.9075,
+        "step": 28480
+      },
+      {
+        "epoch": 0.9085111132370293,
+        "grad_norm": 0.3122044503688812,
+        "learning_rate": 2.2481495980661557e-05,
+        "loss": 0.9134,
+        "step": 28490
+      },
+      {
+        "epoch": 0.9088300009566631,
+        "grad_norm": 0.30028966069221497,
+        "learning_rate": 2.239563751420805e-05,
+        "loss": 0.9082,
+        "step": 28500
+      },
+      {
+        "epoch": 0.9091488886762971,
+        "grad_norm": 0.2991483807563782,
+        "learning_rate": 2.2310106947475637e-05,
+        "loss": 0.9027,
+        "step": 28510
+      },
+      {
+        "epoch": 0.909467776395931,
+        "grad_norm": 0.2894665598869324,
+        "learning_rate": 2.2224903028191445e-05,
+        "loss": 0.913,
+        "step": 28520
+      },
+      {
+        "epoch": 0.9097866641155649,
+        "grad_norm": 0.295378178358078,
+        "learning_rate": 2.2140024508865157e-05,
+        "loss": 0.8794,
+        "step": 28530
+      },
+      {
+        "epoch": 0.9101055518351988,
+        "grad_norm": 0.2910739481449127,
+        "learning_rate": 2.205547014677069e-05,
+        "loss": 0.9054,
+        "step": 28540
+      },
+      {
+        "epoch": 0.9104244395548328,
+        "grad_norm": 0.29597222805023193,
+        "learning_rate": 2.197123870392802e-05,
+        "loss": 0.8965,
+        "step": 28550
+      },
+      {
+        "epoch": 0.9107433272744666,
+        "grad_norm": 0.30073556303977966,
+        "learning_rate": 2.1887328947085065e-05,
+        "loss": 0.9167,
+        "step": 28560
+      },
+      {
+        "epoch": 0.9110622149941006,
+        "grad_norm": 0.2994864583015442,
+        "learning_rate": 2.1803739647699623e-05,
+        "loss": 0.9093,
+        "step": 28570
+      },
+      {
+        "epoch": 0.9113811027137345,
+        "grad_norm": 0.2985394597053528,
+        "learning_rate": 2.172046958192138e-05,
+        "loss": 0.9105,
+        "step": 28580
+      },
+      {
+        "epoch": 0.9116999904333684,
+        "grad_norm": 0.2986948490142822,
+        "learning_rate": 2.1637517530574002e-05,
+        "loss": 0.9166,
+        "step": 28590
+      },
+      {
+        "epoch": 0.9120188781530023,
+        "grad_norm": 0.3221426010131836,
+        "learning_rate": 2.155488227913727e-05,
+        "loss": 0.9234,
+        "step": 28600
+      },
+      {
+        "epoch": 0.9123377658726363,
+        "grad_norm": 0.31154584884643555,
+        "learning_rate": 2.14725626177293e-05,
+        "loss": 0.9182,
+        "step": 28610
+      },
+      {
+        "epoch": 0.9126566535922701,
+        "grad_norm": 0.2910720705986023,
+        "learning_rate": 2.139055734108886e-05,
+        "loss": 0.9048,
+        "step": 28620
+      },
+      {
+        "epoch": 0.9129755413119041,
+        "grad_norm": 0.29777389764785767,
+        "learning_rate": 2.1308865248557674e-05,
+        "loss": 0.8945,
+        "step": 28630
+      },
+      {
+        "epoch": 0.913294429031538,
+        "grad_norm": 0.2937983572483063,
+        "learning_rate": 2.122748514406288e-05,
+        "loss": 0.9024,
+        "step": 28640
+      },
+      {
+        "epoch": 0.913613316751172,
+        "grad_norm": 0.3081187605857849,
+        "learning_rate": 2.1146415836099496e-05,
+        "loss": 0.8989,
+        "step": 28650
+      },
+      {
+        "epoch": 0.9139322044708058,
+        "grad_norm": 0.3129999339580536,
+        "learning_rate": 2.1065656137712995e-05,
+        "loss": 0.905,
+        "step": 28660
+      },
+      {
+        "epoch": 0.9142510921904398,
+        "grad_norm": 0.29838865995407104,
+        "learning_rate": 2.0985204866481902e-05,
+        "loss": 0.9085,
+        "step": 28670
+      },
+      {
+        "epoch": 0.9145699799100737,
+        "grad_norm": 0.3048774302005768,
+        "learning_rate": 2.0905060844500506e-05,
+        "loss": 0.9148,
+        "step": 28680
+      },
+      {
+        "epoch": 0.9148888676297076,
+        "grad_norm": 0.30030128359794617,
+        "learning_rate": 2.0825222898361594e-05,
+        "loss": 0.9164,
+        "step": 28690
+      },
+      {
+        "epoch": 0.9152077553493415,
+        "grad_norm": 0.30484727025032043,
+        "learning_rate": 2.07456898591393e-05,
+        "loss": 0.9038,
+        "step": 28700
+      },
+      {
+        "epoch": 0.9155266430689755,
+        "grad_norm": 0.3036182224750519,
+        "learning_rate": 2.0666460562371957e-05,
+        "loss": 0.9036,
+        "step": 28710
+      },
+      {
+        "epoch": 0.9158455307886093,
+        "grad_norm": 0.30354034900665283,
+        "learning_rate": 2.0587533848045053e-05,
+        "loss": 0.9091,
+        "step": 28720
+      },
+      {
+        "epoch": 0.9161644185082433,
+        "grad_norm": 0.29555216431617737,
+        "learning_rate": 2.050890856057427e-05,
+        "loss": 0.9229,
+        "step": 28730
+      },
+      {
+        "epoch": 0.9164833062278772,
+        "grad_norm": 0.3115293085575104,
+        "learning_rate": 2.0430583548788544e-05,
+        "loss": 0.9274,
+        "step": 28740
+      },
+      {
+        "epoch": 0.9168021939475111,
+        "grad_norm": 0.29061922430992126,
+        "learning_rate": 2.0352557665913218e-05,
+        "loss": 0.915,
+        "step": 28750
+      },
+      {
+        "epoch": 0.917121081667145,
+        "grad_norm": 0.2968040704727173,
+        "learning_rate": 2.027482976955325e-05,
+        "loss": 0.9089,
+        "step": 28760
+      },
+      {
+        "epoch": 0.917439969386779,
+        "grad_norm": 0.3121855556964874,
+        "learning_rate": 2.0197398721676495e-05,
+        "loss": 0.9059,
+        "step": 28770
+      },
+      {
+        "epoch": 0.9177588571064128,
+        "grad_norm": 0.3086029589176178,
+        "learning_rate": 2.0120263388597025e-05,
+        "loss": 0.9078,
+        "step": 28780
+      },
+      {
+        "epoch": 0.9180777448260468,
+        "grad_norm": 0.29410770535469055,
+        "learning_rate": 2.004342264095854e-05,
+        "loss": 0.9042,
+        "step": 28790
+      },
+      {
+        "epoch": 0.9183966325456807,
+        "grad_norm": 0.2909134030342102,
+        "learning_rate": 1.9966875353717854e-05,
+        "loss": 0.9014,
+        "step": 28800
+      },
+      {
+        "epoch": 0.9187155202653146,
+        "grad_norm": 0.305385023355484,
+        "learning_rate": 1.9890620406128368e-05,
+        "loss": 0.9003,
+        "step": 28810
+      },
+      {
+        "epoch": 0.9190344079849485,
+        "grad_norm": 0.3027626574039459,
+        "learning_rate": 1.981465668172373e-05,
+        "loss": 0.886,
+        "step": 28820
+      },
+      {
+        "epoch": 0.9193532957045825,
+        "grad_norm": 0.29952168464660645,
+        "learning_rate": 1.9738983068301432e-05,
+        "loss": 0.9006,
+        "step": 28830
+      },
+      {
+        "epoch": 0.9196721834242163,
+        "grad_norm": 0.2944650948047638,
+        "learning_rate": 1.966359845790656e-05,
+        "loss": 0.904,
+        "step": 28840
+      },
+      {
+        "epoch": 0.9199910711438503,
+        "grad_norm": 0.2949775457382202,
+        "learning_rate": 1.9588501746815556e-05,
+        "loss": 0.9068,
+        "step": 28850
+      },
+      {
+        "epoch": 0.9203099588634842,
+        "grad_norm": 0.29960328340530396,
+        "learning_rate": 1.9513691835520046e-05,
+        "loss": 0.9094,
+        "step": 28860
+      },
+      {
+        "epoch": 0.920628846583118,
+        "grad_norm": 0.3011898398399353,
+        "learning_rate": 1.9439167628710803e-05,
+        "loss": 0.8934,
+        "step": 28870
+      },
+      {
+        "epoch": 0.920947734302752,
+        "grad_norm": 0.3027596175670624,
+        "learning_rate": 1.936492803526162e-05,
+        "loss": 0.902,
+        "step": 28880
+      },
+      {
+        "epoch": 0.921266622022386,
+        "grad_norm": 0.3028440773487091,
+        "learning_rate": 1.9290971968213404e-05,
+        "loss": 0.9131,
+        "step": 28890
+      },
+      {
+        "epoch": 0.9215855097420198,
+        "grad_norm": 0.3038038909435272,
+        "learning_rate": 1.9217298344758223e-05,
+        "loss": 0.9135,
+        "step": 28900
+      },
+      {
+        "epoch": 0.9219043974616538,
+        "grad_norm": 0.30496832728385925,
+        "learning_rate": 1.9143906086223483e-05,
+        "loss": 0.9035,
+        "step": 28910
+      },
+      {
+        "epoch": 0.9222232851812877,
+        "grad_norm": 0.2991454601287842,
+        "learning_rate": 1.907079411805611e-05,
+        "loss": 0.912,
+        "step": 28920
+      },
+      {
+        "epoch": 0.9225421729009216,
+        "grad_norm": 0.3076494038105011,
+        "learning_rate": 1.8997961369806826e-05,
+        "loss": 0.9167,
+        "step": 28930
+      },
+      {
+        "epoch": 0.9228610606205555,
+        "grad_norm": 0.3043328821659088,
+        "learning_rate": 1.8925406775114476e-05,
+        "loss": 0.9028,
+        "step": 28940
+      },
+      {
+        "epoch": 0.9231799483401895,
+        "grad_norm": 0.293874591588974,
+        "learning_rate": 1.8853129271690422e-05,
+        "loss": 0.9062,
+        "step": 28950
+      },
+      {
+        "epoch": 0.9234988360598233,
+        "grad_norm": 0.2940611243247986,
+        "learning_rate": 1.878112780130298e-05,
+        "loss": 0.905,
+        "step": 28960
+      },
+      {
+        "epoch": 0.9238177237794573,
+        "grad_norm": 0.30397751927375793,
+        "learning_rate": 1.8709401309761924e-05,
+        "loss": 0.9002,
+        "step": 28970
+      },
+      {
+        "epoch": 0.9241366114990912,
+        "grad_norm": 0.3081769049167633,
+        "learning_rate": 1.863794874690307e-05,
+        "loss": 0.8932,
+        "step": 28980
+      },
+      {
+        "epoch": 0.924455499218725,
+        "grad_norm": 0.2985953092575073,
+        "learning_rate": 1.8566769066572868e-05,
+        "loss": 0.9033,
+        "step": 28990
+      },
+      {
+        "epoch": 0.924774386938359,
+        "grad_norm": 0.31341174244880676,
+        "learning_rate": 1.849586122661313e-05,
+        "loss": 0.8932,
+        "step": 29000
+      },
+      {
+        "epoch": 0.925093274657993,
+        "grad_norm": 0.30072101950645447,
+        "learning_rate": 1.842522418884572e-05,
+        "loss": 0.9038,
+        "step": 29010
+      },
+      {
+        "epoch": 0.9254121623776268,
+        "grad_norm": 0.29787227511405945,
+        "learning_rate": 1.8354856919057388e-05,
+        "loss": 0.9065,
+        "step": 29020
+      },
+      {
+        "epoch": 0.9257310500972608,
+        "grad_norm": 0.297067254781723,
+        "learning_rate": 1.8284758386984637e-05,
+        "loss": 0.8989,
+        "step": 29030
+      },
+      {
+        "epoch": 0.9260499378168947,
+        "grad_norm": 0.29680055379867554,
+        "learning_rate": 1.8214927566298603e-05,
+        "loss": 0.8947,
+        "step": 29040
+      },
+      {
+        "epoch": 0.9263688255365286,
+        "grad_norm": 0.29479241371154785,
+        "learning_rate": 1.814536343459005e-05,
+        "loss": 0.9007,
+        "step": 29050
+      },
+      {
+        "epoch": 0.9266877132561625,
+        "grad_norm": 0.2918523848056793,
+        "learning_rate": 1.8076064973354396e-05,
+        "loss": 0.912,
+        "step": 29060
+      },
+      {
+        "epoch": 0.9270066009757965,
+        "grad_norm": 0.296645849943161,
+        "learning_rate": 1.80070311679768e-05,
+        "loss": 0.9015,
+        "step": 29070
+      },
+      {
+        "epoch": 0.9273254886954303,
+        "grad_norm": 0.2984611690044403,
+        "learning_rate": 1.793826100771732e-05,
+        "loss": 0.886,
+        "step": 29080
+      },
+      {
+        "epoch": 0.9276443764150643,
+        "grad_norm": 0.2978875935077667,
+        "learning_rate": 1.7869753485696093e-05,
+        "loss": 0.8939,
+        "step": 29090
+      },
+      {
+        "epoch": 0.9279632641346982,
+        "grad_norm": 0.2979802191257477,
+        "learning_rate": 1.7801507598878604e-05,
+        "loss": 0.9128,
+        "step": 29100
+      },
+      {
+        "epoch": 0.928282151854332,
+        "grad_norm": 0.30805638432502747,
+        "learning_rate": 1.7733522348061006e-05,
+        "loss": 0.9297,
+        "step": 29110
+      },
+      {
+        "epoch": 0.928601039573966,
+        "grad_norm": 0.3010908365249634,
+        "learning_rate": 1.7665796737855474e-05,
+        "loss": 0.9166,
+        "step": 29120
+      },
+      {
+        "epoch": 0.9289199272936,
+        "grad_norm": 0.3014841079711914,
+        "learning_rate": 1.7598329776675653e-05,
+        "loss": 0.8999,
+        "step": 29130
+      },
+      {
+        "epoch": 0.9292388150132338,
+        "grad_norm": 0.303438663482666,
+        "learning_rate": 1.753112047672212e-05,
+        "loss": 0.9029,
+        "step": 29140
+      },
+      {
+        "epoch": 0.9295577027328678,
+        "grad_norm": 0.2883744239807129,
+        "learning_rate": 1.7464167853967936e-05,
+        "loss": 0.8886,
+        "step": 29150
+      },
+      {
+        "epoch": 0.9298765904525017,
+        "grad_norm": 0.29751503467559814,
+        "learning_rate": 1.739747092814422e-05,
+        "loss": 0.8856,
+        "step": 29160
+      },
+      {
+        "epoch": 0.9301954781721355,
+        "grad_norm": 0.2995840907096863,
+        "learning_rate": 1.7331028722725825e-05,
+        "loss": 0.9102,
+        "step": 29170
+      },
+      {
+        "epoch": 0.9305143658917695,
+        "grad_norm": 0.30100563168525696,
+        "learning_rate": 1.726484026491702e-05,
+        "loss": 0.9028,
+        "step": 29180
+      },
+      {
+        "epoch": 0.9308332536114035,
+        "grad_norm": 0.2966121733188629,
+        "learning_rate": 1.7198904585637235e-05,
+        "loss": 0.8812,
+        "step": 29190
+      },
+      {
+        "epoch": 0.9311521413310373,
+        "grad_norm": 0.30071768164634705,
+        "learning_rate": 1.713322071950692e-05,
+        "loss": 0.906,
+        "step": 29200
+      },
+      {
+        "epoch": 0.9314710290506713,
+        "grad_norm": 0.30014944076538086,
+        "learning_rate": 1.7067787704833355e-05,
+        "loss": 0.9213,
+        "step": 29210
+      },
+      {
+        "epoch": 0.9317899167703052,
+        "grad_norm": 0.31035640835762024,
+        "learning_rate": 1.7002604583596607e-05,
+        "loss": 0.9063,
+        "step": 29220
+      },
+      {
+        "epoch": 0.932108804489939,
+        "grad_norm": 0.2940179705619812,
+        "learning_rate": 1.6937670401435477e-05,
+        "loss": 0.8975,
+        "step": 29230
+      },
+      {
+        "epoch": 0.932427692209573,
+        "grad_norm": 0.31050607562065125,
+        "learning_rate": 1.6872984207633558e-05,
+        "loss": 0.8947,
+        "step": 29240
+      },
+      {
+        "epoch": 0.932746579929207,
+        "grad_norm": 0.3037688136100769,
+        "learning_rate": 1.6808545055105283e-05,
+        "loss": 0.9016,
+        "step": 29250
+      },
+      {
+        "epoch": 0.9330654676488408,
+        "grad_norm": 0.3068945109844208,
+        "learning_rate": 1.6744352000382084e-05,
+        "loss": 0.903,
+        "step": 29260
+      },
+      {
+        "epoch": 0.9333843553684747,
+        "grad_norm": 0.3037906289100647,
+        "learning_rate": 1.6680404103598565e-05,
+        "loss": 0.8951,
+        "step": 29270
+      },
+      {
+        "epoch": 0.9337032430881087,
+        "grad_norm": 0.3063509166240692,
+        "learning_rate": 1.6616700428478738e-05,
+        "loss": 0.8969,
+        "step": 29280
+      },
+      {
+        "epoch": 0.9340221308077425,
+        "grad_norm": 0.3033375144004822,
+        "learning_rate": 1.6553240042322332e-05,
+        "loss": 0.8981,
+        "step": 29290
+      },
+      {
+        "epoch": 0.9343410185273765,
+        "grad_norm": 0.29124125838279724,
+        "learning_rate": 1.6490022015991115e-05,
+        "loss": 0.8943,
+        "step": 29300
+      },
+      {
+        "epoch": 0.9346599062470105,
+        "grad_norm": 0.297909140586853,
+        "learning_rate": 1.6427045423895318e-05,
+        "loss": 0.9071,
+        "step": 29310
+      },
+      {
+        "epoch": 0.9349787939666443,
+        "grad_norm": 0.29066595435142517,
+        "learning_rate": 1.636430934398004e-05,
+        "loss": 0.9035,
+        "step": 29320
+      },
+      {
+        "epoch": 0.9352976816862782,
+        "grad_norm": 0.3028257191181183,
+        "learning_rate": 1.63018128577118e-05,
+        "loss": 0.9195,
+        "step": 29330
+      },
+      {
+        "epoch": 0.9356165694059122,
+        "grad_norm": 0.30161142349243164,
+        "learning_rate": 1.623955505006505e-05,
+        "loss": 0.8899,
+        "step": 29340
+      },
+      {
+        "epoch": 0.935935457125546,
+        "grad_norm": 0.30503955483436584,
+        "learning_rate": 1.61775350095088e-05,
+        "loss": 0.8928,
+        "step": 29350
+      },
+      {
+        "epoch": 0.93625434484518,
+        "grad_norm": 0.3002879023551941,
+        "learning_rate": 1.611575182799324e-05,
+        "loss": 0.8799,
+        "step": 29360
+      },
+      {
+        "epoch": 0.936573232564814,
+        "grad_norm": 0.31323641538619995,
+        "learning_rate": 1.605420460093651e-05,
+        "loss": 0.9107,
+        "step": 29370
+      },
+      {
+        "epoch": 0.9368921202844478,
+        "grad_norm": 0.29981517791748047,
+        "learning_rate": 1.5992892427211376e-05,
+        "loss": 0.9071,
+        "step": 29380
+      },
+      {
+        "epoch": 0.9372110080040817,
+        "grad_norm": 0.3079449534416199,
+        "learning_rate": 1.5931814409132093e-05,
+        "loss": 0.9044,
+        "step": 29390
+      },
+      {
+        "epoch": 0.9375298957237157,
+        "grad_norm": 0.2934107184410095,
+        "learning_rate": 1.587096965244123e-05,
+        "loss": 0.9099,
+        "step": 29400
+      },
+      {
+        "epoch": 0.9378487834433495,
+        "grad_norm": 0.28969040513038635,
+        "learning_rate": 1.5810357266296604e-05,
+        "loss": 0.9033,
+        "step": 29410
+      },
+      {
+        "epoch": 0.9381676711629835,
+        "grad_norm": 0.30076563358306885,
+        "learning_rate": 1.5749976363258215e-05,
+        "loss": 0.9081,
+        "step": 29420
+      },
+      {
+        "epoch": 0.9384865588826174,
+        "grad_norm": 0.30778390169143677,
+        "learning_rate": 1.5689826059275262e-05,
+        "loss": 0.8972,
+        "step": 29430
+      },
+      {
+        "epoch": 0.9388054466022513,
+        "grad_norm": 0.30305200815200806,
+        "learning_rate": 1.5629905473673202e-05,
+        "loss": 0.8905,
+        "step": 29440
+      },
+      {
+        "epoch": 0.9391243343218852,
+        "grad_norm": 0.2984049916267395,
+        "learning_rate": 1.5570213729140846e-05,
+        "loss": 0.8958,
+        "step": 29450
+      },
+      {
+        "epoch": 0.9394432220415192,
+        "grad_norm": 0.297616571187973,
+        "learning_rate": 1.5510749951717523e-05,
+        "loss": 0.9065,
+        "step": 29460
+      },
+      {
+        "epoch": 0.9397621097611532,
+        "grad_norm": 0.29675933718681335,
+        "learning_rate": 1.5451513270780287e-05,
+        "loss": 0.9012,
+        "step": 29470
+      },
+      {
+        "epoch": 0.940080997480787,
+        "grad_norm": 0.2959672510623932,
+        "learning_rate": 1.539250281903115e-05,
+        "loss": 0.8988,
+        "step": 29480
+      },
+      {
+        "epoch": 0.940399885200421,
+        "grad_norm": 0.29955780506134033,
+        "learning_rate": 1.5333717732484415e-05,
+        "loss": 0.9001,
+        "step": 29490
+      },
+      {
+        "epoch": 0.9407187729200549,
+        "grad_norm": 0.3054618239402771,
+        "learning_rate": 1.5275157150454e-05,
+        "loss": 0.9055,
+        "step": 29500
+      },
+      {
+        "epoch": 0.9410376606396887,
+        "grad_norm": 0.293949156999588,
+        "learning_rate": 1.5216820215540841e-05,
+        "loss": 0.8992,
+        "step": 29510
+      },
+      {
+        "epoch": 0.9413565483593227,
+        "grad_norm": 0.3134571611881256,
+        "learning_rate": 1.5158706073620354e-05,
+        "loss": 0.8935,
+        "step": 29520
+      },
+      {
+        "epoch": 0.9416754360789567,
+        "grad_norm": 0.30413368344306946,
+        "learning_rate": 1.5100813873829904e-05,
+        "loss": 0.8834,
+        "step": 29530
+      },
+      {
+        "epoch": 0.9419943237985905,
+        "grad_norm": 0.3008265495300293,
+        "learning_rate": 1.5043142768556388e-05,
+        "loss": 0.9111,
+        "step": 29540
+      },
+      {
+        "epoch": 0.9423132115182244,
+        "grad_norm": 0.3124244213104248,
+        "learning_rate": 1.4985691913423778e-05,
+        "loss": 0.907,
+        "step": 29550
+      },
+      {
+        "epoch": 0.9426320992378584,
+        "grad_norm": 0.3013421595096588,
+        "learning_rate": 1.4928460467280777e-05,
+        "loss": 0.893,
+        "step": 29560
+      },
+      {
+        "epoch": 0.9429509869574922,
+        "grad_norm": 0.30568447709083557,
+        "learning_rate": 1.4871447592188524e-05,
+        "loss": 0.9057,
+        "step": 29570
+      },
+      {
+        "epoch": 0.9432698746771262,
+        "grad_norm": 0.2980208098888397,
+        "learning_rate": 1.4814652453408288e-05,
+        "loss": 0.8886,
+        "step": 29580
+      },
+      {
+        "epoch": 0.9435887623967602,
+        "grad_norm": 0.30180996656417847,
+        "learning_rate": 1.4758074219389278e-05,
+        "loss": 0.8927,
+        "step": 29590
+      },
+      {
+        "epoch": 0.943907650116394,
+        "grad_norm": 0.3070039451122284,
+        "learning_rate": 1.4701712061756454e-05,
+        "loss": 0.9086,
+        "step": 29600
+      },
+      {
+        "epoch": 0.944226537836028,
+        "grad_norm": 0.3042294383049011,
+        "learning_rate": 1.4645565155298395e-05,
+        "loss": 0.9102,
+        "step": 29610
+      },
+      {
+        "epoch": 0.9445454255556619,
+        "grad_norm": 0.28559359908103943,
+        "learning_rate": 1.458963267795523e-05,
+        "loss": 0.9113,
+        "step": 29620
+      },
+      {
+        "epoch": 0.9448643132752957,
+        "grad_norm": 0.30219465494155884,
+        "learning_rate": 1.4533913810806589e-05,
+        "loss": 0.9138,
+        "step": 29630
+      },
+      {
+        "epoch": 0.9451832009949297,
+        "grad_norm": 0.3139053285121918,
+        "learning_rate": 1.4478407738059622e-05,
+        "loss": 0.8951,
+        "step": 29640
+      },
+      {
+        "epoch": 0.9455020887145636,
+        "grad_norm": 0.2970397472381592,
+        "learning_rate": 1.4423113647037045e-05,
+        "loss": 0.8881,
+        "step": 29650
+      },
+      {
+        "epoch": 0.9458209764341975,
+        "grad_norm": 0.2954052686691284,
+        "learning_rate": 1.4368030728165257e-05,
+        "loss": 0.8962,
+        "step": 29660
+      },
+      {
+        "epoch": 0.9461398641538314,
+        "grad_norm": 0.30081456899642944,
+        "learning_rate": 1.4313158174962467e-05,
+        "loss": 0.9137,
+        "step": 29670
+      },
+      {
+        "epoch": 0.9464587518734654,
+        "grad_norm": 0.29948690533638,
+        "learning_rate": 1.4258495184026909e-05,
+        "loss": 0.9047,
+        "step": 29680
+      },
+      {
+        "epoch": 0.9467776395930992,
+        "grad_norm": 0.2990148067474365,
+        "learning_rate": 1.4204040955025053e-05,
+        "loss": 0.9023,
+        "step": 29690
+      },
+      {
+        "epoch": 0.9470965273127332,
+        "grad_norm": 0.3068961501121521,
+        "learning_rate": 1.4149794690679904e-05,
+        "loss": 0.898,
+        "step": 29700
+      },
+      {
+        "epoch": 0.9474154150323671,
+        "grad_norm": 0.2919014096260071,
+        "learning_rate": 1.409575559675934e-05,
+        "loss": 0.879,
+        "step": 29710
+      },
+      {
+        "epoch": 0.947734302752001,
+        "grad_norm": 0.30254507064819336,
+        "learning_rate": 1.4041922882064455e-05,
+        "loss": 0.8929,
+        "step": 29720
+      },
+      {
+        "epoch": 0.9480531904716349,
+        "grad_norm": 0.3149068057537079,
+        "learning_rate": 1.398829575841799e-05,
+        "loss": 0.9133,
+        "step": 29730
+      },
+      {
+        "epoch": 0.9483720781912689,
+        "grad_norm": 0.30008769035339355,
+        "learning_rate": 1.3934873440652796e-05,
+        "loss": 0.8871,
+        "step": 29740
+      },
+      {
+        "epoch": 0.9486909659109027,
+        "grad_norm": 0.2980942726135254,
+        "learning_rate": 1.3881655146600332e-05,
+        "loss": 0.9036,
+        "step": 29750
+      },
+      {
+        "epoch": 0.9490098536305367,
+        "grad_norm": 0.2967151999473572,
+        "learning_rate": 1.3828640097079218e-05,
+        "loss": 0.9042,
+        "step": 29760
+      },
+      {
+        "epoch": 0.9493287413501706,
+        "grad_norm": 0.306095153093338,
+        "learning_rate": 1.377582751588382e-05,
+        "loss": 0.8936,
+        "step": 29770
+      },
+      {
+        "epoch": 0.9496476290698045,
+        "grad_norm": 0.30544817447662354,
+        "learning_rate": 1.3723216629772897e-05,
+        "loss": 0.8961,
+        "step": 29780
+      },
+      {
+        "epoch": 0.9499665167894384,
+        "grad_norm": 0.3010407090187073,
+        "learning_rate": 1.3670806668458264e-05,
+        "loss": 0.9043,
+        "step": 29790
+      },
+      {
+        "epoch": 0.9502854045090724,
+        "grad_norm": 0.3018208146095276,
+        "learning_rate": 1.3618596864593529e-05,
+        "loss": 0.8995,
+        "step": 29800
+      },
+      {
+        "epoch": 0.9506042922287062,
+        "grad_norm": 0.29537490010261536,
+        "learning_rate": 1.3566586453762848e-05,
+        "loss": 0.8846,
+        "step": 29810
+      },
+      {
+        "epoch": 0.9509231799483402,
+        "grad_norm": 0.29739969968795776,
+        "learning_rate": 1.3514774674469737e-05,
+        "loss": 0.9081,
+        "step": 29820
+      },
+      {
+        "epoch": 0.9512420676679741,
+        "grad_norm": 0.3044932186603546,
+        "learning_rate": 1.3463160768125927e-05,
+        "loss": 0.914,
+        "step": 29830
+      },
+      {
+        "epoch": 0.951560955387608,
+        "grad_norm": 0.30703359842300415,
+        "learning_rate": 1.3411743979040244e-05,
+        "loss": 0.9054,
+        "step": 29840
+      },
+      {
+        "epoch": 0.9518798431072419,
+        "grad_norm": 0.3007422685623169,
+        "learning_rate": 1.3360523554407562e-05,
+        "loss": 0.9064,
+        "step": 29850
+      },
+      {
+        "epoch": 0.9521987308268759,
+        "grad_norm": 0.3024388253688812,
+        "learning_rate": 1.3309498744297768e-05,
+        "loss": 0.8915,
+        "step": 29860
+      },
+      {
+        "epoch": 0.9525176185465097,
+        "grad_norm": 0.30477872490882874,
+        "learning_rate": 1.3258668801644778e-05,
+        "loss": 0.9077,
+        "step": 29870
+      },
+      {
+        "epoch": 0.9528365062661437,
+        "grad_norm": 0.29049551486968994,
+        "learning_rate": 1.3208032982235637e-05,
+        "loss": 0.8962,
+        "step": 29880
+      },
+      {
+        "epoch": 0.9531553939857776,
+        "grad_norm": 0.2952941954135895,
+        "learning_rate": 1.315759054469956e-05,
+        "loss": 0.8911,
+        "step": 29890
+      },
+      {
+        "epoch": 0.9534742817054115,
+        "grad_norm": 0.30060091614723206,
+        "learning_rate": 1.3107340750497128e-05,
+        "loss": 0.8945,
+        "step": 29900
+      },
+      {
+        "epoch": 0.9537931694250454,
+        "grad_norm": 0.29785463213920593,
+        "learning_rate": 1.3057282863909452e-05,
+        "loss": 0.9044,
+        "step": 29910
+      },
+      {
+        "epoch": 0.9541120571446794,
+        "grad_norm": 0.29402804374694824,
+        "learning_rate": 1.3007416152027412e-05,
+        "loss": 0.9048,
+        "step": 29920
+      },
+      {
+        "epoch": 0.9544309448643132,
+        "grad_norm": 0.29524412751197815,
+        "learning_rate": 1.2957739884740917e-05,
+        "loss": 0.8943,
+        "step": 29930
+      },
+      {
+        "epoch": 0.9547498325839472,
+        "grad_norm": 0.31547248363494873,
+        "learning_rate": 1.290825333472822e-05,
+        "loss": 0.9,
+        "step": 29940
+      },
+      {
+        "epoch": 0.9550687203035811,
+        "grad_norm": 0.30710914731025696,
+        "learning_rate": 1.2858955777445266e-05,
+        "loss": 0.9259,
+        "step": 29950
+      },
+      {
+        "epoch": 0.955387608023215,
+        "grad_norm": 0.3070859909057617,
+        "learning_rate": 1.2809846491115096e-05,
+        "loss": 0.8958,
+        "step": 29960
+      },
+      {
+        "epoch": 0.9557064957428489,
+        "grad_norm": 0.2996968924999237,
+        "learning_rate": 1.276092475671726e-05,
+        "loss": 0.908,
+        "step": 29970
+      },
+      {
+        "epoch": 0.9560253834624829,
+        "grad_norm": 0.3038216531276703,
+        "learning_rate": 1.27121898579773e-05,
+        "loss": 0.896,
+        "step": 29980
+      },
+      {
+        "epoch": 0.9563442711821167,
+        "grad_norm": 0.30883684754371643,
+        "learning_rate": 1.266364108135627e-05,
+        "loss": 0.9177,
+        "step": 29990
+      },
+      {
+        "epoch": 0.9566631589017507,
+        "grad_norm": 0.31187060475349426,
+        "learning_rate": 1.2615277716040274e-05,
+        "loss": 0.9133,
+        "step": 30000
+      },
+      {
+        "epoch": 0.9569820466213846,
+        "grad_norm": 0.3062373697757721,
+        "learning_rate": 1.2567099053930065e-05,
+        "loss": 0.9176,
+        "step": 30010
+      },
+      {
+        "epoch": 0.9573009343410185,
+        "grad_norm": 0.31157636642456055,
+        "learning_rate": 1.2519104389630684e-05,
+        "loss": 0.8886,
+        "step": 30020
+      },
+      {
+        "epoch": 0.9576198220606524,
+        "grad_norm": 0.2978048622608185,
+        "learning_rate": 1.2471293020441117e-05,
+        "loss": 0.8989,
+        "step": 30030
+      },
+      {
+        "epoch": 0.9579387097802864,
+        "grad_norm": 0.2935866117477417,
+        "learning_rate": 1.2423664246344036e-05,
+        "loss": 0.8981,
+        "step": 30040
+      },
+      {
+        "epoch": 0.9582575974999202,
+        "grad_norm": 0.30189335346221924,
+        "learning_rate": 1.2376217369995511e-05,
+        "loss": 0.8878,
+        "step": 30050
+      },
+      {
+        "epoch": 0.9585764852195542,
+        "grad_norm": 0.30390262603759766,
+        "learning_rate": 1.2328951696714822e-05,
+        "loss": 0.9185,
+        "step": 30060
+      },
+      {
+        "epoch": 0.9588953729391881,
+        "grad_norm": 0.2978833019733429,
+        "learning_rate": 1.2281866534474292e-05,
+        "loss": 0.9151,
+        "step": 30070
+      },
+      {
+        "epoch": 0.959214260658822,
+        "grad_norm": 0.3033181130886078,
+        "learning_rate": 1.2234961193889144e-05,
+        "loss": 0.8983,
+        "step": 30080
+      },
+      {
+        "epoch": 0.9595331483784559,
+        "grad_norm": 0.2994658648967743,
+        "learning_rate": 1.218823498820741e-05,
+        "loss": 0.9006,
+        "step": 30090
+      },
+      {
+        "epoch": 0.9598520360980899,
+        "grad_norm": 0.30001530051231384,
+        "learning_rate": 1.214168723329988e-05,
+        "loss": 0.9078,
+        "step": 30100
+      },
+      {
+        "epoch": 0.9601709238177237,
+        "grad_norm": 0.2979891896247864,
+        "learning_rate": 1.2095317247650083e-05,
+        "loss": 0.9055,
+        "step": 30110
+      },
+      {
+        "epoch": 0.9604898115373577,
+        "grad_norm": 0.3081294298171997,
+        "learning_rate": 1.204912435234431e-05,
+        "loss": 0.9004,
+        "step": 30120
+      },
+      {
+        "epoch": 0.9608086992569916,
+        "grad_norm": 0.2981680929660797,
+        "learning_rate": 1.200310787106167e-05,
+        "loss": 0.8889,
+        "step": 30130
+      },
+      {
+        "epoch": 0.9611275869766255,
+        "grad_norm": 0.31134116649627686,
+        "learning_rate": 1.195726713006419e-05,
+        "loss": 0.9044,
+        "step": 30140
+      },
+      {
+        "epoch": 0.9614464746962594,
+        "grad_norm": 0.30942800641059875,
+        "learning_rate": 1.1911601458186958e-05,
+        "loss": 0.9033,
+        "step": 30150
+      },
+      {
+        "epoch": 0.9617653624158934,
+        "grad_norm": 0.30746373534202576,
+        "learning_rate": 1.186611018682828e-05,
+        "loss": 0.8926,
+        "step": 30160
+      },
+      {
+        "epoch": 0.9620842501355272,
+        "grad_norm": 0.30481740832328796,
+        "learning_rate": 1.1820792649939912e-05,
+        "loss": 0.8913,
+        "step": 30170
+      },
+      {
+        "epoch": 0.9624031378551612,
+        "grad_norm": 0.2997003495693207,
+        "learning_rate": 1.1775648184017282e-05,
+        "loss": 0.9051,
+        "step": 30180
+      },
+      {
+        "epoch": 0.9627220255747951,
+        "grad_norm": 0.3015143871307373,
+        "learning_rate": 1.1730676128089802e-05,
+        "loss": 0.9117,
+        "step": 30190
+      },
+      {
+        "epoch": 0.963040913294429,
+        "grad_norm": 0.3021433651447296,
+        "learning_rate": 1.1685875823711168e-05,
+        "loss": 0.8969,
+        "step": 30200
+      },
+      {
+        "epoch": 0.9633598010140629,
+        "grad_norm": 0.3064768314361572,
+        "learning_rate": 1.164124661494975e-05,
+        "loss": 0.9047,
+        "step": 30210
+      },
+      {
+        "epoch": 0.9636786887336969,
+        "grad_norm": 0.30161044001579285,
+        "learning_rate": 1.1596787848378949e-05,
+        "loss": 0.9124,
+        "step": 30220
+      },
+      {
+        "epoch": 0.9639975764533307,
+        "grad_norm": 0.30823206901550293,
+        "learning_rate": 1.1552498873067655e-05,
+        "loss": 0.9101,
+        "step": 30230
+      },
+      {
+        "epoch": 0.9643164641729647,
+        "grad_norm": 0.3106628358364105,
+        "learning_rate": 1.1508379040570714e-05,
+        "loss": 0.9075,
+        "step": 30240
+      },
+      {
+        "epoch": 0.9646353518925986,
+        "grad_norm": 0.29570499062538147,
+        "learning_rate": 1.146442770491943e-05,
+        "loss": 0.878,
+        "step": 30250
+      },
+      {
+        "epoch": 0.9649542396122325,
+        "grad_norm": 0.30375078320503235,
+        "learning_rate": 1.1420644222612106e-05,
+        "loss": 0.8909,
+        "step": 30260
+      },
+      {
+        "epoch": 0.9652731273318664,
+        "grad_norm": 0.2965579330921173,
+        "learning_rate": 1.1377027952604628e-05,
+        "loss": 0.8977,
+        "step": 30270
+      },
+      {
+        "epoch": 0.9655920150515004,
+        "grad_norm": 0.30961713194847107,
+        "learning_rate": 1.1333578256301075e-05,
+        "loss": 0.8859,
+        "step": 30280
+      },
+      {
+        "epoch": 0.9659109027711343,
+        "grad_norm": 0.2947857081890106,
+        "learning_rate": 1.129029449754437e-05,
+        "loss": 0.8957,
+        "step": 30290
+      },
+      {
+        "epoch": 0.9662297904907682,
+        "grad_norm": 0.2956250011920929,
+        "learning_rate": 1.1247176042606964e-05,
+        "loss": 0.9079,
+        "step": 30300
+      },
+      {
+        "epoch": 0.9665486782104021,
+        "grad_norm": 0.306384801864624,
+        "learning_rate": 1.1204222260181564e-05,
+        "loss": 0.8966,
+        "step": 30310
+      },
+      {
+        "epoch": 0.9668675659300361,
+        "grad_norm": 0.2974628210067749,
+        "learning_rate": 1.1161432521371883e-05,
+        "loss": 0.8998,
+        "step": 30320
+      },
+      {
+        "epoch": 0.9671864536496699,
+        "grad_norm": 0.3047086298465729,
+        "learning_rate": 1.1118806199683434e-05,
+        "loss": 0.8988,
+        "step": 30330
+      },
+      {
+        "epoch": 0.9675053413693039,
+        "grad_norm": 0.3030594289302826,
+        "learning_rate": 1.1076342671014357e-05,
+        "loss": 0.9229,
+        "step": 30340
+      },
+      {
+        "epoch": 0.9678242290889378,
+        "grad_norm": 0.32489851117134094,
+        "learning_rate": 1.1034041313646285e-05,
+        "loss": 0.9173,
+        "step": 30350
+      },
+      {
+        "epoch": 0.9681431168085717,
+        "grad_norm": 0.2998977601528168,
+        "learning_rate": 1.099190150823524e-05,
+        "loss": 0.9021,
+        "step": 30360
+      },
+      {
+        "epoch": 0.9684620045282056,
+        "grad_norm": 0.29817602038383484,
+        "learning_rate": 1.0949922637802553e-05,
+        "loss": 0.9189,
+        "step": 30370
+      },
+      {
+        "epoch": 0.9687808922478396,
+        "grad_norm": 0.30538299679756165,
+        "learning_rate": 1.0908104087725861e-05,
+        "loss": 0.8946,
+        "step": 30380
+      },
+      {
+        "epoch": 0.9690997799674734,
+        "grad_norm": 0.3044208884239197,
+        "learning_rate": 1.0866445245730072e-05,
+        "loss": 0.8986,
+        "step": 30390
+      },
+      {
+        "epoch": 0.9694186676871074,
+        "grad_norm": 0.29799777269363403,
+        "learning_rate": 1.0824945501878422e-05,
+        "loss": 0.9099,
+        "step": 30400
+      },
+      {
+        "epoch": 0.9697375554067413,
+        "grad_norm": 0.2967960238456726,
+        "learning_rate": 1.0783604248563538e-05,
+        "loss": 0.8932,
+        "step": 30410
+      },
+      {
+        "epoch": 0.9700564431263752,
+        "grad_norm": 0.30792462825775146,
+        "learning_rate": 1.0742420880498549e-05,
+        "loss": 0.9087,
+        "step": 30420
+      },
+      {
+        "epoch": 0.9703753308460091,
+        "grad_norm": 0.297896146774292,
+        "learning_rate": 1.0701394794708213e-05,
+        "loss": 0.9056,
+        "step": 30430
+      },
+      {
+        "epoch": 0.9706942185656431,
+        "grad_norm": 0.2979851961135864,
+        "learning_rate": 1.0660525390520096e-05,
+        "loss": 0.8883,
+        "step": 30440
+      },
+      {
+        "epoch": 0.9710131062852769,
+        "grad_norm": 0.2927461266517639,
+        "learning_rate": 1.0619812069555778e-05,
+        "loss": 0.906,
+        "step": 30450
+      },
+      {
+        "epoch": 0.9713319940049109,
+        "grad_norm": 0.2955889403820038,
+        "learning_rate": 1.0579254235722086e-05,
+        "loss": 0.9076,
+        "step": 30460
+      },
+      {
+        "epoch": 0.9716508817245448,
+        "grad_norm": 0.2992059290409088,
+        "learning_rate": 1.0538851295202372e-05,
+        "loss": 0.9145,
+        "step": 30470
+      },
+      {
+        "epoch": 0.9719697694441787,
+        "grad_norm": 0.30370664596557617,
+        "learning_rate": 1.0498602656447817e-05,
+        "loss": 0.9125,
+        "step": 30480
+      },
+      {
+        "epoch": 0.9722886571638126,
+        "grad_norm": 0.30064305663108826,
+        "learning_rate": 1.0458507730168771e-05,
+        "loss": 0.9096,
+        "step": 30490
+      },
+      {
+        "epoch": 0.9726075448834466,
+        "grad_norm": 0.28889572620391846,
+        "learning_rate": 1.0418565929326121e-05,
+        "loss": 0.9028,
+        "step": 30500
+      },
+      {
+        "epoch": 0.9729264326030804,
+        "grad_norm": 0.30345210433006287,
+        "learning_rate": 1.0378776669122702e-05,
+        "loss": 0.904,
+        "step": 30510
+      },
+      {
+        "epoch": 0.9732453203227144,
+        "grad_norm": 0.30365508794784546,
+        "learning_rate": 1.0339139366994728e-05,
+        "loss": 0.9115,
+        "step": 30520
+      },
+      {
+        "epoch": 0.9735642080423483,
+        "grad_norm": 0.3063391149044037,
+        "learning_rate": 1.0299653442603272e-05,
+        "loss": 0.9011,
+        "step": 30530
+      },
+      {
+        "epoch": 0.9738830957619822,
+        "grad_norm": 0.2992682456970215,
+        "learning_rate": 1.0260318317825752e-05,
+        "loss": 0.8967,
+        "step": 30540
+      },
+      {
+        "epoch": 0.9742019834816161,
+        "grad_norm": 0.31188657879829407,
+        "learning_rate": 1.0221133416747503e-05,
+        "loss": 0.9105,
+        "step": 30550
+      },
+      {
+        "epoch": 0.9745208712012501,
+        "grad_norm": 0.29710718989372253,
+        "learning_rate": 1.0182098165653291e-05,
+        "loss": 0.9099,
+        "step": 30560
+      },
+      {
+        "epoch": 0.9748397589208839,
+        "grad_norm": 0.30273959040641785,
+        "learning_rate": 1.014321199301896e-05,
+        "loss": 0.9047,
+        "step": 30570
+      },
+      {
+        "epoch": 0.9751586466405179,
+        "grad_norm": 0.3000819683074951,
+        "learning_rate": 1.0104474329503038e-05,
+        "loss": 0.9013,
+        "step": 30580
+      },
+      {
+        "epoch": 0.9754775343601518,
+        "grad_norm": 0.299858033657074,
+        "learning_rate": 1.0065884607938414e-05,
+        "loss": 0.9122,
+        "step": 30590
+      },
+      {
+        "epoch": 0.9757964220797857,
+        "grad_norm": 0.31173914670944214,
+        "learning_rate": 1.0027442263324029e-05,
+        "loss": 0.927,
+        "step": 30600
+      },
+      {
+        "epoch": 0.9761153097994196,
+        "grad_norm": 0.2916642427444458,
+        "learning_rate": 9.989146732816599e-06,
+        "loss": 0.9044,
+        "step": 30610
+      },
+      {
+        "epoch": 0.9764341975190536,
+        "grad_norm": 0.29491642117500305,
+        "learning_rate": 9.950997455722386e-06,
+        "loss": 0.902,
+        "step": 30620
+      },
+      {
+        "epoch": 0.9767530852386874,
+        "grad_norm": 0.29398080706596375,
+        "learning_rate": 9.912993873488982e-06,
+        "loss": 0.8944,
+        "step": 30630
+      },
+      {
+        "epoch": 0.9770719729583214,
+        "grad_norm": 0.3057700991630554,
+        "learning_rate": 9.875135429697123e-06,
+        "loss": 0.9103,
+        "step": 30640
+      },
+      {
+        "epoch": 0.9773908606779553,
+        "grad_norm": 0.3077187240123749,
+        "learning_rate": 9.83742157005256e-06,
+        "loss": 0.9091,
+        "step": 30650
+      },
+      {
+        "epoch": 0.9777097483975892,
+        "grad_norm": 0.30761876702308655,
+        "learning_rate": 9.79985174237793e-06,
+        "loss": 0.8964,
+        "step": 30660
+      },
+      {
+        "epoch": 0.9780286361172231,
+        "grad_norm": 0.3030130863189697,
+        "learning_rate": 9.762425396604675e-06,
+        "loss": 0.899,
+        "step": 30670
+      },
+      {
+        "epoch": 0.9783475238368571,
+        "grad_norm": 0.30985233187675476,
+        "learning_rate": 9.72514198476499e-06,
+        "loss": 0.9015,
+        "step": 30680
+      },
+      {
+        "epoch": 0.9786664115564909,
+        "grad_norm": 0.305759072303772,
+        "learning_rate": 9.688000960983798e-06,
+        "loss": 0.9086,
+        "step": 30690
+      },
+      {
+        "epoch": 0.9789852992761249,
+        "grad_norm": 0.2983715236186981,
+        "learning_rate": 9.65100178147076e-06,
+        "loss": 0.9056,
+        "step": 30700
+      },
+      {
+        "epoch": 0.9793041869957588,
+        "grad_norm": 0.30125829577445984,
+        "learning_rate": 9.614143904512304e-06,
+        "loss": 0.912,
+        "step": 30710
+      },
+      {
+        "epoch": 0.9796230747153927,
+        "grad_norm": 0.29553210735321045,
+        "learning_rate": 9.577426790463718e-06,
+        "loss": 0.8988,
+        "step": 30720
+      },
+      {
+        "epoch": 0.9799419624350266,
+        "grad_norm": 0.2971706688404083,
+        "learning_rate": 9.540849901741222e-06,
+        "loss": 0.8887,
+        "step": 30730
+      },
+      {
+        "epoch": 0.9802608501546606,
+        "grad_norm": 0.2950115501880646,
+        "learning_rate": 9.504412702814105e-06,
+        "loss": 0.8844,
+        "step": 30740
+      },
+      {
+        "epoch": 0.9805797378742944,
+        "grad_norm": 0.3017497956752777,
+        "learning_rate": 9.468114660196888e-06,
+        "loss": 0.9029,
+        "step": 30750
+      },
+      {
+        "epoch": 0.9808986255939284,
+        "grad_norm": 0.30914148688316345,
+        "learning_rate": 9.431955242441515e-06,
+        "loss": 0.9113,
+        "step": 30760
+      },
+      {
+        "epoch": 0.9812175133135623,
+        "grad_norm": 0.3126702308654785,
+        "learning_rate": 9.395933920129571e-06,
+        "loss": 0.8945,
+        "step": 30770
+      },
+      {
+        "epoch": 0.9815364010331962,
+        "grad_norm": 0.29880809783935547,
+        "learning_rate": 9.360050165864518e-06,
+        "loss": 0.9026,
+        "step": 30780
+      },
+      {
+        "epoch": 0.9818552887528301,
+        "grad_norm": 0.3105475604534149,
+        "learning_rate": 9.324303454263998e-06,
+        "loss": 0.8953,
+        "step": 30790
+      },
+      {
+        "epoch": 0.9821741764724641,
+        "grad_norm": 0.30883198976516724,
+        "learning_rate": 9.288693261952112e-06,
+        "loss": 0.8923,
+        "step": 30800
+      },
+      {
+        "epoch": 0.9824930641920979,
+        "grad_norm": 0.3052988350391388,
+        "learning_rate": 9.253219067551782e-06,
+        "loss": 0.8886,
+        "step": 30810
+      },
+      {
+        "epoch": 0.9828119519117319,
+        "grad_norm": 0.2951613664627075,
+        "learning_rate": 9.217880351677104e-06,
+        "loss": 0.891,
+        "step": 30820
+      },
+      {
+        "epoch": 0.9831308396313658,
+        "grad_norm": 0.29393279552459717,
+        "learning_rate": 9.182676596925743e-06,
+        "loss": 0.8915,
+        "step": 30830
+      },
+      {
+        "epoch": 0.9834497273509997,
+        "grad_norm": 0.2976394295692444,
+        "learning_rate": 9.147607287871367e-06,
+        "loss": 0.8896,
+        "step": 30840
+      },
+      {
+        "epoch": 0.9837686150706336,
+        "grad_norm": 0.30196166038513184,
+        "learning_rate": 9.112671911056089e-06,
+        "loss": 0.9013,
+        "step": 30850
+      },
+      {
+        "epoch": 0.9840875027902676,
+        "grad_norm": 0.29965201020240784,
+        "learning_rate": 9.077869954982961e-06,
+        "loss": 0.8945,
+        "step": 30860
+      },
+      {
+        "epoch": 0.9844063905099014,
+        "grad_norm": 0.2930929660797119,
+        "learning_rate": 9.043200910108472e-06,
+        "loss": 0.8928,
+        "step": 30870
+      },
+      {
+        "epoch": 0.9847252782295354,
+        "grad_norm": 0.3056459128856659,
+        "learning_rate": 9.008664268835097e-06,
+        "loss": 0.894,
+        "step": 30880
+      },
+      {
+        "epoch": 0.9850441659491693,
+        "grad_norm": 0.3035235106945038,
+        "learning_rate": 8.97425952550387e-06,
+        "loss": 0.9012,
+        "step": 30890
+      },
+      {
+        "epoch": 0.9853630536688032,
+        "grad_norm": 0.2955482602119446,
+        "learning_rate": 8.939986176386965e-06,
+        "loss": 0.9011,
+        "step": 30900
+      },
+      {
+        "epoch": 0.9856819413884371,
+        "grad_norm": 0.3025868535041809,
+        "learning_rate": 8.90584371968033e-06,
+        "loss": 0.8887,
+        "step": 30910
+      },
+      {
+        "epoch": 0.9860008291080711,
+        "grad_norm": 0.2991355061531067,
+        "learning_rate": 8.871831655496341e-06,
+        "loss": 0.9031,
+        "step": 30920
+      },
+      {
+        "epoch": 0.9863197168277049,
+        "grad_norm": 0.3063535690307617,
+        "learning_rate": 8.837949485856484e-06,
+        "loss": 0.8958,
+        "step": 30930
+      },
+      {
+        "epoch": 0.9866386045473389,
+        "grad_norm": 0.29752489924430847,
+        "learning_rate": 8.804196714684051e-06,
+        "loss": 0.8858,
+        "step": 30940
+      },
+      {
+        "epoch": 0.9869574922669728,
+        "grad_norm": 0.30858591198921204,
+        "learning_rate": 8.770572847796902e-06,
+        "loss": 0.8985,
+        "step": 30950
+      },
+      {
+        "epoch": 0.9872763799866067,
+        "grad_norm": 0.3076741099357605,
+        "learning_rate": 8.737077392900202e-06,
+        "loss": 0.9049,
+        "step": 30960
+      },
+      {
+        "epoch": 0.9875952677062406,
+        "grad_norm": 0.3044413924217224,
+        "learning_rate": 8.703709859579232e-06,
+        "loss": 0.9002,
+        "step": 30970
+      },
+      {
+        "epoch": 0.9879141554258746,
+        "grad_norm": 0.309249609708786,
+        "learning_rate": 8.670469759292197e-06,
+        "loss": 0.9028,
+        "step": 30980
+      },
+      {
+        "epoch": 0.9882330431455084,
+        "grad_norm": 0.3098140060901642,
+        "learning_rate": 8.637356605363086e-06,
+        "loss": 0.92,
+        "step": 30990
+      },
+      {
+        "epoch": 0.9885519308651424,
+        "grad_norm": 0.29409000277519226,
+        "learning_rate": 8.604369912974537e-06,
+        "loss": 0.895,
+        "step": 31000
+      },
+      {
+        "epoch": 0.9888708185847763,
+        "grad_norm": 0.3101252615451813,
+        "learning_rate": 8.571509199160733e-06,
+        "loss": 0.9301,
+        "step": 31010
+      },
+      {
+        "epoch": 0.9891897063044102,
+        "grad_norm": 0.316646009683609,
+        "learning_rate": 8.53877398280035e-06,
+        "loss": 0.8972,
+        "step": 31020
+      },
+      {
+        "epoch": 0.9895085940240441,
+        "grad_norm": 0.294486403465271,
+        "learning_rate": 8.506163784609493e-06,
+        "loss": 0.8991,
+        "step": 31030
+      },
+      {
+        "epoch": 0.9898274817436781,
+        "grad_norm": 0.29972952604293823,
+        "learning_rate": 8.47367812713469e-06,
+        "loss": 0.913,
+        "step": 31040
+      },
+      {
+        "epoch": 0.9901463694633119,
+        "grad_norm": 0.30866244435310364,
+        "learning_rate": 8.441316534745896e-06,
+        "loss": 0.9004,
+        "step": 31050
+      },
+      {
+        "epoch": 0.9904652571829459,
+        "grad_norm": 0.29569417238235474,
+        "learning_rate": 8.409078533629542e-06,
+        "loss": 0.8924,
+        "step": 31060
+      },
+      {
+        "epoch": 0.9907841449025798,
+        "grad_norm": 0.3136916160583496,
+        "learning_rate": 8.376963651781573e-06,
+        "loss": 0.8984,
+        "step": 31070
+      },
+      {
+        "epoch": 0.9911030326222137,
+        "grad_norm": 0.3088937997817993,
+        "learning_rate": 8.34497141900056e-06,
+        "loss": 0.8946,
+        "step": 31080
+      },
+      {
+        "epoch": 0.9914219203418476,
+        "grad_norm": 0.3017382025718689,
+        "learning_rate": 8.313101366880808e-06,
+        "loss": 0.895,
+        "step": 31090
+      },
+      {
+        "epoch": 0.9917408080614816,
+        "grad_norm": 0.30170178413391113,
+        "learning_rate": 8.28135302880549e-06,
+        "loss": 0.8961,
+        "step": 31100
+      },
+      {
+        "epoch": 0.9920596957811154,
+        "grad_norm": 0.2898147702217102,
+        "learning_rate": 8.24972593993983e-06,
+        "loss": 0.884,
+        "step": 31110
+      },
+      {
+        "epoch": 0.9923785835007494,
+        "grad_norm": 0.29774415493011475,
+        "learning_rate": 8.21821963722429e-06,
+        "loss": 0.8996,
+        "step": 31120
+      },
+      {
+        "epoch": 0.9926974712203833,
+        "grad_norm": 0.30368444323539734,
+        "learning_rate": 8.186833659367786e-06,
+        "loss": 0.8988,
+        "step": 31130
+      },
+      {
+        "epoch": 0.9930163589400173,
+        "grad_norm": 0.3021566569805145,
+        "learning_rate": 8.155567546840937e-06,
+        "loss": 0.9072,
+        "step": 31140
+      },
+      {
+        "epoch": 0.9933352466596511,
+        "grad_norm": 0.30956295132637024,
+        "learning_rate": 8.124420841869342e-06,
+        "loss": 0.8995,
+        "step": 31150
+      },
+      {
+        "epoch": 0.9936541343792851,
+        "grad_norm": 0.3030258119106293,
+        "learning_rate": 8.093393088426874e-06,
+        "loss": 0.894,
+        "step": 31160
+      },
+      {
+        "epoch": 0.993973022098919,
+        "grad_norm": 0.2955738306045532,
+        "learning_rate": 8.062483832229001e-06,
+        "loss": 0.8893,
+        "step": 31170
+      },
+      {
+        "epoch": 0.9942919098185529,
+        "grad_norm": 0.29659897089004517,
+        "learning_rate": 8.031692620726136e-06,
+        "loss": 0.8884,
+        "step": 31180
+      },
+      {
+        "epoch": 0.9946107975381868,
+        "grad_norm": 0.3124343156814575,
+        "learning_rate": 8.001019003097014e-06,
+        "loss": 0.8917,
+        "step": 31190
+      },
+      {
+        "epoch": 0.9949296852578208,
+        "grad_norm": 0.3097270727157593,
+        "learning_rate": 7.970462530242083e-06,
+        "loss": 0.894,
+        "step": 31200
+      },
+      {
+        "epoch": 0.9952485729774546,
+        "grad_norm": 0.31330564618110657,
+        "learning_rate": 7.940022754776945e-06,
+        "loss": 0.9184,
+        "step": 31210
+      },
+      {
+        "epoch": 0.9955674606970886,
+        "grad_norm": 0.296106219291687,
+        "learning_rate": 7.90969923102579e-06,
+        "loss": 0.8857,
+        "step": 31220
+      },
+      {
+        "epoch": 0.9958863484167225,
+        "grad_norm": 0.28833168745040894,
+        "learning_rate": 7.879491515014875e-06,
+        "loss": 0.8868,
+        "step": 31230
+      },
+      {
+        "epoch": 0.9962052361363564,
+        "grad_norm": 0.30367282032966614,
+        "learning_rate": 7.849399164466028e-06,
+        "loss": 0.8864,
+        "step": 31240
+      },
+      {
+        "epoch": 0.9965241238559903,
+        "grad_norm": 0.30643120408058167,
+        "learning_rate": 7.819421738790164e-06,
+        "loss": 0.9024,
+        "step": 31250
+      },
+      {
+        "epoch": 0.9968430115756243,
+        "grad_norm": 0.2968752086162567,
+        "learning_rate": 7.78955879908085e-06,
+        "loss": 0.8872,
+        "step": 31260
+      },
+      {
+        "epoch": 0.9971618992952581,
+        "grad_norm": 0.31228750944137573,
+        "learning_rate": 7.759809908107856e-06,
+        "loss": 0.9089,
+        "step": 31270
+      },
+      {
+        "epoch": 0.9974807870148921,
+        "grad_norm": 0.3094320595264435,
+        "learning_rate": 7.730174630310774e-06,
+        "loss": 0.8925,
+        "step": 31280
+      },
+      {
+        "epoch": 0.997799674734526,
+        "grad_norm": 0.2979593873023987,
+        "learning_rate": 7.700652531792634e-06,
+        "loss": 0.886,
+        "step": 31290
+      },
+      {
+        "epoch": 0.9981185624541599,
+        "grad_norm": 0.3050536513328552,
+        "learning_rate": 7.671243180313544e-06,
+        "loss": 0.9029,
+        "step": 31300
+      },
+      {
+        "epoch": 0.9984374501737938,
+        "grad_norm": 0.30217939615249634,
+        "learning_rate": 7.641946145284373e-06,
+        "loss": 0.8811,
+        "step": 31310
+      },
+      {
+        "epoch": 0.9987563378934278,
+        "grad_norm": 0.30297768115997314,
+        "learning_rate": 7.612760997760436e-06,
+        "loss": 0.8971,
+        "step": 31320
+      },
+      {
+        "epoch": 0.9990752256130616,
+        "grad_norm": 0.30359673500061035,
+        "learning_rate": 7.583687310435224e-06,
+        "loss": 0.9074,
+        "step": 31330
+      },
+      {
+        "epoch": 0.9993941133326956,
+        "grad_norm": 0.29697471857070923,
+        "learning_rate": 7.5547246576341395e-06,
+        "loss": 0.8844,
+        "step": 31340
+      },
+      {
+        "epoch": 0.9997130010523295,
+        "grad_norm": 0.30605244636535645,
+        "learning_rate": 7.525872615308268e-06,
+        "loss": 0.908,
+        "step": 31350
+      }
+    ],
+    "logging_steps": 10,
+    "max_steps": 31359,
+    "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": 8.828103613319793e+17,
+    "train_batch_size": 512,
+    "trial_name": null,
+    "trial_params": null
+  }