diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21979 @@
+{
+    "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": 2.681196928024292,
+        "learning_rate": 1.5944385981695843e-05,
+        "loss": 9.6714,
+        "step": 10
+      },
+      {
+        "epoch": 0.0006377754392678338,
+        "grad_norm": 2.1921708583831787,
+        "learning_rate": 3.1888771963391685e-05,
+        "loss": 9.1417,
+        "step": 20
+      },
+      {
+        "epoch": 0.0009566631589017507,
+        "grad_norm": 2.14097261428833,
+        "learning_rate": 4.7833157945087525e-05,
+        "loss": 8.6391,
+        "step": 30
+      },
+      {
+        "epoch": 0.0012755508785356677,
+        "grad_norm": 1.5710511207580566,
+        "learning_rate": 6.377754392678337e-05,
+        "loss": 8.037,
+        "step": 40
+      },
+      {
+        "epoch": 0.0015944385981695845,
+        "grad_norm": 1.208488941192627,
+        "learning_rate": 7.972192990847922e-05,
+        "loss": 7.4423,
+        "step": 50
+      },
+      {
+        "epoch": 0.0019133263178035013,
+        "grad_norm": 0.9196066856384277,
+        "learning_rate": 9.566631589017505e-05,
+        "loss": 7.0456,
+        "step": 60
+      },
+      {
+        "epoch": 0.0022322140374374183,
+        "grad_norm": 0.9510890245437622,
+        "learning_rate": 0.00011161070187187091,
+        "loss": 6.6663,
+        "step": 70
+      },
+      {
+        "epoch": 0.0025511017570713354,
+        "grad_norm": 0.6426137089729309,
+        "learning_rate": 0.00012755508785356674,
+        "loss": 6.3935,
+        "step": 80
+      },
+      {
+        "epoch": 0.002869989476705252,
+        "grad_norm": 0.5806668996810913,
+        "learning_rate": 0.0001434994738352626,
+        "loss": 6.2414,
+        "step": 90
+      },
+      {
+        "epoch": 0.003188877196339169,
+        "grad_norm": 0.7059349417686462,
+        "learning_rate": 0.00015944385981695843,
+        "loss": 6.1057,
+        "step": 100
+      },
+      {
+        "epoch": 0.003507764915973086,
+        "grad_norm": 0.8364503383636475,
+        "learning_rate": 0.00017538824579865428,
+        "loss": 5.9646,
+        "step": 110
+      },
+      {
+        "epoch": 0.0038266526356070026,
+        "grad_norm": 0.9244943857192993,
+        "learning_rate": 0.0001913326317803501,
+        "loss": 5.8405,
+        "step": 120
+      },
+      {
+        "epoch": 0.00414554035524092,
+        "grad_norm": 0.7110545039176941,
+        "learning_rate": 0.00020727701776204597,
+        "loss": 5.7436,
+        "step": 130
+      },
+      {
+        "epoch": 0.004464428074874837,
+        "grad_norm": 0.7149806618690491,
+        "learning_rate": 0.00022322140374374182,
+        "loss": 5.606,
+        "step": 140
+      },
+      {
+        "epoch": 0.004783315794508754,
+        "grad_norm": 0.8666573166847229,
+        "learning_rate": 0.00023916578972543764,
+        "loss": 5.4951,
+        "step": 150
+      },
+      {
+        "epoch": 0.005102203514142671,
+        "grad_norm": 0.8655577898025513,
+        "learning_rate": 0.0002551101757071335,
+        "loss": 5.3836,
+        "step": 160
+      },
+      {
+        "epoch": 0.005421091233776587,
+        "grad_norm": 0.9628815650939941,
+        "learning_rate": 0.00027105456168882936,
+        "loss": 5.308,
+        "step": 170
+      },
+      {
+        "epoch": 0.005739978953410504,
+        "grad_norm": 0.9321894645690918,
+        "learning_rate": 0.0002869989476705252,
+        "loss": 5.194,
+        "step": 180
+      },
+      {
+        "epoch": 0.006058866673044421,
+        "grad_norm": 1.1412914991378784,
+        "learning_rate": 0.00030294333365222105,
+        "loss": 5.1022,
+        "step": 190
+      },
+      {
+        "epoch": 0.006377754392678338,
+        "grad_norm": 0.960950493812561,
+        "learning_rate": 0.00031888771963391687,
+        "loss": 4.9813,
+        "step": 200
+      },
+      {
+        "epoch": 0.006696642112312255,
+        "grad_norm": 0.8928449749946594,
+        "learning_rate": 0.0003348321056156127,
+        "loss": 4.8914,
+        "step": 210
+      },
+      {
+        "epoch": 0.007015529831946172,
+        "grad_norm": 0.876811146736145,
+        "learning_rate": 0.00035077649159730856,
+        "loss": 4.7856,
+        "step": 220
+      },
+      {
+        "epoch": 0.007334417551580088,
+        "grad_norm": 1.1284464597702026,
+        "learning_rate": 0.00036672087757900443,
+        "loss": 4.7037,
+        "step": 230
+      },
+      {
+        "epoch": 0.007653305271214005,
+        "grad_norm": 0.9718582630157471,
+        "learning_rate": 0.0003826652635607002,
+        "loss": 4.6136,
+        "step": 240
+      },
+      {
+        "epoch": 0.007972192990847922,
+        "grad_norm": 1.113908052444458,
+        "learning_rate": 0.00039860964954239607,
+        "loss": 4.5741,
+        "step": 250
+      },
+      {
+        "epoch": 0.00829108071048184,
+        "grad_norm": 1.2405577898025513,
+        "learning_rate": 0.00041455403552409194,
+        "loss": 4.4705,
+        "step": 260
+      },
+      {
+        "epoch": 0.008609968430115756,
+        "grad_norm": 0.8793233633041382,
+        "learning_rate": 0.0004304984215057878,
+        "loss": 4.4333,
+        "step": 270
+      },
+      {
+        "epoch": 0.008928856149749673,
+        "grad_norm": 0.8734621405601501,
+        "learning_rate": 0.00044644280748748364,
+        "loss": 4.3445,
+        "step": 280
+      },
+      {
+        "epoch": 0.00924774386938359,
+        "grad_norm": 0.9244428277015686,
+        "learning_rate": 0.0004623871934691794,
+        "loss": 4.3066,
+        "step": 290
+      },
+      {
+        "epoch": 0.009566631589017507,
+        "grad_norm": 0.8140283226966858,
+        "learning_rate": 0.0004783315794508753,
+        "loss": 4.2334,
+        "step": 300
+      },
+      {
+        "epoch": 0.009885519308651424,
+        "grad_norm": 1.2293978929519653,
+        "learning_rate": 0.0004942759654325711,
+        "loss": 4.1764,
+        "step": 310
+      },
+      {
+        "epoch": 0.010204407028285341,
+        "grad_norm": 1.0468558073043823,
+        "learning_rate": 0.0005,
+        "loss": 4.1287,
+        "step": 320
+      },
+      {
+        "epoch": 0.010523294747919257,
+        "grad_norm": 1.0311610698699951,
+        "learning_rate": 0.0005,
+        "loss": 4.0477,
+        "step": 330
+      },
+      {
+        "epoch": 0.010842182467553174,
+        "grad_norm": 0.8697277307510376,
+        "learning_rate": 0.0005,
+        "loss": 4.0209,
+        "step": 340
+      },
+      {
+        "epoch": 0.01116107018718709,
+        "grad_norm": 0.8703891634941101,
+        "learning_rate": 0.0005,
+        "loss": 3.9472,
+        "step": 350
+      },
+      {
+        "epoch": 0.011479957906821008,
+        "grad_norm": 0.8960039019584656,
+        "learning_rate": 0.0005,
+        "loss": 3.9306,
+        "step": 360
+      },
+      {
+        "epoch": 0.011798845626454925,
+        "grad_norm": 0.8864967823028564,
+        "learning_rate": 0.0005,
+        "loss": 3.8554,
+        "step": 370
+      },
+      {
+        "epoch": 0.012117733346088842,
+        "grad_norm": 0.9100608825683594,
+        "learning_rate": 0.0005,
+        "loss": 3.8203,
+        "step": 380
+      },
+      {
+        "epoch": 0.012436621065722759,
+        "grad_norm": 0.9239365458488464,
+        "learning_rate": 0.0005,
+        "loss": 3.7934,
+        "step": 390
+      },
+      {
+        "epoch": 0.012755508785356676,
+        "grad_norm": 1.0264390707015991,
+        "learning_rate": 0.0005,
+        "loss": 3.7561,
+        "step": 400
+      },
+      {
+        "epoch": 0.013074396504990593,
+        "grad_norm": 0.9991441369056702,
+        "learning_rate": 0.0005,
+        "loss": 3.704,
+        "step": 410
+      },
+      {
+        "epoch": 0.01339328422462451,
+        "grad_norm": 1.0525001287460327,
+        "learning_rate": 0.0005,
+        "loss": 3.6787,
+        "step": 420
+      },
+      {
+        "epoch": 0.013712171944258427,
+        "grad_norm": 0.9851738214492798,
+        "learning_rate": 0.0005,
+        "loss": 3.6323,
+        "step": 430
+      },
+      {
+        "epoch": 0.014031059663892344,
+        "grad_norm": 1.1018930673599243,
+        "learning_rate": 0.0005,
+        "loss": 3.5947,
+        "step": 440
+      },
+      {
+        "epoch": 0.014349947383526261,
+        "grad_norm": 0.9886248111724854,
+        "learning_rate": 0.0005,
+        "loss": 3.5393,
+        "step": 450
+      },
+      {
+        "epoch": 0.014668835103160176,
+        "grad_norm": 0.9339925050735474,
+        "learning_rate": 0.0005,
+        "loss": 3.5243,
+        "step": 460
+      },
+      {
+        "epoch": 0.014987722822794093,
+        "grad_norm": 0.9865897297859192,
+        "learning_rate": 0.0005,
+        "loss": 3.5089,
+        "step": 470
+      },
+      {
+        "epoch": 0.01530661054242801,
+        "grad_norm": 1.0084065198898315,
+        "learning_rate": 0.0005,
+        "loss": 3.4716,
+        "step": 480
+      },
+      {
+        "epoch": 0.01562549826206193,
+        "grad_norm": 1.1506084203720093,
+        "learning_rate": 0.0005,
+        "loss": 3.4007,
+        "step": 490
+      },
+      {
+        "epoch": 0.015944385981695845,
+        "grad_norm": 1.1100022792816162,
+        "learning_rate": 0.0005,
+        "loss": 3.4015,
+        "step": 500
+      },
+      {
+        "epoch": 0.016263273701329763,
+        "grad_norm": 1.1417492628097534,
+        "learning_rate": 0.0005,
+        "loss": 3.3523,
+        "step": 510
+      },
+      {
+        "epoch": 0.01658216142096368,
+        "grad_norm": 1.2564970254898071,
+        "learning_rate": 0.0005,
+        "loss": 3.3233,
+        "step": 520
+      },
+      {
+        "epoch": 0.016901049140597594,
+        "grad_norm": 1.175044298171997,
+        "learning_rate": 0.0005,
+        "loss": 3.3223,
+        "step": 530
+      },
+      {
+        "epoch": 0.017219936860231513,
+        "grad_norm": 1.1329243183135986,
+        "learning_rate": 0.0005,
+        "loss": 3.2792,
+        "step": 540
+      },
+      {
+        "epoch": 0.017538824579865428,
+        "grad_norm": 0.9119124412536621,
+        "learning_rate": 0.0005,
+        "loss": 3.2103,
+        "step": 550
+      },
+      {
+        "epoch": 0.017857712299499347,
+        "grad_norm": 1.143915057182312,
+        "learning_rate": 0.0005,
+        "loss": 3.2056,
+        "step": 560
+      },
+      {
+        "epoch": 0.018176600019133262,
+        "grad_norm": 1.0789998769760132,
+        "learning_rate": 0.0005,
+        "loss": 3.18,
+        "step": 570
+      },
+      {
+        "epoch": 0.01849548773876718,
+        "grad_norm": 1.4629336595535278,
+        "learning_rate": 0.0005,
+        "loss": 3.1569,
+        "step": 580
+      },
+      {
+        "epoch": 0.018814375458401096,
+        "grad_norm": 1.2161282300949097,
+        "learning_rate": 0.0005,
+        "loss": 3.1021,
+        "step": 590
+      },
+      {
+        "epoch": 0.019133263178035015,
+        "grad_norm": 1.2440675497055054,
+        "learning_rate": 0.0005,
+        "loss": 3.1033,
+        "step": 600
+      },
+      {
+        "epoch": 0.01945215089766893,
+        "grad_norm": 1.496854543685913,
+        "learning_rate": 0.0005,
+        "loss": 3.0736,
+        "step": 610
+      },
+      {
+        "epoch": 0.01977103861730285,
+        "grad_norm": 1.3418179750442505,
+        "learning_rate": 0.0005,
+        "loss": 3.0,
+        "step": 620
+      },
+      {
+        "epoch": 0.020089926336936764,
+        "grad_norm": 1.0596439838409424,
+        "learning_rate": 0.0005,
+        "loss": 2.9936,
+        "step": 630
+      },
+      {
+        "epoch": 0.020408814056570683,
+        "grad_norm": 1.0319935083389282,
+        "learning_rate": 0.0005,
+        "loss": 2.9885,
+        "step": 640
+      },
+      {
+        "epoch": 0.0207277017762046,
+        "grad_norm": 1.1878321170806885,
+        "learning_rate": 0.0005,
+        "loss": 2.9625,
+        "step": 650
+      },
+      {
+        "epoch": 0.021046589495838514,
+        "grad_norm": 1.1297259330749512,
+        "learning_rate": 0.0005,
+        "loss": 2.9487,
+        "step": 660
+      },
+      {
+        "epoch": 0.021365477215472432,
+        "grad_norm": 1.0326006412506104,
+        "learning_rate": 0.0005,
+        "loss": 2.9224,
+        "step": 670
+      },
+      {
+        "epoch": 0.021684364935106348,
+        "grad_norm": 0.9553630352020264,
+        "learning_rate": 0.0005,
+        "loss": 2.8889,
+        "step": 680
+      },
+      {
+        "epoch": 0.022003252654740266,
+        "grad_norm": 1.092453956604004,
+        "learning_rate": 0.0005,
+        "loss": 2.8851,
+        "step": 690
+      },
+      {
+        "epoch": 0.02232214037437418,
+        "grad_norm": 1.2691729068756104,
+        "learning_rate": 0.0005,
+        "loss": 2.8428,
+        "step": 700
+      },
+      {
+        "epoch": 0.0226410280940081,
+        "grad_norm": 1.1289976835250854,
+        "learning_rate": 0.0005,
+        "loss": 2.8537,
+        "step": 710
+      },
+      {
+        "epoch": 0.022959915813642016,
+        "grad_norm": 1.206107258796692,
+        "learning_rate": 0.0005,
+        "loss": 2.7994,
+        "step": 720
+      },
+      {
+        "epoch": 0.023278803533275935,
+        "grad_norm": 1.0612897872924805,
+        "learning_rate": 0.0005,
+        "loss": 2.7953,
+        "step": 730
+      },
+      {
+        "epoch": 0.02359769125290985,
+        "grad_norm": 1.0839524269104004,
+        "learning_rate": 0.0005,
+        "loss": 2.7583,
+        "step": 740
+      },
+      {
+        "epoch": 0.02391657897254377,
+        "grad_norm": 0.9679558277130127,
+        "learning_rate": 0.0005,
+        "loss": 2.7797,
+        "step": 750
+      },
+      {
+        "epoch": 0.024235466692177684,
+        "grad_norm": 1.139195442199707,
+        "learning_rate": 0.0005,
+        "loss": 2.7543,
+        "step": 760
+      },
+      {
+        "epoch": 0.024554354411811603,
+        "grad_norm": 1.2014336585998535,
+        "learning_rate": 0.0005,
+        "loss": 2.7655,
+        "step": 770
+      },
+      {
+        "epoch": 0.024873242131445518,
+        "grad_norm": 1.0808125734329224,
+        "learning_rate": 0.0005,
+        "loss": 2.6973,
+        "step": 780
+      },
+      {
+        "epoch": 0.025192129851079433,
+        "grad_norm": 1.2096153497695923,
+        "learning_rate": 0.0005,
+        "loss": 2.7071,
+        "step": 790
+      },
+      {
+        "epoch": 0.025511017570713352,
+        "grad_norm": 1.1014368534088135,
+        "learning_rate": 0.0005,
+        "loss": 2.7005,
+        "step": 800
+      },
+      {
+        "epoch": 0.025829905290347267,
+        "grad_norm": 1.1239562034606934,
+        "learning_rate": 0.0005,
+        "loss": 2.6559,
+        "step": 810
+      },
+      {
+        "epoch": 0.026148793009981186,
+        "grad_norm": 1.0517128705978394,
+        "learning_rate": 0.0005,
+        "loss": 2.6809,
+        "step": 820
+      },
+      {
+        "epoch": 0.0264676807296151,
+        "grad_norm": 1.086920142173767,
+        "learning_rate": 0.0005,
+        "loss": 2.6294,
+        "step": 830
+      },
+      {
+        "epoch": 0.02678656844924902,
+        "grad_norm": 1.189956784248352,
+        "learning_rate": 0.0005,
+        "loss": 2.6253,
+        "step": 840
+      },
+      {
+        "epoch": 0.027105456168882935,
+        "grad_norm": 1.3793237209320068,
+        "learning_rate": 0.0005,
+        "loss": 2.6066,
+        "step": 850
+      },
+      {
+        "epoch": 0.027424343888516854,
+        "grad_norm": 1.0952047109603882,
+        "learning_rate": 0.0005,
+        "loss": 2.6004,
+        "step": 860
+      },
+      {
+        "epoch": 0.02774323160815077,
+        "grad_norm": 1.0428546667099,
+        "learning_rate": 0.0005,
+        "loss": 2.5666,
+        "step": 870
+      },
+      {
+        "epoch": 0.028062119327784688,
+        "grad_norm": 1.1217527389526367,
+        "learning_rate": 0.0005,
+        "loss": 2.5656,
+        "step": 880
+      },
+      {
+        "epoch": 0.028381007047418604,
+        "grad_norm": 1.186268925666809,
+        "learning_rate": 0.0005,
+        "loss": 2.5476,
+        "step": 890
+      },
+      {
+        "epoch": 0.028699894767052522,
+        "grad_norm": 1.3410747051239014,
+        "learning_rate": 0.0005,
+        "loss": 2.5374,
+        "step": 900
+      },
+      {
+        "epoch": 0.029018782486686438,
+        "grad_norm": 1.0802406072616577,
+        "learning_rate": 0.0005,
+        "loss": 2.4922,
+        "step": 910
+      },
+      {
+        "epoch": 0.029337670206320353,
+        "grad_norm": 1.1122218370437622,
+        "learning_rate": 0.0005,
+        "loss": 2.5051,
+        "step": 920
+      },
+      {
+        "epoch": 0.02965655792595427,
+        "grad_norm": 1.207636833190918,
+        "learning_rate": 0.0005,
+        "loss": 2.5154,
+        "step": 930
+      },
+      {
+        "epoch": 0.029975445645588187,
+        "grad_norm": 1.0092180967330933,
+        "learning_rate": 0.0005,
+        "loss": 2.5097,
+        "step": 940
+      },
+      {
+        "epoch": 0.030294333365222106,
+        "grad_norm": 0.9906138777732849,
+        "learning_rate": 0.0005,
+        "loss": 2.4812,
+        "step": 950
+      },
+      {
+        "epoch": 0.03061322108485602,
+        "grad_norm": 1.0704057216644287,
+        "learning_rate": 0.0005,
+        "loss": 2.475,
+        "step": 960
+      },
+      {
+        "epoch": 0.03093210880448994,
+        "grad_norm": 1.0282485485076904,
+        "learning_rate": 0.0005,
+        "loss": 2.4569,
+        "step": 970
+      },
+      {
+        "epoch": 0.03125099652412386,
+        "grad_norm": 1.147625207901001,
+        "learning_rate": 0.0005,
+        "loss": 2.4775,
+        "step": 980
+      },
+      {
+        "epoch": 0.03156988424375777,
+        "grad_norm": 1.1026866436004639,
+        "learning_rate": 0.0005,
+        "loss": 2.4594,
+        "step": 990
+      },
+      {
+        "epoch": 0.03188877196339169,
+        "grad_norm": 1.0910698175430298,
+        "learning_rate": 0.0005,
+        "loss": 2.4308,
+        "step": 1000
+      },
+      {
+        "epoch": 0.03220765968302561,
+        "grad_norm": 0.9975622296333313,
+        "learning_rate": 0.0005,
+        "loss": 2.422,
+        "step": 1010
+      },
+      {
+        "epoch": 0.03252654740265953,
+        "grad_norm": 1.1360799074172974,
+        "learning_rate": 0.0005,
+        "loss": 2.4286,
+        "step": 1020
+      },
+      {
+        "epoch": 0.03284543512229344,
+        "grad_norm": 0.971255362033844,
+        "learning_rate": 0.0005,
+        "loss": 2.4026,
+        "step": 1030
+      },
+      {
+        "epoch": 0.03316432284192736,
+        "grad_norm": 0.9612915515899658,
+        "learning_rate": 0.0005,
+        "loss": 2.3425,
+        "step": 1040
+      },
+      {
+        "epoch": 0.033483210561561276,
+        "grad_norm": 0.983697772026062,
+        "learning_rate": 0.0005,
+        "loss": 2.3794,
+        "step": 1050
+      },
+      {
+        "epoch": 0.03380209828119519,
+        "grad_norm": 1.234744668006897,
+        "learning_rate": 0.0005,
+        "loss": 2.3309,
+        "step": 1060
+      },
+      {
+        "epoch": 0.03412098600082911,
+        "grad_norm": 1.040803074836731,
+        "learning_rate": 0.0005,
+        "loss": 2.3611,
+        "step": 1070
+      },
+      {
+        "epoch": 0.034439873720463025,
+        "grad_norm": 0.9408783912658691,
+        "learning_rate": 0.0005,
+        "loss": 2.352,
+        "step": 1080
+      },
+      {
+        "epoch": 0.034758761440096944,
+        "grad_norm": 0.9767993688583374,
+        "learning_rate": 0.0005,
+        "loss": 2.3517,
+        "step": 1090
+      },
+      {
+        "epoch": 0.035077649159730856,
+        "grad_norm": 1.0135048627853394,
+        "learning_rate": 0.0005,
+        "loss": 2.3566,
+        "step": 1100
+      },
+      {
+        "epoch": 0.035396536879364775,
+        "grad_norm": 1.091168761253357,
+        "learning_rate": 0.0005,
+        "loss": 2.3013,
+        "step": 1110
+      },
+      {
+        "epoch": 0.035715424598998693,
+        "grad_norm": 1.0776761770248413,
+        "learning_rate": 0.0005,
+        "loss": 2.3057,
+        "step": 1120
+      },
+      {
+        "epoch": 0.03603431231863261,
+        "grad_norm": 1.0987141132354736,
+        "learning_rate": 0.0005,
+        "loss": 2.3045,
+        "step": 1130
+      },
+      {
+        "epoch": 0.036353200038266524,
+        "grad_norm": 1.1936005353927612,
+        "learning_rate": 0.0005,
+        "loss": 2.3158,
+        "step": 1140
+      },
+      {
+        "epoch": 0.03667208775790044,
+        "grad_norm": 0.9826390147209167,
+        "learning_rate": 0.0005,
+        "loss": 2.2981,
+        "step": 1150
+      },
+      {
+        "epoch": 0.03699097547753436,
+        "grad_norm": 1.0962214469909668,
+        "learning_rate": 0.0005,
+        "loss": 2.2778,
+        "step": 1160
+      },
+      {
+        "epoch": 0.03730986319716828,
+        "grad_norm": 1.0983978509902954,
+        "learning_rate": 0.0005,
+        "loss": 2.2461,
+        "step": 1170
+      },
+      {
+        "epoch": 0.03762875091680219,
+        "grad_norm": 1.0521897077560425,
+        "learning_rate": 0.0005,
+        "loss": 2.2119,
+        "step": 1180
+      },
+      {
+        "epoch": 0.03794763863643611,
+        "grad_norm": 0.9314139485359192,
+        "learning_rate": 0.0005,
+        "loss": 2.2284,
+        "step": 1190
+      },
+      {
+        "epoch": 0.03826652635607003,
+        "grad_norm": 0.9761852622032166,
+        "learning_rate": 0.0005,
+        "loss": 2.2425,
+        "step": 1200
+      },
+      {
+        "epoch": 0.03858541407570394,
+        "grad_norm": 0.9609031081199646,
+        "learning_rate": 0.0005,
+        "loss": 2.2046,
+        "step": 1210
+      },
+      {
+        "epoch": 0.03890430179533786,
+        "grad_norm": 1.3822377920150757,
+        "learning_rate": 0.0005,
+        "loss": 2.2251,
+        "step": 1220
+      },
+      {
+        "epoch": 0.03922318951497178,
+        "grad_norm": 0.9732682108879089,
+        "learning_rate": 0.0005,
+        "loss": 2.2095,
+        "step": 1230
+      },
+      {
+        "epoch": 0.0395420772346057,
+        "grad_norm": 1.0873984098434448,
+        "learning_rate": 0.0005,
+        "loss": 2.2142,
+        "step": 1240
+      },
+      {
+        "epoch": 0.03986096495423961,
+        "grad_norm": 1.0694262981414795,
+        "learning_rate": 0.0005,
+        "loss": 2.179,
+        "step": 1250
+      },
+      {
+        "epoch": 0.04017985267387353,
+        "grad_norm": 1.1795161962509155,
+        "learning_rate": 0.0005,
+        "loss": 2.1788,
+        "step": 1260
+      },
+      {
+        "epoch": 0.04049874039350745,
+        "grad_norm": 0.96063631772995,
+        "learning_rate": 0.0005,
+        "loss": 2.1825,
+        "step": 1270
+      },
+      {
+        "epoch": 0.040817628113141366,
+        "grad_norm": 0.9153825044631958,
+        "learning_rate": 0.0005,
+        "loss": 2.1649,
+        "step": 1280
+      },
+      {
+        "epoch": 0.04113651583277528,
+        "grad_norm": 1.3356082439422607,
+        "learning_rate": 0.0005,
+        "loss": 2.1648,
+        "step": 1290
+      },
+      {
+        "epoch": 0.0414554035524092,
+        "grad_norm": 0.9260676503181458,
+        "learning_rate": 0.0005,
+        "loss": 2.1341,
+        "step": 1300
+      },
+      {
+        "epoch": 0.041774291272043115,
+        "grad_norm": 0.9380478262901306,
+        "learning_rate": 0.0005,
+        "loss": 2.117,
+        "step": 1310
+      },
+      {
+        "epoch": 0.04209317899167703,
+        "grad_norm": 0.9406986832618713,
+        "learning_rate": 0.0005,
+        "loss": 2.1629,
+        "step": 1320
+      },
+      {
+        "epoch": 0.042412066711310946,
+        "grad_norm": 1.151871681213379,
+        "learning_rate": 0.0005,
+        "loss": 2.1444,
+        "step": 1330
+      },
+      {
+        "epoch": 0.042730954430944865,
+        "grad_norm": 1.0236380100250244,
+        "learning_rate": 0.0005,
+        "loss": 2.1186,
+        "step": 1340
+      },
+      {
+        "epoch": 0.04304984215057878,
+        "grad_norm": 0.9131529331207275,
+        "learning_rate": 0.0005,
+        "loss": 2.1427,
+        "step": 1350
+      },
+      {
+        "epoch": 0.043368729870212695,
+        "grad_norm": 0.9835687279701233,
+        "learning_rate": 0.0005,
+        "loss": 2.1181,
+        "step": 1360
+      },
+      {
+        "epoch": 0.043687617589846614,
+        "grad_norm": 0.9309093356132507,
+        "learning_rate": 0.0005,
+        "loss": 2.1202,
+        "step": 1370
+      },
+      {
+        "epoch": 0.04400650530948053,
+        "grad_norm": 0.9279321432113647,
+        "learning_rate": 0.0005,
+        "loss": 2.1264,
+        "step": 1380
+      },
+      {
+        "epoch": 0.04432539302911445,
+        "grad_norm": 0.8885673880577087,
+        "learning_rate": 0.0005,
+        "loss": 2.0761,
+        "step": 1390
+      },
+      {
+        "epoch": 0.04464428074874836,
+        "grad_norm": 0.9342918992042542,
+        "learning_rate": 0.0005,
+        "loss": 2.1036,
+        "step": 1400
+      },
+      {
+        "epoch": 0.04496316846838228,
+        "grad_norm": 0.9131391644477844,
+        "learning_rate": 0.0005,
+        "loss": 2.104,
+        "step": 1410
+      },
+      {
+        "epoch": 0.0452820561880162,
+        "grad_norm": 0.974519670009613,
+        "learning_rate": 0.0005,
+        "loss": 2.0614,
+        "step": 1420
+      },
+      {
+        "epoch": 0.04560094390765012,
+        "grad_norm": 0.9178261756896973,
+        "learning_rate": 0.0005,
+        "loss": 2.0875,
+        "step": 1430
+      },
+      {
+        "epoch": 0.04591983162728403,
+        "grad_norm": 0.9592669606208801,
+        "learning_rate": 0.0005,
+        "loss": 2.0546,
+        "step": 1440
+      },
+      {
+        "epoch": 0.04623871934691795,
+        "grad_norm": 0.9199544191360474,
+        "learning_rate": 0.0005,
+        "loss": 2.1031,
+        "step": 1450
+      },
+      {
+        "epoch": 0.04655760706655187,
+        "grad_norm": 1.066640019416809,
+        "learning_rate": 0.0005,
+        "loss": 2.0426,
+        "step": 1460
+      },
+      {
+        "epoch": 0.04687649478618578,
+        "grad_norm": 0.919484555721283,
+        "learning_rate": 0.0005,
+        "loss": 2.0447,
+        "step": 1470
+      },
+      {
+        "epoch": 0.0471953825058197,
+        "grad_norm": 0.9945185780525208,
+        "learning_rate": 0.0005,
+        "loss": 2.0507,
+        "step": 1480
+      },
+      {
+        "epoch": 0.04751427022545362,
+        "grad_norm": 0.9459637403488159,
+        "learning_rate": 0.0005,
+        "loss": 2.0422,
+        "step": 1490
+      },
+      {
+        "epoch": 0.04783315794508754,
+        "grad_norm": 0.9146264791488647,
+        "learning_rate": 0.0005,
+        "loss": 2.0266,
+        "step": 1500
+      },
+      {
+        "epoch": 0.04815204566472145,
+        "grad_norm": 0.912291407585144,
+        "learning_rate": 0.0005,
+        "loss": 2.0012,
+        "step": 1510
+      },
+      {
+        "epoch": 0.04847093338435537,
+        "grad_norm": 0.9094820618629456,
+        "learning_rate": 0.0005,
+        "loss": 2.0285,
+        "step": 1520
+      },
+      {
+        "epoch": 0.048789821103989287,
+        "grad_norm": 0.9189174771308899,
+        "learning_rate": 0.0005,
+        "loss": 2.0342,
+        "step": 1530
+      },
+      {
+        "epoch": 0.049108708823623205,
+        "grad_norm": 0.985211193561554,
+        "learning_rate": 0.0005,
+        "loss": 1.9833,
+        "step": 1540
+      },
+      {
+        "epoch": 0.04942759654325712,
+        "grad_norm": 0.9880468845367432,
+        "learning_rate": 0.0005,
+        "loss": 2.0074,
+        "step": 1550
+      },
+      {
+        "epoch": 0.049746484262891036,
+        "grad_norm": 1.0333455801010132,
+        "learning_rate": 0.0005,
+        "loss": 1.9974,
+        "step": 1560
+      },
+      {
+        "epoch": 0.050065371982524955,
+        "grad_norm": 1.1017787456512451,
+        "learning_rate": 0.0005,
+        "loss": 2.001,
+        "step": 1570
+      },
+      {
+        "epoch": 0.050384259702158866,
+        "grad_norm": 0.8883527517318726,
+        "learning_rate": 0.0005,
+        "loss": 1.968,
+        "step": 1580
+      },
+      {
+        "epoch": 0.050703147421792785,
+        "grad_norm": 0.8553420901298523,
+        "learning_rate": 0.0005,
+        "loss": 1.9722,
+        "step": 1590
+      },
+      {
+        "epoch": 0.051022035141426704,
+        "grad_norm": 1.0707831382751465,
+        "learning_rate": 0.0005,
+        "loss": 1.9713,
+        "step": 1600
+      },
+      {
+        "epoch": 0.05134092286106062,
+        "grad_norm": 1.1193410158157349,
+        "learning_rate": 0.0005,
+        "loss": 1.9747,
+        "step": 1610
+      },
+      {
+        "epoch": 0.051659810580694535,
+        "grad_norm": 1.1910367012023926,
+        "learning_rate": 0.0005,
+        "loss": 1.9869,
+        "step": 1620
+      },
+      {
+        "epoch": 0.05197869830032845,
+        "grad_norm": 1.4951293468475342,
+        "learning_rate": 0.0005,
+        "loss": 1.9484,
+        "step": 1630
+      },
+      {
+        "epoch": 0.05229758601996237,
+        "grad_norm": 0.9338844418525696,
+        "learning_rate": 0.0005,
+        "loss": 1.9554,
+        "step": 1640
+      },
+      {
+        "epoch": 0.05261647373959629,
+        "grad_norm": 1.0976824760437012,
+        "learning_rate": 0.0005,
+        "loss": 1.9719,
+        "step": 1650
+      },
+      {
+        "epoch": 0.0529353614592302,
+        "grad_norm": 1.0786077976226807,
+        "learning_rate": 0.0005,
+        "loss": 1.9556,
+        "step": 1660
+      },
+      {
+        "epoch": 0.05325424917886412,
+        "grad_norm": 0.9365764260292053,
+        "learning_rate": 0.0005,
+        "loss": 1.9511,
+        "step": 1670
+      },
+      {
+        "epoch": 0.05357313689849804,
+        "grad_norm": 0.8856982588768005,
+        "learning_rate": 0.0005,
+        "loss": 1.903,
+        "step": 1680
+      },
+      {
+        "epoch": 0.05389202461813196,
+        "grad_norm": 0.8451250791549683,
+        "learning_rate": 0.0005,
+        "loss": 1.9223,
+        "step": 1690
+      },
+      {
+        "epoch": 0.05421091233776587,
+        "grad_norm": 0.8898938894271851,
+        "learning_rate": 0.0005,
+        "loss": 1.9358,
+        "step": 1700
+      },
+      {
+        "epoch": 0.05452980005739979,
+        "grad_norm": 0.9035238027572632,
+        "learning_rate": 0.0005,
+        "loss": 1.9235,
+        "step": 1710
+      },
+      {
+        "epoch": 0.05484868777703371,
+        "grad_norm": 0.8509615659713745,
+        "learning_rate": 0.0005,
+        "loss": 1.9259,
+        "step": 1720
+      },
+      {
+        "epoch": 0.05516757549666762,
+        "grad_norm": 0.8624535799026489,
+        "learning_rate": 0.0005,
+        "loss": 1.9197,
+        "step": 1730
+      },
+      {
+        "epoch": 0.05548646321630154,
+        "grad_norm": 1.032548427581787,
+        "learning_rate": 0.0005,
+        "loss": 1.9275,
+        "step": 1740
+      },
+      {
+        "epoch": 0.05580535093593546,
+        "grad_norm": 0.8983070254325867,
+        "learning_rate": 0.0005,
+        "loss": 1.9079,
+        "step": 1750
+      },
+      {
+        "epoch": 0.056124238655569376,
+        "grad_norm": 0.8418633341789246,
+        "learning_rate": 0.0005,
+        "loss": 1.9123,
+        "step": 1760
+      },
+      {
+        "epoch": 0.05644312637520329,
+        "grad_norm": 0.9653736352920532,
+        "learning_rate": 0.0005,
+        "loss": 1.8994,
+        "step": 1770
+      },
+      {
+        "epoch": 0.05676201409483721,
+        "grad_norm": 0.9427167177200317,
+        "learning_rate": 0.0005,
+        "loss": 1.9248,
+        "step": 1780
+      },
+      {
+        "epoch": 0.057080901814471126,
+        "grad_norm": 1.0386487245559692,
+        "learning_rate": 0.0005,
+        "loss": 1.895,
+        "step": 1790
+      },
+      {
+        "epoch": 0.057399789534105045,
+        "grad_norm": 1.0085606575012207,
+        "learning_rate": 0.0005,
+        "loss": 1.8912,
+        "step": 1800
+      },
+      {
+        "epoch": 0.057718677253738956,
+        "grad_norm": 1.0900158882141113,
+        "learning_rate": 0.0005,
+        "loss": 1.9122,
+        "step": 1810
+      },
+      {
+        "epoch": 0.058037564973372875,
+        "grad_norm": 0.8093986511230469,
+        "learning_rate": 0.0005,
+        "loss": 1.8888,
+        "step": 1820
+      },
+      {
+        "epoch": 0.058356452693006794,
+        "grad_norm": 1.0416687726974487,
+        "learning_rate": 0.0005,
+        "loss": 1.8784,
+        "step": 1830
+      },
+      {
+        "epoch": 0.058675340412640706,
+        "grad_norm": 0.825915515422821,
+        "learning_rate": 0.0005,
+        "loss": 1.8781,
+        "step": 1840
+      },
+      {
+        "epoch": 0.058994228132274625,
+        "grad_norm": 0.896038293838501,
+        "learning_rate": 0.0005,
+        "loss": 1.8479,
+        "step": 1850
+      },
+      {
+        "epoch": 0.05931311585190854,
+        "grad_norm": 0.9250373840332031,
+        "learning_rate": 0.0005,
+        "loss": 1.8707,
+        "step": 1860
+      },
+      {
+        "epoch": 0.05963200357154246,
+        "grad_norm": 0.8338049054145813,
+        "learning_rate": 0.0005,
+        "loss": 1.8715,
+        "step": 1870
+      },
+      {
+        "epoch": 0.059950891291176374,
+        "grad_norm": 0.8659617900848389,
+        "learning_rate": 0.0005,
+        "loss": 1.8735,
+        "step": 1880
+      },
+      {
+        "epoch": 0.06026977901081029,
+        "grad_norm": 1.0219095945358276,
+        "learning_rate": 0.0005,
+        "loss": 1.8642,
+        "step": 1890
+      },
+      {
+        "epoch": 0.06058866673044421,
+        "grad_norm": 0.8440826535224915,
+        "learning_rate": 0.0005,
+        "loss": 1.8672,
+        "step": 1900
+      },
+      {
+        "epoch": 0.06090755445007813,
+        "grad_norm": 0.8432860970497131,
+        "learning_rate": 0.0005,
+        "loss": 1.8538,
+        "step": 1910
+      },
+      {
+        "epoch": 0.06122644216971204,
+        "grad_norm": 0.8388710021972656,
+        "learning_rate": 0.0005,
+        "loss": 1.8308,
+        "step": 1920
+      },
+      {
+        "epoch": 0.06154532988934596,
+        "grad_norm": 0.9477251768112183,
+        "learning_rate": 0.0005,
+        "loss": 1.8575,
+        "step": 1930
+      },
+      {
+        "epoch": 0.06186421760897988,
+        "grad_norm": 0.8459080457687378,
+        "learning_rate": 0.0005,
+        "loss": 1.8327,
+        "step": 1940
+      },
+      {
+        "epoch": 0.0621831053286138,
+        "grad_norm": 0.8614171743392944,
+        "learning_rate": 0.0005,
+        "loss": 1.8447,
+        "step": 1950
+      },
+      {
+        "epoch": 0.06250199304824772,
+        "grad_norm": 0.8767685294151306,
+        "learning_rate": 0.0005,
+        "loss": 1.8365,
+        "step": 1960
+      },
+      {
+        "epoch": 0.06282088076788163,
+        "grad_norm": 0.856082558631897,
+        "learning_rate": 0.0005,
+        "loss": 1.847,
+        "step": 1970
+      },
+      {
+        "epoch": 0.06313976848751554,
+        "grad_norm": 0.8638346791267395,
+        "learning_rate": 0.0005,
+        "loss": 1.8228,
+        "step": 1980
+      },
+      {
+        "epoch": 0.06345865620714947,
+        "grad_norm": 0.8353111743927002,
+        "learning_rate": 0.0005,
+        "loss": 1.8124,
+        "step": 1990
+      },
+      {
+        "epoch": 0.06377754392678338,
+        "grad_norm": 0.8533226847648621,
+        "learning_rate": 0.0005,
+        "loss": 1.8204,
+        "step": 2000
+      },
+      {
+        "epoch": 0.06409643164641729,
+        "grad_norm": 0.7952243089675903,
+        "learning_rate": 0.0005,
+        "loss": 1.8185,
+        "step": 2010
+      },
+      {
+        "epoch": 0.06441531936605122,
+        "grad_norm": 0.8410241007804871,
+        "learning_rate": 0.0005,
+        "loss": 1.8369,
+        "step": 2020
+      },
+      {
+        "epoch": 0.06473420708568513,
+        "grad_norm": 1.3236920833587646,
+        "learning_rate": 0.0005,
+        "loss": 1.8089,
+        "step": 2030
+      },
+      {
+        "epoch": 0.06505309480531905,
+        "grad_norm": 0.9753334522247314,
+        "learning_rate": 0.0005,
+        "loss": 1.8173,
+        "step": 2040
+      },
+      {
+        "epoch": 0.06537198252495297,
+        "grad_norm": 0.811911940574646,
+        "learning_rate": 0.0005,
+        "loss": 1.7852,
+        "step": 2050
+      },
+      {
+        "epoch": 0.06569087024458688,
+        "grad_norm": 0.8044120669364929,
+        "learning_rate": 0.0005,
+        "loss": 1.7763,
+        "step": 2060
+      },
+      {
+        "epoch": 0.0660097579642208,
+        "grad_norm": 0.8142799735069275,
+        "learning_rate": 0.0005,
+        "loss": 1.8074,
+        "step": 2070
+      },
+      {
+        "epoch": 0.06632864568385471,
+        "grad_norm": 0.8504486680030823,
+        "learning_rate": 0.0005,
+        "loss": 1.7901,
+        "step": 2080
+      },
+      {
+        "epoch": 0.06664753340348863,
+        "grad_norm": 0.9170723557472229,
+        "learning_rate": 0.0005,
+        "loss": 1.8103,
+        "step": 2090
+      },
+      {
+        "epoch": 0.06696642112312255,
+        "grad_norm": 0.8451781272888184,
+        "learning_rate": 0.0005,
+        "loss": 1.8068,
+        "step": 2100
+      },
+      {
+        "epoch": 0.06728530884275646,
+        "grad_norm": 0.8535498976707458,
+        "learning_rate": 0.0005,
+        "loss": 1.7775,
+        "step": 2110
+      },
+      {
+        "epoch": 0.06760419656239038,
+        "grad_norm": 0.852803111076355,
+        "learning_rate": 0.0005,
+        "loss": 1.7843,
+        "step": 2120
+      },
+      {
+        "epoch": 0.0679230842820243,
+        "grad_norm": 0.8039234280586243,
+        "learning_rate": 0.0005,
+        "loss": 1.785,
+        "step": 2130
+      },
+      {
+        "epoch": 0.06824197200165821,
+        "grad_norm": 0.8931598663330078,
+        "learning_rate": 0.0005,
+        "loss": 1.7751,
+        "step": 2140
+      },
+      {
+        "epoch": 0.06856085972129214,
+        "grad_norm": 0.8621709942817688,
+        "learning_rate": 0.0005,
+        "loss": 1.7848,
+        "step": 2150
+      },
+      {
+        "epoch": 0.06887974744092605,
+        "grad_norm": 0.8556991815567017,
+        "learning_rate": 0.0005,
+        "loss": 1.7845,
+        "step": 2160
+      },
+      {
+        "epoch": 0.06919863516055996,
+        "grad_norm": 0.8265063166618347,
+        "learning_rate": 0.0005,
+        "loss": 1.7709,
+        "step": 2170
+      },
+      {
+        "epoch": 0.06951752288019389,
+        "grad_norm": 0.9165987372398376,
+        "learning_rate": 0.0005,
+        "loss": 1.7663,
+        "step": 2180
+      },
+      {
+        "epoch": 0.0698364105998278,
+        "grad_norm": 0.8655464053153992,
+        "learning_rate": 0.0005,
+        "loss": 1.7503,
+        "step": 2190
+      },
+      {
+        "epoch": 0.07015529831946171,
+        "grad_norm": 0.8358378410339355,
+        "learning_rate": 0.0005,
+        "loss": 1.767,
+        "step": 2200
+      },
+      {
+        "epoch": 0.07047418603909564,
+        "grad_norm": 0.7970449328422546,
+        "learning_rate": 0.0005,
+        "loss": 1.7433,
+        "step": 2210
+      },
+      {
+        "epoch": 0.07079307375872955,
+        "grad_norm": 0.8039758205413818,
+        "learning_rate": 0.0005,
+        "loss": 1.7564,
+        "step": 2220
+      },
+      {
+        "epoch": 0.07111196147836348,
+        "grad_norm": 0.7957394123077393,
+        "learning_rate": 0.0005,
+        "loss": 1.7397,
+        "step": 2230
+      },
+      {
+        "epoch": 0.07143084919799739,
+        "grad_norm": 0.8606018424034119,
+        "learning_rate": 0.0005,
+        "loss": 1.7447,
+        "step": 2240
+      },
+      {
+        "epoch": 0.0717497369176313,
+        "grad_norm": 0.7819696664810181,
+        "learning_rate": 0.0005,
+        "loss": 1.7233,
+        "step": 2250
+      },
+      {
+        "epoch": 0.07206862463726522,
+        "grad_norm": 0.775804877281189,
+        "learning_rate": 0.0005,
+        "loss": 1.7132,
+        "step": 2260
+      },
+      {
+        "epoch": 0.07238751235689914,
+        "grad_norm": 0.9183884859085083,
+        "learning_rate": 0.0005,
+        "loss": 1.7403,
+        "step": 2270
+      },
+      {
+        "epoch": 0.07270640007653305,
+        "grad_norm": 0.8144393563270569,
+        "learning_rate": 0.0005,
+        "loss": 1.7333,
+        "step": 2280
+      },
+      {
+        "epoch": 0.07302528779616697,
+        "grad_norm": 0.8179895281791687,
+        "learning_rate": 0.0005,
+        "loss": 1.714,
+        "step": 2290
+      },
+      {
+        "epoch": 0.07334417551580089,
+        "grad_norm": 0.7930556535720825,
+        "learning_rate": 0.0005,
+        "loss": 1.7198,
+        "step": 2300
+      },
+      {
+        "epoch": 0.0736630632354348,
+        "grad_norm": 0.7923799157142639,
+        "learning_rate": 0.0005,
+        "loss": 1.7112,
+        "step": 2310
+      },
+      {
+        "epoch": 0.07398195095506872,
+        "grad_norm": 0.8188552856445312,
+        "learning_rate": 0.0005,
+        "loss": 1.7261,
+        "step": 2320
+      },
+      {
+        "epoch": 0.07430083867470264,
+        "grad_norm": 0.7819402813911438,
+        "learning_rate": 0.0005,
+        "loss": 1.7271,
+        "step": 2330
+      },
+      {
+        "epoch": 0.07461972639433656,
+        "grad_norm": 0.7663988471031189,
+        "learning_rate": 0.0005,
+        "loss": 1.7076,
+        "step": 2340
+      },
+      {
+        "epoch": 0.07493861411397047,
+        "grad_norm": 0.7735258340835571,
+        "learning_rate": 0.0005,
+        "loss": 1.7281,
+        "step": 2350
+      },
+      {
+        "epoch": 0.07525750183360438,
+        "grad_norm": 0.7962582111358643,
+        "learning_rate": 0.0005,
+        "loss": 1.717,
+        "step": 2360
+      },
+      {
+        "epoch": 0.07557638955323831,
+        "grad_norm": 0.7835531830787659,
+        "learning_rate": 0.0005,
+        "loss": 1.6795,
+        "step": 2370
+      },
+      {
+        "epoch": 0.07589527727287222,
+        "grad_norm": 0.7508684396743774,
+        "learning_rate": 0.0005,
+        "loss": 1.6853,
+        "step": 2380
+      },
+      {
+        "epoch": 0.07621416499250613,
+        "grad_norm": 0.9591211080551147,
+        "learning_rate": 0.0005,
+        "loss": 1.722,
+        "step": 2390
+      },
+      {
+        "epoch": 0.07653305271214006,
+        "grad_norm": 0.7573085427284241,
+        "learning_rate": 0.0005,
+        "loss": 1.7091,
+        "step": 2400
+      },
+      {
+        "epoch": 0.07685194043177397,
+        "grad_norm": 0.791395902633667,
+        "learning_rate": 0.0005,
+        "loss": 1.7169,
+        "step": 2410
+      },
+      {
+        "epoch": 0.07717082815140788,
+        "grad_norm": 0.7925633788108826,
+        "learning_rate": 0.0005,
+        "loss": 1.6926,
+        "step": 2420
+      },
+      {
+        "epoch": 0.07748971587104181,
+        "grad_norm": 0.7916731238365173,
+        "learning_rate": 0.0005,
+        "loss": 1.6913,
+        "step": 2430
+      },
+      {
+        "epoch": 0.07780860359067572,
+        "grad_norm": 0.7813157439231873,
+        "learning_rate": 0.0005,
+        "loss": 1.6846,
+        "step": 2440
+      },
+      {
+        "epoch": 0.07812749131030965,
+        "grad_norm": 0.9070016145706177,
+        "learning_rate": 0.0005,
+        "loss": 1.6771,
+        "step": 2450
+      },
+      {
+        "epoch": 0.07844637902994356,
+        "grad_norm": 0.8243926763534546,
+        "learning_rate": 0.0005,
+        "loss": 1.6888,
+        "step": 2460
+      },
+      {
+        "epoch": 0.07876526674957747,
+        "grad_norm": 0.8692016005516052,
+        "learning_rate": 0.0005,
+        "loss": 1.6894,
+        "step": 2470
+      },
+      {
+        "epoch": 0.0790841544692114,
+        "grad_norm": 0.8020526170730591,
+        "learning_rate": 0.0005,
+        "loss": 1.6661,
+        "step": 2480
+      },
+      {
+        "epoch": 0.07940304218884531,
+        "grad_norm": 0.739294171333313,
+        "learning_rate": 0.0005,
+        "loss": 1.7,
+        "step": 2490
+      },
+      {
+        "epoch": 0.07972192990847922,
+        "grad_norm": 0.7623685598373413,
+        "learning_rate": 0.0005,
+        "loss": 1.6818,
+        "step": 2500
+      },
+      {
+        "epoch": 0.08004081762811315,
+        "grad_norm": 0.7918365597724915,
+        "learning_rate": 0.0005,
+        "loss": 1.6575,
+        "step": 2510
+      },
+      {
+        "epoch": 0.08035970534774706,
+        "grad_norm": 0.8381354212760925,
+        "learning_rate": 0.0005,
+        "loss": 1.6799,
+        "step": 2520
+      },
+      {
+        "epoch": 0.08067859306738097,
+        "grad_norm": 0.7383261919021606,
+        "learning_rate": 0.0005,
+        "loss": 1.6723,
+        "step": 2530
+      },
+      {
+        "epoch": 0.0809974807870149,
+        "grad_norm": 0.74249267578125,
+        "learning_rate": 0.0005,
+        "loss": 1.6751,
+        "step": 2540
+      },
+      {
+        "epoch": 0.0813163685066488,
+        "grad_norm": 0.8358879685401917,
+        "learning_rate": 0.0005,
+        "loss": 1.6605,
+        "step": 2550
+      },
+      {
+        "epoch": 0.08163525622628273,
+        "grad_norm": 0.8204383850097656,
+        "learning_rate": 0.0005,
+        "loss": 1.6672,
+        "step": 2560
+      },
+      {
+        "epoch": 0.08195414394591664,
+        "grad_norm": 0.7568264007568359,
+        "learning_rate": 0.0005,
+        "loss": 1.6383,
+        "step": 2570
+      },
+      {
+        "epoch": 0.08227303166555056,
+        "grad_norm": 0.7871072292327881,
+        "learning_rate": 0.0005,
+        "loss": 1.6717,
+        "step": 2580
+      },
+      {
+        "epoch": 0.08259191938518448,
+        "grad_norm": 0.7703539133071899,
+        "learning_rate": 0.0005,
+        "loss": 1.6437,
+        "step": 2590
+      },
+      {
+        "epoch": 0.0829108071048184,
+        "grad_norm": 0.7762648463249207,
+        "learning_rate": 0.0005,
+        "loss": 1.6584,
+        "step": 2600
+      },
+      {
+        "epoch": 0.0832296948244523,
+        "grad_norm": 0.8607839345932007,
+        "learning_rate": 0.0005,
+        "loss": 1.6598,
+        "step": 2610
+      },
+      {
+        "epoch": 0.08354858254408623,
+        "grad_norm": 0.7817122936248779,
+        "learning_rate": 0.0005,
+        "loss": 1.6381,
+        "step": 2620
+      },
+      {
+        "epoch": 0.08386747026372014,
+        "grad_norm": 0.9235209226608276,
+        "learning_rate": 0.0005,
+        "loss": 1.6452,
+        "step": 2630
+      },
+      {
+        "epoch": 0.08418635798335405,
+        "grad_norm": 0.7671082019805908,
+        "learning_rate": 0.0005,
+        "loss": 1.6419,
+        "step": 2640
+      },
+      {
+        "epoch": 0.08450524570298798,
+        "grad_norm": 0.7085024118423462,
+        "learning_rate": 0.0005,
+        "loss": 1.6466,
+        "step": 2650
+      },
+      {
+        "epoch": 0.08482413342262189,
+        "grad_norm": 0.7406449913978577,
+        "learning_rate": 0.0005,
+        "loss": 1.6507,
+        "step": 2660
+      },
+      {
+        "epoch": 0.08514302114225582,
+        "grad_norm": 0.8121161460876465,
+        "learning_rate": 0.0005,
+        "loss": 1.6288,
+        "step": 2670
+      },
+      {
+        "epoch": 0.08546190886188973,
+        "grad_norm": 0.766755998134613,
+        "learning_rate": 0.0005,
+        "loss": 1.6614,
+        "step": 2680
+      },
+      {
+        "epoch": 0.08578079658152364,
+        "grad_norm": 0.8837841153144836,
+        "learning_rate": 0.0005,
+        "loss": 1.6366,
+        "step": 2690
+      },
+      {
+        "epoch": 0.08609968430115757,
+        "grad_norm": 0.7676401734352112,
+        "learning_rate": 0.0005,
+        "loss": 1.6415,
+        "step": 2700
+      },
+      {
+        "epoch": 0.08641857202079148,
+        "grad_norm": 0.8545228838920593,
+        "learning_rate": 0.0005,
+        "loss": 1.6338,
+        "step": 2710
+      },
+      {
+        "epoch": 0.08673745974042539,
+        "grad_norm": 0.8082299828529358,
+        "learning_rate": 0.0005,
+        "loss": 1.6483,
+        "step": 2720
+      },
+      {
+        "epoch": 0.08705634746005932,
+        "grad_norm": 0.802437961101532,
+        "learning_rate": 0.0005,
+        "loss": 1.6479,
+        "step": 2730
+      },
+      {
+        "epoch": 0.08737523517969323,
+        "grad_norm": 0.8295302391052246,
+        "learning_rate": 0.0005,
+        "loss": 1.6678,
+        "step": 2740
+      },
+      {
+        "epoch": 0.08769412289932715,
+        "grad_norm": 0.8102624416351318,
+        "learning_rate": 0.0005,
+        "loss": 1.6217,
+        "step": 2750
+      },
+      {
+        "epoch": 0.08801301061896107,
+        "grad_norm": 0.7187849879264832,
+        "learning_rate": 0.0005,
+        "loss": 1.6188,
+        "step": 2760
+      },
+      {
+        "epoch": 0.08833189833859498,
+        "grad_norm": 0.810950756072998,
+        "learning_rate": 0.0005,
+        "loss": 1.6191,
+        "step": 2770
+      },
+      {
+        "epoch": 0.0886507860582289,
+        "grad_norm": 0.7218251824378967,
+        "learning_rate": 0.0005,
+        "loss": 1.6483,
+        "step": 2780
+      },
+      {
+        "epoch": 0.08896967377786281,
+        "grad_norm": 0.7232185006141663,
+        "learning_rate": 0.0005,
+        "loss": 1.5982,
+        "step": 2790
+      },
+      {
+        "epoch": 0.08928856149749673,
+        "grad_norm": 0.7348928451538086,
+        "learning_rate": 0.0005,
+        "loss": 1.6335,
+        "step": 2800
+      },
+      {
+        "epoch": 0.08960744921713065,
+        "grad_norm": 0.793999195098877,
+        "learning_rate": 0.0005,
+        "loss": 1.5883,
+        "step": 2810
+      },
+      {
+        "epoch": 0.08992633693676456,
+        "grad_norm": 0.7381916642189026,
+        "learning_rate": 0.0005,
+        "loss": 1.6178,
+        "step": 2820
+      },
+      {
+        "epoch": 0.09024522465639848,
+        "grad_norm": 0.735526442527771,
+        "learning_rate": 0.0005,
+        "loss": 1.6145,
+        "step": 2830
+      },
+      {
+        "epoch": 0.0905641123760324,
+        "grad_norm": 0.7207257747650146,
+        "learning_rate": 0.0005,
+        "loss": 1.6203,
+        "step": 2840
+      },
+      {
+        "epoch": 0.09088300009566631,
+        "grad_norm": 0.7474569082260132,
+        "learning_rate": 0.0005,
+        "loss": 1.596,
+        "step": 2850
+      },
+      {
+        "epoch": 0.09120188781530024,
+        "grad_norm": 0.7070997953414917,
+        "learning_rate": 0.0005,
+        "loss": 1.6022,
+        "step": 2860
+      },
+      {
+        "epoch": 0.09152077553493415,
+        "grad_norm": 0.7811879515647888,
+        "learning_rate": 0.0005,
+        "loss": 1.6031,
+        "step": 2870
+      },
+      {
+        "epoch": 0.09183966325456806,
+        "grad_norm": 0.7370976209640503,
+        "learning_rate": 0.0005,
+        "loss": 1.6109,
+        "step": 2880
+      },
+      {
+        "epoch": 0.09215855097420199,
+        "grad_norm": 0.733187198638916,
+        "learning_rate": 0.0005,
+        "loss": 1.5982,
+        "step": 2890
+      },
+      {
+        "epoch": 0.0924774386938359,
+        "grad_norm": 0.7401499152183533,
+        "learning_rate": 0.0005,
+        "loss": 1.6011,
+        "step": 2900
+      },
+      {
+        "epoch": 0.09279632641346981,
+        "grad_norm": 0.7699506282806396,
+        "learning_rate": 0.0005,
+        "loss": 1.6043,
+        "step": 2910
+      },
+      {
+        "epoch": 0.09311521413310374,
+        "grad_norm": 0.7591161727905273,
+        "learning_rate": 0.0005,
+        "loss": 1.6081,
+        "step": 2920
+      },
+      {
+        "epoch": 0.09343410185273765,
+        "grad_norm": 0.7230544090270996,
+        "learning_rate": 0.0005,
+        "loss": 1.6115,
+        "step": 2930
+      },
+      {
+        "epoch": 0.09375298957237156,
+        "grad_norm": 0.7355446219444275,
+        "learning_rate": 0.0005,
+        "loss": 1.6163,
+        "step": 2940
+      },
+      {
+        "epoch": 0.09407187729200549,
+        "grad_norm": 0.7396198511123657,
+        "learning_rate": 0.0005,
+        "loss": 1.5915,
+        "step": 2950
+      },
+      {
+        "epoch": 0.0943907650116394,
+        "grad_norm": 0.7075987458229065,
+        "learning_rate": 0.0005,
+        "loss": 1.5886,
+        "step": 2960
+      },
+      {
+        "epoch": 0.09470965273127332,
+        "grad_norm": 0.7205303311347961,
+        "learning_rate": 0.0005,
+        "loss": 1.6006,
+        "step": 2970
+      },
+      {
+        "epoch": 0.09502854045090724,
+        "grad_norm": 0.7048115730285645,
+        "learning_rate": 0.0005,
+        "loss": 1.5907,
+        "step": 2980
+      },
+      {
+        "epoch": 0.09534742817054115,
+        "grad_norm": 0.7107388377189636,
+        "learning_rate": 0.0005,
+        "loss": 1.5916,
+        "step": 2990
+      },
+      {
+        "epoch": 0.09566631589017507,
+        "grad_norm": 0.7625985741615295,
+        "learning_rate": 0.0005,
+        "loss": 1.6023,
+        "step": 3000
+      },
+      {
+        "epoch": 0.09598520360980899,
+        "grad_norm": 0.7164791822433472,
+        "learning_rate": 0.0005,
+        "loss": 1.5883,
+        "step": 3010
+      },
+      {
+        "epoch": 0.0963040913294429,
+        "grad_norm": 0.7049804925918579,
+        "learning_rate": 0.0005,
+        "loss": 1.5835,
+        "step": 3020
+      },
+      {
+        "epoch": 0.09662297904907682,
+        "grad_norm": 0.7338706254959106,
+        "learning_rate": 0.0005,
+        "loss": 1.5845,
+        "step": 3030
+      },
+      {
+        "epoch": 0.09694186676871074,
+        "grad_norm": 0.7438425421714783,
+        "learning_rate": 0.0005,
+        "loss": 1.5869,
+        "step": 3040
+      },
+      {
+        "epoch": 0.09726075448834465,
+        "grad_norm": 0.7559459209442139,
+        "learning_rate": 0.0005,
+        "loss": 1.6004,
+        "step": 3050
+      },
+      {
+        "epoch": 0.09757964220797857,
+        "grad_norm": 0.7225502729415894,
+        "learning_rate": 0.0005,
+        "loss": 1.5889,
+        "step": 3060
+      },
+      {
+        "epoch": 0.09789852992761248,
+        "grad_norm": 0.7186091542243958,
+        "learning_rate": 0.0005,
+        "loss": 1.5739,
+        "step": 3070
+      },
+      {
+        "epoch": 0.09821741764724641,
+        "grad_norm": 0.7074151635169983,
+        "learning_rate": 0.0005,
+        "loss": 1.5841,
+        "step": 3080
+      },
+      {
+        "epoch": 0.09853630536688032,
+        "grad_norm": 0.7287281155586243,
+        "learning_rate": 0.0005,
+        "loss": 1.5719,
+        "step": 3090
+      },
+      {
+        "epoch": 0.09885519308651423,
+        "grad_norm": 0.700652539730072,
+        "learning_rate": 0.0005,
+        "loss": 1.5617,
+        "step": 3100
+      },
+      {
+        "epoch": 0.09917408080614816,
+        "grad_norm": 0.7762500047683716,
+        "learning_rate": 0.0005,
+        "loss": 1.5547,
+        "step": 3110
+      },
+      {
+        "epoch": 0.09949296852578207,
+        "grad_norm": 0.7286017537117004,
+        "learning_rate": 0.0005,
+        "loss": 1.5845,
+        "step": 3120
+      },
+      {
+        "epoch": 0.09981185624541598,
+        "grad_norm": 0.7398425340652466,
+        "learning_rate": 0.0005,
+        "loss": 1.5928,
+        "step": 3130
+      },
+      {
+        "epoch": 0.10013074396504991,
+        "grad_norm": 0.6964101195335388,
+        "learning_rate": 0.0005,
+        "loss": 1.5711,
+        "step": 3140
+      },
+      {
+        "epoch": 0.10044963168468382,
+        "grad_norm": 0.7937759160995483,
+        "learning_rate": 0.0005,
+        "loss": 1.5783,
+        "step": 3150
+      },
+      {
+        "epoch": 0.10076851940431773,
+        "grad_norm": 0.7517949342727661,
+        "learning_rate": 0.0005,
+        "loss": 1.5573,
+        "step": 3160
+      },
+      {
+        "epoch": 0.10108740712395166,
+        "grad_norm": 0.7467934489250183,
+        "learning_rate": 0.0005,
+        "loss": 1.5653,
+        "step": 3170
+      },
+      {
+        "epoch": 0.10140629484358557,
+        "grad_norm": 0.686786949634552,
+        "learning_rate": 0.0005,
+        "loss": 1.5551,
+        "step": 3180
+      },
+      {
+        "epoch": 0.1017251825632195,
+        "grad_norm": 0.7808279991149902,
+        "learning_rate": 0.0005,
+        "loss": 1.5439,
+        "step": 3190
+      },
+      {
+        "epoch": 0.10204407028285341,
+        "grad_norm": 0.7051215767860413,
+        "learning_rate": 0.0005,
+        "loss": 1.5463,
+        "step": 3200
+      },
+      {
+        "epoch": 0.10236295800248732,
+        "grad_norm": 0.7086171507835388,
+        "learning_rate": 0.0005,
+        "loss": 1.5796,
+        "step": 3210
+      },
+      {
+        "epoch": 0.10268184572212125,
+        "grad_norm": 0.7291361093521118,
+        "learning_rate": 0.0005,
+        "loss": 1.5741,
+        "step": 3220
+      },
+      {
+        "epoch": 0.10300073344175516,
+        "grad_norm": 0.7728156447410583,
+        "learning_rate": 0.0005,
+        "loss": 1.5405,
+        "step": 3230
+      },
+      {
+        "epoch": 0.10331962116138907,
+        "grad_norm": 0.7691132426261902,
+        "learning_rate": 0.0005,
+        "loss": 1.5261,
+        "step": 3240
+      },
+      {
+        "epoch": 0.103638508881023,
+        "grad_norm": 0.7408841252326965,
+        "learning_rate": 0.0005,
+        "loss": 1.5516,
+        "step": 3250
+      },
+      {
+        "epoch": 0.1039573966006569,
+        "grad_norm": 0.7222586870193481,
+        "learning_rate": 0.0005,
+        "loss": 1.5657,
+        "step": 3260
+      },
+      {
+        "epoch": 0.10427628432029083,
+        "grad_norm": 0.7752788066864014,
+        "learning_rate": 0.0005,
+        "loss": 1.5175,
+        "step": 3270
+      },
+      {
+        "epoch": 0.10459517203992474,
+        "grad_norm": 0.7558481097221375,
+        "learning_rate": 0.0005,
+        "loss": 1.5609,
+        "step": 3280
+      },
+      {
+        "epoch": 0.10491405975955866,
+        "grad_norm": 0.6986261010169983,
+        "learning_rate": 0.0005,
+        "loss": 1.5228,
+        "step": 3290
+      },
+      {
+        "epoch": 0.10523294747919258,
+        "grad_norm": 0.7217074632644653,
+        "learning_rate": 0.0005,
+        "loss": 1.5521,
+        "step": 3300
+      },
+      {
+        "epoch": 0.1055518351988265,
+        "grad_norm": 0.7102519869804382,
+        "learning_rate": 0.0005,
+        "loss": 1.542,
+        "step": 3310
+      },
+      {
+        "epoch": 0.1058707229184604,
+        "grad_norm": 0.7186601161956787,
+        "learning_rate": 0.0005,
+        "loss": 1.5246,
+        "step": 3320
+      },
+      {
+        "epoch": 0.10618961063809433,
+        "grad_norm": 0.7427906394004822,
+        "learning_rate": 0.0005,
+        "loss": 1.5897,
+        "step": 3330
+      },
+      {
+        "epoch": 0.10650849835772824,
+        "grad_norm": 0.6721486449241638,
+        "learning_rate": 0.0005,
+        "loss": 1.555,
+        "step": 3340
+      },
+      {
+        "epoch": 0.10682738607736215,
+        "grad_norm": 0.7453601360321045,
+        "learning_rate": 0.0005,
+        "loss": 1.5611,
+        "step": 3350
+      },
+      {
+        "epoch": 0.10714627379699608,
+        "grad_norm": 0.733101487159729,
+        "learning_rate": 0.0005,
+        "loss": 1.5424,
+        "step": 3360
+      },
+      {
+        "epoch": 0.10746516151662999,
+        "grad_norm": 0.6932695508003235,
+        "learning_rate": 0.0005,
+        "loss": 1.5567,
+        "step": 3370
+      },
+      {
+        "epoch": 0.10778404923626392,
+        "grad_norm": 0.7290396690368652,
+        "learning_rate": 0.0005,
+        "loss": 1.5409,
+        "step": 3380
+      },
+      {
+        "epoch": 0.10810293695589783,
+        "grad_norm": 0.749410092830658,
+        "learning_rate": 0.0005,
+        "loss": 1.5056,
+        "step": 3390
+      },
+      {
+        "epoch": 0.10842182467553174,
+        "grad_norm": 0.7252135276794434,
+        "learning_rate": 0.0005,
+        "loss": 1.526,
+        "step": 3400
+      },
+      {
+        "epoch": 0.10874071239516567,
+        "grad_norm": 0.6941826939582825,
+        "learning_rate": 0.0005,
+        "loss": 1.5365,
+        "step": 3410
+      },
+      {
+        "epoch": 0.10905960011479958,
+        "grad_norm": 0.6843234896659851,
+        "learning_rate": 0.0005,
+        "loss": 1.5437,
+        "step": 3420
+      },
+      {
+        "epoch": 0.10937848783443349,
+        "grad_norm": 0.7058086395263672,
+        "learning_rate": 0.0005,
+        "loss": 1.5404,
+        "step": 3430
+      },
+      {
+        "epoch": 0.10969737555406742,
+        "grad_norm": 0.7085135579109192,
+        "learning_rate": 0.0005,
+        "loss": 1.5432,
+        "step": 3440
+      },
+      {
+        "epoch": 0.11001626327370133,
+        "grad_norm": 0.6775384545326233,
+        "learning_rate": 0.0005,
+        "loss": 1.5483,
+        "step": 3450
+      },
+      {
+        "epoch": 0.11033515099333524,
+        "grad_norm": 0.6983676552772522,
+        "learning_rate": 0.0005,
+        "loss": 1.5062,
+        "step": 3460
+      },
+      {
+        "epoch": 0.11065403871296917,
+        "grad_norm": 0.7024519443511963,
+        "learning_rate": 0.0005,
+        "loss": 1.5343,
+        "step": 3470
+      },
+      {
+        "epoch": 0.11097292643260308,
+        "grad_norm": 0.7096553444862366,
+        "learning_rate": 0.0005,
+        "loss": 1.5145,
+        "step": 3480
+      },
+      {
+        "epoch": 0.111291814152237,
+        "grad_norm": 0.7284001111984253,
+        "learning_rate": 0.0005,
+        "loss": 1.5124,
+        "step": 3490
+      },
+      {
+        "epoch": 0.11161070187187092,
+        "grad_norm": 0.6964322328567505,
+        "learning_rate": 0.0005,
+        "loss": 1.5246,
+        "step": 3500
+      },
+      {
+        "epoch": 0.11192958959150483,
+        "grad_norm": 0.6839665174484253,
+        "learning_rate": 0.0005,
+        "loss": 1.5239,
+        "step": 3510
+      },
+      {
+        "epoch": 0.11224847731113875,
+        "grad_norm": 0.665112316608429,
+        "learning_rate": 0.0005,
+        "loss": 1.5158,
+        "step": 3520
+      },
+      {
+        "epoch": 0.11256736503077266,
+        "grad_norm": 0.7121447920799255,
+        "learning_rate": 0.0005,
+        "loss": 1.5107,
+        "step": 3530
+      },
+      {
+        "epoch": 0.11288625275040658,
+        "grad_norm": 0.6824846267700195,
+        "learning_rate": 0.0005,
+        "loss": 1.5239,
+        "step": 3540
+      },
+      {
+        "epoch": 0.1132051404700405,
+        "grad_norm": 0.6695664525032043,
+        "learning_rate": 0.0005,
+        "loss": 1.5204,
+        "step": 3550
+      },
+      {
+        "epoch": 0.11352402818967441,
+        "grad_norm": 0.6862232089042664,
+        "learning_rate": 0.0005,
+        "loss": 1.5173,
+        "step": 3560
+      },
+      {
+        "epoch": 0.11384291590930833,
+        "grad_norm": 0.6916778087615967,
+        "learning_rate": 0.0005,
+        "loss": 1.4952,
+        "step": 3570
+      },
+      {
+        "epoch": 0.11416180362894225,
+        "grad_norm": 0.6827712655067444,
+        "learning_rate": 0.0005,
+        "loss": 1.4932,
+        "step": 3580
+      },
+      {
+        "epoch": 0.11448069134857616,
+        "grad_norm": 0.7031466960906982,
+        "learning_rate": 0.0005,
+        "loss": 1.5347,
+        "step": 3590
+      },
+      {
+        "epoch": 0.11479957906821009,
+        "grad_norm": 0.7130076289176941,
+        "learning_rate": 0.0005,
+        "loss": 1.5229,
+        "step": 3600
+      },
+      {
+        "epoch": 0.115118466787844,
+        "grad_norm": 0.6662548184394836,
+        "learning_rate": 0.0005,
+        "loss": 1.4887,
+        "step": 3610
+      },
+      {
+        "epoch": 0.11543735450747791,
+        "grad_norm": 0.6988325715065002,
+        "learning_rate": 0.0005,
+        "loss": 1.4849,
+        "step": 3620
+      },
+      {
+        "epoch": 0.11575624222711184,
+        "grad_norm": 0.7220348715782166,
+        "learning_rate": 0.0005,
+        "loss": 1.5171,
+        "step": 3630
+      },
+      {
+        "epoch": 0.11607512994674575,
+        "grad_norm": 0.7088020443916321,
+        "learning_rate": 0.0005,
+        "loss": 1.5283,
+        "step": 3640
+      },
+      {
+        "epoch": 0.11639401766637966,
+        "grad_norm": 0.7373560070991516,
+        "learning_rate": 0.0005,
+        "loss": 1.5044,
+        "step": 3650
+      },
+      {
+        "epoch": 0.11671290538601359,
+        "grad_norm": 0.6741558313369751,
+        "learning_rate": 0.0005,
+        "loss": 1.4895,
+        "step": 3660
+      },
+      {
+        "epoch": 0.1170317931056475,
+        "grad_norm": 0.6753981113433838,
+        "learning_rate": 0.0005,
+        "loss": 1.5154,
+        "step": 3670
+      },
+      {
+        "epoch": 0.11735068082528141,
+        "grad_norm": 0.6715815663337708,
+        "learning_rate": 0.0005,
+        "loss": 1.5322,
+        "step": 3680
+      },
+      {
+        "epoch": 0.11766956854491534,
+        "grad_norm": 0.6624891757965088,
+        "learning_rate": 0.0005,
+        "loss": 1.4974,
+        "step": 3690
+      },
+      {
+        "epoch": 0.11798845626454925,
+        "grad_norm": 0.6958360075950623,
+        "learning_rate": 0.0005,
+        "loss": 1.5063,
+        "step": 3700
+      },
+      {
+        "epoch": 0.11830734398418317,
+        "grad_norm": 0.6877066493034363,
+        "learning_rate": 0.0005,
+        "loss": 1.5088,
+        "step": 3710
+      },
+      {
+        "epoch": 0.11862623170381709,
+        "grad_norm": 0.6606530547142029,
+        "learning_rate": 0.0005,
+        "loss": 1.4886,
+        "step": 3720
+      },
+      {
+        "epoch": 0.118945119423451,
+        "grad_norm": 0.6680471897125244,
+        "learning_rate": 0.0005,
+        "loss": 1.5079,
+        "step": 3730
+      },
+      {
+        "epoch": 0.11926400714308492,
+        "grad_norm": 0.6819968819618225,
+        "learning_rate": 0.0005,
+        "loss": 1.5104,
+        "step": 3740
+      },
+      {
+        "epoch": 0.11958289486271884,
+        "grad_norm": 0.6881066560745239,
+        "learning_rate": 0.0005,
+        "loss": 1.4819,
+        "step": 3750
+      },
+      {
+        "epoch": 0.11990178258235275,
+        "grad_norm": 0.6798304915428162,
+        "learning_rate": 0.0005,
+        "loss": 1.4933,
+        "step": 3760
+      },
+      {
+        "epoch": 0.12022067030198667,
+        "grad_norm": 0.7182119488716125,
+        "learning_rate": 0.0005,
+        "loss": 1.506,
+        "step": 3770
+      },
+      {
+        "epoch": 0.12053955802162059,
+        "grad_norm": 0.6950844526290894,
+        "learning_rate": 0.0005,
+        "loss": 1.5218,
+        "step": 3780
+      },
+      {
+        "epoch": 0.12085844574125451,
+        "grad_norm": 0.692387580871582,
+        "learning_rate": 0.0005,
+        "loss": 1.4787,
+        "step": 3790
+      },
+      {
+        "epoch": 0.12117733346088842,
+        "grad_norm": 0.7074518203735352,
+        "learning_rate": 0.0005,
+        "loss": 1.4774,
+        "step": 3800
+      },
+      {
+        "epoch": 0.12149622118052233,
+        "grad_norm": 0.6878761649131775,
+        "learning_rate": 0.0005,
+        "loss": 1.4782,
+        "step": 3810
+      },
+      {
+        "epoch": 0.12181510890015626,
+        "grad_norm": 0.6472737789154053,
+        "learning_rate": 0.0005,
+        "loss": 1.4832,
+        "step": 3820
+      },
+      {
+        "epoch": 0.12213399661979017,
+        "grad_norm": 0.6680033802986145,
+        "learning_rate": 0.0005,
+        "loss": 1.4781,
+        "step": 3830
+      },
+      {
+        "epoch": 0.12245288433942408,
+        "grad_norm": 0.7118170857429504,
+        "learning_rate": 0.0005,
+        "loss": 1.4986,
+        "step": 3840
+      },
+      {
+        "epoch": 0.12277177205905801,
+        "grad_norm": 0.6713599562644958,
+        "learning_rate": 0.0005,
+        "loss": 1.4971,
+        "step": 3850
+      },
+      {
+        "epoch": 0.12309065977869192,
+        "grad_norm": 0.6707881093025208,
+        "learning_rate": 0.0005,
+        "loss": 1.4985,
+        "step": 3860
+      },
+      {
+        "epoch": 0.12340954749832583,
+        "grad_norm": 0.7143165469169617,
+        "learning_rate": 0.0005,
+        "loss": 1.5075,
+        "step": 3870
+      },
+      {
+        "epoch": 0.12372843521795976,
+        "grad_norm": 0.6780462265014648,
+        "learning_rate": 0.0005,
+        "loss": 1.4691,
+        "step": 3880
+      },
+      {
+        "epoch": 0.12404732293759367,
+        "grad_norm": 0.7073563933372498,
+        "learning_rate": 0.0005,
+        "loss": 1.4906,
+        "step": 3890
+      },
+      {
+        "epoch": 0.1243662106572276,
+        "grad_norm": 0.6700391173362732,
+        "learning_rate": 0.0005,
+        "loss": 1.4866,
+        "step": 3900
+      },
+      {
+        "epoch": 0.12468509837686151,
+        "grad_norm": 0.6709848046302795,
+        "learning_rate": 0.0005,
+        "loss": 1.4813,
+        "step": 3910
+      },
+      {
+        "epoch": 0.12500398609649543,
+        "grad_norm": 0.6658124923706055,
+        "learning_rate": 0.0005,
+        "loss": 1.4889,
+        "step": 3920
+      },
+      {
+        "epoch": 0.12532287381612933,
+        "grad_norm": 0.6852616667747498,
+        "learning_rate": 0.0005,
+        "loss": 1.4599,
+        "step": 3930
+      },
+      {
+        "epoch": 0.12564176153576326,
+        "grad_norm": 0.6966326832771301,
+        "learning_rate": 0.0005,
+        "loss": 1.4588,
+        "step": 3940
+      },
+      {
+        "epoch": 0.12596064925539718,
+        "grad_norm": 0.6924886703491211,
+        "learning_rate": 0.0005,
+        "loss": 1.4731,
+        "step": 3950
+      },
+      {
+        "epoch": 0.12627953697503108,
+        "grad_norm": 0.7122734785079956,
+        "learning_rate": 0.0005,
+        "loss": 1.4586,
+        "step": 3960
+      },
+      {
+        "epoch": 0.126598424694665,
+        "grad_norm": 0.6784149408340454,
+        "learning_rate": 0.0005,
+        "loss": 1.4796,
+        "step": 3970
+      },
+      {
+        "epoch": 0.12691731241429893,
+        "grad_norm": 0.6619938015937805,
+        "learning_rate": 0.0005,
+        "loss": 1.4619,
+        "step": 3980
+      },
+      {
+        "epoch": 0.12723620013393283,
+        "grad_norm": 0.6966860890388489,
+        "learning_rate": 0.0005,
+        "loss": 1.4702,
+        "step": 3990
+      },
+      {
+        "epoch": 0.12755508785356676,
+        "grad_norm": 0.6390230655670166,
+        "learning_rate": 0.0005,
+        "loss": 1.4623,
+        "step": 4000
+      },
+      {
+        "epoch": 0.12787397557320068,
+        "grad_norm": 0.6684625744819641,
+        "learning_rate": 0.0005,
+        "loss": 1.48,
+        "step": 4010
+      },
+      {
+        "epoch": 0.12819286329283458,
+        "grad_norm": 0.796931266784668,
+        "learning_rate": 0.0005,
+        "loss": 1.4648,
+        "step": 4020
+      },
+      {
+        "epoch": 0.1285117510124685,
+        "grad_norm": 0.6581517457962036,
+        "learning_rate": 0.0005,
+        "loss": 1.4751,
+        "step": 4030
+      },
+      {
+        "epoch": 0.12883063873210243,
+        "grad_norm": 0.6862014532089233,
+        "learning_rate": 0.0005,
+        "loss": 1.4666,
+        "step": 4040
+      },
+      {
+        "epoch": 0.12914952645173636,
+        "grad_norm": 0.7265594005584717,
+        "learning_rate": 0.0005,
+        "loss": 1.4742,
+        "step": 4050
+      },
+      {
+        "epoch": 0.12946841417137026,
+        "grad_norm": 0.667931079864502,
+        "learning_rate": 0.0005,
+        "loss": 1.4865,
+        "step": 4060
+      },
+      {
+        "epoch": 0.12978730189100418,
+        "grad_norm": 0.6380496025085449,
+        "learning_rate": 0.0005,
+        "loss": 1.4761,
+        "step": 4070
+      },
+      {
+        "epoch": 0.1301061896106381,
+        "grad_norm": 0.688275158405304,
+        "learning_rate": 0.0005,
+        "loss": 1.4506,
+        "step": 4080
+      },
+      {
+        "epoch": 0.130425077330272,
+        "grad_norm": 0.6321500539779663,
+        "learning_rate": 0.0005,
+        "loss": 1.4676,
+        "step": 4090
+      },
+      {
+        "epoch": 0.13074396504990593,
+        "grad_norm": 0.6487618088722229,
+        "learning_rate": 0.0005,
+        "loss": 1.464,
+        "step": 4100
+      },
+      {
+        "epoch": 0.13106285276953986,
+        "grad_norm": 0.674380898475647,
+        "learning_rate": 0.0005,
+        "loss": 1.4641,
+        "step": 4110
+      },
+      {
+        "epoch": 0.13138174048917375,
+        "grad_norm": 0.6674288511276245,
+        "learning_rate": 0.0005,
+        "loss": 1.4433,
+        "step": 4120
+      },
+      {
+        "epoch": 0.13170062820880768,
+        "grad_norm": 0.611838161945343,
+        "learning_rate": 0.0005,
+        "loss": 1.4598,
+        "step": 4130
+      },
+      {
+        "epoch": 0.1320195159284416,
+        "grad_norm": 0.6699022650718689,
+        "learning_rate": 0.0005,
+        "loss": 1.4584,
+        "step": 4140
+      },
+      {
+        "epoch": 0.1323384036480755,
+        "grad_norm": 0.6623623371124268,
+        "learning_rate": 0.0005,
+        "loss": 1.4421,
+        "step": 4150
+      },
+      {
+        "epoch": 0.13265729136770943,
+        "grad_norm": 0.648499608039856,
+        "learning_rate": 0.0005,
+        "loss": 1.4512,
+        "step": 4160
+      },
+      {
+        "epoch": 0.13297617908734335,
+        "grad_norm": 0.6404657363891602,
+        "learning_rate": 0.0005,
+        "loss": 1.4548,
+        "step": 4170
+      },
+      {
+        "epoch": 0.13329506680697725,
+        "grad_norm": 0.6383042335510254,
+        "learning_rate": 0.0005,
+        "loss": 1.431,
+        "step": 4180
+      },
+      {
+        "epoch": 0.13361395452661118,
+        "grad_norm": 0.6430786848068237,
+        "learning_rate": 0.0005,
+        "loss": 1.4438,
+        "step": 4190
+      },
+      {
+        "epoch": 0.1339328422462451,
+        "grad_norm": 0.643415629863739,
+        "learning_rate": 0.0005,
+        "loss": 1.4543,
+        "step": 4200
+      },
+      {
+        "epoch": 0.134251729965879,
+        "grad_norm": 0.644809365272522,
+        "learning_rate": 0.0005,
+        "loss": 1.4467,
+        "step": 4210
+      },
+      {
+        "epoch": 0.13457061768551293,
+        "grad_norm": 0.6084710955619812,
+        "learning_rate": 0.0005,
+        "loss": 1.4426,
+        "step": 4220
+      },
+      {
+        "epoch": 0.13488950540514685,
+        "grad_norm": 0.6093063950538635,
+        "learning_rate": 0.0005,
+        "loss": 1.4329,
+        "step": 4230
+      },
+      {
+        "epoch": 0.13520839312478075,
+        "grad_norm": 0.6573523879051208,
+        "learning_rate": 0.0005,
+        "loss": 1.4752,
+        "step": 4240
+      },
+      {
+        "epoch": 0.13552728084441468,
+        "grad_norm": 0.6293638944625854,
+        "learning_rate": 0.0005,
+        "loss": 1.4492,
+        "step": 4250
+      },
+      {
+        "epoch": 0.1358461685640486,
+        "grad_norm": 0.6632397174835205,
+        "learning_rate": 0.0005,
+        "loss": 1.4643,
+        "step": 4260
+      },
+      {
+        "epoch": 0.13616505628368253,
+        "grad_norm": 0.6814329028129578,
+        "learning_rate": 0.0005,
+        "loss": 1.4011,
+        "step": 4270
+      },
+      {
+        "epoch": 0.13648394400331643,
+        "grad_norm": 0.7331423163414001,
+        "learning_rate": 0.0005,
+        "loss": 1.4466,
+        "step": 4280
+      },
+      {
+        "epoch": 0.13680283172295035,
+        "grad_norm": 0.645703136920929,
+        "learning_rate": 0.0005,
+        "loss": 1.4247,
+        "step": 4290
+      },
+      {
+        "epoch": 0.13712171944258428,
+        "grad_norm": 0.6080441474914551,
+        "learning_rate": 0.0005,
+        "loss": 1.4398,
+        "step": 4300
+      },
+      {
+        "epoch": 0.13744060716221818,
+        "grad_norm": 0.6814742684364319,
+        "learning_rate": 0.0005,
+        "loss": 1.4523,
+        "step": 4310
+      },
+      {
+        "epoch": 0.1377594948818521,
+        "grad_norm": 0.6626353859901428,
+        "learning_rate": 0.0005,
+        "loss": 1.4323,
+        "step": 4320
+      },
+      {
+        "epoch": 0.13807838260148603,
+        "grad_norm": 0.6317815184593201,
+        "learning_rate": 0.0005,
+        "loss": 1.4403,
+        "step": 4330
+      },
+      {
+        "epoch": 0.13839727032111993,
+        "grad_norm": 0.6251382827758789,
+        "learning_rate": 0.0005,
+        "loss": 1.4377,
+        "step": 4340
+      },
+      {
+        "epoch": 0.13871615804075385,
+        "grad_norm": 0.6542896032333374,
+        "learning_rate": 0.0005,
+        "loss": 1.44,
+        "step": 4350
+      },
+      {
+        "epoch": 0.13903504576038778,
+        "grad_norm": 0.6480959057807922,
+        "learning_rate": 0.0005,
+        "loss": 1.4399,
+        "step": 4360
+      },
+      {
+        "epoch": 0.13935393348002167,
+        "grad_norm": 0.6671744585037231,
+        "learning_rate": 0.0005,
+        "loss": 1.4552,
+        "step": 4370
+      },
+      {
+        "epoch": 0.1396728211996556,
+        "grad_norm": 0.6613544225692749,
+        "learning_rate": 0.0005,
+        "loss": 1.4311,
+        "step": 4380
+      },
+      {
+        "epoch": 0.13999170891928953,
+        "grad_norm": 0.6300452947616577,
+        "learning_rate": 0.0005,
+        "loss": 1.4346,
+        "step": 4390
+      },
+      {
+        "epoch": 0.14031059663892342,
+        "grad_norm": 0.6960680484771729,
+        "learning_rate": 0.0005,
+        "loss": 1.4496,
+        "step": 4400
+      },
+      {
+        "epoch": 0.14062948435855735,
+        "grad_norm": 0.6525735259056091,
+        "learning_rate": 0.0005,
+        "loss": 1.4524,
+        "step": 4410
+      },
+      {
+        "epoch": 0.14094837207819128,
+        "grad_norm": 0.6144914031028748,
+        "learning_rate": 0.0005,
+        "loss": 1.4161,
+        "step": 4420
+      },
+      {
+        "epoch": 0.14126725979782517,
+        "grad_norm": 0.6260567307472229,
+        "learning_rate": 0.0005,
+        "loss": 1.431,
+        "step": 4430
+      },
+      {
+        "epoch": 0.1415861475174591,
+        "grad_norm": 0.6403346657752991,
+        "learning_rate": 0.0005,
+        "loss": 1.4194,
+        "step": 4440
+      },
+      {
+        "epoch": 0.14190503523709302,
+        "grad_norm": 0.6182661652565002,
+        "learning_rate": 0.0005,
+        "loss": 1.4047,
+        "step": 4450
+      },
+      {
+        "epoch": 0.14222392295672695,
+        "grad_norm": 0.6593475341796875,
+        "learning_rate": 0.0005,
+        "loss": 1.4063,
+        "step": 4460
+      },
+      {
+        "epoch": 0.14254281067636085,
+        "grad_norm": 0.661416232585907,
+        "learning_rate": 0.0005,
+        "loss": 1.4229,
+        "step": 4470
+      },
+      {
+        "epoch": 0.14286169839599477,
+        "grad_norm": 0.6097050905227661,
+        "learning_rate": 0.0005,
+        "loss": 1.4242,
+        "step": 4480
+      },
+      {
+        "epoch": 0.1431805861156287,
+        "grad_norm": 0.6463915109634399,
+        "learning_rate": 0.0005,
+        "loss": 1.4195,
+        "step": 4490
+      },
+      {
+        "epoch": 0.1434994738352626,
+        "grad_norm": 0.7001182436943054,
+        "learning_rate": 0.0005,
+        "loss": 1.4472,
+        "step": 4500
+      },
+      {
+        "epoch": 0.14381836155489652,
+        "grad_norm": 0.6229003667831421,
+        "learning_rate": 0.0005,
+        "loss": 1.4525,
+        "step": 4510
+      },
+      {
+        "epoch": 0.14413724927453045,
+        "grad_norm": 0.6146069169044495,
+        "learning_rate": 0.0005,
+        "loss": 1.4217,
+        "step": 4520
+      },
+      {
+        "epoch": 0.14445613699416435,
+        "grad_norm": 0.6353657245635986,
+        "learning_rate": 0.0005,
+        "loss": 1.4177,
+        "step": 4530
+      },
+      {
+        "epoch": 0.14477502471379827,
+        "grad_norm": 0.646435558795929,
+        "learning_rate": 0.0005,
+        "loss": 1.43,
+        "step": 4540
+      },
+      {
+        "epoch": 0.1450939124334322,
+        "grad_norm": 0.6327551603317261,
+        "learning_rate": 0.0005,
+        "loss": 1.4269,
+        "step": 4550
+      },
+      {
+        "epoch": 0.1454128001530661,
+        "grad_norm": 0.63446044921875,
+        "learning_rate": 0.0005,
+        "loss": 1.4469,
+        "step": 4560
+      },
+      {
+        "epoch": 0.14573168787270002,
+        "grad_norm": 0.6193106770515442,
+        "learning_rate": 0.0005,
+        "loss": 1.4215,
+        "step": 4570
+      },
+      {
+        "epoch": 0.14605057559233395,
+        "grad_norm": 0.607028067111969,
+        "learning_rate": 0.0005,
+        "loss": 1.4083,
+        "step": 4580
+      },
+      {
+        "epoch": 0.14636946331196785,
+        "grad_norm": 0.6031425595283508,
+        "learning_rate": 0.0005,
+        "loss": 1.4102,
+        "step": 4590
+      },
+      {
+        "epoch": 0.14668835103160177,
+        "grad_norm": 0.633690595626831,
+        "learning_rate": 0.0005,
+        "loss": 1.4287,
+        "step": 4600
+      },
+      {
+        "epoch": 0.1470072387512357,
+        "grad_norm": 0.6302832365036011,
+        "learning_rate": 0.0005,
+        "loss": 1.4223,
+        "step": 4610
+      },
+      {
+        "epoch": 0.1473261264708696,
+        "grad_norm": 0.6090842485427856,
+        "learning_rate": 0.0005,
+        "loss": 1.4273,
+        "step": 4620
+      },
+      {
+        "epoch": 0.14764501419050352,
+        "grad_norm": 0.6385555863380432,
+        "learning_rate": 0.0005,
+        "loss": 1.4256,
+        "step": 4630
+      },
+      {
+        "epoch": 0.14796390191013745,
+        "grad_norm": 0.6406751871109009,
+        "learning_rate": 0.0005,
+        "loss": 1.4308,
+        "step": 4640
+      },
+      {
+        "epoch": 0.14828278962977134,
+        "grad_norm": 0.6482434868812561,
+        "learning_rate": 0.0005,
+        "loss": 1.4267,
+        "step": 4650
+      },
+      {
+        "epoch": 0.14860167734940527,
+        "grad_norm": 0.7152848839759827,
+        "learning_rate": 0.0005,
+        "loss": 1.4185,
+        "step": 4660
+      },
+      {
+        "epoch": 0.1489205650690392,
+        "grad_norm": 0.6866331100463867,
+        "learning_rate": 0.0005,
+        "loss": 1.4083,
+        "step": 4670
+      },
+      {
+        "epoch": 0.14923945278867312,
+        "grad_norm": 0.6515335440635681,
+        "learning_rate": 0.0005,
+        "loss": 1.4259,
+        "step": 4680
+      },
+      {
+        "epoch": 0.14955834050830702,
+        "grad_norm": 0.6101216673851013,
+        "learning_rate": 0.0005,
+        "loss": 1.4207,
+        "step": 4690
+      },
+      {
+        "epoch": 0.14987722822794095,
+        "grad_norm": 0.6153760552406311,
+        "learning_rate": 0.0005,
+        "loss": 1.3919,
+        "step": 4700
+      },
+      {
+        "epoch": 0.15019611594757487,
+        "grad_norm": 0.6293302178382874,
+        "learning_rate": 0.0005,
+        "loss": 1.4083,
+        "step": 4710
+      },
+      {
+        "epoch": 0.15051500366720877,
+        "grad_norm": 0.6889540553092957,
+        "learning_rate": 0.0005,
+        "loss": 1.3989,
+        "step": 4720
+      },
+      {
+        "epoch": 0.1508338913868427,
+        "grad_norm": 0.6246590614318848,
+        "learning_rate": 0.0005,
+        "loss": 1.4081,
+        "step": 4730
+      },
+      {
+        "epoch": 0.15115277910647662,
+        "grad_norm": 0.6354957818984985,
+        "learning_rate": 0.0005,
+        "loss": 1.4072,
+        "step": 4740
+      },
+      {
+        "epoch": 0.15147166682611052,
+        "grad_norm": 0.6348922252655029,
+        "learning_rate": 0.0005,
+        "loss": 1.3956,
+        "step": 4750
+      },
+      {
+        "epoch": 0.15179055454574444,
+        "grad_norm": 0.6126857995986938,
+        "learning_rate": 0.0005,
+        "loss": 1.3893,
+        "step": 4760
+      },
+      {
+        "epoch": 0.15210944226537837,
+        "grad_norm": 0.6216022372245789,
+        "learning_rate": 0.0005,
+        "loss": 1.4103,
+        "step": 4770
+      },
+      {
+        "epoch": 0.15242832998501227,
+        "grad_norm": 0.5942848920822144,
+        "learning_rate": 0.0005,
+        "loss": 1.3894,
+        "step": 4780
+      },
+      {
+        "epoch": 0.1527472177046462,
+        "grad_norm": 0.6769245862960815,
+        "learning_rate": 0.0005,
+        "loss": 1.4215,
+        "step": 4790
+      },
+      {
+        "epoch": 0.15306610542428012,
+        "grad_norm": 0.6551666855812073,
+        "learning_rate": 0.0005,
+        "loss": 1.3935,
+        "step": 4800
+      },
+      {
+        "epoch": 0.15338499314391402,
+        "grad_norm": 0.6036399006843567,
+        "learning_rate": 0.0005,
+        "loss": 1.4144,
+        "step": 4810
+      },
+      {
+        "epoch": 0.15370388086354794,
+        "grad_norm": 0.5946652889251709,
+        "learning_rate": 0.0005,
+        "loss": 1.4069,
+        "step": 4820
+      },
+      {
+        "epoch": 0.15402276858318187,
+        "grad_norm": 0.577434778213501,
+        "learning_rate": 0.0005,
+        "loss": 1.3909,
+        "step": 4830
+      },
+      {
+        "epoch": 0.15434165630281577,
+        "grad_norm": 0.581184446811676,
+        "learning_rate": 0.0005,
+        "loss": 1.4093,
+        "step": 4840
+      },
+      {
+        "epoch": 0.1546605440224497,
+        "grad_norm": 0.6142603158950806,
+        "learning_rate": 0.0005,
+        "loss": 1.4001,
+        "step": 4850
+      },
+      {
+        "epoch": 0.15497943174208362,
+        "grad_norm": 0.6196522116661072,
+        "learning_rate": 0.0005,
+        "loss": 1.3984,
+        "step": 4860
+      },
+      {
+        "epoch": 0.15529831946171752,
+        "grad_norm": 0.6150513291358948,
+        "learning_rate": 0.0005,
+        "loss": 1.3812,
+        "step": 4870
+      },
+      {
+        "epoch": 0.15561720718135144,
+        "grad_norm": 0.6181188821792603,
+        "learning_rate": 0.0005,
+        "loss": 1.3767,
+        "step": 4880
+      },
+      {
+        "epoch": 0.15593609490098537,
+        "grad_norm": 0.5967665910720825,
+        "learning_rate": 0.0005,
+        "loss": 1.4041,
+        "step": 4890
+      },
+      {
+        "epoch": 0.1562549826206193,
+        "grad_norm": 0.6157842874526978,
+        "learning_rate": 0.0005,
+        "loss": 1.4045,
+        "step": 4900
+      },
+      {
+        "epoch": 0.1565738703402532,
+        "grad_norm": 0.5986465215682983,
+        "learning_rate": 0.0005,
+        "loss": 1.4117,
+        "step": 4910
+      },
+      {
+        "epoch": 0.15689275805988712,
+        "grad_norm": 0.6068465113639832,
+        "learning_rate": 0.0005,
+        "loss": 1.3989,
+        "step": 4920
+      },
+      {
+        "epoch": 0.15721164577952104,
+        "grad_norm": 0.6013087630271912,
+        "learning_rate": 0.0005,
+        "loss": 1.4044,
+        "step": 4930
+      },
+      {
+        "epoch": 0.15753053349915494,
+        "grad_norm": 0.5771439671516418,
+        "learning_rate": 0.0005,
+        "loss": 1.3809,
+        "step": 4940
+      },
+      {
+        "epoch": 0.15784942121878887,
+        "grad_norm": 0.62718665599823,
+        "learning_rate": 0.0005,
+        "loss": 1.3753,
+        "step": 4950
+      },
+      {
+        "epoch": 0.1581683089384228,
+        "grad_norm": 0.6010807752609253,
+        "learning_rate": 0.0005,
+        "loss": 1.4032,
+        "step": 4960
+      },
+      {
+        "epoch": 0.1584871966580567,
+        "grad_norm": 0.5989909768104553,
+        "learning_rate": 0.0005,
+        "loss": 1.3964,
+        "step": 4970
+      },
+      {
+        "epoch": 0.15880608437769062,
+        "grad_norm": 0.6056174635887146,
+        "learning_rate": 0.0005,
+        "loss": 1.3824,
+        "step": 4980
+      },
+      {
+        "epoch": 0.15912497209732454,
+        "grad_norm": 0.645205557346344,
+        "learning_rate": 0.0005,
+        "loss": 1.385,
+        "step": 4990
+      },
+      {
+        "epoch": 0.15944385981695844,
+        "grad_norm": 0.5958195328712463,
+        "learning_rate": 0.0005,
+        "loss": 1.3599,
+        "step": 5000
+      },
+      {
+        "epoch": 0.15976274753659236,
+        "grad_norm": 0.6125824451446533,
+        "learning_rate": 0.0005,
+        "loss": 1.3972,
+        "step": 5010
+      },
+      {
+        "epoch": 0.1600816352562263,
+        "grad_norm": 0.5644070506095886,
+        "learning_rate": 0.0005,
+        "loss": 1.3807,
+        "step": 5020
+      },
+      {
+        "epoch": 0.1604005229758602,
+        "grad_norm": 0.6514316201210022,
+        "learning_rate": 0.0005,
+        "loss": 1.4205,
+        "step": 5030
+      },
+      {
+        "epoch": 0.1607194106954941,
+        "grad_norm": 0.5961187481880188,
+        "learning_rate": 0.0005,
+        "loss": 1.3991,
+        "step": 5040
+      },
+      {
+        "epoch": 0.16103829841512804,
+        "grad_norm": 0.5987887978553772,
+        "learning_rate": 0.0005,
+        "loss": 1.3844,
+        "step": 5050
+      },
+      {
+        "epoch": 0.16135718613476194,
+        "grad_norm": 0.6017197370529175,
+        "learning_rate": 0.0005,
+        "loss": 1.3868,
+        "step": 5060
+      },
+      {
+        "epoch": 0.16167607385439586,
+        "grad_norm": 0.6067407131195068,
+        "learning_rate": 0.0005,
+        "loss": 1.3791,
+        "step": 5070
+      },
+      {
+        "epoch": 0.1619949615740298,
+        "grad_norm": 0.579651951789856,
+        "learning_rate": 0.0005,
+        "loss": 1.3845,
+        "step": 5080
+      },
+      {
+        "epoch": 0.16231384929366371,
+        "grad_norm": 0.571557879447937,
+        "learning_rate": 0.0005,
+        "loss": 1.4063,
+        "step": 5090
+      },
+      {
+        "epoch": 0.1626327370132976,
+        "grad_norm": 0.5880619287490845,
+        "learning_rate": 0.0005,
+        "loss": 1.3924,
+        "step": 5100
+      },
+      {
+        "epoch": 0.16295162473293154,
+        "grad_norm": 0.6160964965820312,
+        "learning_rate": 0.0005,
+        "loss": 1.3861,
+        "step": 5110
+      },
+      {
+        "epoch": 0.16327051245256546,
+        "grad_norm": 0.6288329362869263,
+        "learning_rate": 0.0005,
+        "loss": 1.3661,
+        "step": 5120
+      },
+      {
+        "epoch": 0.16358940017219936,
+        "grad_norm": 0.5810098052024841,
+        "learning_rate": 0.0005,
+        "loss": 1.3856,
+        "step": 5130
+      },
+      {
+        "epoch": 0.1639082878918333,
+        "grad_norm": 0.6061198115348816,
+        "learning_rate": 0.0005,
+        "loss": 1.3792,
+        "step": 5140
+      },
+      {
+        "epoch": 0.1642271756114672,
+        "grad_norm": 0.5987685918807983,
+        "learning_rate": 0.0005,
+        "loss": 1.3813,
+        "step": 5150
+      },
+      {
+        "epoch": 0.1645460633311011,
+        "grad_norm": 0.6192308068275452,
+        "learning_rate": 0.0005,
+        "loss": 1.3805,
+        "step": 5160
+      },
+      {
+        "epoch": 0.16486495105073504,
+        "grad_norm": 0.6115210056304932,
+        "learning_rate": 0.0005,
+        "loss": 1.3873,
+        "step": 5170
+      },
+      {
+        "epoch": 0.16518383877036896,
+        "grad_norm": 0.5993800759315491,
+        "learning_rate": 0.0005,
+        "loss": 1.3931,
+        "step": 5180
+      },
+      {
+        "epoch": 0.16550272649000286,
+        "grad_norm": 0.5932030081748962,
+        "learning_rate": 0.0005,
+        "loss": 1.4059,
+        "step": 5190
+      },
+      {
+        "epoch": 0.1658216142096368,
+        "grad_norm": 0.5708052515983582,
+        "learning_rate": 0.0005,
+        "loss": 1.361,
+        "step": 5200
+      },
+      {
+        "epoch": 0.1661405019292707,
+        "grad_norm": 0.5772805213928223,
+        "learning_rate": 0.0005,
+        "loss": 1.3641,
+        "step": 5210
+      },
+      {
+        "epoch": 0.1664593896489046,
+        "grad_norm": 0.5902208685874939,
+        "learning_rate": 0.0005,
+        "loss": 1.3697,
+        "step": 5220
+      },
+      {
+        "epoch": 0.16677827736853854,
+        "grad_norm": 0.6566534042358398,
+        "learning_rate": 0.0005,
+        "loss": 1.3673,
+        "step": 5230
+      },
+      {
+        "epoch": 0.16709716508817246,
+        "grad_norm": 0.5957956314086914,
+        "learning_rate": 0.0005,
+        "loss": 1.3819,
+        "step": 5240
+      },
+      {
+        "epoch": 0.16741605280780636,
+        "grad_norm": 0.5835301280021667,
+        "learning_rate": 0.0005,
+        "loss": 1.3761,
+        "step": 5250
+      },
+      {
+        "epoch": 0.16773494052744028,
+        "grad_norm": 0.6204524636268616,
+        "learning_rate": 0.0005,
+        "loss": 1.3636,
+        "step": 5260
+      },
+      {
+        "epoch": 0.1680538282470742,
+        "grad_norm": 0.6161551475524902,
+        "learning_rate": 0.0005,
+        "loss": 1.3584,
+        "step": 5270
+      },
+      {
+        "epoch": 0.1683727159667081,
+        "grad_norm": 0.5636275410652161,
+        "learning_rate": 0.0005,
+        "loss": 1.3727,
+        "step": 5280
+      },
+      {
+        "epoch": 0.16869160368634203,
+        "grad_norm": 0.5892273187637329,
+        "learning_rate": 0.0005,
+        "loss": 1.3688,
+        "step": 5290
+      },
+      {
+        "epoch": 0.16901049140597596,
+        "grad_norm": 0.5970210433006287,
+        "learning_rate": 0.0005,
+        "loss": 1.3697,
+        "step": 5300
+      },
+      {
+        "epoch": 0.16932937912560989,
+        "grad_norm": 0.5779142379760742,
+        "learning_rate": 0.0005,
+        "loss": 1.3734,
+        "step": 5310
+      },
+      {
+        "epoch": 0.16964826684524378,
+        "grad_norm": 0.6144216060638428,
+        "learning_rate": 0.0005,
+        "loss": 1.3671,
+        "step": 5320
+      },
+      {
+        "epoch": 0.1699671545648777,
+        "grad_norm": 0.5949397683143616,
+        "learning_rate": 0.0005,
+        "loss": 1.383,
+        "step": 5330
+      },
+      {
+        "epoch": 0.17028604228451164,
+        "grad_norm": 0.6158933043479919,
+        "learning_rate": 0.0005,
+        "loss": 1.385,
+        "step": 5340
+      },
+      {
+        "epoch": 0.17060493000414553,
+        "grad_norm": 0.6164098381996155,
+        "learning_rate": 0.0005,
+        "loss": 1.3707,
+        "step": 5350
+      },
+      {
+        "epoch": 0.17092381772377946,
+        "grad_norm": 0.6229051351547241,
+        "learning_rate": 0.0005,
+        "loss": 1.3695,
+        "step": 5360
+      },
+      {
+        "epoch": 0.17124270544341338,
+        "grad_norm": 0.5980965495109558,
+        "learning_rate": 0.0005,
+        "loss": 1.3562,
+        "step": 5370
+      },
+      {
+        "epoch": 0.17156159316304728,
+        "grad_norm": 0.5617382526397705,
+        "learning_rate": 0.0005,
+        "loss": 1.3436,
+        "step": 5380
+      },
+      {
+        "epoch": 0.1718804808826812,
+        "grad_norm": 0.5663665533065796,
+        "learning_rate": 0.0005,
+        "loss": 1.3608,
+        "step": 5390
+      },
+      {
+        "epoch": 0.17219936860231513,
+        "grad_norm": 0.5852437019348145,
+        "learning_rate": 0.0005,
+        "loss": 1.3885,
+        "step": 5400
+      },
+      {
+        "epoch": 0.17251825632194903,
+        "grad_norm": 0.5675665140151978,
+        "learning_rate": 0.0005,
+        "loss": 1.3581,
+        "step": 5410
+      },
+      {
+        "epoch": 0.17283714404158296,
+        "grad_norm": 0.5745238065719604,
+        "learning_rate": 0.0005,
+        "loss": 1.3736,
+        "step": 5420
+      },
+      {
+        "epoch": 0.17315603176121688,
+        "grad_norm": 0.5912173390388489,
+        "learning_rate": 0.0005,
+        "loss": 1.3652,
+        "step": 5430
+      },
+      {
+        "epoch": 0.17347491948085078,
+        "grad_norm": 0.5719927549362183,
+        "learning_rate": 0.0005,
+        "loss": 1.3722,
+        "step": 5440
+      },
+      {
+        "epoch": 0.1737938072004847,
+        "grad_norm": 0.5705735683441162,
+        "learning_rate": 0.0005,
+        "loss": 1.3752,
+        "step": 5450
+      },
+      {
+        "epoch": 0.17411269492011863,
+        "grad_norm": 0.6265610456466675,
+        "learning_rate": 0.0005,
+        "loss": 1.3506,
+        "step": 5460
+      },
+      {
+        "epoch": 0.17443158263975253,
+        "grad_norm": 0.5637654662132263,
+        "learning_rate": 0.0005,
+        "loss": 1.3734,
+        "step": 5470
+      },
+      {
+        "epoch": 0.17475047035938646,
+        "grad_norm": 0.5880090594291687,
+        "learning_rate": 0.0005,
+        "loss": 1.3779,
+        "step": 5480
+      },
+      {
+        "epoch": 0.17506935807902038,
+        "grad_norm": 0.5817082524299622,
+        "learning_rate": 0.0005,
+        "loss": 1.3343,
+        "step": 5490
+      },
+      {
+        "epoch": 0.1753882457986543,
+        "grad_norm": 0.5926219820976257,
+        "learning_rate": 0.0005,
+        "loss": 1.3365,
+        "step": 5500
+      },
+      {
+        "epoch": 0.1757071335182882,
+        "grad_norm": 0.5910172462463379,
+        "learning_rate": 0.0005,
+        "loss": 1.3721,
+        "step": 5510
+      },
+      {
+        "epoch": 0.17602602123792213,
+        "grad_norm": 0.5813695192337036,
+        "learning_rate": 0.0005,
+        "loss": 1.3704,
+        "step": 5520
+      },
+      {
+        "epoch": 0.17634490895755606,
+        "grad_norm": 0.5902416110038757,
+        "learning_rate": 0.0005,
+        "loss": 1.3513,
+        "step": 5530
+      },
+      {
+        "epoch": 0.17666379667718995,
+        "grad_norm": 0.6056077480316162,
+        "learning_rate": 0.0005,
+        "loss": 1.3543,
+        "step": 5540
+      },
+      {
+        "epoch": 0.17698268439682388,
+        "grad_norm": 0.6056801080703735,
+        "learning_rate": 0.0005,
+        "loss": 1.3714,
+        "step": 5550
+      },
+      {
+        "epoch": 0.1773015721164578,
+        "grad_norm": 0.5975509285926819,
+        "learning_rate": 0.0005,
+        "loss": 1.3634,
+        "step": 5560
+      },
+      {
+        "epoch": 0.1776204598360917,
+        "grad_norm": 0.5552816390991211,
+        "learning_rate": 0.0005,
+        "loss": 1.3502,
+        "step": 5570
+      },
+      {
+        "epoch": 0.17793934755572563,
+        "grad_norm": 0.5838658809661865,
+        "learning_rate": 0.0005,
+        "loss": 1.3397,
+        "step": 5580
+      },
+      {
+        "epoch": 0.17825823527535956,
+        "grad_norm": 0.5876995921134949,
+        "learning_rate": 0.0005,
+        "loss": 1.3776,
+        "step": 5590
+      },
+      {
+        "epoch": 0.17857712299499345,
+        "grad_norm": 0.5738474130630493,
+        "learning_rate": 0.0005,
+        "loss": 1.3667,
+        "step": 5600
+      },
+      {
+        "epoch": 0.17889601071462738,
+        "grad_norm": 0.5683244466781616,
+        "learning_rate": 0.0005,
+        "loss": 1.3628,
+        "step": 5610
+      },
+      {
+        "epoch": 0.1792148984342613,
+        "grad_norm": 0.5847506523132324,
+        "learning_rate": 0.0005,
+        "loss": 1.3759,
+        "step": 5620
+      },
+      {
+        "epoch": 0.1795337861538952,
+        "grad_norm": 0.5738845467567444,
+        "learning_rate": 0.0005,
+        "loss": 1.3629,
+        "step": 5630
+      },
+      {
+        "epoch": 0.17985267387352913,
+        "grad_norm": 0.5815554261207581,
+        "learning_rate": 0.0005,
+        "loss": 1.3511,
+        "step": 5640
+      },
+      {
+        "epoch": 0.18017156159316305,
+        "grad_norm": 0.5679082870483398,
+        "learning_rate": 0.0005,
+        "loss": 1.37,
+        "step": 5650
+      },
+      {
+        "epoch": 0.18049044931279695,
+        "grad_norm": 0.5690569877624512,
+        "learning_rate": 0.0005,
+        "loss": 1.3632,
+        "step": 5660
+      },
+      {
+        "epoch": 0.18080933703243088,
+        "grad_norm": 0.5927101969718933,
+        "learning_rate": 0.0005,
+        "loss": 1.362,
+        "step": 5670
+      },
+      {
+        "epoch": 0.1811282247520648,
+        "grad_norm": 0.5851690769195557,
+        "learning_rate": 0.0005,
+        "loss": 1.3414,
+        "step": 5680
+      },
+      {
+        "epoch": 0.1814471124716987,
+        "grad_norm": 0.5578193664550781,
+        "learning_rate": 0.0005,
+        "loss": 1.3451,
+        "step": 5690
+      },
+      {
+        "epoch": 0.18176600019133263,
+        "grad_norm": 0.5566641688346863,
+        "learning_rate": 0.0005,
+        "loss": 1.3481,
+        "step": 5700
+      },
+      {
+        "epoch": 0.18208488791096655,
+        "grad_norm": 0.5803577899932861,
+        "learning_rate": 0.0005,
+        "loss": 1.3584,
+        "step": 5710
+      },
+      {
+        "epoch": 0.18240377563060048,
+        "grad_norm": 0.574792206287384,
+        "learning_rate": 0.0005,
+        "loss": 1.3742,
+        "step": 5720
+      },
+      {
+        "epoch": 0.18272266335023438,
+        "grad_norm": 0.5856817960739136,
+        "learning_rate": 0.0005,
+        "loss": 1.337,
+        "step": 5730
+      },
+      {
+        "epoch": 0.1830415510698683,
+        "grad_norm": 0.6036618947982788,
+        "learning_rate": 0.0005,
+        "loss": 1.3275,
+        "step": 5740
+      },
+      {
+        "epoch": 0.18336043878950223,
+        "grad_norm": 0.5599528551101685,
+        "learning_rate": 0.0005,
+        "loss": 1.3694,
+        "step": 5750
+      },
+      {
+        "epoch": 0.18367932650913613,
+        "grad_norm": 0.573541522026062,
+        "learning_rate": 0.0005,
+        "loss": 1.3507,
+        "step": 5760
+      },
+      {
+        "epoch": 0.18399821422877005,
+        "grad_norm": 0.617546558380127,
+        "learning_rate": 0.0005,
+        "loss": 1.3508,
+        "step": 5770
+      },
+      {
+        "epoch": 0.18431710194840398,
+        "grad_norm": 0.5568141341209412,
+        "learning_rate": 0.0005,
+        "loss": 1.3327,
+        "step": 5780
+      },
+      {
+        "epoch": 0.18463598966803788,
+        "grad_norm": 0.548579216003418,
+        "learning_rate": 0.0005,
+        "loss": 1.3349,
+        "step": 5790
+      },
+      {
+        "epoch": 0.1849548773876718,
+        "grad_norm": 0.5528845191001892,
+        "learning_rate": 0.0005,
+        "loss": 1.3442,
+        "step": 5800
+      },
+      {
+        "epoch": 0.18527376510730573,
+        "grad_norm": 0.5503314137458801,
+        "learning_rate": 0.0005,
+        "loss": 1.3472,
+        "step": 5810
+      },
+      {
+        "epoch": 0.18559265282693962,
+        "grad_norm": 0.5958365797996521,
+        "learning_rate": 0.0005,
+        "loss": 1.3391,
+        "step": 5820
+      },
+      {
+        "epoch": 0.18591154054657355,
+        "grad_norm": 0.5654858350753784,
+        "learning_rate": 0.0005,
+        "loss": 1.3448,
+        "step": 5830
+      },
+      {
+        "epoch": 0.18623042826620748,
+        "grad_norm": 0.6166044473648071,
+        "learning_rate": 0.0005,
+        "loss": 1.3442,
+        "step": 5840
+      },
+      {
+        "epoch": 0.18654931598584137,
+        "grad_norm": 0.5609534382820129,
+        "learning_rate": 0.0005,
+        "loss": 1.3509,
+        "step": 5850
+      },
+      {
+        "epoch": 0.1868682037054753,
+        "grad_norm": 0.5806050300598145,
+        "learning_rate": 0.0005,
+        "loss": 1.3431,
+        "step": 5860
+      },
+      {
+        "epoch": 0.18718709142510923,
+        "grad_norm": 0.6101890206336975,
+        "learning_rate": 0.0005,
+        "loss": 1.3339,
+        "step": 5870
+      },
+      {
+        "epoch": 0.18750597914474312,
+        "grad_norm": 0.5922510027885437,
+        "learning_rate": 0.0005,
+        "loss": 1.3578,
+        "step": 5880
+      },
+      {
+        "epoch": 0.18782486686437705,
+        "grad_norm": 0.5860453248023987,
+        "learning_rate": 0.0005,
+        "loss": 1.3468,
+        "step": 5890
+      },
+      {
+        "epoch": 0.18814375458401097,
+        "grad_norm": 0.5527572631835938,
+        "learning_rate": 0.0005,
+        "loss": 1.3397,
+        "step": 5900
+      },
+      {
+        "epoch": 0.18846264230364487,
+        "grad_norm": 0.5668513178825378,
+        "learning_rate": 0.0005,
+        "loss": 1.3356,
+        "step": 5910
+      },
+      {
+        "epoch": 0.1887815300232788,
+        "grad_norm": 0.536500871181488,
+        "learning_rate": 0.0005,
+        "loss": 1.3372,
+        "step": 5920
+      },
+      {
+        "epoch": 0.18910041774291272,
+        "grad_norm": 0.5639796853065491,
+        "learning_rate": 0.0005,
+        "loss": 1.3358,
+        "step": 5930
+      },
+      {
+        "epoch": 0.18941930546254665,
+        "grad_norm": 0.5703504085540771,
+        "learning_rate": 0.0005,
+        "loss": 1.349,
+        "step": 5940
+      },
+      {
+        "epoch": 0.18973819318218055,
+        "grad_norm": 0.573939859867096,
+        "learning_rate": 0.0005,
+        "loss": 1.3361,
+        "step": 5950
+      },
+      {
+        "epoch": 0.19005708090181447,
+        "grad_norm": 0.5942139029502869,
+        "learning_rate": 0.0005,
+        "loss": 1.3656,
+        "step": 5960
+      },
+      {
+        "epoch": 0.1903759686214484,
+        "grad_norm": 0.5792593359947205,
+        "learning_rate": 0.0005,
+        "loss": 1.3463,
+        "step": 5970
+      },
+      {
+        "epoch": 0.1906948563410823,
+        "grad_norm": 0.5661564469337463,
+        "learning_rate": 0.0005,
+        "loss": 1.3356,
+        "step": 5980
+      },
+      {
+        "epoch": 0.19101374406071622,
+        "grad_norm": 0.6026549339294434,
+        "learning_rate": 0.0005,
+        "loss": 1.3432,
+        "step": 5990
+      },
+      {
+        "epoch": 0.19133263178035015,
+        "grad_norm": 0.5632230043411255,
+        "learning_rate": 0.0005,
+        "loss": 1.3223,
+        "step": 6000
+      },
+      {
+        "epoch": 0.19165151949998405,
+        "grad_norm": 0.5511482357978821,
+        "learning_rate": 0.0005,
+        "loss": 1.3489,
+        "step": 6010
+      },
+      {
+        "epoch": 0.19197040721961797,
+        "grad_norm": 0.5755233764648438,
+        "learning_rate": 0.0005,
+        "loss": 1.3262,
+        "step": 6020
+      },
+      {
+        "epoch": 0.1922892949392519,
+        "grad_norm": 0.5578722953796387,
+        "learning_rate": 0.0005,
+        "loss": 1.3175,
+        "step": 6030
+      },
+      {
+        "epoch": 0.1926081826588858,
+        "grad_norm": 0.5360584259033203,
+        "learning_rate": 0.0005,
+        "loss": 1.3301,
+        "step": 6040
+      },
+      {
+        "epoch": 0.19292707037851972,
+        "grad_norm": 0.5476540923118591,
+        "learning_rate": 0.0005,
+        "loss": 1.3303,
+        "step": 6050
+      },
+      {
+        "epoch": 0.19324595809815365,
+        "grad_norm": 0.5518861413002014,
+        "learning_rate": 0.0005,
+        "loss": 1.3216,
+        "step": 6060
+      },
+      {
+        "epoch": 0.19356484581778755,
+        "grad_norm": 0.5956515669822693,
+        "learning_rate": 0.0005,
+        "loss": 1.3313,
+        "step": 6070
+      },
+      {
+        "epoch": 0.19388373353742147,
+        "grad_norm": 0.5531336665153503,
+        "learning_rate": 0.0005,
+        "loss": 1.3255,
+        "step": 6080
+      },
+      {
+        "epoch": 0.1942026212570554,
+        "grad_norm": 0.5420064330101013,
+        "learning_rate": 0.0005,
+        "loss": 1.3311,
+        "step": 6090
+      },
+      {
+        "epoch": 0.1945215089766893,
+        "grad_norm": 0.539386510848999,
+        "learning_rate": 0.0005,
+        "loss": 1.3359,
+        "step": 6100
+      },
+      {
+        "epoch": 0.19484039669632322,
+        "grad_norm": 0.5622890591621399,
+        "learning_rate": 0.0005,
+        "loss": 1.3406,
+        "step": 6110
+      },
+      {
+        "epoch": 0.19515928441595715,
+        "grad_norm": 0.6167592406272888,
+        "learning_rate": 0.0005,
+        "loss": 1.3323,
+        "step": 6120
+      },
+      {
+        "epoch": 0.19547817213559107,
+        "grad_norm": 0.5873464345932007,
+        "learning_rate": 0.0005,
+        "loss": 1.3123,
+        "step": 6130
+      },
+      {
+        "epoch": 0.19579705985522497,
+        "grad_norm": 0.5778335928916931,
+        "learning_rate": 0.0005,
+        "loss": 1.3387,
+        "step": 6140
+      },
+      {
+        "epoch": 0.1961159475748589,
+        "grad_norm": 0.5280761122703552,
+        "learning_rate": 0.0005,
+        "loss": 1.3042,
+        "step": 6150
+      },
+      {
+        "epoch": 0.19643483529449282,
+        "grad_norm": 0.5577932000160217,
+        "learning_rate": 0.0005,
+        "loss": 1.3082,
+        "step": 6160
+      },
+      {
+        "epoch": 0.19675372301412672,
+        "grad_norm": 0.554754376411438,
+        "learning_rate": 0.0005,
+        "loss": 1.3451,
+        "step": 6170
+      },
+      {
+        "epoch": 0.19707261073376064,
+        "grad_norm": 0.5626440048217773,
+        "learning_rate": 0.0005,
+        "loss": 1.3236,
+        "step": 6180
+      },
+      {
+        "epoch": 0.19739149845339457,
+        "grad_norm": 0.5483393669128418,
+        "learning_rate": 0.0005,
+        "loss": 1.3454,
+        "step": 6190
+      },
+      {
+        "epoch": 0.19771038617302847,
+        "grad_norm": 0.5727918148040771,
+        "learning_rate": 0.0005,
+        "loss": 1.3288,
+        "step": 6200
+      },
+      {
+        "epoch": 0.1980292738926624,
+        "grad_norm": 0.5533772706985474,
+        "learning_rate": 0.0005,
+        "loss": 1.3344,
+        "step": 6210
+      },
+      {
+        "epoch": 0.19834816161229632,
+        "grad_norm": 0.5411686301231384,
+        "learning_rate": 0.0005,
+        "loss": 1.3095,
+        "step": 6220
+      },
+      {
+        "epoch": 0.19866704933193022,
+        "grad_norm": 0.5486521124839783,
+        "learning_rate": 0.0005,
+        "loss": 1.3103,
+        "step": 6230
+      },
+      {
+        "epoch": 0.19898593705156414,
+        "grad_norm": 0.5434985160827637,
+        "learning_rate": 0.0005,
+        "loss": 1.3179,
+        "step": 6240
+      },
+      {
+        "epoch": 0.19930482477119807,
+        "grad_norm": 0.5697535872459412,
+        "learning_rate": 0.0005,
+        "loss": 1.3321,
+        "step": 6250
+      },
+      {
+        "epoch": 0.19962371249083197,
+        "grad_norm": 0.5627325177192688,
+        "learning_rate": 0.0005,
+        "loss": 1.311,
+        "step": 6260
+      },
+      {
+        "epoch": 0.1999426002104659,
+        "grad_norm": 0.5584254264831543,
+        "learning_rate": 0.0005,
+        "loss": 1.326,
+        "step": 6270
+      },
+      {
+        "epoch": 0.20026148793009982,
+        "grad_norm": 0.5627668499946594,
+        "learning_rate": 0.0005,
+        "loss": 1.3252,
+        "step": 6280
+      },
+      {
+        "epoch": 0.20058037564973372,
+        "grad_norm": 0.5576115250587463,
+        "learning_rate": 0.0005,
+        "loss": 1.3356,
+        "step": 6290
+      },
+      {
+        "epoch": 0.20089926336936764,
+        "grad_norm": 0.5420365333557129,
+        "learning_rate": 0.0005,
+        "loss": 1.3271,
+        "step": 6300
+      },
+      {
+        "epoch": 0.20121815108900157,
+        "grad_norm": 0.5424570441246033,
+        "learning_rate": 0.0005,
+        "loss": 1.322,
+        "step": 6310
+      },
+      {
+        "epoch": 0.20153703880863547,
+        "grad_norm": 0.5431638360023499,
+        "learning_rate": 0.0005,
+        "loss": 1.3184,
+        "step": 6320
+      },
+      {
+        "epoch": 0.2018559265282694,
+        "grad_norm": 0.5798773765563965,
+        "learning_rate": 0.0005,
+        "loss": 1.3165,
+        "step": 6330
+      },
+      {
+        "epoch": 0.20217481424790332,
+        "grad_norm": 0.5577746033668518,
+        "learning_rate": 0.0005,
+        "loss": 1.3209,
+        "step": 6340
+      },
+      {
+        "epoch": 0.20249370196753724,
+        "grad_norm": 0.5490031242370605,
+        "learning_rate": 0.0005,
+        "loss": 1.3127,
+        "step": 6350
+      },
+      {
+        "epoch": 0.20281258968717114,
+        "grad_norm": 0.5594183802604675,
+        "learning_rate": 0.0005,
+        "loss": 1.3175,
+        "step": 6360
+      },
+      {
+        "epoch": 0.20313147740680507,
+        "grad_norm": 0.5752307772636414,
+        "learning_rate": 0.0005,
+        "loss": 1.3382,
+        "step": 6370
+      },
+      {
+        "epoch": 0.203450365126439,
+        "grad_norm": 0.5838162302970886,
+        "learning_rate": 0.0005,
+        "loss": 1.3274,
+        "step": 6380
+      },
+      {
+        "epoch": 0.2037692528460729,
+        "grad_norm": 0.5440608859062195,
+        "learning_rate": 0.0005,
+        "loss": 1.314,
+        "step": 6390
+      },
+      {
+        "epoch": 0.20408814056570682,
+        "grad_norm": 0.5637199282646179,
+        "learning_rate": 0.0005,
+        "loss": 1.301,
+        "step": 6400
+      },
+      {
+        "epoch": 0.20440702828534074,
+        "grad_norm": 0.583897054195404,
+        "learning_rate": 0.0005,
+        "loss": 1.331,
+        "step": 6410
+      },
+      {
+        "epoch": 0.20472591600497464,
+        "grad_norm": 0.5349193215370178,
+        "learning_rate": 0.0005,
+        "loss": 1.3085,
+        "step": 6420
+      },
+      {
+        "epoch": 0.20504480372460857,
+        "grad_norm": 0.572115957736969,
+        "learning_rate": 0.0005,
+        "loss": 1.3353,
+        "step": 6430
+      },
+      {
+        "epoch": 0.2053636914442425,
+        "grad_norm": 0.5371711254119873,
+        "learning_rate": 0.0005,
+        "loss": 1.3115,
+        "step": 6440
+      },
+      {
+        "epoch": 0.2056825791638764,
+        "grad_norm": 0.5468522310256958,
+        "learning_rate": 0.0005,
+        "loss": 1.3058,
+        "step": 6450
+      },
+      {
+        "epoch": 0.20600146688351031,
+        "grad_norm": 0.513854444026947,
+        "learning_rate": 0.0005,
+        "loss": 1.3067,
+        "step": 6460
+      },
+      {
+        "epoch": 0.20632035460314424,
+        "grad_norm": 0.5438308715820312,
+        "learning_rate": 0.0005,
+        "loss": 1.3245,
+        "step": 6470
+      },
+      {
+        "epoch": 0.20663924232277814,
+        "grad_norm": 0.5419757962226868,
+        "learning_rate": 0.0005,
+        "loss": 1.317,
+        "step": 6480
+      },
+      {
+        "epoch": 0.20695813004241206,
+        "grad_norm": 0.5453127026557922,
+        "learning_rate": 0.0005,
+        "loss": 1.3188,
+        "step": 6490
+      },
+      {
+        "epoch": 0.207277017762046,
+        "grad_norm": 0.5806515216827393,
+        "learning_rate": 0.0005,
+        "loss": 1.3101,
+        "step": 6500
+      },
+      {
+        "epoch": 0.2075959054816799,
+        "grad_norm": 0.5650323033332825,
+        "learning_rate": 0.0005,
+        "loss": 1.31,
+        "step": 6510
+      },
+      {
+        "epoch": 0.2079147932013138,
+        "grad_norm": 0.5299544334411621,
+        "learning_rate": 0.0005,
+        "loss": 1.3122,
+        "step": 6520
+      },
+      {
+        "epoch": 0.20823368092094774,
+        "grad_norm": 0.5299350619316101,
+        "learning_rate": 0.0005,
+        "loss": 1.3182,
+        "step": 6530
+      },
+      {
+        "epoch": 0.20855256864058166,
+        "grad_norm": 0.544809877872467,
+        "learning_rate": 0.0005,
+        "loss": 1.3349,
+        "step": 6540
+      },
+      {
+        "epoch": 0.20887145636021556,
+        "grad_norm": 0.5426610112190247,
+        "learning_rate": 0.0005,
+        "loss": 1.3163,
+        "step": 6550
+      },
+      {
+        "epoch": 0.2091903440798495,
+        "grad_norm": 0.5313812494277954,
+        "learning_rate": 0.0005,
+        "loss": 1.3082,
+        "step": 6560
+      },
+      {
+        "epoch": 0.20950923179948341,
+        "grad_norm": 0.5384641289710999,
+        "learning_rate": 0.0005,
+        "loss": 1.3154,
+        "step": 6570
+      },
+      {
+        "epoch": 0.2098281195191173,
+        "grad_norm": 0.5682090520858765,
+        "learning_rate": 0.0005,
+        "loss": 1.3104,
+        "step": 6580
+      },
+      {
+        "epoch": 0.21014700723875124,
+        "grad_norm": 0.5284824371337891,
+        "learning_rate": 0.0005,
+        "loss": 1.3181,
+        "step": 6590
+      },
+      {
+        "epoch": 0.21046589495838516,
+        "grad_norm": 0.546111524105072,
+        "learning_rate": 0.0005,
+        "loss": 1.3002,
+        "step": 6600
+      },
+      {
+        "epoch": 0.21078478267801906,
+        "grad_norm": 0.5103371143341064,
+        "learning_rate": 0.0005,
+        "loss": 1.3159,
+        "step": 6610
+      },
+      {
+        "epoch": 0.211103670397653,
+        "grad_norm": 0.5273628830909729,
+        "learning_rate": 0.0005,
+        "loss": 1.3016,
+        "step": 6620
+      },
+      {
+        "epoch": 0.2114225581172869,
+        "grad_norm": 0.5440049767494202,
+        "learning_rate": 0.0005,
+        "loss": 1.3133,
+        "step": 6630
+      },
+      {
+        "epoch": 0.2117414458369208,
+        "grad_norm": 0.5308232307434082,
+        "learning_rate": 0.0005,
+        "loss": 1.3247,
+        "step": 6640
+      },
+      {
+        "epoch": 0.21206033355655474,
+        "grad_norm": 0.5194700360298157,
+        "learning_rate": 0.0005,
+        "loss": 1.2996,
+        "step": 6650
+      },
+      {
+        "epoch": 0.21237922127618866,
+        "grad_norm": 0.5374641418457031,
+        "learning_rate": 0.0005,
+        "loss": 1.312,
+        "step": 6660
+      },
+      {
+        "epoch": 0.21269810899582256,
+        "grad_norm": 0.5642796754837036,
+        "learning_rate": 0.0005,
+        "loss": 1.3089,
+        "step": 6670
+      },
+      {
+        "epoch": 0.21301699671545649,
+        "grad_norm": 0.5580105781555176,
+        "learning_rate": 0.0005,
+        "loss": 1.3046,
+        "step": 6680
+      },
+      {
+        "epoch": 0.2133358844350904,
+        "grad_norm": 0.5559312105178833,
+        "learning_rate": 0.0005,
+        "loss": 1.3164,
+        "step": 6690
+      },
+      {
+        "epoch": 0.2136547721547243,
+        "grad_norm": 0.5587300062179565,
+        "learning_rate": 0.0005,
+        "loss": 1.3083,
+        "step": 6700
+      },
+      {
+        "epoch": 0.21397365987435824,
+        "grad_norm": 0.5269460678100586,
+        "learning_rate": 0.0005,
+        "loss": 1.2957,
+        "step": 6710
+      },
+      {
+        "epoch": 0.21429254759399216,
+        "grad_norm": 0.5704488158226013,
+        "learning_rate": 0.0005,
+        "loss": 1.3109,
+        "step": 6720
+      },
+      {
+        "epoch": 0.21461143531362606,
+        "grad_norm": 0.5360540747642517,
+        "learning_rate": 0.0005,
+        "loss": 1.3009,
+        "step": 6730
+      },
+      {
+        "epoch": 0.21493032303325998,
+        "grad_norm": 0.5284568071365356,
+        "learning_rate": 0.0005,
+        "loss": 1.3047,
+        "step": 6740
+      },
+      {
+        "epoch": 0.2152492107528939,
+        "grad_norm": 0.5489094257354736,
+        "learning_rate": 0.0005,
+        "loss": 1.3223,
+        "step": 6750
+      },
+      {
+        "epoch": 0.21556809847252784,
+        "grad_norm": 0.524774968624115,
+        "learning_rate": 0.0005,
+        "loss": 1.3144,
+        "step": 6760
+      },
+      {
+        "epoch": 0.21588698619216173,
+        "grad_norm": 0.5359005928039551,
+        "learning_rate": 0.0005,
+        "loss": 1.3012,
+        "step": 6770
+      },
+      {
+        "epoch": 0.21620587391179566,
+        "grad_norm": 0.5301958322525024,
+        "learning_rate": 0.0005,
+        "loss": 1.2909,
+        "step": 6780
+      },
+      {
+        "epoch": 0.21652476163142959,
+        "grad_norm": 0.5353237390518188,
+        "learning_rate": 0.0005,
+        "loss": 1.2899,
+        "step": 6790
+      },
+      {
+        "epoch": 0.21684364935106348,
+        "grad_norm": 0.551356852054596,
+        "learning_rate": 0.0005,
+        "loss": 1.2839,
+        "step": 6800
+      },
+      {
+        "epoch": 0.2171625370706974,
+        "grad_norm": 0.5137448906898499,
+        "learning_rate": 0.0005,
+        "loss": 1.3146,
+        "step": 6810
+      },
+      {
+        "epoch": 0.21748142479033133,
+        "grad_norm": 0.5484623312950134,
+        "learning_rate": 0.0005,
+        "loss": 1.3036,
+        "step": 6820
+      },
+      {
+        "epoch": 0.21780031250996523,
+        "grad_norm": 0.5172665119171143,
+        "learning_rate": 0.0005,
+        "loss": 1.3037,
+        "step": 6830
+      },
+      {
+        "epoch": 0.21811920022959916,
+        "grad_norm": 0.5307798981666565,
+        "learning_rate": 0.0005,
+        "loss": 1.3003,
+        "step": 6840
+      },
+      {
+        "epoch": 0.21843808794923308,
+        "grad_norm": 0.5149516463279724,
+        "learning_rate": 0.0005,
+        "loss": 1.2833,
+        "step": 6850
+      },
+      {
+        "epoch": 0.21875697566886698,
+        "grad_norm": 0.5275344252586365,
+        "learning_rate": 0.0005,
+        "loss": 1.2849,
+        "step": 6860
+      },
+      {
+        "epoch": 0.2190758633885009,
+        "grad_norm": 0.5206189155578613,
+        "learning_rate": 0.0005,
+        "loss": 1.3039,
+        "step": 6870
+      },
+      {
+        "epoch": 0.21939475110813483,
+        "grad_norm": 0.5311983823776245,
+        "learning_rate": 0.0005,
+        "loss": 1.3034,
+        "step": 6880
+      },
+      {
+        "epoch": 0.21971363882776873,
+        "grad_norm": 0.5286182165145874,
+        "learning_rate": 0.0005,
+        "loss": 1.2961,
+        "step": 6890
+      },
+      {
+        "epoch": 0.22003252654740266,
+        "grad_norm": 0.6045443415641785,
+        "learning_rate": 0.0005,
+        "loss": 1.3086,
+        "step": 6900
+      },
+      {
+        "epoch": 0.22035141426703658,
+        "grad_norm": 0.5496406555175781,
+        "learning_rate": 0.0005,
+        "loss": 1.3021,
+        "step": 6910
+      },
+      {
+        "epoch": 0.22067030198667048,
+        "grad_norm": 0.5401288270950317,
+        "learning_rate": 0.0005,
+        "loss": 1.3183,
+        "step": 6920
+      },
+      {
+        "epoch": 0.2209891897063044,
+        "grad_norm": 0.5402335524559021,
+        "learning_rate": 0.0005,
+        "loss": 1.3063,
+        "step": 6930
+      },
+      {
+        "epoch": 0.22130807742593833,
+        "grad_norm": 0.5452973246574402,
+        "learning_rate": 0.0005,
+        "loss": 1.2922,
+        "step": 6940
+      },
+      {
+        "epoch": 0.22162696514557226,
+        "grad_norm": 0.5425865650177002,
+        "learning_rate": 0.0005,
+        "loss": 1.2902,
+        "step": 6950
+      },
+      {
+        "epoch": 0.22194585286520616,
+        "grad_norm": 0.5472378134727478,
+        "learning_rate": 0.0005,
+        "loss": 1.2816,
+        "step": 6960
+      },
+      {
+        "epoch": 0.22226474058484008,
+        "grad_norm": 0.5230280160903931,
+        "learning_rate": 0.0005,
+        "loss": 1.2981,
+        "step": 6970
+      },
+      {
+        "epoch": 0.222583628304474,
+        "grad_norm": 0.5512002110481262,
+        "learning_rate": 0.0005,
+        "loss": 1.2964,
+        "step": 6980
+      },
+      {
+        "epoch": 0.2229025160241079,
+        "grad_norm": 0.5232899785041809,
+        "learning_rate": 0.0005,
+        "loss": 1.3066,
+        "step": 6990
+      },
+      {
+        "epoch": 0.22322140374374183,
+        "grad_norm": 0.5482252836227417,
+        "learning_rate": 0.0005,
+        "loss": 1.3177,
+        "step": 7000
+      },
+      {
+        "epoch": 0.22354029146337576,
+        "grad_norm": 0.5405844449996948,
+        "learning_rate": 0.0005,
+        "loss": 1.2851,
+        "step": 7010
+      },
+      {
+        "epoch": 0.22385917918300965,
+        "grad_norm": 0.5282372236251831,
+        "learning_rate": 0.0005,
+        "loss": 1.3024,
+        "step": 7020
+      },
+      {
+        "epoch": 0.22417806690264358,
+        "grad_norm": 0.5448574423789978,
+        "learning_rate": 0.0005,
+        "loss": 1.2734,
+        "step": 7030
+      },
+      {
+        "epoch": 0.2244969546222775,
+        "grad_norm": 0.5466793775558472,
+        "learning_rate": 0.0005,
+        "loss": 1.2765,
+        "step": 7040
+      },
+      {
+        "epoch": 0.2248158423419114,
+        "grad_norm": 0.5259258151054382,
+        "learning_rate": 0.0005,
+        "loss": 1.2955,
+        "step": 7050
+      },
+      {
+        "epoch": 0.22513473006154533,
+        "grad_norm": 0.507414698600769,
+        "learning_rate": 0.0005,
+        "loss": 1.29,
+        "step": 7060
+      },
+      {
+        "epoch": 0.22545361778117926,
+        "grad_norm": 0.5376728177070618,
+        "learning_rate": 0.0005,
+        "loss": 1.2839,
+        "step": 7070
+      },
+      {
+        "epoch": 0.22577250550081315,
+        "grad_norm": 0.5190555453300476,
+        "learning_rate": 0.0005,
+        "loss": 1.3078,
+        "step": 7080
+      },
+      {
+        "epoch": 0.22609139322044708,
+        "grad_norm": 0.5176852941513062,
+        "learning_rate": 0.0005,
+        "loss": 1.3132,
+        "step": 7090
+      },
+      {
+        "epoch": 0.226410280940081,
+        "grad_norm": 0.5273293852806091,
+        "learning_rate": 0.0005,
+        "loss": 1.2938,
+        "step": 7100
+      },
+      {
+        "epoch": 0.2267291686597149,
+        "grad_norm": 0.5152580142021179,
+        "learning_rate": 0.0005,
+        "loss": 1.3058,
+        "step": 7110
+      },
+      {
+        "epoch": 0.22704805637934883,
+        "grad_norm": 0.5310482382774353,
+        "learning_rate": 0.0005,
+        "loss": 1.2821,
+        "step": 7120
+      },
+      {
+        "epoch": 0.22736694409898275,
+        "grad_norm": 0.5562300086021423,
+        "learning_rate": 0.0005,
+        "loss": 1.2986,
+        "step": 7130
+      },
+      {
+        "epoch": 0.22768583181861665,
+        "grad_norm": 0.529692530632019,
+        "learning_rate": 0.0005,
+        "loss": 1.2916,
+        "step": 7140
+      },
+      {
+        "epoch": 0.22800471953825058,
+        "grad_norm": 0.5387162566184998,
+        "learning_rate": 0.0005,
+        "loss": 1.2918,
+        "step": 7150
+      },
+      {
+        "epoch": 0.2283236072578845,
+        "grad_norm": 0.5448839664459229,
+        "learning_rate": 0.0005,
+        "loss": 1.2934,
+        "step": 7160
+      },
+      {
+        "epoch": 0.22864249497751843,
+        "grad_norm": 0.5419529676437378,
+        "learning_rate": 0.0005,
+        "loss": 1.3105,
+        "step": 7170
+      },
+      {
+        "epoch": 0.22896138269715233,
+        "grad_norm": 0.5596801042556763,
+        "learning_rate": 0.0005,
+        "loss": 1.2759,
+        "step": 7180
+      },
+      {
+        "epoch": 0.22928027041678625,
+        "grad_norm": 0.5463248491287231,
+        "learning_rate": 0.0005,
+        "loss": 1.285,
+        "step": 7190
+      },
+      {
+        "epoch": 0.22959915813642018,
+        "grad_norm": 0.5241756439208984,
+        "learning_rate": 0.0005,
+        "loss": 1.2835,
+        "step": 7200
+      },
+      {
+        "epoch": 0.22991804585605408,
+        "grad_norm": 0.5329312682151794,
+        "learning_rate": 0.0005,
+        "loss": 1.2818,
+        "step": 7210
+      },
+      {
+        "epoch": 0.230236933575688,
+        "grad_norm": 0.49857091903686523,
+        "learning_rate": 0.0005,
+        "loss": 1.2621,
+        "step": 7220
+      },
+      {
+        "epoch": 0.23055582129532193,
+        "grad_norm": 0.5254444479942322,
+        "learning_rate": 0.0005,
+        "loss": 1.2955,
+        "step": 7230
+      },
+      {
+        "epoch": 0.23087470901495583,
+        "grad_norm": 0.5368390679359436,
+        "learning_rate": 0.0005,
+        "loss": 1.2988,
+        "step": 7240
+      },
+      {
+        "epoch": 0.23119359673458975,
+        "grad_norm": 0.5087143778800964,
+        "learning_rate": 0.0005,
+        "loss": 1.2879,
+        "step": 7250
+      },
+      {
+        "epoch": 0.23151248445422368,
+        "grad_norm": 0.5278485417366028,
+        "learning_rate": 0.0005,
+        "loss": 1.2773,
+        "step": 7260
+      },
+      {
+        "epoch": 0.23183137217385758,
+        "grad_norm": 0.5035619735717773,
+        "learning_rate": 0.0005,
+        "loss": 1.2755,
+        "step": 7270
+      },
+      {
+        "epoch": 0.2321502598934915,
+        "grad_norm": 0.5509461760520935,
+        "learning_rate": 0.0005,
+        "loss": 1.3007,
+        "step": 7280
+      },
+      {
+        "epoch": 0.23246914761312543,
+        "grad_norm": 0.5227276682853699,
+        "learning_rate": 0.0005,
+        "loss": 1.2881,
+        "step": 7290
+      },
+      {
+        "epoch": 0.23278803533275932,
+        "grad_norm": 0.5183719396591187,
+        "learning_rate": 0.0005,
+        "loss": 1.2799,
+        "step": 7300
+      },
+      {
+        "epoch": 0.23310692305239325,
+        "grad_norm": 0.5770544409751892,
+        "learning_rate": 0.0005,
+        "loss": 1.2969,
+        "step": 7310
+      },
+      {
+        "epoch": 0.23342581077202718,
+        "grad_norm": 0.5247523784637451,
+        "learning_rate": 0.0005,
+        "loss": 1.2864,
+        "step": 7320
+      },
+      {
+        "epoch": 0.23374469849166107,
+        "grad_norm": 0.5339850187301636,
+        "learning_rate": 0.0005,
+        "loss": 1.271,
+        "step": 7330
+      },
+      {
+        "epoch": 0.234063586211295,
+        "grad_norm": 0.513947069644928,
+        "learning_rate": 0.0005,
+        "loss": 1.2904,
+        "step": 7340
+      },
+      {
+        "epoch": 0.23438247393092893,
+        "grad_norm": 0.507355809211731,
+        "learning_rate": 0.0005,
+        "loss": 1.2875,
+        "step": 7350
+      },
+      {
+        "epoch": 0.23470136165056282,
+        "grad_norm": 0.5152395367622375,
+        "learning_rate": 0.0005,
+        "loss": 1.2773,
+        "step": 7360
+      },
+      {
+        "epoch": 0.23502024937019675,
+        "grad_norm": 0.5163211822509766,
+        "learning_rate": 0.0005,
+        "loss": 1.2803,
+        "step": 7370
+      },
+      {
+        "epoch": 0.23533913708983067,
+        "grad_norm": 0.5329576134681702,
+        "learning_rate": 0.0005,
+        "loss": 1.2877,
+        "step": 7380
+      },
+      {
+        "epoch": 0.2356580248094646,
+        "grad_norm": 0.515319287776947,
+        "learning_rate": 0.0005,
+        "loss": 1.2851,
+        "step": 7390
+      },
+      {
+        "epoch": 0.2359769125290985,
+        "grad_norm": 0.5044041872024536,
+        "learning_rate": 0.0005,
+        "loss": 1.2724,
+        "step": 7400
+      },
+      {
+        "epoch": 0.23629580024873242,
+        "grad_norm": 0.5218659043312073,
+        "learning_rate": 0.0005,
+        "loss": 1.2765,
+        "step": 7410
+      },
+      {
+        "epoch": 0.23661468796836635,
+        "grad_norm": 0.5328546166419983,
+        "learning_rate": 0.0005,
+        "loss": 1.3,
+        "step": 7420
+      },
+      {
+        "epoch": 0.23693357568800025,
+        "grad_norm": 0.5284678339958191,
+        "learning_rate": 0.0005,
+        "loss": 1.2623,
+        "step": 7430
+      },
+      {
+        "epoch": 0.23725246340763417,
+        "grad_norm": 0.5512279272079468,
+        "learning_rate": 0.0005,
+        "loss": 1.2746,
+        "step": 7440
+      },
+      {
+        "epoch": 0.2375713511272681,
+        "grad_norm": 0.5096049308776855,
+        "learning_rate": 0.0005,
+        "loss": 1.298,
+        "step": 7450
+      },
+      {
+        "epoch": 0.237890238846902,
+        "grad_norm": 0.5141339898109436,
+        "learning_rate": 0.0005,
+        "loss": 1.2937,
+        "step": 7460
+      },
+      {
+        "epoch": 0.23820912656653592,
+        "grad_norm": 0.5120203495025635,
+        "learning_rate": 0.0005,
+        "loss": 1.2836,
+        "step": 7470
+      },
+      {
+        "epoch": 0.23852801428616985,
+        "grad_norm": 0.5968315005302429,
+        "learning_rate": 0.0005,
+        "loss": 1.3015,
+        "step": 7480
+      },
+      {
+        "epoch": 0.23884690200580375,
+        "grad_norm": 0.49347633123397827,
+        "learning_rate": 0.0005,
+        "loss": 1.2705,
+        "step": 7490
+      },
+      {
+        "epoch": 0.23916578972543767,
+        "grad_norm": 0.50834721326828,
+        "learning_rate": 0.0005,
+        "loss": 1.3048,
+        "step": 7500
+      },
+      {
+        "epoch": 0.2394846774450716,
+        "grad_norm": 0.5340325236320496,
+        "learning_rate": 0.0005,
+        "loss": 1.2792,
+        "step": 7510
+      },
+      {
+        "epoch": 0.2398035651647055,
+        "grad_norm": 0.5159140825271606,
+        "learning_rate": 0.0005,
+        "loss": 1.2863,
+        "step": 7520
+      },
+      {
+        "epoch": 0.24012245288433942,
+        "grad_norm": 0.5006542801856995,
+        "learning_rate": 0.0005,
+        "loss": 1.2877,
+        "step": 7530
+      },
+      {
+        "epoch": 0.24044134060397335,
+        "grad_norm": 0.5034393072128296,
+        "learning_rate": 0.0005,
+        "loss": 1.2783,
+        "step": 7540
+      },
+      {
+        "epoch": 0.24076022832360724,
+        "grad_norm": 0.5194485187530518,
+        "learning_rate": 0.0005,
+        "loss": 1.2617,
+        "step": 7550
+      },
+      {
+        "epoch": 0.24107911604324117,
+        "grad_norm": 0.4872956871986389,
+        "learning_rate": 0.0005,
+        "loss": 1.2691,
+        "step": 7560
+      },
+      {
+        "epoch": 0.2413980037628751,
+        "grad_norm": 0.5291786193847656,
+        "learning_rate": 0.0005,
+        "loss": 1.2955,
+        "step": 7570
+      },
+      {
+        "epoch": 0.24171689148250902,
+        "grad_norm": 0.5270318388938904,
+        "learning_rate": 0.0005,
+        "loss": 1.2738,
+        "step": 7580
+      },
+      {
+        "epoch": 0.24203577920214292,
+        "grad_norm": 0.5061559081077576,
+        "learning_rate": 0.0005,
+        "loss": 1.2767,
+        "step": 7590
+      },
+      {
+        "epoch": 0.24235466692177685,
+        "grad_norm": 0.5161083936691284,
+        "learning_rate": 0.0005,
+        "loss": 1.2884,
+        "step": 7600
+      },
+      {
+        "epoch": 0.24267355464141077,
+        "grad_norm": 0.5098254084587097,
+        "learning_rate": 0.0005,
+        "loss": 1.254,
+        "step": 7610
+      },
+      {
+        "epoch": 0.24299244236104467,
+        "grad_norm": 0.5221891403198242,
+        "learning_rate": 0.0005,
+        "loss": 1.2629,
+        "step": 7620
+      },
+      {
+        "epoch": 0.2433113300806786,
+        "grad_norm": 0.4903581142425537,
+        "learning_rate": 0.0005,
+        "loss": 1.2831,
+        "step": 7630
+      },
+      {
+        "epoch": 0.24363021780031252,
+        "grad_norm": 0.5214158892631531,
+        "learning_rate": 0.0005,
+        "loss": 1.306,
+        "step": 7640
+      },
+      {
+        "epoch": 0.24394910551994642,
+        "grad_norm": 0.5176328420639038,
+        "learning_rate": 0.0005,
+        "loss": 1.2712,
+        "step": 7650
+      },
+      {
+        "epoch": 0.24426799323958034,
+        "grad_norm": 0.5082030296325684,
+        "learning_rate": 0.0005,
+        "loss": 1.2707,
+        "step": 7660
+      },
+      {
+        "epoch": 0.24458688095921427,
+        "grad_norm": 0.5280209183692932,
+        "learning_rate": 0.0005,
+        "loss": 1.2687,
+        "step": 7670
+      },
+      {
+        "epoch": 0.24490576867884817,
+        "grad_norm": 0.5040051341056824,
+        "learning_rate": 0.0005,
+        "loss": 1.2586,
+        "step": 7680
+      },
+      {
+        "epoch": 0.2452246563984821,
+        "grad_norm": 0.5304771065711975,
+        "learning_rate": 0.0005,
+        "loss": 1.2823,
+        "step": 7690
+      },
+      {
+        "epoch": 0.24554354411811602,
+        "grad_norm": 0.5364067554473877,
+        "learning_rate": 0.0005,
+        "loss": 1.2667,
+        "step": 7700
+      },
+      {
+        "epoch": 0.24586243183774992,
+        "grad_norm": 0.5052260160446167,
+        "learning_rate": 0.0005,
+        "loss": 1.2513,
+        "step": 7710
+      },
+      {
+        "epoch": 0.24618131955738384,
+        "grad_norm": 0.5323502421379089,
+        "learning_rate": 0.0005,
+        "loss": 1.2708,
+        "step": 7720
+      },
+      {
+        "epoch": 0.24650020727701777,
+        "grad_norm": 0.49328145384788513,
+        "learning_rate": 0.0005,
+        "loss": 1.2792,
+        "step": 7730
+      },
+      {
+        "epoch": 0.24681909499665167,
+        "grad_norm": 0.5066396594047546,
+        "learning_rate": 0.0005,
+        "loss": 1.2724,
+        "step": 7740
+      },
+      {
+        "epoch": 0.2471379827162856,
+        "grad_norm": 0.5193277597427368,
+        "learning_rate": 0.0005,
+        "loss": 1.2937,
+        "step": 7750
+      },
+      {
+        "epoch": 0.24745687043591952,
+        "grad_norm": 0.506821870803833,
+        "learning_rate": 0.0005,
+        "loss": 1.264,
+        "step": 7760
+      },
+      {
+        "epoch": 0.24777575815555342,
+        "grad_norm": 0.4907330572605133,
+        "learning_rate": 0.0005,
+        "loss": 1.2785,
+        "step": 7770
+      },
+      {
+        "epoch": 0.24809464587518734,
+        "grad_norm": 0.53122478723526,
+        "learning_rate": 0.0005,
+        "loss": 1.2495,
+        "step": 7780
+      },
+      {
+        "epoch": 0.24841353359482127,
+        "grad_norm": 0.5209131836891174,
+        "learning_rate": 0.0005,
+        "loss": 1.2884,
+        "step": 7790
+      },
+      {
+        "epoch": 0.2487324213144552,
+        "grad_norm": 0.5227766036987305,
+        "learning_rate": 0.0005,
+        "loss": 1.2676,
+        "step": 7800
+      },
+      {
+        "epoch": 0.2490513090340891,
+        "grad_norm": 0.5297362208366394,
+        "learning_rate": 0.0005,
+        "loss": 1.2781,
+        "step": 7810
+      },
+      {
+        "epoch": 0.24937019675372302,
+        "grad_norm": 0.5168874859809875,
+        "learning_rate": 0.0005,
+        "loss": 1.2637,
+        "step": 7820
+      },
+      {
+        "epoch": 0.24968908447335694,
+        "grad_norm": 0.5064091682434082,
+        "learning_rate": 0.0005,
+        "loss": 1.2512,
+        "step": 7830
+      },
+      {
+        "epoch": 0.25000797219299087,
+        "grad_norm": 0.49836963415145874,
+        "learning_rate": 0.0005,
+        "loss": 1.2539,
+        "step": 7840
+      },
+      {
+        "epoch": 0.25032685991262477,
+        "grad_norm": 0.49791836738586426,
+        "learning_rate": 0.0005,
+        "loss": 1.2421,
+        "step": 7850
+      },
+      {
+        "epoch": 0.25064574763225866,
+        "grad_norm": 0.5025899410247803,
+        "learning_rate": 0.0005,
+        "loss": 1.2452,
+        "step": 7860
+      },
+      {
+        "epoch": 0.2509646353518926,
+        "grad_norm": 0.5179859399795532,
+        "learning_rate": 0.0005,
+        "loss": 1.2486,
+        "step": 7870
+      },
+      {
+        "epoch": 0.2512835230715265,
+        "grad_norm": 0.47681480646133423,
+        "learning_rate": 0.0005,
+        "loss": 1.2545,
+        "step": 7880
+      },
+      {
+        "epoch": 0.2516024107911604,
+        "grad_norm": 0.5042859315872192,
+        "learning_rate": 0.0005,
+        "loss": 1.2929,
+        "step": 7890
+      },
+      {
+        "epoch": 0.25192129851079437,
+        "grad_norm": 0.49867138266563416,
+        "learning_rate": 0.0005,
+        "loss": 1.267,
+        "step": 7900
+      },
+      {
+        "epoch": 0.25224018623042826,
+        "grad_norm": 0.5240023732185364,
+        "learning_rate": 0.0005,
+        "loss": 1.2651,
+        "step": 7910
+      },
+      {
+        "epoch": 0.25255907395006216,
+        "grad_norm": 0.5321725606918335,
+        "learning_rate": 0.0005,
+        "loss": 1.2684,
+        "step": 7920
+      },
+      {
+        "epoch": 0.2528779616696961,
+        "grad_norm": 0.534798800945282,
+        "learning_rate": 0.0005,
+        "loss": 1.2606,
+        "step": 7930
+      },
+      {
+        "epoch": 0.25319684938933,
+        "grad_norm": 0.5044108033180237,
+        "learning_rate": 0.0005,
+        "loss": 1.2564,
+        "step": 7940
+      },
+      {
+        "epoch": 0.2535157371089639,
+        "grad_norm": 0.48750004172325134,
+        "learning_rate": 0.0005,
+        "loss": 1.2609,
+        "step": 7950
+      },
+      {
+        "epoch": 0.25383462482859787,
+        "grad_norm": 0.5071402192115784,
+        "learning_rate": 0.0005,
+        "loss": 1.2736,
+        "step": 7960
+      },
+      {
+        "epoch": 0.25415351254823176,
+        "grad_norm": 0.49629124999046326,
+        "learning_rate": 0.0005,
+        "loss": 1.266,
+        "step": 7970
+      },
+      {
+        "epoch": 0.25447240026786566,
+        "grad_norm": 0.5239920020103455,
+        "learning_rate": 0.0005,
+        "loss": 1.2882,
+        "step": 7980
+      },
+      {
+        "epoch": 0.2547912879874996,
+        "grad_norm": 0.5081344842910767,
+        "learning_rate": 0.0005,
+        "loss": 1.2813,
+        "step": 7990
+      },
+      {
+        "epoch": 0.2551101757071335,
+        "grad_norm": 0.49004441499710083,
+        "learning_rate": 0.0005,
+        "loss": 1.2561,
+        "step": 8000
+      },
+      {
+        "epoch": 0.2554290634267674,
+        "grad_norm": 0.5031710267066956,
+        "learning_rate": 0.0005,
+        "loss": 1.249,
+        "step": 8010
+      },
+      {
+        "epoch": 0.25574795114640136,
+        "grad_norm": 0.5036451816558838,
+        "learning_rate": 0.0005,
+        "loss": 1.2641,
+        "step": 8020
+      },
+      {
+        "epoch": 0.25606683886603526,
+        "grad_norm": 0.4846326410770416,
+        "learning_rate": 0.0005,
+        "loss": 1.249,
+        "step": 8030
+      },
+      {
+        "epoch": 0.25638572658566916,
+        "grad_norm": 0.5058959722518921,
+        "learning_rate": 0.0005,
+        "loss": 1.2644,
+        "step": 8040
+      },
+      {
+        "epoch": 0.2567046143053031,
+        "grad_norm": 0.5171712636947632,
+        "learning_rate": 0.0005,
+        "loss": 1.2767,
+        "step": 8050
+      },
+      {
+        "epoch": 0.257023502024937,
+        "grad_norm": 0.5237571597099304,
+        "learning_rate": 0.0005,
+        "loss": 1.2619,
+        "step": 8060
+      },
+      {
+        "epoch": 0.2573423897445709,
+        "grad_norm": 0.4908455014228821,
+        "learning_rate": 0.0005,
+        "loss": 1.2743,
+        "step": 8070
+      },
+      {
+        "epoch": 0.25766127746420486,
+        "grad_norm": 0.504867434501648,
+        "learning_rate": 0.0005,
+        "loss": 1.259,
+        "step": 8080
+      },
+      {
+        "epoch": 0.25798016518383876,
+        "grad_norm": 0.5199716687202454,
+        "learning_rate": 0.0005,
+        "loss": 1.2925,
+        "step": 8090
+      },
+      {
+        "epoch": 0.2582990529034727,
+        "grad_norm": 0.49130043387413025,
+        "learning_rate": 0.0005,
+        "loss": 1.2742,
+        "step": 8100
+      },
+      {
+        "epoch": 0.2586179406231066,
+        "grad_norm": 0.517345666885376,
+        "learning_rate": 0.0005,
+        "loss": 1.2719,
+        "step": 8110
+      },
+      {
+        "epoch": 0.2589368283427405,
+        "grad_norm": 0.49142032861709595,
+        "learning_rate": 0.0005,
+        "loss": 1.2468,
+        "step": 8120
+      },
+      {
+        "epoch": 0.25925571606237446,
+        "grad_norm": 0.49685269594192505,
+        "learning_rate": 0.0005,
+        "loss": 1.2534,
+        "step": 8130
+      },
+      {
+        "epoch": 0.25957460378200836,
+        "grad_norm": 0.48098668456077576,
+        "learning_rate": 0.0005,
+        "loss": 1.2317,
+        "step": 8140
+      },
+      {
+        "epoch": 0.25989349150164226,
+        "grad_norm": 0.5050351619720459,
+        "learning_rate": 0.0005,
+        "loss": 1.2576,
+        "step": 8150
+      },
+      {
+        "epoch": 0.2602123792212762,
+        "grad_norm": 0.506672739982605,
+        "learning_rate": 0.0005,
+        "loss": 1.2539,
+        "step": 8160
+      },
+      {
+        "epoch": 0.2605312669409101,
+        "grad_norm": 0.49320903420448303,
+        "learning_rate": 0.0005,
+        "loss": 1.2421,
+        "step": 8170
+      },
+      {
+        "epoch": 0.260850154660544,
+        "grad_norm": 0.5003274083137512,
+        "learning_rate": 0.0005,
+        "loss": 1.2658,
+        "step": 8180
+      },
+      {
+        "epoch": 0.26116904238017796,
+        "grad_norm": 0.498389333486557,
+        "learning_rate": 0.0005,
+        "loss": 1.2512,
+        "step": 8190
+      },
+      {
+        "epoch": 0.26148793009981186,
+        "grad_norm": 0.49021822214126587,
+        "learning_rate": 0.0005,
+        "loss": 1.2569,
+        "step": 8200
+      },
+      {
+        "epoch": 0.26180681781944576,
+        "grad_norm": 0.49076321721076965,
+        "learning_rate": 0.0005,
+        "loss": 1.2751,
+        "step": 8210
+      },
+      {
+        "epoch": 0.2621257055390797,
+        "grad_norm": 0.5223627686500549,
+        "learning_rate": 0.0005,
+        "loss": 1.2742,
+        "step": 8220
+      },
+      {
+        "epoch": 0.2624445932587136,
+        "grad_norm": 0.4935097396373749,
+        "learning_rate": 0.0005,
+        "loss": 1.2523,
+        "step": 8230
+      },
+      {
+        "epoch": 0.2627634809783475,
+        "grad_norm": 0.47696298360824585,
+        "learning_rate": 0.0005,
+        "loss": 1.246,
+        "step": 8240
+      },
+      {
+        "epoch": 0.26308236869798146,
+        "grad_norm": 0.5192391276359558,
+        "learning_rate": 0.0005,
+        "loss": 1.2377,
+        "step": 8250
+      },
+      {
+        "epoch": 0.26340125641761536,
+        "grad_norm": 0.47619664669036865,
+        "learning_rate": 0.0005,
+        "loss": 1.2697,
+        "step": 8260
+      },
+      {
+        "epoch": 0.26372014413724926,
+        "grad_norm": 0.4984939992427826,
+        "learning_rate": 0.0005,
+        "loss": 1.2588,
+        "step": 8270
+      },
+      {
+        "epoch": 0.2640390318568832,
+        "grad_norm": 0.49781128764152527,
+        "learning_rate": 0.0005,
+        "loss": 1.265,
+        "step": 8280
+      },
+      {
+        "epoch": 0.2643579195765171,
+        "grad_norm": 0.5310728549957275,
+        "learning_rate": 0.0005,
+        "loss": 1.2626,
+        "step": 8290
+      },
+      {
+        "epoch": 0.264676807296151,
+        "grad_norm": 0.4933306574821472,
+        "learning_rate": 0.0005,
+        "loss": 1.2506,
+        "step": 8300
+      },
+      {
+        "epoch": 0.26499569501578496,
+        "grad_norm": 0.5005725622177124,
+        "learning_rate": 0.0005,
+        "loss": 1.2512,
+        "step": 8310
+      },
+      {
+        "epoch": 0.26531458273541886,
+        "grad_norm": 0.4766061305999756,
+        "learning_rate": 0.0005,
+        "loss": 1.2401,
+        "step": 8320
+      },
+      {
+        "epoch": 0.26563347045505276,
+        "grad_norm": 0.5212782025337219,
+        "learning_rate": 0.0005,
+        "loss": 1.2389,
+        "step": 8330
+      },
+      {
+        "epoch": 0.2659523581746867,
+        "grad_norm": 0.49564340710639954,
+        "learning_rate": 0.0005,
+        "loss": 1.2662,
+        "step": 8340
+      },
+      {
+        "epoch": 0.2662712458943206,
+        "grad_norm": 0.5133705139160156,
+        "learning_rate": 0.0005,
+        "loss": 1.2501,
+        "step": 8350
+      },
+      {
+        "epoch": 0.2665901336139545,
+        "grad_norm": 0.4940152168273926,
+        "learning_rate": 0.0005,
+        "loss": 1.2794,
+        "step": 8360
+      },
+      {
+        "epoch": 0.26690902133358846,
+        "grad_norm": 0.5306583642959595,
+        "learning_rate": 0.0005,
+        "loss": 1.2737,
+        "step": 8370
+      },
+      {
+        "epoch": 0.26722790905322236,
+        "grad_norm": 0.4824816584587097,
+        "learning_rate": 0.0005,
+        "loss": 1.2582,
+        "step": 8380
+      },
+      {
+        "epoch": 0.26754679677285625,
+        "grad_norm": 0.5017719268798828,
+        "learning_rate": 0.0005,
+        "loss": 1.2583,
+        "step": 8390
+      },
+      {
+        "epoch": 0.2678656844924902,
+        "grad_norm": 0.5216484665870667,
+        "learning_rate": 0.0005,
+        "loss": 1.2525,
+        "step": 8400
+      },
+      {
+        "epoch": 0.2681845722121241,
+        "grad_norm": 0.5009568333625793,
+        "learning_rate": 0.0005,
+        "loss": 1.2473,
+        "step": 8410
+      },
+      {
+        "epoch": 0.268503459931758,
+        "grad_norm": 0.4954254627227783,
+        "learning_rate": 0.0005,
+        "loss": 1.2477,
+        "step": 8420
+      },
+      {
+        "epoch": 0.26882234765139196,
+        "grad_norm": 0.4804369807243347,
+        "learning_rate": 0.0005,
+        "loss": 1.2432,
+        "step": 8430
+      },
+      {
+        "epoch": 0.26914123537102586,
+        "grad_norm": 0.4917435050010681,
+        "learning_rate": 0.0005,
+        "loss": 1.2392,
+        "step": 8440
+      },
+      {
+        "epoch": 0.26946012309065975,
+        "grad_norm": 0.49070221185684204,
+        "learning_rate": 0.0005,
+        "loss": 1.2515,
+        "step": 8450
+      },
+      {
+        "epoch": 0.2697790108102937,
+        "grad_norm": 0.4955213665962219,
+        "learning_rate": 0.0005,
+        "loss": 1.2485,
+        "step": 8460
+      },
+      {
+        "epoch": 0.2700978985299276,
+        "grad_norm": 0.4929785430431366,
+        "learning_rate": 0.0005,
+        "loss": 1.2606,
+        "step": 8470
+      },
+      {
+        "epoch": 0.2704167862495615,
+        "grad_norm": 0.47558659315109253,
+        "learning_rate": 0.0005,
+        "loss": 1.2425,
+        "step": 8480
+      },
+      {
+        "epoch": 0.27073567396919546,
+        "grad_norm": 0.5124787092208862,
+        "learning_rate": 0.0005,
+        "loss": 1.2464,
+        "step": 8490
+      },
+      {
+        "epoch": 0.27105456168882935,
+        "grad_norm": 0.47737860679626465,
+        "learning_rate": 0.0005,
+        "loss": 1.2731,
+        "step": 8500
+      },
+      {
+        "epoch": 0.2713734494084633,
+        "grad_norm": 0.4930284917354584,
+        "learning_rate": 0.0005,
+        "loss": 1.2417,
+        "step": 8510
+      },
+      {
+        "epoch": 0.2716923371280972,
+        "grad_norm": 0.48435503244400024,
+        "learning_rate": 0.0005,
+        "loss": 1.238,
+        "step": 8520
+      },
+      {
+        "epoch": 0.2720112248477311,
+        "grad_norm": 0.48783987760543823,
+        "learning_rate": 0.0005,
+        "loss": 1.2385,
+        "step": 8530
+      },
+      {
+        "epoch": 0.27233011256736506,
+        "grad_norm": 0.48623397946357727,
+        "learning_rate": 0.0005,
+        "loss": 1.2498,
+        "step": 8540
+      },
+      {
+        "epoch": 0.27264900028699895,
+        "grad_norm": 0.5029369592666626,
+        "learning_rate": 0.0005,
+        "loss": 1.2725,
+        "step": 8550
+      },
+      {
+        "epoch": 0.27296788800663285,
+        "grad_norm": 0.4850176274776459,
+        "learning_rate": 0.0005,
+        "loss": 1.2601,
+        "step": 8560
+      },
+      {
+        "epoch": 0.2732867757262668,
+        "grad_norm": 0.4850327968597412,
+        "learning_rate": 0.0005,
+        "loss": 1.2708,
+        "step": 8570
+      },
+      {
+        "epoch": 0.2736056634459007,
+        "grad_norm": 0.48380863666534424,
+        "learning_rate": 0.0005,
+        "loss": 1.2322,
+        "step": 8580
+      },
+      {
+        "epoch": 0.2739245511655346,
+        "grad_norm": 0.48448917269706726,
+        "learning_rate": 0.0005,
+        "loss": 1.2386,
+        "step": 8590
+      },
+      {
+        "epoch": 0.27424343888516856,
+        "grad_norm": 0.5155131220817566,
+        "learning_rate": 0.0005,
+        "loss": 1.2579,
+        "step": 8600
+      },
+      {
+        "epoch": 0.27456232660480245,
+        "grad_norm": 0.4893667697906494,
+        "learning_rate": 0.0005,
+        "loss": 1.2625,
+        "step": 8610
+      },
+      {
+        "epoch": 0.27488121432443635,
+        "grad_norm": 0.4755896329879761,
+        "learning_rate": 0.0005,
+        "loss": 1.23,
+        "step": 8620
+      },
+      {
+        "epoch": 0.2752001020440703,
+        "grad_norm": 0.4704909920692444,
+        "learning_rate": 0.0005,
+        "loss": 1.2391,
+        "step": 8630
+      },
+      {
+        "epoch": 0.2755189897637042,
+        "grad_norm": 0.48995453119277954,
+        "learning_rate": 0.0005,
+        "loss": 1.2358,
+        "step": 8640
+      },
+      {
+        "epoch": 0.2758378774833381,
+        "grad_norm": 0.4795413911342621,
+        "learning_rate": 0.0005,
+        "loss": 1.2648,
+        "step": 8650
+      },
+      {
+        "epoch": 0.27615676520297205,
+        "grad_norm": 0.4703845977783203,
+        "learning_rate": 0.0005,
+        "loss": 1.2421,
+        "step": 8660
+      },
+      {
+        "epoch": 0.27647565292260595,
+        "grad_norm": 0.48536381125450134,
+        "learning_rate": 0.0005,
+        "loss": 1.2467,
+        "step": 8670
+      },
+      {
+        "epoch": 0.27679454064223985,
+        "grad_norm": 0.4863326847553253,
+        "learning_rate": 0.0005,
+        "loss": 1.2332,
+        "step": 8680
+      },
+      {
+        "epoch": 0.2771134283618738,
+        "grad_norm": 0.4987434148788452,
+        "learning_rate": 0.0005,
+        "loss": 1.2286,
+        "step": 8690
+      },
+      {
+        "epoch": 0.2774323160815077,
+        "grad_norm": 0.4930117726325989,
+        "learning_rate": 0.0005,
+        "loss": 1.2462,
+        "step": 8700
+      },
+      {
+        "epoch": 0.2777512038011416,
+        "grad_norm": 0.47064122557640076,
+        "learning_rate": 0.0005,
+        "loss": 1.2494,
+        "step": 8710
+      },
+      {
+        "epoch": 0.27807009152077555,
+        "grad_norm": 0.5006730556488037,
+        "learning_rate": 0.0005,
+        "loss": 1.2479,
+        "step": 8720
+      },
+      {
+        "epoch": 0.27838897924040945,
+        "grad_norm": 0.4980914294719696,
+        "learning_rate": 0.0005,
+        "loss": 1.2202,
+        "step": 8730
+      },
+      {
+        "epoch": 0.27870786696004335,
+        "grad_norm": 0.49676623940467834,
+        "learning_rate": 0.0005,
+        "loss": 1.2397,
+        "step": 8740
+      },
+      {
+        "epoch": 0.2790267546796773,
+        "grad_norm": 0.5021325945854187,
+        "learning_rate": 0.0005,
+        "loss": 1.2487,
+        "step": 8750
+      },
+      {
+        "epoch": 0.2793456423993112,
+        "grad_norm": 0.4874837100505829,
+        "learning_rate": 0.0005,
+        "loss": 1.2396,
+        "step": 8760
+      },
+      {
+        "epoch": 0.2796645301189451,
+        "grad_norm": 0.4959019422531128,
+        "learning_rate": 0.0005,
+        "loss": 1.2452,
+        "step": 8770
+      },
+      {
+        "epoch": 0.27998341783857905,
+        "grad_norm": 0.4884132146835327,
+        "learning_rate": 0.0005,
+        "loss": 1.2535,
+        "step": 8780
+      },
+      {
+        "epoch": 0.28030230555821295,
+        "grad_norm": 0.47749945521354675,
+        "learning_rate": 0.0005,
+        "loss": 1.2397,
+        "step": 8790
+      },
+      {
+        "epoch": 0.28062119327784685,
+        "grad_norm": 0.49802863597869873,
+        "learning_rate": 0.0005,
+        "loss": 1.2327,
+        "step": 8800
+      },
+      {
+        "epoch": 0.2809400809974808,
+        "grad_norm": 0.4573667645454407,
+        "learning_rate": 0.0005,
+        "loss": 1.257,
+        "step": 8810
+      },
+      {
+        "epoch": 0.2812589687171147,
+        "grad_norm": 0.47507792711257935,
+        "learning_rate": 0.0005,
+        "loss": 1.2424,
+        "step": 8820
+      },
+      {
+        "epoch": 0.2815778564367486,
+        "grad_norm": 0.4829874336719513,
+        "learning_rate": 0.0005,
+        "loss": 1.2699,
+        "step": 8830
+      },
+      {
+        "epoch": 0.28189674415638255,
+        "grad_norm": 0.48574453592300415,
+        "learning_rate": 0.0005,
+        "loss": 1.2402,
+        "step": 8840
+      },
+      {
+        "epoch": 0.28221563187601645,
+        "grad_norm": 0.47372007369995117,
+        "learning_rate": 0.0005,
+        "loss": 1.2436,
+        "step": 8850
+      },
+      {
+        "epoch": 0.28253451959565035,
+        "grad_norm": 0.4779689908027649,
+        "learning_rate": 0.0005,
+        "loss": 1.247,
+        "step": 8860
+      },
+      {
+        "epoch": 0.2828534073152843,
+        "grad_norm": 0.5082475543022156,
+        "learning_rate": 0.0005,
+        "loss": 1.2308,
+        "step": 8870
+      },
+      {
+        "epoch": 0.2831722950349182,
+        "grad_norm": 0.48237717151641846,
+        "learning_rate": 0.0005,
+        "loss": 1.2502,
+        "step": 8880
+      },
+      {
+        "epoch": 0.2834911827545521,
+        "grad_norm": 0.4815846383571625,
+        "learning_rate": 0.0005,
+        "loss": 1.2288,
+        "step": 8890
+      },
+      {
+        "epoch": 0.28381007047418605,
+        "grad_norm": 0.5010101199150085,
+        "learning_rate": 0.0005,
+        "loss": 1.2487,
+        "step": 8900
+      },
+      {
+        "epoch": 0.28412895819381995,
+        "grad_norm": 0.49066150188446045,
+        "learning_rate": 0.0005,
+        "loss": 1.2343,
+        "step": 8910
+      },
+      {
+        "epoch": 0.2844478459134539,
+        "grad_norm": 0.48220255970954895,
+        "learning_rate": 0.0005,
+        "loss": 1.2351,
+        "step": 8920
+      },
+      {
+        "epoch": 0.2847667336330878,
+        "grad_norm": 0.4660911560058594,
+        "learning_rate": 0.0005,
+        "loss": 1.2439,
+        "step": 8930
+      },
+      {
+        "epoch": 0.2850856213527217,
+        "grad_norm": 0.5226732492446899,
+        "learning_rate": 0.0005,
+        "loss": 1.239,
+        "step": 8940
+      },
+      {
+        "epoch": 0.28540450907235565,
+        "grad_norm": 0.4776681065559387,
+        "learning_rate": 0.0005,
+        "loss": 1.2482,
+        "step": 8950
+      },
+      {
+        "epoch": 0.28572339679198955,
+        "grad_norm": 0.4984127879142761,
+        "learning_rate": 0.0005,
+        "loss": 1.2289,
+        "step": 8960
+      },
+      {
+        "epoch": 0.28604228451162345,
+        "grad_norm": 0.484502375125885,
+        "learning_rate": 0.0005,
+        "loss": 1.2339,
+        "step": 8970
+      },
+      {
+        "epoch": 0.2863611722312574,
+        "grad_norm": 0.48515409231185913,
+        "learning_rate": 0.0005,
+        "loss": 1.2377,
+        "step": 8980
+      },
+      {
+        "epoch": 0.2866800599508913,
+        "grad_norm": 0.48801496624946594,
+        "learning_rate": 0.0005,
+        "loss": 1.2372,
+        "step": 8990
+      },
+      {
+        "epoch": 0.2869989476705252,
+        "grad_norm": 0.503757655620575,
+        "learning_rate": 0.0005,
+        "loss": 1.2642,
+        "step": 9000
+      },
+      {
+        "epoch": 0.28731783539015915,
+        "grad_norm": 0.47584301233291626,
+        "learning_rate": 0.0005,
+        "loss": 1.2172,
+        "step": 9010
+      },
+      {
+        "epoch": 0.28763672310979305,
+        "grad_norm": 0.48166683316230774,
+        "learning_rate": 0.0005,
+        "loss": 1.238,
+        "step": 9020
+      },
+      {
+        "epoch": 0.28795561082942694,
+        "grad_norm": 0.47695672512054443,
+        "learning_rate": 0.0005,
+        "loss": 1.2382,
+        "step": 9030
+      },
+      {
+        "epoch": 0.2882744985490609,
+        "grad_norm": 0.48136505484580994,
+        "learning_rate": 0.0005,
+        "loss": 1.2453,
+        "step": 9040
+      },
+      {
+        "epoch": 0.2885933862686948,
+        "grad_norm": 0.4705105721950531,
+        "learning_rate": 0.0005,
+        "loss": 1.2316,
+        "step": 9050
+      },
+      {
+        "epoch": 0.2889122739883287,
+        "grad_norm": 0.47545409202575684,
+        "learning_rate": 0.0005,
+        "loss": 1.2235,
+        "step": 9060
+      },
+      {
+        "epoch": 0.28923116170796265,
+        "grad_norm": 0.48143270611763,
+        "learning_rate": 0.0005,
+        "loss": 1.2318,
+        "step": 9070
+      },
+      {
+        "epoch": 0.28955004942759655,
+        "grad_norm": 0.4943726360797882,
+        "learning_rate": 0.0005,
+        "loss": 1.2389,
+        "step": 9080
+      },
+      {
+        "epoch": 0.28986893714723044,
+        "grad_norm": 0.4849379062652588,
+        "learning_rate": 0.0005,
+        "loss": 1.2589,
+        "step": 9090
+      },
+      {
+        "epoch": 0.2901878248668644,
+        "grad_norm": 0.4788464307785034,
+        "learning_rate": 0.0005,
+        "loss": 1.2294,
+        "step": 9100
+      },
+      {
+        "epoch": 0.2905067125864983,
+        "grad_norm": 0.4739382266998291,
+        "learning_rate": 0.0005,
+        "loss": 1.2329,
+        "step": 9110
+      },
+      {
+        "epoch": 0.2908256003061322,
+        "grad_norm": 0.5209506750106812,
+        "learning_rate": 0.0005,
+        "loss": 1.2454,
+        "step": 9120
+      },
+      {
+        "epoch": 0.29114448802576615,
+        "grad_norm": 0.4819193184375763,
+        "learning_rate": 0.0005,
+        "loss": 1.2409,
+        "step": 9130
+      },
+      {
+        "epoch": 0.29146337574540004,
+        "grad_norm": 0.4805019199848175,
+        "learning_rate": 0.0005,
+        "loss": 1.2297,
+        "step": 9140
+      },
+      {
+        "epoch": 0.29178226346503394,
+        "grad_norm": 0.4745943248271942,
+        "learning_rate": 0.0005,
+        "loss": 1.2266,
+        "step": 9150
+      },
+      {
+        "epoch": 0.2921011511846679,
+        "grad_norm": 0.47815170884132385,
+        "learning_rate": 0.0005,
+        "loss": 1.2464,
+        "step": 9160
+      },
+      {
+        "epoch": 0.2924200389043018,
+        "grad_norm": 0.47933608293533325,
+        "learning_rate": 0.0005,
+        "loss": 1.2507,
+        "step": 9170
+      },
+      {
+        "epoch": 0.2927389266239357,
+        "grad_norm": 0.4662920832633972,
+        "learning_rate": 0.0005,
+        "loss": 1.2414,
+        "step": 9180
+      },
+      {
+        "epoch": 0.29305781434356964,
+        "grad_norm": 0.4747319519519806,
+        "learning_rate": 0.0005,
+        "loss": 1.2223,
+        "step": 9190
+      },
+      {
+        "epoch": 0.29337670206320354,
+        "grad_norm": 0.4738784730434418,
+        "learning_rate": 0.0005,
+        "loss": 1.2254,
+        "step": 9200
+      },
+      {
+        "epoch": 0.29369558978283744,
+        "grad_norm": 0.4642687141895294,
+        "learning_rate": 0.0005,
+        "loss": 1.2193,
+        "step": 9210
+      },
+      {
+        "epoch": 0.2940144775024714,
+        "grad_norm": 0.48891007900238037,
+        "learning_rate": 0.0005,
+        "loss": 1.2275,
+        "step": 9220
+      },
+      {
+        "epoch": 0.2943333652221053,
+        "grad_norm": 0.4732115864753723,
+        "learning_rate": 0.0005,
+        "loss": 1.2375,
+        "step": 9230
+      },
+      {
+        "epoch": 0.2946522529417392,
+        "grad_norm": 0.4736773669719696,
+        "learning_rate": 0.0005,
+        "loss": 1.216,
+        "step": 9240
+      },
+      {
+        "epoch": 0.29497114066137314,
+        "grad_norm": 0.4922406077384949,
+        "learning_rate": 0.0005,
+        "loss": 1.2357,
+        "step": 9250
+      },
+      {
+        "epoch": 0.29529002838100704,
+        "grad_norm": 0.4949341416358948,
+        "learning_rate": 0.0005,
+        "loss": 1.2299,
+        "step": 9260
+      },
+      {
+        "epoch": 0.29560891610064094,
+        "grad_norm": 0.4667598009109497,
+        "learning_rate": 0.0005,
+        "loss": 1.2184,
+        "step": 9270
+      },
+      {
+        "epoch": 0.2959278038202749,
+        "grad_norm": 0.4855829179286957,
+        "learning_rate": 0.0005,
+        "loss": 1.2289,
+        "step": 9280
+      },
+      {
+        "epoch": 0.2962466915399088,
+        "grad_norm": 0.4694145917892456,
+        "learning_rate": 0.0005,
+        "loss": 1.2085,
+        "step": 9290
+      },
+      {
+        "epoch": 0.2965655792595427,
+        "grad_norm": 0.466633141040802,
+        "learning_rate": 0.0005,
+        "loss": 1.228,
+        "step": 9300
+      },
+      {
+        "epoch": 0.29688446697917664,
+        "grad_norm": 0.4754934012889862,
+        "learning_rate": 0.0005,
+        "loss": 1.2398,
+        "step": 9310
+      },
+      {
+        "epoch": 0.29720335469881054,
+        "grad_norm": 0.47169917821884155,
+        "learning_rate": 0.0005,
+        "loss": 1.2096,
+        "step": 9320
+      },
+      {
+        "epoch": 0.29752224241844444,
+        "grad_norm": 0.476678729057312,
+        "learning_rate": 0.0005,
+        "loss": 1.2385,
+        "step": 9330
+      },
+      {
+        "epoch": 0.2978411301380784,
+        "grad_norm": 0.4720135033130646,
+        "learning_rate": 0.0005,
+        "loss": 1.2181,
+        "step": 9340
+      },
+      {
+        "epoch": 0.2981600178577123,
+        "grad_norm": 0.4739975035190582,
+        "learning_rate": 0.0005,
+        "loss": 1.218,
+        "step": 9350
+      },
+      {
+        "epoch": 0.29847890557734624,
+        "grad_norm": 0.475588321685791,
+        "learning_rate": 0.0005,
+        "loss": 1.2289,
+        "step": 9360
+      },
+      {
+        "epoch": 0.29879779329698014,
+        "grad_norm": 0.4978891909122467,
+        "learning_rate": 0.0005,
+        "loss": 1.233,
+        "step": 9370
+      },
+      {
+        "epoch": 0.29911668101661404,
+        "grad_norm": 0.47417309880256653,
+        "learning_rate": 0.0005,
+        "loss": 1.2259,
+        "step": 9380
+      },
+      {
+        "epoch": 0.299435568736248,
+        "grad_norm": 0.4477439522743225,
+        "learning_rate": 0.0005,
+        "loss": 1.2146,
+        "step": 9390
+      },
+      {
+        "epoch": 0.2997544564558819,
+        "grad_norm": 0.4888128340244293,
+        "learning_rate": 0.0005,
+        "loss": 1.2264,
+        "step": 9400
+      },
+      {
+        "epoch": 0.3000733441755158,
+        "grad_norm": 0.45956334471702576,
+        "learning_rate": 0.0005,
+        "loss": 1.2328,
+        "step": 9410
+      },
+      {
+        "epoch": 0.30039223189514974,
+        "grad_norm": 0.4596233367919922,
+        "learning_rate": 0.0005,
+        "loss": 1.228,
+        "step": 9420
+      },
+      {
+        "epoch": 0.30071111961478364,
+        "grad_norm": 0.4777589738368988,
+        "learning_rate": 0.0005,
+        "loss": 1.2207,
+        "step": 9430
+      },
+      {
+        "epoch": 0.30103000733441754,
+        "grad_norm": 0.4956303834915161,
+        "learning_rate": 0.0005,
+        "loss": 1.2327,
+        "step": 9440
+      },
+      {
+        "epoch": 0.3013488950540515,
+        "grad_norm": 0.4638616144657135,
+        "learning_rate": 0.0005,
+        "loss": 1.2341,
+        "step": 9450
+      },
+      {
+        "epoch": 0.3016677827736854,
+        "grad_norm": 0.4854125678539276,
+        "learning_rate": 0.0005,
+        "loss": 1.2394,
+        "step": 9460
+      },
+      {
+        "epoch": 0.3019866704933193,
+        "grad_norm": 0.4745231568813324,
+        "learning_rate": 0.0005,
+        "loss": 1.2479,
+        "step": 9470
+      },
+      {
+        "epoch": 0.30230555821295324,
+        "grad_norm": 0.47391799092292786,
+        "learning_rate": 0.0005,
+        "loss": 1.2291,
+        "step": 9480
+      },
+      {
+        "epoch": 0.30262444593258714,
+        "grad_norm": 0.4805321991443634,
+        "learning_rate": 0.0005,
+        "loss": 1.2325,
+        "step": 9490
+      },
+      {
+        "epoch": 0.30294333365222104,
+        "grad_norm": 0.4893067479133606,
+        "learning_rate": 0.0005,
+        "loss": 1.2322,
+        "step": 9500
+      },
+      {
+        "epoch": 0.303262221371855,
+        "grad_norm": 0.4834955036640167,
+        "learning_rate": 0.0005,
+        "loss": 1.2159,
+        "step": 9510
+      },
+      {
+        "epoch": 0.3035811090914889,
+        "grad_norm": 0.4895293414592743,
+        "learning_rate": 0.0005,
+        "loss": 1.2223,
+        "step": 9520
+      },
+      {
+        "epoch": 0.3038999968111228,
+        "grad_norm": 0.4753477871417999,
+        "learning_rate": 0.0005,
+        "loss": 1.2063,
+        "step": 9530
+      },
+      {
+        "epoch": 0.30421888453075674,
+        "grad_norm": 0.47976627945899963,
+        "learning_rate": 0.0005,
+        "loss": 1.2204,
+        "step": 9540
+      },
+      {
+        "epoch": 0.30453777225039064,
+        "grad_norm": 0.4710116386413574,
+        "learning_rate": 0.0005,
+        "loss": 1.1968,
+        "step": 9550
+      },
+      {
+        "epoch": 0.30485665997002453,
+        "grad_norm": 0.46347519755363464,
+        "learning_rate": 0.0005,
+        "loss": 1.2224,
+        "step": 9560
+      },
+      {
+        "epoch": 0.3051755476896585,
+        "grad_norm": 0.4651038646697998,
+        "learning_rate": 0.0005,
+        "loss": 1.2198,
+        "step": 9570
+      },
+      {
+        "epoch": 0.3054944354092924,
+        "grad_norm": 0.4524023234844208,
+        "learning_rate": 0.0005,
+        "loss": 1.236,
+        "step": 9580
+      },
+      {
+        "epoch": 0.3058133231289263,
+        "grad_norm": 0.4605560600757599,
+        "learning_rate": 0.0005,
+        "loss": 1.222,
+        "step": 9590
+      },
+      {
+        "epoch": 0.30613221084856024,
+        "grad_norm": 0.5155693888664246,
+        "learning_rate": 0.0005,
+        "loss": 1.2167,
+        "step": 9600
+      },
+      {
+        "epoch": 0.30645109856819414,
+        "grad_norm": 0.48311251401901245,
+        "learning_rate": 0.0005,
+        "loss": 1.2207,
+        "step": 9610
+      },
+      {
+        "epoch": 0.30676998628782803,
+        "grad_norm": 0.47260165214538574,
+        "learning_rate": 0.0005,
+        "loss": 1.2305,
+        "step": 9620
+      },
+      {
+        "epoch": 0.307088874007462,
+        "grad_norm": 0.4718555510044098,
+        "learning_rate": 0.0005,
+        "loss": 1.1954,
+        "step": 9630
+      },
+      {
+        "epoch": 0.3074077617270959,
+        "grad_norm": 0.46655380725860596,
+        "learning_rate": 0.0005,
+        "loss": 1.2258,
+        "step": 9640
+      },
+      {
+        "epoch": 0.3077266494467298,
+        "grad_norm": 0.4614108204841614,
+        "learning_rate": 0.0005,
+        "loss": 1.2205,
+        "step": 9650
+      },
+      {
+        "epoch": 0.30804553716636374,
+        "grad_norm": 0.45910724997520447,
+        "learning_rate": 0.0005,
+        "loss": 1.2105,
+        "step": 9660
+      },
+      {
+        "epoch": 0.30836442488599763,
+        "grad_norm": 0.47271105647087097,
+        "learning_rate": 0.0005,
+        "loss": 1.2095,
+        "step": 9670
+      },
+      {
+        "epoch": 0.30868331260563153,
+        "grad_norm": 0.45809072256088257,
+        "learning_rate": 0.0005,
+        "loss": 1.2192,
+        "step": 9680
+      },
+      {
+        "epoch": 0.3090022003252655,
+        "grad_norm": 0.4718875288963318,
+        "learning_rate": 0.0005,
+        "loss": 1.2293,
+        "step": 9690
+      },
+      {
+        "epoch": 0.3093210880448994,
+        "grad_norm": 0.44941264390945435,
+        "learning_rate": 0.0005,
+        "loss": 1.2226,
+        "step": 9700
+      },
+      {
+        "epoch": 0.3096399757645333,
+        "grad_norm": 0.46634870767593384,
+        "learning_rate": 0.0005,
+        "loss": 1.2337,
+        "step": 9710
+      },
+      {
+        "epoch": 0.30995886348416724,
+        "grad_norm": 0.4822189211845398,
+        "learning_rate": 0.0005,
+        "loss": 1.2324,
+        "step": 9720
+      },
+      {
+        "epoch": 0.31027775120380113,
+        "grad_norm": 0.47028934955596924,
+        "learning_rate": 0.0005,
+        "loss": 1.2305,
+        "step": 9730
+      },
+      {
+        "epoch": 0.31059663892343503,
+        "grad_norm": 0.48364537954330444,
+        "learning_rate": 0.0005,
+        "loss": 1.2244,
+        "step": 9740
+      },
+      {
+        "epoch": 0.310915526643069,
+        "grad_norm": 0.469500333070755,
+        "learning_rate": 0.0005,
+        "loss": 1.2086,
+        "step": 9750
+      },
+      {
+        "epoch": 0.3112344143627029,
+        "grad_norm": 0.44105562567710876,
+        "learning_rate": 0.0005,
+        "loss": 1.222,
+        "step": 9760
+      },
+      {
+        "epoch": 0.31155330208233684,
+        "grad_norm": 0.4658164978027344,
+        "learning_rate": 0.0005,
+        "loss": 1.2396,
+        "step": 9770
+      },
+      {
+        "epoch": 0.31187218980197073,
+        "grad_norm": 0.46329885721206665,
+        "learning_rate": 0.0005,
+        "loss": 1.203,
+        "step": 9780
+      },
+      {
+        "epoch": 0.31219107752160463,
+        "grad_norm": 0.46826115250587463,
+        "learning_rate": 0.0005,
+        "loss": 1.2125,
+        "step": 9790
+      },
+      {
+        "epoch": 0.3125099652412386,
+        "grad_norm": 0.45736852288246155,
+        "learning_rate": 0.0005,
+        "loss": 1.2275,
+        "step": 9800
+      },
+      {
+        "epoch": 0.3128288529608725,
+        "grad_norm": 0.4739522635936737,
+        "learning_rate": 0.0005,
+        "loss": 1.2263,
+        "step": 9810
+      },
+      {
+        "epoch": 0.3131477406805064,
+        "grad_norm": 0.4789111316204071,
+        "learning_rate": 0.0005,
+        "loss": 1.2277,
+        "step": 9820
+      },
+      {
+        "epoch": 0.31346662840014033,
+        "grad_norm": 0.46404367685317993,
+        "learning_rate": 0.0005,
+        "loss": 1.2127,
+        "step": 9830
+      },
+      {
+        "epoch": 0.31378551611977423,
+        "grad_norm": 0.4566570520401001,
+        "learning_rate": 0.0005,
+        "loss": 1.2174,
+        "step": 9840
+      },
+      {
+        "epoch": 0.31410440383940813,
+        "grad_norm": 0.46813470125198364,
+        "learning_rate": 0.0005,
+        "loss": 1.1996,
+        "step": 9850
+      },
+      {
+        "epoch": 0.3144232915590421,
+        "grad_norm": 0.4678925573825836,
+        "learning_rate": 0.0005,
+        "loss": 1.2285,
+        "step": 9860
+      },
+      {
+        "epoch": 0.314742179278676,
+        "grad_norm": 0.4692615866661072,
+        "learning_rate": 0.0005,
+        "loss": 1.2266,
+        "step": 9870
+      },
+      {
+        "epoch": 0.3150610669983099,
+        "grad_norm": 0.46353766322135925,
+        "learning_rate": 0.0005,
+        "loss": 1.2248,
+        "step": 9880
+      },
+      {
+        "epoch": 0.31537995471794383,
+        "grad_norm": 0.4622415006160736,
+        "learning_rate": 0.0005,
+        "loss": 1.2079,
+        "step": 9890
+      },
+      {
+        "epoch": 0.31569884243757773,
+        "grad_norm": 0.44016581773757935,
+        "learning_rate": 0.0005,
+        "loss": 1.2117,
+        "step": 9900
+      },
+      {
+        "epoch": 0.31601773015721163,
+        "grad_norm": 0.44085219502449036,
+        "learning_rate": 0.0005,
+        "loss": 1.2008,
+        "step": 9910
+      },
+      {
+        "epoch": 0.3163366178768456,
+        "grad_norm": 0.473407119512558,
+        "learning_rate": 0.0005,
+        "loss": 1.2106,
+        "step": 9920
+      },
+      {
+        "epoch": 0.3166555055964795,
+        "grad_norm": 0.4463644027709961,
+        "learning_rate": 0.0005,
+        "loss": 1.2212,
+        "step": 9930
+      },
+      {
+        "epoch": 0.3169743933161134,
+        "grad_norm": 0.47022417187690735,
+        "learning_rate": 0.0005,
+        "loss": 1.2217,
+        "step": 9940
+      },
+      {
+        "epoch": 0.31729328103574733,
+        "grad_norm": 0.45228978991508484,
+        "learning_rate": 0.0005,
+        "loss": 1.2263,
+        "step": 9950
+      },
+      {
+        "epoch": 0.31761216875538123,
+        "grad_norm": 0.45226162672042847,
+        "learning_rate": 0.0005,
+        "loss": 1.2165,
+        "step": 9960
+      },
+      {
+        "epoch": 0.31793105647501513,
+        "grad_norm": 0.46043747663497925,
+        "learning_rate": 0.0005,
+        "loss": 1.2378,
+        "step": 9970
+      },
+      {
+        "epoch": 0.3182499441946491,
+        "grad_norm": 0.4622530937194824,
+        "learning_rate": 0.0005,
+        "loss": 1.2129,
+        "step": 9980
+      },
+      {
+        "epoch": 0.318568831914283,
+        "grad_norm": 0.4723272919654846,
+        "learning_rate": 0.0005,
+        "loss": 1.2202,
+        "step": 9990
+      },
+      {
+        "epoch": 0.3188877196339169,
+        "grad_norm": 0.4637812077999115,
+        "learning_rate": 0.0005,
+        "loss": 1.2209,
+        "step": 10000
+      },
+      {
+        "epoch": 0.31920660735355083,
+        "grad_norm": 0.4611336886882782,
+        "learning_rate": 0.0005,
+        "loss": 1.2244,
+        "step": 10010
+      },
+      {
+        "epoch": 0.31952549507318473,
+        "grad_norm": 0.47000154852867126,
+        "learning_rate": 0.0005,
+        "loss": 1.2047,
+        "step": 10020
+      },
+      {
+        "epoch": 0.3198443827928186,
+        "grad_norm": 0.451267272233963,
+        "learning_rate": 0.0005,
+        "loss": 1.2125,
+        "step": 10030
+      },
+      {
+        "epoch": 0.3201632705124526,
+        "grad_norm": 0.4722524881362915,
+        "learning_rate": 0.0005,
+        "loss": 1.2124,
+        "step": 10040
+      },
+      {
+        "epoch": 0.3204821582320865,
+        "grad_norm": 0.48461806774139404,
+        "learning_rate": 0.0005,
+        "loss": 1.2339,
+        "step": 10050
+      },
+      {
+        "epoch": 0.3208010459517204,
+        "grad_norm": 0.4575521647930145,
+        "learning_rate": 0.0005,
+        "loss": 1.2186,
+        "step": 10060
+      },
+      {
+        "epoch": 0.32111993367135433,
+        "grad_norm": 0.4630667269229889,
+        "learning_rate": 0.0005,
+        "loss": 1.2015,
+        "step": 10070
+      },
+      {
+        "epoch": 0.3214388213909882,
+        "grad_norm": 0.452785462141037,
+        "learning_rate": 0.0005,
+        "loss": 1.2178,
+        "step": 10080
+      },
+      {
+        "epoch": 0.3217577091106221,
+        "grad_norm": 0.4500238001346588,
+        "learning_rate": 0.0005,
+        "loss": 1.2138,
+        "step": 10090
+      },
+      {
+        "epoch": 0.3220765968302561,
+        "grad_norm": 0.46511325240135193,
+        "learning_rate": 0.0005,
+        "loss": 1.2234,
+        "step": 10100
+      },
+      {
+        "epoch": 0.32239548454989,
+        "grad_norm": 0.4618381857872009,
+        "learning_rate": 0.0005,
+        "loss": 1.2198,
+        "step": 10110
+      },
+      {
+        "epoch": 0.3227143722695239,
+        "grad_norm": 0.4588926434516907,
+        "learning_rate": 0.0005,
+        "loss": 1.1924,
+        "step": 10120
+      },
+      {
+        "epoch": 0.32303325998915783,
+        "grad_norm": 0.49134373664855957,
+        "learning_rate": 0.0005,
+        "loss": 1.2022,
+        "step": 10130
+      },
+      {
+        "epoch": 0.3233521477087917,
+        "grad_norm": 0.4511118233203888,
+        "learning_rate": 0.0005,
+        "loss": 1.2077,
+        "step": 10140
+      },
+      {
+        "epoch": 0.3236710354284256,
+        "grad_norm": 0.45254966616630554,
+        "learning_rate": 0.0005,
+        "loss": 1.219,
+        "step": 10150
+      },
+      {
+        "epoch": 0.3239899231480596,
+        "grad_norm": 0.4560336470603943,
+        "learning_rate": 0.0005,
+        "loss": 1.2097,
+        "step": 10160
+      },
+      {
+        "epoch": 0.3243088108676935,
+        "grad_norm": 0.467703640460968,
+        "learning_rate": 0.0005,
+        "loss": 1.2075,
+        "step": 10170
+      },
+      {
+        "epoch": 0.32462769858732743,
+        "grad_norm": 0.46083834767341614,
+        "learning_rate": 0.0005,
+        "loss": 1.2012,
+        "step": 10180
+      },
+      {
+        "epoch": 0.3249465863069613,
+        "grad_norm": 0.5108883380889893,
+        "learning_rate": 0.0005,
+        "loss": 1.1997,
+        "step": 10190
+      },
+      {
+        "epoch": 0.3252654740265952,
+        "grad_norm": 0.46303167939186096,
+        "learning_rate": 0.0005,
+        "loss": 1.209,
+        "step": 10200
+      },
+      {
+        "epoch": 0.3255843617462292,
+        "grad_norm": 0.46116575598716736,
+        "learning_rate": 0.0005,
+        "loss": 1.2078,
+        "step": 10210
+      },
+      {
+        "epoch": 0.3259032494658631,
+        "grad_norm": 0.46415525674819946,
+        "learning_rate": 0.0005,
+        "loss": 1.2107,
+        "step": 10220
+      },
+      {
+        "epoch": 0.326222137185497,
+        "grad_norm": 0.4536409378051758,
+        "learning_rate": 0.0005,
+        "loss": 1.2397,
+        "step": 10230
+      },
+      {
+        "epoch": 0.3265410249051309,
+        "grad_norm": 0.4473925530910492,
+        "learning_rate": 0.0005,
+        "loss": 1.2301,
+        "step": 10240
+      },
+      {
+        "epoch": 0.3268599126247648,
+        "grad_norm": 0.46480435132980347,
+        "learning_rate": 0.0005,
+        "loss": 1.185,
+        "step": 10250
+      },
+      {
+        "epoch": 0.3271788003443987,
+        "grad_norm": 0.4542619287967682,
+        "learning_rate": 0.0005,
+        "loss": 1.2121,
+        "step": 10260
+      },
+      {
+        "epoch": 0.3274976880640327,
+        "grad_norm": 0.44392162561416626,
+        "learning_rate": 0.0005,
+        "loss": 1.2085,
+        "step": 10270
+      },
+      {
+        "epoch": 0.3278165757836666,
+        "grad_norm": 0.465130478143692,
+        "learning_rate": 0.0005,
+        "loss": 1.2149,
+        "step": 10280
+      },
+      {
+        "epoch": 0.3281354635033005,
+        "grad_norm": 0.4558584988117218,
+        "learning_rate": 0.0005,
+        "loss": 1.2166,
+        "step": 10290
+      },
+      {
+        "epoch": 0.3284543512229344,
+        "grad_norm": 0.45904284715652466,
+        "learning_rate": 0.0005,
+        "loss": 1.2063,
+        "step": 10300
+      },
+      {
+        "epoch": 0.3287732389425683,
+        "grad_norm": 0.48073023557662964,
+        "learning_rate": 0.0005,
+        "loss": 1.2313,
+        "step": 10310
+      },
+      {
+        "epoch": 0.3290921266622022,
+        "grad_norm": 0.4648429751396179,
+        "learning_rate": 0.0005,
+        "loss": 1.2006,
+        "step": 10320
+      },
+      {
+        "epoch": 0.3294110143818362,
+        "grad_norm": 0.4585430920124054,
+        "learning_rate": 0.0005,
+        "loss": 1.2345,
+        "step": 10330
+      },
+      {
+        "epoch": 0.3297299021014701,
+        "grad_norm": 0.46392467617988586,
+        "learning_rate": 0.0005,
+        "loss": 1.2242,
+        "step": 10340
+      },
+      {
+        "epoch": 0.33004878982110397,
+        "grad_norm": 0.4722246527671814,
+        "learning_rate": 0.0005,
+        "loss": 1.1988,
+        "step": 10350
+      },
+      {
+        "epoch": 0.3303676775407379,
+        "grad_norm": 0.49293071031570435,
+        "learning_rate": 0.0005,
+        "loss": 1.2072,
+        "step": 10360
+      },
+      {
+        "epoch": 0.3306865652603718,
+        "grad_norm": 0.4574640393257141,
+        "learning_rate": 0.0005,
+        "loss": 1.2284,
+        "step": 10370
+      },
+      {
+        "epoch": 0.3310054529800057,
+        "grad_norm": 0.4533388614654541,
+        "learning_rate": 0.0005,
+        "loss": 1.2003,
+        "step": 10380
+      },
+      {
+        "epoch": 0.3313243406996397,
+        "grad_norm": 0.461215078830719,
+        "learning_rate": 0.0005,
+        "loss": 1.1933,
+        "step": 10390
+      },
+      {
+        "epoch": 0.3316432284192736,
+        "grad_norm": 0.4396399259567261,
+        "learning_rate": 0.0005,
+        "loss": 1.2125,
+        "step": 10400
+      },
+      {
+        "epoch": 0.33196211613890747,
+        "grad_norm": 0.439049631357193,
+        "learning_rate": 0.0005,
+        "loss": 1.1929,
+        "step": 10410
+      },
+      {
+        "epoch": 0.3322810038585414,
+        "grad_norm": 0.46211275458335876,
+        "learning_rate": 0.0005,
+        "loss": 1.2061,
+        "step": 10420
+      },
+      {
+        "epoch": 0.3325998915781753,
+        "grad_norm": 0.4407380223274231,
+        "learning_rate": 0.0005,
+        "loss": 1.1955,
+        "step": 10430
+      },
+      {
+        "epoch": 0.3329187792978092,
+        "grad_norm": 0.45931512117385864,
+        "learning_rate": 0.0005,
+        "loss": 1.2269,
+        "step": 10440
+      },
+      {
+        "epoch": 0.3332376670174432,
+        "grad_norm": 0.4495917558670044,
+        "learning_rate": 0.0005,
+        "loss": 1.202,
+        "step": 10450
+      },
+      {
+        "epoch": 0.33355655473707707,
+        "grad_norm": 0.4619958698749542,
+        "learning_rate": 0.0005,
+        "loss": 1.1961,
+        "step": 10460
+      },
+      {
+        "epoch": 0.33387544245671097,
+        "grad_norm": 0.4782272279262543,
+        "learning_rate": 0.0005,
+        "loss": 1.217,
+        "step": 10470
+      },
+      {
+        "epoch": 0.3341943301763449,
+        "grad_norm": 0.4582221210002899,
+        "learning_rate": 0.0005,
+        "loss": 1.2012,
+        "step": 10480
+      },
+      {
+        "epoch": 0.3345132178959788,
+        "grad_norm": 0.44622093439102173,
+        "learning_rate": 0.0005,
+        "loss": 1.1942,
+        "step": 10490
+      },
+      {
+        "epoch": 0.3348321056156127,
+        "grad_norm": 0.4802570343017578,
+        "learning_rate": 0.0005,
+        "loss": 1.2005,
+        "step": 10500
+      },
+      {
+        "epoch": 0.33515099333524667,
+        "grad_norm": 0.46790650486946106,
+        "learning_rate": 0.0005,
+        "loss": 1.2128,
+        "step": 10510
+      },
+      {
+        "epoch": 0.33546988105488057,
+        "grad_norm": 0.45268651843070984,
+        "learning_rate": 0.0005,
+        "loss": 1.2211,
+        "step": 10520
+      },
+      {
+        "epoch": 0.33578876877451447,
+        "grad_norm": 0.45402753353118896,
+        "learning_rate": 0.0005,
+        "loss": 1.2109,
+        "step": 10530
+      },
+      {
+        "epoch": 0.3361076564941484,
+        "grad_norm": 0.4552929401397705,
+        "learning_rate": 0.0005,
+        "loss": 1.1862,
+        "step": 10540
+      },
+      {
+        "epoch": 0.3364265442137823,
+        "grad_norm": 0.4635280668735504,
+        "learning_rate": 0.0005,
+        "loss": 1.1947,
+        "step": 10550
+      },
+      {
+        "epoch": 0.3367454319334162,
+        "grad_norm": 0.4471782147884369,
+        "learning_rate": 0.0005,
+        "loss": 1.2033,
+        "step": 10560
+      },
+      {
+        "epoch": 0.33706431965305017,
+        "grad_norm": 0.46447694301605225,
+        "learning_rate": 0.0005,
+        "loss": 1.192,
+        "step": 10570
+      },
+      {
+        "epoch": 0.33738320737268407,
+        "grad_norm": 0.4564967453479767,
+        "learning_rate": 0.0005,
+        "loss": 1.2017,
+        "step": 10580
+      },
+      {
+        "epoch": 0.337702095092318,
+        "grad_norm": 0.451744019985199,
+        "learning_rate": 0.0005,
+        "loss": 1.1689,
+        "step": 10590
+      },
+      {
+        "epoch": 0.3380209828119519,
+        "grad_norm": 0.46503663063049316,
+        "learning_rate": 0.0005,
+        "loss": 1.1923,
+        "step": 10600
+      },
+      {
+        "epoch": 0.3383398705315858,
+        "grad_norm": 0.45748820900917053,
+        "learning_rate": 0.0005,
+        "loss": 1.2015,
+        "step": 10610
+      },
+      {
+        "epoch": 0.33865875825121977,
+        "grad_norm": 0.43980541825294495,
+        "learning_rate": 0.0005,
+        "loss": 1.2261,
+        "step": 10620
+      },
+      {
+        "epoch": 0.33897764597085367,
+        "grad_norm": 0.4510059058666229,
+        "learning_rate": 0.0005,
+        "loss": 1.1988,
+        "step": 10630
+      },
+      {
+        "epoch": 0.33929653369048757,
+        "grad_norm": 0.45995068550109863,
+        "learning_rate": 0.0005,
+        "loss": 1.2083,
+        "step": 10640
+      },
+      {
+        "epoch": 0.3396154214101215,
+        "grad_norm": 0.44615572690963745,
+        "learning_rate": 0.0005,
+        "loss": 1.2075,
+        "step": 10650
+      },
+      {
+        "epoch": 0.3399343091297554,
+        "grad_norm": 0.4503238797187805,
+        "learning_rate": 0.0005,
+        "loss": 1.2034,
+        "step": 10660
+      },
+      {
+        "epoch": 0.3402531968493893,
+        "grad_norm": 0.43003490567207336,
+        "learning_rate": 0.0005,
+        "loss": 1.207,
+        "step": 10670
+      },
+      {
+        "epoch": 0.34057208456902327,
+        "grad_norm": 0.4652049243450165,
+        "learning_rate": 0.0005,
+        "loss": 1.2131,
+        "step": 10680
+      },
+      {
+        "epoch": 0.34089097228865717,
+        "grad_norm": 0.4483892619609833,
+        "learning_rate": 0.0005,
+        "loss": 1.2003,
+        "step": 10690
+      },
+      {
+        "epoch": 0.34120986000829107,
+        "grad_norm": 0.4728235602378845,
+        "learning_rate": 0.0005,
+        "loss": 1.2066,
+        "step": 10700
+      },
+      {
+        "epoch": 0.341528747727925,
+        "grad_norm": 0.44875046610832214,
+        "learning_rate": 0.0005,
+        "loss": 1.2013,
+        "step": 10710
+      },
+      {
+        "epoch": 0.3418476354475589,
+        "grad_norm": 0.4573599696159363,
+        "learning_rate": 0.0005,
+        "loss": 1.2079,
+        "step": 10720
+      },
+      {
+        "epoch": 0.3421665231671928,
+        "grad_norm": 0.45679569244384766,
+        "learning_rate": 0.0005,
+        "loss": 1.1825,
+        "step": 10730
+      },
+      {
+        "epoch": 0.34248541088682677,
+        "grad_norm": 0.4661482572555542,
+        "learning_rate": 0.0005,
+        "loss": 1.217,
+        "step": 10740
+      },
+      {
+        "epoch": 0.34280429860646067,
+        "grad_norm": 0.44722360372543335,
+        "learning_rate": 0.0005,
+        "loss": 1.1993,
+        "step": 10750
+      },
+      {
+        "epoch": 0.34312318632609456,
+        "grad_norm": 0.4567379653453827,
+        "learning_rate": 0.0005,
+        "loss": 1.1921,
+        "step": 10760
+      },
+      {
+        "epoch": 0.3434420740457285,
+        "grad_norm": 0.4360160231590271,
+        "learning_rate": 0.0005,
+        "loss": 1.1815,
+        "step": 10770
+      },
+      {
+        "epoch": 0.3437609617653624,
+        "grad_norm": 0.45101627707481384,
+        "learning_rate": 0.0005,
+        "loss": 1.2016,
+        "step": 10780
+      },
+      {
+        "epoch": 0.3440798494849963,
+        "grad_norm": 0.4734707474708557,
+        "learning_rate": 0.0005,
+        "loss": 1.1938,
+        "step": 10790
+      },
+      {
+        "epoch": 0.34439873720463027,
+        "grad_norm": 0.4517531096935272,
+        "learning_rate": 0.0005,
+        "loss": 1.2056,
+        "step": 10800
+      },
+      {
+        "epoch": 0.34471762492426417,
+        "grad_norm": 0.46230632066726685,
+        "learning_rate": 0.0005,
+        "loss": 1.2213,
+        "step": 10810
+      },
+      {
+        "epoch": 0.34503651264389806,
+        "grad_norm": 0.4498429000377655,
+        "learning_rate": 0.0005,
+        "loss": 1.21,
+        "step": 10820
+      },
+      {
+        "epoch": 0.345355400363532,
+        "grad_norm": 0.4520126283168793,
+        "learning_rate": 0.0005,
+        "loss": 1.196,
+        "step": 10830
+      },
+      {
+        "epoch": 0.3456742880831659,
+        "grad_norm": 0.4503188133239746,
+        "learning_rate": 0.0005,
+        "loss": 1.1855,
+        "step": 10840
+      },
+      {
+        "epoch": 0.3459931758027998,
+        "grad_norm": 0.4529699683189392,
+        "learning_rate": 0.0005,
+        "loss": 1.1873,
+        "step": 10850
+      },
+      {
+        "epoch": 0.34631206352243377,
+        "grad_norm": 0.4275098741054535,
+        "learning_rate": 0.0005,
+        "loss": 1.1872,
+        "step": 10860
+      },
+      {
+        "epoch": 0.34663095124206766,
+        "grad_norm": 0.4628719985485077,
+        "learning_rate": 0.0005,
+        "loss": 1.2048,
+        "step": 10870
+      },
+      {
+        "epoch": 0.34694983896170156,
+        "grad_norm": 0.4489217698574066,
+        "learning_rate": 0.0005,
+        "loss": 1.2196,
+        "step": 10880
+      },
+      {
+        "epoch": 0.3472687266813355,
+        "grad_norm": 0.4392339885234833,
+        "learning_rate": 0.0005,
+        "loss": 1.204,
+        "step": 10890
+      },
+      {
+        "epoch": 0.3475876144009694,
+        "grad_norm": 0.45925724506378174,
+        "learning_rate": 0.0005,
+        "loss": 1.21,
+        "step": 10900
+      },
+      {
+        "epoch": 0.3479065021206033,
+        "grad_norm": 0.46018514037132263,
+        "learning_rate": 0.0005,
+        "loss": 1.1791,
+        "step": 10910
+      },
+      {
+        "epoch": 0.34822538984023726,
+        "grad_norm": 0.46905317902565,
+        "learning_rate": 0.0005,
+        "loss": 1.204,
+        "step": 10920
+      },
+      {
+        "epoch": 0.34854427755987116,
+        "grad_norm": 0.4454002380371094,
+        "learning_rate": 0.0005,
+        "loss": 1.209,
+        "step": 10930
+      },
+      {
+        "epoch": 0.34886316527950506,
+        "grad_norm": 0.4870285987854004,
+        "learning_rate": 0.0005,
+        "loss": 1.189,
+        "step": 10940
+      },
+      {
+        "epoch": 0.349182052999139,
+        "grad_norm": 0.45176053047180176,
+        "learning_rate": 0.0005,
+        "loss": 1.2131,
+        "step": 10950
+      },
+      {
+        "epoch": 0.3495009407187729,
+        "grad_norm": 0.4306580722332001,
+        "learning_rate": 0.0005,
+        "loss": 1.1981,
+        "step": 10960
+      },
+      {
+        "epoch": 0.3498198284384068,
+        "grad_norm": 0.4551098644733429,
+        "learning_rate": 0.0005,
+        "loss": 1.2007,
+        "step": 10970
+      },
+      {
+        "epoch": 0.35013871615804076,
+        "grad_norm": 0.4566822350025177,
+        "learning_rate": 0.0005,
+        "loss": 1.2013,
+        "step": 10980
+      },
+      {
+        "epoch": 0.35045760387767466,
+        "grad_norm": 0.46287766098976135,
+        "learning_rate": 0.0005,
+        "loss": 1.1762,
+        "step": 10990
+      },
+      {
+        "epoch": 0.3507764915973086,
+        "grad_norm": 0.45999109745025635,
+        "learning_rate": 0.0005,
+        "loss": 1.196,
+        "step": 11000
+      },
+      {
+        "epoch": 0.3510953793169425,
+        "grad_norm": 0.4592169523239136,
+        "learning_rate": 0.0005,
+        "loss": 1.2074,
+        "step": 11010
+      },
+      {
+        "epoch": 0.3514142670365764,
+        "grad_norm": 0.4633693993091583,
+        "learning_rate": 0.0005,
+        "loss": 1.1902,
+        "step": 11020
+      },
+      {
+        "epoch": 0.35173315475621036,
+        "grad_norm": 0.43769994378089905,
+        "learning_rate": 0.0005,
+        "loss": 1.1725,
+        "step": 11030
+      },
+      {
+        "epoch": 0.35205204247584426,
+        "grad_norm": 0.4633389413356781,
+        "learning_rate": 0.0005,
+        "loss": 1.1902,
+        "step": 11040
+      },
+      {
+        "epoch": 0.35237093019547816,
+        "grad_norm": 0.4522418975830078,
+        "learning_rate": 0.0005,
+        "loss": 1.1719,
+        "step": 11050
+      },
+      {
+        "epoch": 0.3526898179151121,
+        "grad_norm": 0.4321923553943634,
+        "learning_rate": 0.0005,
+        "loss": 1.1958,
+        "step": 11060
+      },
+      {
+        "epoch": 0.353008705634746,
+        "grad_norm": 0.4340358376502991,
+        "learning_rate": 0.0005,
+        "loss": 1.1893,
+        "step": 11070
+      },
+      {
+        "epoch": 0.3533275933543799,
+        "grad_norm": 0.4427970051765442,
+        "learning_rate": 0.0005,
+        "loss": 1.184,
+        "step": 11080
+      },
+      {
+        "epoch": 0.35364648107401386,
+        "grad_norm": 0.4384293854236603,
+        "learning_rate": 0.0005,
+        "loss": 1.174,
+        "step": 11090
+      },
+      {
+        "epoch": 0.35396536879364776,
+        "grad_norm": 0.45408880710601807,
+        "learning_rate": 0.0005,
+        "loss": 1.177,
+        "step": 11100
+      },
+      {
+        "epoch": 0.35428425651328166,
+        "grad_norm": 0.4425502121448517,
+        "learning_rate": 0.0005,
+        "loss": 1.2027,
+        "step": 11110
+      },
+      {
+        "epoch": 0.3546031442329156,
+        "grad_norm": 0.4402698576450348,
+        "learning_rate": 0.0005,
+        "loss": 1.2067,
+        "step": 11120
+      },
+      {
+        "epoch": 0.3549220319525495,
+        "grad_norm": 0.45279955863952637,
+        "learning_rate": 0.0005,
+        "loss": 1.2,
+        "step": 11130
+      },
+      {
+        "epoch": 0.3552409196721834,
+        "grad_norm": 0.4827028214931488,
+        "learning_rate": 0.0005,
+        "loss": 1.1948,
+        "step": 11140
+      },
+      {
+        "epoch": 0.35555980739181736,
+        "grad_norm": 0.4452892243862152,
+        "learning_rate": 0.0005,
+        "loss": 1.1916,
+        "step": 11150
+      },
+      {
+        "epoch": 0.35587869511145126,
+        "grad_norm": 0.4474916160106659,
+        "learning_rate": 0.0005,
+        "loss": 1.199,
+        "step": 11160
+      },
+      {
+        "epoch": 0.35619758283108516,
+        "grad_norm": 0.43313753604888916,
+        "learning_rate": 0.0005,
+        "loss": 1.1972,
+        "step": 11170
+      },
+      {
+        "epoch": 0.3565164705507191,
+        "grad_norm": 0.4603961110115051,
+        "learning_rate": 0.0005,
+        "loss": 1.1911,
+        "step": 11180
+      },
+      {
+        "epoch": 0.356835358270353,
+        "grad_norm": 0.4489038586616516,
+        "learning_rate": 0.0005,
+        "loss": 1.1858,
+        "step": 11190
+      },
+      {
+        "epoch": 0.3571542459899869,
+        "grad_norm": 0.4428690969944,
+        "learning_rate": 0.0005,
+        "loss": 1.1666,
+        "step": 11200
+      },
+      {
+        "epoch": 0.35747313370962086,
+        "grad_norm": 0.43662405014038086,
+        "learning_rate": 0.0005,
+        "loss": 1.2024,
+        "step": 11210
+      },
+      {
+        "epoch": 0.35779202142925476,
+        "grad_norm": 0.42232009768486023,
+        "learning_rate": 0.0005,
+        "loss": 1.1872,
+        "step": 11220
+      },
+      {
+        "epoch": 0.35811090914888866,
+        "grad_norm": 0.44732868671417236,
+        "learning_rate": 0.0005,
+        "loss": 1.2081,
+        "step": 11230
+      },
+      {
+        "epoch": 0.3584297968685226,
+        "grad_norm": 0.44722238183021545,
+        "learning_rate": 0.0005,
+        "loss": 1.2073,
+        "step": 11240
+      },
+      {
+        "epoch": 0.3587486845881565,
+        "grad_norm": 0.46360519528388977,
+        "learning_rate": 0.0005,
+        "loss": 1.1852,
+        "step": 11250
+      },
+      {
+        "epoch": 0.3590675723077904,
+        "grad_norm": 0.46256133913993835,
+        "learning_rate": 0.0005,
+        "loss": 1.1761,
+        "step": 11260
+      },
+      {
+        "epoch": 0.35938646002742436,
+        "grad_norm": 0.4595744013786316,
+        "learning_rate": 0.0005,
+        "loss": 1.2061,
+        "step": 11270
+      },
+      {
+        "epoch": 0.35970534774705826,
+        "grad_norm": 0.4376492500305176,
+        "learning_rate": 0.0005,
+        "loss": 1.175,
+        "step": 11280
+      },
+      {
+        "epoch": 0.36002423546669216,
+        "grad_norm": 0.47340521216392517,
+        "learning_rate": 0.0005,
+        "loss": 1.1821,
+        "step": 11290
+      },
+      {
+        "epoch": 0.3603431231863261,
+        "grad_norm": 0.47642752528190613,
+        "learning_rate": 0.0005,
+        "loss": 1.1941,
+        "step": 11300
+      },
+      {
+        "epoch": 0.36066201090596,
+        "grad_norm": 0.44139501452445984,
+        "learning_rate": 0.0005,
+        "loss": 1.1818,
+        "step": 11310
+      },
+      {
+        "epoch": 0.3609808986255939,
+        "grad_norm": 0.4428239166736603,
+        "learning_rate": 0.0005,
+        "loss": 1.178,
+        "step": 11320
+      },
+      {
+        "epoch": 0.36129978634522786,
+        "grad_norm": 0.4435330331325531,
+        "learning_rate": 0.0005,
+        "loss": 1.1774,
+        "step": 11330
+      },
+      {
+        "epoch": 0.36161867406486176,
+        "grad_norm": 0.4315539300441742,
+        "learning_rate": 0.0005,
+        "loss": 1.2015,
+        "step": 11340
+      },
+      {
+        "epoch": 0.36193756178449565,
+        "grad_norm": 0.4380136728286743,
+        "learning_rate": 0.0005,
+        "loss": 1.1874,
+        "step": 11350
+      },
+      {
+        "epoch": 0.3622564495041296,
+        "grad_norm": 0.44613757729530334,
+        "learning_rate": 0.0005,
+        "loss": 1.1948,
+        "step": 11360
+      },
+      {
+        "epoch": 0.3625753372237635,
+        "grad_norm": 0.439974844455719,
+        "learning_rate": 0.0005,
+        "loss": 1.1864,
+        "step": 11370
+      },
+      {
+        "epoch": 0.3628942249433974,
+        "grad_norm": 0.4617900848388672,
+        "learning_rate": 0.0005,
+        "loss": 1.185,
+        "step": 11380
+      },
+      {
+        "epoch": 0.36321311266303136,
+        "grad_norm": 0.4472865164279938,
+        "learning_rate": 0.0005,
+        "loss": 1.1888,
+        "step": 11390
+      },
+      {
+        "epoch": 0.36353200038266525,
+        "grad_norm": 0.44490471482276917,
+        "learning_rate": 0.0005,
+        "loss": 1.1955,
+        "step": 11400
+      },
+      {
+        "epoch": 0.3638508881022992,
+        "grad_norm": 0.43914979696273804,
+        "learning_rate": 0.0005,
+        "loss": 1.1781,
+        "step": 11410
+      },
+      {
+        "epoch": 0.3641697758219331,
+        "grad_norm": 0.4476803243160248,
+        "learning_rate": 0.0005,
+        "loss": 1.1889,
+        "step": 11420
+      },
+      {
+        "epoch": 0.364488663541567,
+        "grad_norm": 0.438223659992218,
+        "learning_rate": 0.0005,
+        "loss": 1.1659,
+        "step": 11430
+      },
+      {
+        "epoch": 0.36480755126120096,
+        "grad_norm": 0.4376436173915863,
+        "learning_rate": 0.0005,
+        "loss": 1.1903,
+        "step": 11440
+      },
+      {
+        "epoch": 0.36512643898083486,
+        "grad_norm": 0.44571664929389954,
+        "learning_rate": 0.0005,
+        "loss": 1.188,
+        "step": 11450
+      },
+      {
+        "epoch": 0.36544532670046875,
+        "grad_norm": 0.44138994812965393,
+        "learning_rate": 0.0005,
+        "loss": 1.171,
+        "step": 11460
+      },
+      {
+        "epoch": 0.3657642144201027,
+        "grad_norm": 0.458737313747406,
+        "learning_rate": 0.0005,
+        "loss": 1.1912,
+        "step": 11470
+      },
+      {
+        "epoch": 0.3660831021397366,
+        "grad_norm": 0.4489744007587433,
+        "learning_rate": 0.0005,
+        "loss": 1.1821,
+        "step": 11480
+      },
+      {
+        "epoch": 0.3664019898593705,
+        "grad_norm": 0.45018211007118225,
+        "learning_rate": 0.0005,
+        "loss": 1.1903,
+        "step": 11490
+      },
+      {
+        "epoch": 0.36672087757900446,
+        "grad_norm": 0.45186272263526917,
+        "learning_rate": 0.0005,
+        "loss": 1.2084,
+        "step": 11500
+      },
+      {
+        "epoch": 0.36703976529863835,
+        "grad_norm": 0.44593289494514465,
+        "learning_rate": 0.0005,
+        "loss": 1.1764,
+        "step": 11510
+      },
+      {
+        "epoch": 0.36735865301827225,
+        "grad_norm": 0.45831361413002014,
+        "learning_rate": 0.0005,
+        "loss": 1.2031,
+        "step": 11520
+      },
+      {
+        "epoch": 0.3676775407379062,
+        "grad_norm": 0.44360455870628357,
+        "learning_rate": 0.0005,
+        "loss": 1.1727,
+        "step": 11530
+      },
+      {
+        "epoch": 0.3679964284575401,
+        "grad_norm": 0.4307442009449005,
+        "learning_rate": 0.0005,
+        "loss": 1.1638,
+        "step": 11540
+      },
+      {
+        "epoch": 0.368315316177174,
+        "grad_norm": 0.4518921971321106,
+        "learning_rate": 0.0005,
+        "loss": 1.1875,
+        "step": 11550
+      },
+      {
+        "epoch": 0.36863420389680795,
+        "grad_norm": 0.43437421321868896,
+        "learning_rate": 0.0005,
+        "loss": 1.1774,
+        "step": 11560
+      },
+      {
+        "epoch": 0.36895309161644185,
+        "grad_norm": 0.4409539997577667,
+        "learning_rate": 0.0005,
+        "loss": 1.1971,
+        "step": 11570
+      },
+      {
+        "epoch": 0.36927197933607575,
+        "grad_norm": 0.43100041151046753,
+        "learning_rate": 0.0005,
+        "loss": 1.1633,
+        "step": 11580
+      },
+      {
+        "epoch": 0.3695908670557097,
+        "grad_norm": 0.441154807806015,
+        "learning_rate": 0.0005,
+        "loss": 1.1709,
+        "step": 11590
+      },
+      {
+        "epoch": 0.3699097547753436,
+        "grad_norm": 0.44322308897972107,
+        "learning_rate": 0.0005,
+        "loss": 1.2069,
+        "step": 11600
+      },
+      {
+        "epoch": 0.3702286424949775,
+        "grad_norm": 0.4248104989528656,
+        "learning_rate": 0.0005,
+        "loss": 1.1893,
+        "step": 11610
+      },
+      {
+        "epoch": 0.37054753021461145,
+        "grad_norm": 0.46101078391075134,
+        "learning_rate": 0.0005,
+        "loss": 1.1997,
+        "step": 11620
+      },
+      {
+        "epoch": 0.37086641793424535,
+        "grad_norm": 0.4464070200920105,
+        "learning_rate": 0.0005,
+        "loss": 1.1887,
+        "step": 11630
+      },
+      {
+        "epoch": 0.37118530565387925,
+        "grad_norm": 0.4598291218280792,
+        "learning_rate": 0.0005,
+        "loss": 1.2104,
+        "step": 11640
+      },
+      {
+        "epoch": 0.3715041933735132,
+        "grad_norm": 0.40784159302711487,
+        "learning_rate": 0.0005,
+        "loss": 1.1796,
+        "step": 11650
+      },
+      {
+        "epoch": 0.3718230810931471,
+        "grad_norm": 0.4431672692298889,
+        "learning_rate": 0.0005,
+        "loss": 1.2027,
+        "step": 11660
+      },
+      {
+        "epoch": 0.372141968812781,
+        "grad_norm": 0.45555850863456726,
+        "learning_rate": 0.0005,
+        "loss": 1.2067,
+        "step": 11670
+      },
+      {
+        "epoch": 0.37246085653241495,
+        "grad_norm": 0.43322280049324036,
+        "learning_rate": 0.0005,
+        "loss": 1.1851,
+        "step": 11680
+      },
+      {
+        "epoch": 0.37277974425204885,
+        "grad_norm": 0.4348566234111786,
+        "learning_rate": 0.0005,
+        "loss": 1.1799,
+        "step": 11690
+      },
+      {
+        "epoch": 0.37309863197168275,
+        "grad_norm": 0.42040494084358215,
+        "learning_rate": 0.0005,
+        "loss": 1.1667,
+        "step": 11700
+      },
+      {
+        "epoch": 0.3734175196913167,
+        "grad_norm": 0.4358799159526825,
+        "learning_rate": 0.0005,
+        "loss": 1.1818,
+        "step": 11710
+      },
+      {
+        "epoch": 0.3737364074109506,
+        "grad_norm": 0.4317675828933716,
+        "learning_rate": 0.0005,
+        "loss": 1.1811,
+        "step": 11720
+      },
+      {
+        "epoch": 0.3740552951305845,
+        "grad_norm": 0.4507165849208832,
+        "learning_rate": 0.0005,
+        "loss": 1.1913,
+        "step": 11730
+      },
+      {
+        "epoch": 0.37437418285021845,
+        "grad_norm": 0.43800291419029236,
+        "learning_rate": 0.0005,
+        "loss": 1.1965,
+        "step": 11740
+      },
+      {
+        "epoch": 0.37469307056985235,
+        "grad_norm": 0.45497605204582214,
+        "learning_rate": 0.0005,
+        "loss": 1.2025,
+        "step": 11750
+      },
+      {
+        "epoch": 0.37501195828948625,
+        "grad_norm": 0.4357222020626068,
+        "learning_rate": 0.0005,
+        "loss": 1.1949,
+        "step": 11760
+      },
+      {
+        "epoch": 0.3753308460091202,
+        "grad_norm": 0.46163061261177063,
+        "learning_rate": 0.0005,
+        "loss": 1.1907,
+        "step": 11770
+      },
+      {
+        "epoch": 0.3756497337287541,
+        "grad_norm": 0.43153810501098633,
+        "learning_rate": 0.0005,
+        "loss": 1.1813,
+        "step": 11780
+      },
+      {
+        "epoch": 0.375968621448388,
+        "grad_norm": 0.4308339059352875,
+        "learning_rate": 0.0005,
+        "loss": 1.2098,
+        "step": 11790
+      },
+      {
+        "epoch": 0.37628750916802195,
+        "grad_norm": 0.4210801422595978,
+        "learning_rate": 0.0005,
+        "loss": 1.1825,
+        "step": 11800
+      },
+      {
+        "epoch": 0.37660639688765585,
+        "grad_norm": 0.4504418671131134,
+        "learning_rate": 0.0005,
+        "loss": 1.1906,
+        "step": 11810
+      },
+      {
+        "epoch": 0.37692528460728975,
+        "grad_norm": 0.42673376202583313,
+        "learning_rate": 0.0005,
+        "loss": 1.1925,
+        "step": 11820
+      },
+      {
+        "epoch": 0.3772441723269237,
+        "grad_norm": 0.42398470640182495,
+        "learning_rate": 0.0005,
+        "loss": 1.1663,
+        "step": 11830
+      },
+      {
+        "epoch": 0.3775630600465576,
+        "grad_norm": 0.437023401260376,
+        "learning_rate": 0.0005,
+        "loss": 1.1899,
+        "step": 11840
+      },
+      {
+        "epoch": 0.37788194776619155,
+        "grad_norm": 0.4427235722541809,
+        "learning_rate": 0.0005,
+        "loss": 1.1861,
+        "step": 11850
+      },
+      {
+        "epoch": 0.37820083548582545,
+        "grad_norm": 0.44343510270118713,
+        "learning_rate": 0.0005,
+        "loss": 1.1963,
+        "step": 11860
+      },
+      {
+        "epoch": 0.37851972320545935,
+        "grad_norm": 0.43557801842689514,
+        "learning_rate": 0.0005,
+        "loss": 1.1906,
+        "step": 11870
+      },
+      {
+        "epoch": 0.3788386109250933,
+        "grad_norm": 0.43895554542541504,
+        "learning_rate": 0.0005,
+        "loss": 1.1911,
+        "step": 11880
+      },
+      {
+        "epoch": 0.3791574986447272,
+        "grad_norm": 0.4184170961380005,
+        "learning_rate": 0.0005,
+        "loss": 1.1792,
+        "step": 11890
+      },
+      {
+        "epoch": 0.3794763863643611,
+        "grad_norm": 0.4252275824546814,
+        "learning_rate": 0.0005,
+        "loss": 1.1738,
+        "step": 11900
+      },
+      {
+        "epoch": 0.37979527408399505,
+        "grad_norm": 0.43294379115104675,
+        "learning_rate": 0.0005,
+        "loss": 1.1898,
+        "step": 11910
+      },
+      {
+        "epoch": 0.38011416180362895,
+        "grad_norm": 0.4419662654399872,
+        "learning_rate": 0.0005,
+        "loss": 1.1968,
+        "step": 11920
+      },
+      {
+        "epoch": 0.38043304952326285,
+        "grad_norm": 0.42845383286476135,
+        "learning_rate": 0.0005,
+        "loss": 1.1602,
+        "step": 11930
+      },
+      {
+        "epoch": 0.3807519372428968,
+        "grad_norm": 0.4342953860759735,
+        "learning_rate": 0.0005,
+        "loss": 1.1538,
+        "step": 11940
+      },
+      {
+        "epoch": 0.3810708249625307,
+        "grad_norm": 0.4351995885372162,
+        "learning_rate": 0.0005,
+        "loss": 1.177,
+        "step": 11950
+      },
+      {
+        "epoch": 0.3813897126821646,
+        "grad_norm": 0.43881794810295105,
+        "learning_rate": 0.0005,
+        "loss": 1.1927,
+        "step": 11960
+      },
+      {
+        "epoch": 0.38170860040179855,
+        "grad_norm": 0.4251026213169098,
+        "learning_rate": 0.0005,
+        "loss": 1.1852,
+        "step": 11970
+      },
+      {
+        "epoch": 0.38202748812143245,
+        "grad_norm": 0.4276167154312134,
+        "learning_rate": 0.0005,
+        "loss": 1.166,
+        "step": 11980
+      },
+      {
+        "epoch": 0.38234637584106634,
+        "grad_norm": 0.44587334990501404,
+        "learning_rate": 0.0005,
+        "loss": 1.1749,
+        "step": 11990
+      },
+      {
+        "epoch": 0.3826652635607003,
+        "grad_norm": 0.46067747473716736,
+        "learning_rate": 0.0005,
+        "loss": 1.1795,
+        "step": 12000
+      },
+      {
+        "epoch": 0.3829841512803342,
+        "grad_norm": 0.43104439973831177,
+        "learning_rate": 0.0005,
+        "loss": 1.1913,
+        "step": 12010
+      },
+      {
+        "epoch": 0.3833030389999681,
+        "grad_norm": 0.44496071338653564,
+        "learning_rate": 0.0005,
+        "loss": 1.2024,
+        "step": 12020
+      },
+      {
+        "epoch": 0.38362192671960205,
+        "grad_norm": 0.4469740390777588,
+        "learning_rate": 0.0005,
+        "loss": 1.177,
+        "step": 12030
+      },
+      {
+        "epoch": 0.38394081443923594,
+        "grad_norm": 0.4275093078613281,
+        "learning_rate": 0.0005,
+        "loss": 1.1733,
+        "step": 12040
+      },
+      {
+        "epoch": 0.38425970215886984,
+        "grad_norm": 0.4296175539493561,
+        "learning_rate": 0.0005,
+        "loss": 1.1801,
+        "step": 12050
+      },
+      {
+        "epoch": 0.3845785898785038,
+        "grad_norm": 0.43391743302345276,
+        "learning_rate": 0.0005,
+        "loss": 1.1688,
+        "step": 12060
+      },
+      {
+        "epoch": 0.3848974775981377,
+        "grad_norm": 0.4239729940891266,
+        "learning_rate": 0.0005,
+        "loss": 1.1883,
+        "step": 12070
+      },
+      {
+        "epoch": 0.3852163653177716,
+        "grad_norm": 0.4424794614315033,
+        "learning_rate": 0.0005,
+        "loss": 1.1819,
+        "step": 12080
+      },
+      {
+        "epoch": 0.38553525303740555,
+        "grad_norm": 0.4417058229446411,
+        "learning_rate": 0.0005,
+        "loss": 1.1742,
+        "step": 12090
+      },
+      {
+        "epoch": 0.38585414075703944,
+        "grad_norm": 0.43083709478378296,
+        "learning_rate": 0.0005,
+        "loss": 1.1843,
+        "step": 12100
+      },
+      {
+        "epoch": 0.38617302847667334,
+        "grad_norm": 0.434842973947525,
+        "learning_rate": 0.0005,
+        "loss": 1.1875,
+        "step": 12110
+      },
+      {
+        "epoch": 0.3864919161963073,
+        "grad_norm": 0.4427316188812256,
+        "learning_rate": 0.0005,
+        "loss": 1.1912,
+        "step": 12120
+      },
+      {
+        "epoch": 0.3868108039159412,
+        "grad_norm": 0.44422024488449097,
+        "learning_rate": 0.0005,
+        "loss": 1.1893,
+        "step": 12130
+      },
+      {
+        "epoch": 0.3871296916355751,
+        "grad_norm": 0.42151498794555664,
+        "learning_rate": 0.0005,
+        "loss": 1.1565,
+        "step": 12140
+      },
+      {
+        "epoch": 0.38744857935520904,
+        "grad_norm": 0.42576828598976135,
+        "learning_rate": 0.0005,
+        "loss": 1.1992,
+        "step": 12150
+      },
+      {
+        "epoch": 0.38776746707484294,
+        "grad_norm": 0.4259600341320038,
+        "learning_rate": 0.0005,
+        "loss": 1.1968,
+        "step": 12160
+      },
+      {
+        "epoch": 0.38808635479447684,
+        "grad_norm": 0.41919681429862976,
+        "learning_rate": 0.0005,
+        "loss": 1.1566,
+        "step": 12170
+      },
+      {
+        "epoch": 0.3884052425141108,
+        "grad_norm": 0.43111515045166016,
+        "learning_rate": 0.0005,
+        "loss": 1.1983,
+        "step": 12180
+      },
+      {
+        "epoch": 0.3887241302337447,
+        "grad_norm": 0.4664030373096466,
+        "learning_rate": 0.0005,
+        "loss": 1.1738,
+        "step": 12190
+      },
+      {
+        "epoch": 0.3890430179533786,
+        "grad_norm": 0.4208368957042694,
+        "learning_rate": 0.0005,
+        "loss": 1.1711,
+        "step": 12200
+      },
+      {
+        "epoch": 0.38936190567301254,
+        "grad_norm": 0.4356335401535034,
+        "learning_rate": 0.0005,
+        "loss": 1.1787,
+        "step": 12210
+      },
+      {
+        "epoch": 0.38968079339264644,
+        "grad_norm": 0.42046454548835754,
+        "learning_rate": 0.0005,
+        "loss": 1.1818,
+        "step": 12220
+      },
+      {
+        "epoch": 0.38999968111228034,
+        "grad_norm": 0.41660618782043457,
+        "learning_rate": 0.0005,
+        "loss": 1.1674,
+        "step": 12230
+      },
+      {
+        "epoch": 0.3903185688319143,
+        "grad_norm": 0.43841058015823364,
+        "learning_rate": 0.0005,
+        "loss": 1.1689,
+        "step": 12240
+      },
+      {
+        "epoch": 0.3906374565515482,
+        "grad_norm": 0.4638073146343231,
+        "learning_rate": 0.0005,
+        "loss": 1.1595,
+        "step": 12250
+      },
+      {
+        "epoch": 0.39095634427118214,
+        "grad_norm": 0.4431755840778351,
+        "learning_rate": 0.0005,
+        "loss": 1.1745,
+        "step": 12260
+      },
+      {
+        "epoch": 0.39127523199081604,
+        "grad_norm": 0.4278421401977539,
+        "learning_rate": 0.0005,
+        "loss": 1.1848,
+        "step": 12270
+      },
+      {
+        "epoch": 0.39159411971044994,
+        "grad_norm": 0.4424417316913605,
+        "learning_rate": 0.0005,
+        "loss": 1.1735,
+        "step": 12280
+      },
+      {
+        "epoch": 0.3919130074300839,
+        "grad_norm": 0.43157055974006653,
+        "learning_rate": 0.0005,
+        "loss": 1.1764,
+        "step": 12290
+      },
+      {
+        "epoch": 0.3922318951497178,
+        "grad_norm": 0.43115106225013733,
+        "learning_rate": 0.0005,
+        "loss": 1.1909,
+        "step": 12300
+      },
+      {
+        "epoch": 0.3925507828693517,
+        "grad_norm": 0.4346446394920349,
+        "learning_rate": 0.0005,
+        "loss": 1.1777,
+        "step": 12310
+      },
+      {
+        "epoch": 0.39286967058898564,
+        "grad_norm": 0.43510472774505615,
+        "learning_rate": 0.0005,
+        "loss": 1.1889,
+        "step": 12320
+      },
+      {
+        "epoch": 0.39318855830861954,
+        "grad_norm": 0.4409896433353424,
+        "learning_rate": 0.0005,
+        "loss": 1.187,
+        "step": 12330
+      },
+      {
+        "epoch": 0.39350744602825344,
+        "grad_norm": 0.4372926354408264,
+        "learning_rate": 0.0005,
+        "loss": 1.1625,
+        "step": 12340
+      },
+      {
+        "epoch": 0.3938263337478874,
+        "grad_norm": 0.4218306839466095,
+        "learning_rate": 0.0005,
+        "loss": 1.1609,
+        "step": 12350
+      },
+      {
+        "epoch": 0.3941452214675213,
+        "grad_norm": 0.4451087415218353,
+        "learning_rate": 0.0005,
+        "loss": 1.2059,
+        "step": 12360
+      },
+      {
+        "epoch": 0.3944641091871552,
+        "grad_norm": 0.4136224687099457,
+        "learning_rate": 0.0005,
+        "loss": 1.1685,
+        "step": 12370
+      },
+      {
+        "epoch": 0.39478299690678914,
+        "grad_norm": 0.4264117479324341,
+        "learning_rate": 0.0005,
+        "loss": 1.1674,
+        "step": 12380
+      },
+      {
+        "epoch": 0.39510188462642304,
+        "grad_norm": 0.45373213291168213,
+        "learning_rate": 0.0005,
+        "loss": 1.1692,
+        "step": 12390
+      },
+      {
+        "epoch": 0.39542077234605694,
+        "grad_norm": 0.4304463565349579,
+        "learning_rate": 0.0005,
+        "loss": 1.1847,
+        "step": 12400
+      },
+      {
+        "epoch": 0.3957396600656909,
+        "grad_norm": 0.4379657506942749,
+        "learning_rate": 0.0005,
+        "loss": 1.1655,
+        "step": 12410
+      },
+      {
+        "epoch": 0.3960585477853248,
+        "grad_norm": 0.4436817169189453,
+        "learning_rate": 0.0005,
+        "loss": 1.1784,
+        "step": 12420
+      },
+      {
+        "epoch": 0.3963774355049587,
+        "grad_norm": 0.43348872661590576,
+        "learning_rate": 0.0005,
+        "loss": 1.1819,
+        "step": 12430
+      },
+      {
+        "epoch": 0.39669632322459264,
+        "grad_norm": 0.42717644572257996,
+        "learning_rate": 0.0005,
+        "loss": 1.1569,
+        "step": 12440
+      },
+      {
+        "epoch": 0.39701521094422654,
+        "grad_norm": 0.4594727158546448,
+        "learning_rate": 0.0005,
+        "loss": 1.1825,
+        "step": 12450
+      },
+      {
+        "epoch": 0.39733409866386044,
+        "grad_norm": 0.41570910811424255,
+        "learning_rate": 0.0005,
+        "loss": 1.1639,
+        "step": 12460
+      },
+      {
+        "epoch": 0.3976529863834944,
+        "grad_norm": 0.4306057095527649,
+        "learning_rate": 0.0005,
+        "loss": 1.1904,
+        "step": 12470
+      },
+      {
+        "epoch": 0.3979718741031283,
+        "grad_norm": 0.4283396005630493,
+        "learning_rate": 0.0005,
+        "loss": 1.1881,
+        "step": 12480
+      },
+      {
+        "epoch": 0.3982907618227622,
+        "grad_norm": 0.4277869164943695,
+        "learning_rate": 0.0005,
+        "loss": 1.1743,
+        "step": 12490
+      },
+      {
+        "epoch": 0.39860964954239614,
+        "grad_norm": 0.4295739233493805,
+        "learning_rate": 0.0005,
+        "loss": 1.1919,
+        "step": 12500
+      },
+      {
+        "epoch": 0.39892853726203004,
+        "grad_norm": 0.4237912893295288,
+        "learning_rate": 0.0005,
+        "loss": 1.1662,
+        "step": 12510
+      },
+      {
+        "epoch": 0.39924742498166393,
+        "grad_norm": 0.42816078662872314,
+        "learning_rate": 0.0005,
+        "loss": 1.1684,
+        "step": 12520
+      },
+      {
+        "epoch": 0.3995663127012979,
+        "grad_norm": 0.420441210269928,
+        "learning_rate": 0.0005,
+        "loss": 1.1875,
+        "step": 12530
+      },
+      {
+        "epoch": 0.3998852004209318,
+        "grad_norm": 0.4237239956855774,
+        "learning_rate": 0.0005,
+        "loss": 1.1732,
+        "step": 12540
+      },
+      {
+        "epoch": 0.4002040881405657,
+        "grad_norm": 0.43103668093681335,
+        "learning_rate": 0.0005,
+        "loss": 1.1517,
+        "step": 12550
+      },
+      {
+        "epoch": 0.40052297586019964,
+        "grad_norm": 0.42024657130241394,
+        "learning_rate": 0.0005,
+        "loss": 1.168,
+        "step": 12560
+      },
+      {
+        "epoch": 0.40084186357983353,
+        "grad_norm": 0.43330562114715576,
+        "learning_rate": 0.0005,
+        "loss": 1.1971,
+        "step": 12570
+      },
+      {
+        "epoch": 0.40116075129946743,
+        "grad_norm": 0.4399755895137787,
+        "learning_rate": 0.0005,
+        "loss": 1.1461,
+        "step": 12580
+      },
+      {
+        "epoch": 0.4014796390191014,
+        "grad_norm": 0.4258183538913727,
+        "learning_rate": 0.0005,
+        "loss": 1.1975,
+        "step": 12590
+      },
+      {
+        "epoch": 0.4017985267387353,
+        "grad_norm": 0.4347604513168335,
+        "learning_rate": 0.0005,
+        "loss": 1.1687,
+        "step": 12600
+      },
+      {
+        "epoch": 0.4021174144583692,
+        "grad_norm": 0.424478679895401,
+        "learning_rate": 0.0005,
+        "loss": 1.1737,
+        "step": 12610
+      },
+      {
+        "epoch": 0.40243630217800314,
+        "grad_norm": 0.42958155274391174,
+        "learning_rate": 0.0005,
+        "loss": 1.1814,
+        "step": 12620
+      },
+      {
+        "epoch": 0.40275518989763703,
+        "grad_norm": 0.426096111536026,
+        "learning_rate": 0.0005,
+        "loss": 1.1575,
+        "step": 12630
+      },
+      {
+        "epoch": 0.40307407761727093,
+        "grad_norm": 0.4363478422164917,
+        "learning_rate": 0.0005,
+        "loss": 1.156,
+        "step": 12640
+      },
+      {
+        "epoch": 0.4033929653369049,
+        "grad_norm": 0.41518527269363403,
+        "learning_rate": 0.0005,
+        "loss": 1.1786,
+        "step": 12650
+      },
+      {
+        "epoch": 0.4037118530565388,
+        "grad_norm": 0.42658382654190063,
+        "learning_rate": 0.0005,
+        "loss": 1.1673,
+        "step": 12660
+      },
+      {
+        "epoch": 0.40403074077617274,
+        "grad_norm": 0.42411237955093384,
+        "learning_rate": 0.0005,
+        "loss": 1.1663,
+        "step": 12670
+      },
+      {
+        "epoch": 0.40434962849580663,
+        "grad_norm": 0.4455260634422302,
+        "learning_rate": 0.0005,
+        "loss": 1.1758,
+        "step": 12680
+      },
+      {
+        "epoch": 0.40466851621544053,
+        "grad_norm": 0.428243488073349,
+        "learning_rate": 0.0005,
+        "loss": 1.1588,
+        "step": 12690
+      },
+      {
+        "epoch": 0.4049874039350745,
+        "grad_norm": 0.4175855815410614,
+        "learning_rate": 0.0005,
+        "loss": 1.1768,
+        "step": 12700
+      },
+      {
+        "epoch": 0.4053062916547084,
+        "grad_norm": 0.42678576707839966,
+        "learning_rate": 0.0005,
+        "loss": 1.1582,
+        "step": 12710
+      },
+      {
+        "epoch": 0.4056251793743423,
+        "grad_norm": 0.44132933020591736,
+        "learning_rate": 0.0005,
+        "loss": 1.1794,
+        "step": 12720
+      },
+      {
+        "epoch": 0.40594406709397624,
+        "grad_norm": 0.44395214319229126,
+        "learning_rate": 0.0005,
+        "loss": 1.1632,
+        "step": 12730
+      },
+      {
+        "epoch": 0.40626295481361013,
+        "grad_norm": 0.44196462631225586,
+        "learning_rate": 0.0005,
+        "loss": 1.1761,
+        "step": 12740
+      },
+      {
+        "epoch": 0.40658184253324403,
+        "grad_norm": 0.4145953059196472,
+        "learning_rate": 0.0005,
+        "loss": 1.1646,
+        "step": 12750
+      },
+      {
+        "epoch": 0.406900730252878,
+        "grad_norm": 0.43945613503456116,
+        "learning_rate": 0.0005,
+        "loss": 1.1682,
+        "step": 12760
+      },
+      {
+        "epoch": 0.4072196179725119,
+        "grad_norm": 0.42901989817619324,
+        "learning_rate": 0.0005,
+        "loss": 1.1838,
+        "step": 12770
+      },
+      {
+        "epoch": 0.4075385056921458,
+        "grad_norm": 0.44105032086372375,
+        "learning_rate": 0.0005,
+        "loss": 1.1648,
+        "step": 12780
+      },
+      {
+        "epoch": 0.40785739341177973,
+        "grad_norm": 0.4337145686149597,
+        "learning_rate": 0.0005,
+        "loss": 1.1676,
+        "step": 12790
+      },
+      {
+        "epoch": 0.40817628113141363,
+        "grad_norm": 0.46507975459098816,
+        "learning_rate": 0.0005,
+        "loss": 1.1839,
+        "step": 12800
+      },
+      {
+        "epoch": 0.40849516885104753,
+        "grad_norm": 0.43042483925819397,
+        "learning_rate": 0.0005,
+        "loss": 1.161,
+        "step": 12810
+      },
+      {
+        "epoch": 0.4088140565706815,
+        "grad_norm": 0.42969202995300293,
+        "learning_rate": 0.0005,
+        "loss": 1.1644,
+        "step": 12820
+      },
+      {
+        "epoch": 0.4091329442903154,
+        "grad_norm": 0.43208038806915283,
+        "learning_rate": 0.0005,
+        "loss": 1.1608,
+        "step": 12830
+      },
+      {
+        "epoch": 0.4094518320099493,
+        "grad_norm": 0.4246330261230469,
+        "learning_rate": 0.0005,
+        "loss": 1.1395,
+        "step": 12840
+      },
+      {
+        "epoch": 0.40977071972958323,
+        "grad_norm": 0.4385770857334137,
+        "learning_rate": 0.0005,
+        "loss": 1.1721,
+        "step": 12850
+      },
+      {
+        "epoch": 0.41008960744921713,
+        "grad_norm": 0.40635982155799866,
+        "learning_rate": 0.0005,
+        "loss": 1.1554,
+        "step": 12860
+      },
+      {
+        "epoch": 0.41040849516885103,
+        "grad_norm": 0.41713228821754456,
+        "learning_rate": 0.0005,
+        "loss": 1.1593,
+        "step": 12870
+      },
+      {
+        "epoch": 0.410727382888485,
+        "grad_norm": 0.4230583608150482,
+        "learning_rate": 0.0005,
+        "loss": 1.172,
+        "step": 12880
+      },
+      {
+        "epoch": 0.4110462706081189,
+        "grad_norm": 0.44719910621643066,
+        "learning_rate": 0.0005,
+        "loss": 1.1767,
+        "step": 12890
+      },
+      {
+        "epoch": 0.4113651583277528,
+        "grad_norm": 0.42063602805137634,
+        "learning_rate": 0.0005,
+        "loss": 1.1557,
+        "step": 12900
+      },
+      {
+        "epoch": 0.41168404604738673,
+        "grad_norm": 0.41372743248939514,
+        "learning_rate": 0.0005,
+        "loss": 1.1572,
+        "step": 12910
+      },
+      {
+        "epoch": 0.41200293376702063,
+        "grad_norm": 0.4369075298309326,
+        "learning_rate": 0.0005,
+        "loss": 1.165,
+        "step": 12920
+      },
+      {
+        "epoch": 0.4123218214866545,
+        "grad_norm": 0.43609854578971863,
+        "learning_rate": 0.0005,
+        "loss": 1.1567,
+        "step": 12930
+      },
+      {
+        "epoch": 0.4126407092062885,
+        "grad_norm": 0.41011208295822144,
+        "learning_rate": 0.0005,
+        "loss": 1.1786,
+        "step": 12940
+      },
+      {
+        "epoch": 0.4129595969259224,
+        "grad_norm": 0.43973419070243835,
+        "learning_rate": 0.0005,
+        "loss": 1.1755,
+        "step": 12950
+      },
+      {
+        "epoch": 0.4132784846455563,
+        "grad_norm": 0.4214501678943634,
+        "learning_rate": 0.0005,
+        "loss": 1.1744,
+        "step": 12960
+      },
+      {
+        "epoch": 0.41359737236519023,
+        "grad_norm": 0.42284494638442993,
+        "learning_rate": 0.0005,
+        "loss": 1.1742,
+        "step": 12970
+      },
+      {
+        "epoch": 0.41391626008482413,
+        "grad_norm": 0.43984121084213257,
+        "learning_rate": 0.0005,
+        "loss": 1.1808,
+        "step": 12980
+      },
+      {
+        "epoch": 0.414235147804458,
+        "grad_norm": 0.4616177976131439,
+        "learning_rate": 0.0005,
+        "loss": 1.1607,
+        "step": 12990
+      },
+      {
+        "epoch": 0.414554035524092,
+        "grad_norm": 0.420259565114975,
+        "learning_rate": 0.0005,
+        "loss": 1.1679,
+        "step": 13000
+      },
+      {
+        "epoch": 0.4148729232437259,
+        "grad_norm": 0.4331679046154022,
+        "learning_rate": 0.0005,
+        "loss": 1.1488,
+        "step": 13010
+      },
+      {
+        "epoch": 0.4151918109633598,
+        "grad_norm": 0.42250847816467285,
+        "learning_rate": 0.0005,
+        "loss": 1.1873,
+        "step": 13020
+      },
+      {
+        "epoch": 0.41551069868299373,
+        "grad_norm": 0.45015767216682434,
+        "learning_rate": 0.0005,
+        "loss": 1.1737,
+        "step": 13030
+      },
+      {
+        "epoch": 0.4158295864026276,
+        "grad_norm": 0.4316963255405426,
+        "learning_rate": 0.0005,
+        "loss": 1.1608,
+        "step": 13040
+      },
+      {
+        "epoch": 0.4161484741222615,
+        "grad_norm": 0.4264684021472931,
+        "learning_rate": 0.0005,
+        "loss": 1.1651,
+        "step": 13050
+      },
+      {
+        "epoch": 0.4164673618418955,
+        "grad_norm": 0.44561827182769775,
+        "learning_rate": 0.0005,
+        "loss": 1.1813,
+        "step": 13060
+      },
+      {
+        "epoch": 0.4167862495615294,
+        "grad_norm": 0.42165082693099976,
+        "learning_rate": 0.0005,
+        "loss": 1.167,
+        "step": 13070
+      },
+      {
+        "epoch": 0.41710513728116333,
+        "grad_norm": 0.4280150532722473,
+        "learning_rate": 0.0005,
+        "loss": 1.1726,
+        "step": 13080
+      },
+      {
+        "epoch": 0.4174240250007972,
+        "grad_norm": 0.43624866008758545,
+        "learning_rate": 0.0005,
+        "loss": 1.1699,
+        "step": 13090
+      },
+      {
+        "epoch": 0.4177429127204311,
+        "grad_norm": 0.429521381855011,
+        "learning_rate": 0.0005,
+        "loss": 1.159,
+        "step": 13100
+      },
+      {
+        "epoch": 0.4180618004400651,
+        "grad_norm": 0.4358668625354767,
+        "learning_rate": 0.0005,
+        "loss": 1.1704,
+        "step": 13110
+      },
+      {
+        "epoch": 0.418380688159699,
+        "grad_norm": 0.42937684059143066,
+        "learning_rate": 0.0005,
+        "loss": 1.1778,
+        "step": 13120
+      },
+      {
+        "epoch": 0.4186995758793329,
+        "grad_norm": 0.4294481575489044,
+        "learning_rate": 0.0005,
+        "loss": 1.1718,
+        "step": 13130
+      },
+      {
+        "epoch": 0.41901846359896683,
+        "grad_norm": 0.42481085658073425,
+        "learning_rate": 0.0005,
+        "loss": 1.1659,
+        "step": 13140
+      },
+      {
+        "epoch": 0.4193373513186007,
+        "grad_norm": 0.41510918736457825,
+        "learning_rate": 0.0005,
+        "loss": 1.1796,
+        "step": 13150
+      },
+      {
+        "epoch": 0.4196562390382346,
+        "grad_norm": 0.43382757902145386,
+        "learning_rate": 0.0005,
+        "loss": 1.1604,
+        "step": 13160
+      },
+      {
+        "epoch": 0.4199751267578686,
+        "grad_norm": 0.4185143709182739,
+        "learning_rate": 0.0005,
+        "loss": 1.1658,
+        "step": 13170
+      },
+      {
+        "epoch": 0.4202940144775025,
+        "grad_norm": 0.4401332437992096,
+        "learning_rate": 0.0005,
+        "loss": 1.1515,
+        "step": 13180
+      },
+      {
+        "epoch": 0.4206129021971364,
+        "grad_norm": 0.4322357773780823,
+        "learning_rate": 0.0005,
+        "loss": 1.1659,
+        "step": 13190
+      },
+      {
+        "epoch": 0.4209317899167703,
+        "grad_norm": 0.42740195989608765,
+        "learning_rate": 0.0005,
+        "loss": 1.1863,
+        "step": 13200
+      },
+      {
+        "epoch": 0.4212506776364042,
+        "grad_norm": 0.4310891032218933,
+        "learning_rate": 0.0005,
+        "loss": 1.1719,
+        "step": 13210
+      },
+      {
+        "epoch": 0.4215695653560381,
+        "grad_norm": 0.44095975160598755,
+        "learning_rate": 0.0005,
+        "loss": 1.1635,
+        "step": 13220
+      },
+      {
+        "epoch": 0.4218884530756721,
+        "grad_norm": 0.4069562554359436,
+        "learning_rate": 0.0005,
+        "loss": 1.1734,
+        "step": 13230
+      },
+      {
+        "epoch": 0.422207340795306,
+        "grad_norm": 0.42801713943481445,
+        "learning_rate": 0.0005,
+        "loss": 1.1558,
+        "step": 13240
+      },
+      {
+        "epoch": 0.42252622851493987,
+        "grad_norm": 0.4137992262840271,
+        "learning_rate": 0.0005,
+        "loss": 1.1467,
+        "step": 13250
+      },
+      {
+        "epoch": 0.4228451162345738,
+        "grad_norm": 0.42337340116500854,
+        "learning_rate": 0.0005,
+        "loss": 1.1541,
+        "step": 13260
+      },
+      {
+        "epoch": 0.4231640039542077,
+        "grad_norm": 0.4387548863887787,
+        "learning_rate": 0.0005,
+        "loss": 1.1738,
+        "step": 13270
+      },
+      {
+        "epoch": 0.4234828916738416,
+        "grad_norm": 0.423705130815506,
+        "learning_rate": 0.0005,
+        "loss": 1.1468,
+        "step": 13280
+      },
+      {
+        "epoch": 0.4238017793934756,
+        "grad_norm": 0.418472021818161,
+        "learning_rate": 0.0005,
+        "loss": 1.1587,
+        "step": 13290
+      },
+      {
+        "epoch": 0.4241206671131095,
+        "grad_norm": 0.4126065671443939,
+        "learning_rate": 0.0005,
+        "loss": 1.179,
+        "step": 13300
+      },
+      {
+        "epoch": 0.42443955483274337,
+        "grad_norm": 0.42028361558914185,
+        "learning_rate": 0.0005,
+        "loss": 1.1812,
+        "step": 13310
+      },
+      {
+        "epoch": 0.4247584425523773,
+        "grad_norm": 0.4209747612476349,
+        "learning_rate": 0.0005,
+        "loss": 1.1574,
+        "step": 13320
+      },
+      {
+        "epoch": 0.4250773302720112,
+        "grad_norm": 0.4240000545978546,
+        "learning_rate": 0.0005,
+        "loss": 1.1689,
+        "step": 13330
+      },
+      {
+        "epoch": 0.4253962179916451,
+        "grad_norm": 0.42597630620002747,
+        "learning_rate": 0.0005,
+        "loss": 1.1598,
+        "step": 13340
+      },
+      {
+        "epoch": 0.4257151057112791,
+        "grad_norm": 0.43621891736984253,
+        "learning_rate": 0.0005,
+        "loss": 1.1601,
+        "step": 13350
+      },
+      {
+        "epoch": 0.42603399343091297,
+        "grad_norm": 0.44976115226745605,
+        "learning_rate": 0.0005,
+        "loss": 1.1533,
+        "step": 13360
+      },
+      {
+        "epoch": 0.42635288115054687,
+        "grad_norm": 0.425903856754303,
+        "learning_rate": 0.0005,
+        "loss": 1.1671,
+        "step": 13370
+      },
+      {
+        "epoch": 0.4266717688701808,
+        "grad_norm": 0.41458699107170105,
+        "learning_rate": 0.0005,
+        "loss": 1.1479,
+        "step": 13380
+      },
+      {
+        "epoch": 0.4269906565898147,
+        "grad_norm": 0.4458717703819275,
+        "learning_rate": 0.0005,
+        "loss": 1.1533,
+        "step": 13390
+      },
+      {
+        "epoch": 0.4273095443094486,
+        "grad_norm": 0.4254353642463684,
+        "learning_rate": 0.0005,
+        "loss": 1.1509,
+        "step": 13400
+      },
+      {
+        "epoch": 0.4276284320290826,
+        "grad_norm": 0.4332568347454071,
+        "learning_rate": 0.0005,
+        "loss": 1.1423,
+        "step": 13410
+      },
+      {
+        "epoch": 0.42794731974871647,
+        "grad_norm": 0.4396291971206665,
+        "learning_rate": 0.0005,
+        "loss": 1.175,
+        "step": 13420
+      },
+      {
+        "epoch": 0.42826620746835037,
+        "grad_norm": 0.41766542196273804,
+        "learning_rate": 0.0005,
+        "loss": 1.1629,
+        "step": 13430
+      },
+      {
+        "epoch": 0.4285850951879843,
+        "grad_norm": 0.44327792525291443,
+        "learning_rate": 0.0005,
+        "loss": 1.1706,
+        "step": 13440
+      },
+      {
+        "epoch": 0.4289039829076182,
+        "grad_norm": 0.4321768283843994,
+        "learning_rate": 0.0005,
+        "loss": 1.1568,
+        "step": 13450
+      },
+      {
+        "epoch": 0.4292228706272521,
+        "grad_norm": 0.4065990746021271,
+        "learning_rate": 0.0005,
+        "loss": 1.1606,
+        "step": 13460
+      },
+      {
+        "epoch": 0.42954175834688607,
+        "grad_norm": 0.41537582874298096,
+        "learning_rate": 0.0005,
+        "loss": 1.1647,
+        "step": 13470
+      },
+      {
+        "epoch": 0.42986064606651997,
+        "grad_norm": 0.4257226586341858,
+        "learning_rate": 0.0005,
+        "loss": 1.1768,
+        "step": 13480
+      },
+      {
+        "epoch": 0.4301795337861539,
+        "grad_norm": 0.3980241119861603,
+        "learning_rate": 0.0005,
+        "loss": 1.154,
+        "step": 13490
+      },
+      {
+        "epoch": 0.4304984215057878,
+        "grad_norm": 0.42201271653175354,
+        "learning_rate": 0.0005,
+        "loss": 1.1632,
+        "step": 13500
+      },
+      {
+        "epoch": 0.4308173092254217,
+        "grad_norm": 0.41972705721855164,
+        "learning_rate": 0.0005,
+        "loss": 1.152,
+        "step": 13510
+      },
+      {
+        "epoch": 0.43113619694505567,
+        "grad_norm": 0.4092382788658142,
+        "learning_rate": 0.0005,
+        "loss": 1.1834,
+        "step": 13520
+      },
+      {
+        "epoch": 0.43145508466468957,
+        "grad_norm": 0.41219362616539,
+        "learning_rate": 0.0005,
+        "loss": 1.1528,
+        "step": 13530
+      },
+      {
+        "epoch": 0.43177397238432347,
+        "grad_norm": 0.41763296723365784,
+        "learning_rate": 0.0005,
+        "loss": 1.1608,
+        "step": 13540
+      },
+      {
+        "epoch": 0.4320928601039574,
+        "grad_norm": 0.4221011698246002,
+        "learning_rate": 0.0005,
+        "loss": 1.1705,
+        "step": 13550
+      },
+      {
+        "epoch": 0.4324117478235913,
+        "grad_norm": 0.4206273853778839,
+        "learning_rate": 0.0005,
+        "loss": 1.1581,
+        "step": 13560
+      },
+      {
+        "epoch": 0.4327306355432252,
+        "grad_norm": 0.42883896827697754,
+        "learning_rate": 0.0005,
+        "loss": 1.1612,
+        "step": 13570
+      },
+      {
+        "epoch": 0.43304952326285917,
+        "grad_norm": 0.4313361942768097,
+        "learning_rate": 0.0005,
+        "loss": 1.1689,
+        "step": 13580
+      },
+      {
+        "epoch": 0.43336841098249307,
+        "grad_norm": 0.41933777928352356,
+        "learning_rate": 0.0005,
+        "loss": 1.1562,
+        "step": 13590
+      },
+      {
+        "epoch": 0.43368729870212697,
+        "grad_norm": 0.44052955508232117,
+        "learning_rate": 0.0005,
+        "loss": 1.1537,
+        "step": 13600
+      },
+      {
+        "epoch": 0.4340061864217609,
+        "grad_norm": 0.4265698194503784,
+        "learning_rate": 0.0005,
+        "loss": 1.1595,
+        "step": 13610
+      },
+      {
+        "epoch": 0.4343250741413948,
+        "grad_norm": 0.4011492133140564,
+        "learning_rate": 0.0005,
+        "loss": 1.1793,
+        "step": 13620
+      },
+      {
+        "epoch": 0.4346439618610287,
+        "grad_norm": 0.4022427499294281,
+        "learning_rate": 0.0005,
+        "loss": 1.1432,
+        "step": 13630
+      },
+      {
+        "epoch": 0.43496284958066267,
+        "grad_norm": 0.4209270775318146,
+        "learning_rate": 0.0005,
+        "loss": 1.165,
+        "step": 13640
+      },
+      {
+        "epoch": 0.43528173730029657,
+        "grad_norm": 0.4407374858856201,
+        "learning_rate": 0.0005,
+        "loss": 1.1607,
+        "step": 13650
+      },
+      {
+        "epoch": 0.43560062501993047,
+        "grad_norm": 0.4113706648349762,
+        "learning_rate": 0.0005,
+        "loss": 1.1457,
+        "step": 13660
+      },
+      {
+        "epoch": 0.4359195127395644,
+        "grad_norm": 0.42725327610969543,
+        "learning_rate": 0.0005,
+        "loss": 1.1668,
+        "step": 13670
+      },
+      {
+        "epoch": 0.4362384004591983,
+        "grad_norm": 0.42654532194137573,
+        "learning_rate": 0.0005,
+        "loss": 1.1522,
+        "step": 13680
+      },
+      {
+        "epoch": 0.4365572881788322,
+        "grad_norm": 0.41542568802833557,
+        "learning_rate": 0.0005,
+        "loss": 1.1496,
+        "step": 13690
+      },
+      {
+        "epoch": 0.43687617589846617,
+        "grad_norm": 0.4379822611808777,
+        "learning_rate": 0.0005,
+        "loss": 1.1668,
+        "step": 13700
+      },
+      {
+        "epoch": 0.43719506361810007,
+        "grad_norm": 0.42327746748924255,
+        "learning_rate": 0.0005,
+        "loss": 1.1498,
+        "step": 13710
+      },
+      {
+        "epoch": 0.43751395133773396,
+        "grad_norm": 0.41067299246788025,
+        "learning_rate": 0.0005,
+        "loss": 1.1342,
+        "step": 13720
+      },
+      {
+        "epoch": 0.4378328390573679,
+        "grad_norm": 0.4195924997329712,
+        "learning_rate": 0.0005,
+        "loss": 1.1741,
+        "step": 13730
+      },
+      {
+        "epoch": 0.4381517267770018,
+        "grad_norm": 0.41079947352409363,
+        "learning_rate": 0.0005,
+        "loss": 1.1619,
+        "step": 13740
+      },
+      {
+        "epoch": 0.4384706144966357,
+        "grad_norm": 0.41595518589019775,
+        "learning_rate": 0.0005,
+        "loss": 1.1534,
+        "step": 13750
+      },
+      {
+        "epoch": 0.43878950221626967,
+        "grad_norm": 0.4261476397514343,
+        "learning_rate": 0.0005,
+        "loss": 1.1495,
+        "step": 13760
+      },
+      {
+        "epoch": 0.43910838993590356,
+        "grad_norm": 0.4288196265697479,
+        "learning_rate": 0.0005,
+        "loss": 1.1637,
+        "step": 13770
+      },
+      {
+        "epoch": 0.43942727765553746,
+        "grad_norm": 0.44342705607414246,
+        "learning_rate": 0.0005,
+        "loss": 1.165,
+        "step": 13780
+      },
+      {
+        "epoch": 0.4397461653751714,
+        "grad_norm": 0.4130489230155945,
+        "learning_rate": 0.0005,
+        "loss": 1.1606,
+        "step": 13790
+      },
+      {
+        "epoch": 0.4400650530948053,
+        "grad_norm": 0.418384313583374,
+        "learning_rate": 0.0005,
+        "loss": 1.1607,
+        "step": 13800
+      },
+      {
+        "epoch": 0.4403839408144392,
+        "grad_norm": 0.4078846573829651,
+        "learning_rate": 0.0005,
+        "loss": 1.146,
+        "step": 13810
+      },
+      {
+        "epoch": 0.44070282853407317,
+        "grad_norm": 0.427979439496994,
+        "learning_rate": 0.0005,
+        "loss": 1.1497,
+        "step": 13820
+      },
+      {
+        "epoch": 0.44102171625370706,
+        "grad_norm": 0.4124217629432678,
+        "learning_rate": 0.0005,
+        "loss": 1.1528,
+        "step": 13830
+      },
+      {
+        "epoch": 0.44134060397334096,
+        "grad_norm": 0.42145681381225586,
+        "learning_rate": 0.0005,
+        "loss": 1.1374,
+        "step": 13840
+      },
+      {
+        "epoch": 0.4416594916929749,
+        "grad_norm": 0.41087475419044495,
+        "learning_rate": 0.0005,
+        "loss": 1.1522,
+        "step": 13850
+      },
+      {
+        "epoch": 0.4419783794126088,
+        "grad_norm": 0.42248791456222534,
+        "learning_rate": 0.0005,
+        "loss": 1.1567,
+        "step": 13860
+      },
+      {
+        "epoch": 0.4422972671322427,
+        "grad_norm": 0.4094279408454895,
+        "learning_rate": 0.0005,
+        "loss": 1.1404,
+        "step": 13870
+      },
+      {
+        "epoch": 0.44261615485187666,
+        "grad_norm": 0.4195246398448944,
+        "learning_rate": 0.0005,
+        "loss": 1.1353,
+        "step": 13880
+      },
+      {
+        "epoch": 0.44293504257151056,
+        "grad_norm": 0.39682644605636597,
+        "learning_rate": 0.0005,
+        "loss": 1.1392,
+        "step": 13890
+      },
+      {
+        "epoch": 0.4432539302911445,
+        "grad_norm": 0.4020282030105591,
+        "learning_rate": 0.0005,
+        "loss": 1.1631,
+        "step": 13900
+      },
+      {
+        "epoch": 0.4435728180107784,
+        "grad_norm": 0.4157191216945648,
+        "learning_rate": 0.0005,
+        "loss": 1.144,
+        "step": 13910
+      },
+      {
+        "epoch": 0.4438917057304123,
+        "grad_norm": 0.4178768992424011,
+        "learning_rate": 0.0005,
+        "loss": 1.1582,
+        "step": 13920
+      },
+      {
+        "epoch": 0.44421059345004626,
+        "grad_norm": 0.4138433039188385,
+        "learning_rate": 0.0005,
+        "loss": 1.1669,
+        "step": 13930
+      },
+      {
+        "epoch": 0.44452948116968016,
+        "grad_norm": 0.4343213737010956,
+        "learning_rate": 0.0005,
+        "loss": 1.1389,
+        "step": 13940
+      },
+      {
+        "epoch": 0.44484836888931406,
+        "grad_norm": 0.413717657327652,
+        "learning_rate": 0.0005,
+        "loss": 1.1524,
+        "step": 13950
+      },
+      {
+        "epoch": 0.445167256608948,
+        "grad_norm": 0.409458190202713,
+        "learning_rate": 0.0005,
+        "loss": 1.1737,
+        "step": 13960
+      },
+      {
+        "epoch": 0.4454861443285819,
+        "grad_norm": 0.4400709569454193,
+        "learning_rate": 0.0005,
+        "loss": 1.1473,
+        "step": 13970
+      },
+      {
+        "epoch": 0.4458050320482158,
+        "grad_norm": 0.4103112518787384,
+        "learning_rate": 0.0005,
+        "loss": 1.152,
+        "step": 13980
+      },
+      {
+        "epoch": 0.44612391976784976,
+        "grad_norm": 0.4030466079711914,
+        "learning_rate": 0.0005,
+        "loss": 1.1472,
+        "step": 13990
+      },
+      {
+        "epoch": 0.44644280748748366,
+        "grad_norm": 0.42554450035095215,
+        "learning_rate": 0.0005,
+        "loss": 1.1455,
+        "step": 14000
+      },
+      {
+        "epoch": 0.44676169520711756,
+        "grad_norm": 0.41038069128990173,
+        "learning_rate": 0.0005,
+        "loss": 1.1594,
+        "step": 14010
+      },
+      {
+        "epoch": 0.4470805829267515,
+        "grad_norm": 0.4226815402507782,
+        "learning_rate": 0.0005,
+        "loss": 1.1603,
+        "step": 14020
+      },
+      {
+        "epoch": 0.4473994706463854,
+        "grad_norm": 0.4165605306625366,
+        "learning_rate": 0.0005,
+        "loss": 1.1596,
+        "step": 14030
+      },
+      {
+        "epoch": 0.4477183583660193,
+        "grad_norm": 0.41343143582344055,
+        "learning_rate": 0.0005,
+        "loss": 1.1435,
+        "step": 14040
+      },
+      {
+        "epoch": 0.44803724608565326,
+        "grad_norm": 0.4170566499233246,
+        "learning_rate": 0.0005,
+        "loss": 1.1341,
+        "step": 14050
+      },
+      {
+        "epoch": 0.44835613380528716,
+        "grad_norm": 0.4268888235092163,
+        "learning_rate": 0.0005,
+        "loss": 1.143,
+        "step": 14060
+      },
+      {
+        "epoch": 0.44867502152492106,
+        "grad_norm": 0.41298648715019226,
+        "learning_rate": 0.0005,
+        "loss": 1.1344,
+        "step": 14070
+      },
+      {
+        "epoch": 0.448993909244555,
+        "grad_norm": 0.4087713658809662,
+        "learning_rate": 0.0005,
+        "loss": 1.1482,
+        "step": 14080
+      },
+      {
+        "epoch": 0.4493127969641889,
+        "grad_norm": 0.40675875544548035,
+        "learning_rate": 0.0005,
+        "loss": 1.1469,
+        "step": 14090
+      },
+      {
+        "epoch": 0.4496316846838228,
+        "grad_norm": 0.418449342250824,
+        "learning_rate": 0.0005,
+        "loss": 1.1489,
+        "step": 14100
+      },
+      {
+        "epoch": 0.44995057240345676,
+        "grad_norm": 0.42261627316474915,
+        "learning_rate": 0.0005,
+        "loss": 1.1529,
+        "step": 14110
+      },
+      {
+        "epoch": 0.45026946012309066,
+        "grad_norm": 0.40711814165115356,
+        "learning_rate": 0.0005,
+        "loss": 1.1418,
+        "step": 14120
+      },
+      {
+        "epoch": 0.45058834784272456,
+        "grad_norm": 0.407012015581131,
+        "learning_rate": 0.0005,
+        "loss": 1.1438,
+        "step": 14130
+      },
+      {
+        "epoch": 0.4509072355623585,
+        "grad_norm": 0.41482147574424744,
+        "learning_rate": 0.0005,
+        "loss": 1.14,
+        "step": 14140
+      },
+      {
+        "epoch": 0.4512261232819924,
+        "grad_norm": 0.4099246561527252,
+        "learning_rate": 0.0005,
+        "loss": 1.1732,
+        "step": 14150
+      },
+      {
+        "epoch": 0.4515450110016263,
+        "grad_norm": 0.4071347713470459,
+        "learning_rate": 0.0005,
+        "loss": 1.1456,
+        "step": 14160
+      },
+      {
+        "epoch": 0.45186389872126026,
+        "grad_norm": 0.4065478444099426,
+        "learning_rate": 0.0005,
+        "loss": 1.1378,
+        "step": 14170
+      },
+      {
+        "epoch": 0.45218278644089416,
+        "grad_norm": 0.40051156282424927,
+        "learning_rate": 0.0005,
+        "loss": 1.1486,
+        "step": 14180
+      },
+      {
+        "epoch": 0.45250167416052806,
+        "grad_norm": 0.40654802322387695,
+        "learning_rate": 0.0005,
+        "loss": 1.1315,
+        "step": 14190
+      },
+      {
+        "epoch": 0.452820561880162,
+        "grad_norm": 0.41403114795684814,
+        "learning_rate": 0.0005,
+        "loss": 1.1609,
+        "step": 14200
+      },
+      {
+        "epoch": 0.4531394495997959,
+        "grad_norm": 0.4000602066516876,
+        "learning_rate": 0.0005,
+        "loss": 1.1606,
+        "step": 14210
+      },
+      {
+        "epoch": 0.4534583373194298,
+        "grad_norm": 0.39985981583595276,
+        "learning_rate": 0.0005,
+        "loss": 1.1508,
+        "step": 14220
+      },
+      {
+        "epoch": 0.45377722503906376,
+        "grad_norm": 0.4208848774433136,
+        "learning_rate": 0.0005,
+        "loss": 1.1375,
+        "step": 14230
+      },
+      {
+        "epoch": 0.45409611275869766,
+        "grad_norm": 0.41846519708633423,
+        "learning_rate": 0.0005,
+        "loss": 1.1314,
+        "step": 14240
+      },
+      {
+        "epoch": 0.45441500047833155,
+        "grad_norm": 0.41302967071533203,
+        "learning_rate": 0.0005,
+        "loss": 1.1531,
+        "step": 14250
+      },
+      {
+        "epoch": 0.4547338881979655,
+        "grad_norm": 0.41688963770866394,
+        "learning_rate": 0.0005,
+        "loss": 1.1639,
+        "step": 14260
+      },
+      {
+        "epoch": 0.4550527759175994,
+        "grad_norm": 0.4071559011936188,
+        "learning_rate": 0.0005,
+        "loss": 1.1474,
+        "step": 14270
+      },
+      {
+        "epoch": 0.4553716636372333,
+        "grad_norm": 0.41632965207099915,
+        "learning_rate": 0.0005,
+        "loss": 1.1551,
+        "step": 14280
+      },
+      {
+        "epoch": 0.45569055135686726,
+        "grad_norm": 0.4219265878200531,
+        "learning_rate": 0.0005,
+        "loss": 1.1587,
+        "step": 14290
+      },
+      {
+        "epoch": 0.45600943907650116,
+        "grad_norm": 0.42590487003326416,
+        "learning_rate": 0.0005,
+        "loss": 1.163,
+        "step": 14300
+      },
+      {
+        "epoch": 0.45632832679613505,
+        "grad_norm": 0.4112999141216278,
+        "learning_rate": 0.0005,
+        "loss": 1.1649,
+        "step": 14310
+      },
+      {
+        "epoch": 0.456647214515769,
+        "grad_norm": 0.4211723208427429,
+        "learning_rate": 0.0005,
+        "loss": 1.1522,
+        "step": 14320
+      },
+      {
+        "epoch": 0.4569661022354029,
+        "grad_norm": 0.4247067868709564,
+        "learning_rate": 0.0005,
+        "loss": 1.145,
+        "step": 14330
+      },
+      {
+        "epoch": 0.45728498995503686,
+        "grad_norm": 0.4237869381904602,
+        "learning_rate": 0.0005,
+        "loss": 1.1372,
+        "step": 14340
+      },
+      {
+        "epoch": 0.45760387767467076,
+        "grad_norm": 0.40753382444381714,
+        "learning_rate": 0.0005,
+        "loss": 1.1495,
+        "step": 14350
+      },
+      {
+        "epoch": 0.45792276539430465,
+        "grad_norm": 0.3949471414089203,
+        "learning_rate": 0.0005,
+        "loss": 1.1419,
+        "step": 14360
+      },
+      {
+        "epoch": 0.4582416531139386,
+        "grad_norm": 0.41907766461372375,
+        "learning_rate": 0.0005,
+        "loss": 1.172,
+        "step": 14370
+      },
+      {
+        "epoch": 0.4585605408335725,
+        "grad_norm": 0.4017127752304077,
+        "learning_rate": 0.0005,
+        "loss": 1.159,
+        "step": 14380
+      },
+      {
+        "epoch": 0.4588794285532064,
+        "grad_norm": 0.4082666337490082,
+        "learning_rate": 0.0005,
+        "loss": 1.1411,
+        "step": 14390
+      },
+      {
+        "epoch": 0.45919831627284036,
+        "grad_norm": 0.4150290787220001,
+        "learning_rate": 0.0005,
+        "loss": 1.166,
+        "step": 14400
+      },
+      {
+        "epoch": 0.45951720399247425,
+        "grad_norm": 0.3951573967933655,
+        "learning_rate": 0.0005,
+        "loss": 1.1188,
+        "step": 14410
+      },
+      {
+        "epoch": 0.45983609171210815,
+        "grad_norm": 0.40606120228767395,
+        "learning_rate": 0.0005,
+        "loss": 1.1425,
+        "step": 14420
+      },
+      {
+        "epoch": 0.4601549794317421,
+        "grad_norm": 0.4190419316291809,
+        "learning_rate": 0.0005,
+        "loss": 1.1442,
+        "step": 14430
+      },
+      {
+        "epoch": 0.460473867151376,
+        "grad_norm": 0.42822325229644775,
+        "learning_rate": 0.0005,
+        "loss": 1.163,
+        "step": 14440
+      },
+      {
+        "epoch": 0.4607927548710099,
+        "grad_norm": 0.40327173471450806,
+        "learning_rate": 0.0005,
+        "loss": 1.1391,
+        "step": 14450
+      },
+      {
+        "epoch": 0.46111164259064386,
+        "grad_norm": 0.42737719416618347,
+        "learning_rate": 0.0005,
+        "loss": 1.1554,
+        "step": 14460
+      },
+      {
+        "epoch": 0.46143053031027775,
+        "grad_norm": 0.4340229034423828,
+        "learning_rate": 0.0005,
+        "loss": 1.1435,
+        "step": 14470
+      },
+      {
+        "epoch": 0.46174941802991165,
+        "grad_norm": 0.4117223024368286,
+        "learning_rate": 0.0005,
+        "loss": 1.1304,
+        "step": 14480
+      },
+      {
+        "epoch": 0.4620683057495456,
+        "grad_norm": 0.4109940826892853,
+        "learning_rate": 0.0005,
+        "loss": 1.1399,
+        "step": 14490
+      },
+      {
+        "epoch": 0.4623871934691795,
+        "grad_norm": 0.4243199825286865,
+        "learning_rate": 0.0005,
+        "loss": 1.1481,
+        "step": 14500
+      },
+      {
+        "epoch": 0.4627060811888134,
+        "grad_norm": 0.43547868728637695,
+        "learning_rate": 0.0005,
+        "loss": 1.14,
+        "step": 14510
+      },
+      {
+        "epoch": 0.46302496890844735,
+        "grad_norm": 0.41224968433380127,
+        "learning_rate": 0.0005,
+        "loss": 1.1398,
+        "step": 14520
+      },
+      {
+        "epoch": 0.46334385662808125,
+        "grad_norm": 0.41750672459602356,
+        "learning_rate": 0.0005,
+        "loss": 1.1661,
+        "step": 14530
+      },
+      {
+        "epoch": 0.46366274434771515,
+        "grad_norm": 0.39538687467575073,
+        "learning_rate": 0.0005,
+        "loss": 1.1334,
+        "step": 14540
+      },
+      {
+        "epoch": 0.4639816320673491,
+        "grad_norm": 0.4226230978965759,
+        "learning_rate": 0.0005,
+        "loss": 1.1621,
+        "step": 14550
+      },
+      {
+        "epoch": 0.464300519786983,
+        "grad_norm": 0.41002535820007324,
+        "learning_rate": 0.0005,
+        "loss": 1.1453,
+        "step": 14560
+      },
+      {
+        "epoch": 0.4646194075066169,
+        "grad_norm": 0.41372770071029663,
+        "learning_rate": 0.0005,
+        "loss": 1.1478,
+        "step": 14570
+      },
+      {
+        "epoch": 0.46493829522625085,
+        "grad_norm": 0.410127192735672,
+        "learning_rate": 0.0005,
+        "loss": 1.1447,
+        "step": 14580
+      },
+      {
+        "epoch": 0.46525718294588475,
+        "grad_norm": 0.40535253286361694,
+        "learning_rate": 0.0005,
+        "loss": 1.1455,
+        "step": 14590
+      },
+      {
+        "epoch": 0.46557607066551865,
+        "grad_norm": 0.4096167981624603,
+        "learning_rate": 0.0005,
+        "loss": 1.1449,
+        "step": 14600
+      },
+      {
+        "epoch": 0.4658949583851526,
+        "grad_norm": 0.41067060828208923,
+        "learning_rate": 0.0005,
+        "loss": 1.155,
+        "step": 14610
+      },
+      {
+        "epoch": 0.4662138461047865,
+        "grad_norm": 0.40875697135925293,
+        "learning_rate": 0.0005,
+        "loss": 1.1477,
+        "step": 14620
+      },
+      {
+        "epoch": 0.4665327338244204,
+        "grad_norm": 0.4016030430793762,
+        "learning_rate": 0.0005,
+        "loss": 1.152,
+        "step": 14630
+      },
+      {
+        "epoch": 0.46685162154405435,
+        "grad_norm": 0.41118061542510986,
+        "learning_rate": 0.0005,
+        "loss": 1.1248,
+        "step": 14640
+      },
+      {
+        "epoch": 0.46717050926368825,
+        "grad_norm": 0.4004136323928833,
+        "learning_rate": 0.0005,
+        "loss": 1.149,
+        "step": 14650
+      },
+      {
+        "epoch": 0.46748939698332215,
+        "grad_norm": 0.4095616042613983,
+        "learning_rate": 0.0005,
+        "loss": 1.1405,
+        "step": 14660
+      },
+      {
+        "epoch": 0.4678082847029561,
+        "grad_norm": 0.41689032316207886,
+        "learning_rate": 0.0005,
+        "loss": 1.144,
+        "step": 14670
+      },
+      {
+        "epoch": 0.46812717242259,
+        "grad_norm": 0.397847980260849,
+        "learning_rate": 0.0005,
+        "loss": 1.1308,
+        "step": 14680
+      },
+      {
+        "epoch": 0.4684460601422239,
+        "grad_norm": 0.410165935754776,
+        "learning_rate": 0.0005,
+        "loss": 1.1371,
+        "step": 14690
+      },
+      {
+        "epoch": 0.46876494786185785,
+        "grad_norm": 0.4155512750148773,
+        "learning_rate": 0.0005,
+        "loss": 1.1417,
+        "step": 14700
+      },
+      {
+        "epoch": 0.46908383558149175,
+        "grad_norm": 0.4086609184741974,
+        "learning_rate": 0.0005,
+        "loss": 1.1567,
+        "step": 14710
+      },
+      {
+        "epoch": 0.46940272330112565,
+        "grad_norm": 0.4168750047683716,
+        "learning_rate": 0.0005,
+        "loss": 1.1477,
+        "step": 14720
+      },
+      {
+        "epoch": 0.4697216110207596,
+        "grad_norm": 0.4083302915096283,
+        "learning_rate": 0.0005,
+        "loss": 1.1445,
+        "step": 14730
+      },
+      {
+        "epoch": 0.4700404987403935,
+        "grad_norm": 0.3982124924659729,
+        "learning_rate": 0.0005,
+        "loss": 1.1555,
+        "step": 14740
+      },
+      {
+        "epoch": 0.47035938646002745,
+        "grad_norm": 0.4231313467025757,
+        "learning_rate": 0.0005,
+        "loss": 1.1589,
+        "step": 14750
+      },
+      {
+        "epoch": 0.47067827417966135,
+        "grad_norm": 0.3778389096260071,
+        "learning_rate": 0.0005,
+        "loss": 1.1284,
+        "step": 14760
+      },
+      {
+        "epoch": 0.47099716189929525,
+        "grad_norm": 0.40142810344696045,
+        "learning_rate": 0.0005,
+        "loss": 1.1164,
+        "step": 14770
+      },
+      {
+        "epoch": 0.4713160496189292,
+        "grad_norm": 0.4071197807788849,
+        "learning_rate": 0.0005,
+        "loss": 1.1509,
+        "step": 14780
+      },
+      {
+        "epoch": 0.4716349373385631,
+        "grad_norm": 0.4330447316169739,
+        "learning_rate": 0.0005,
+        "loss": 1.1559,
+        "step": 14790
+      },
+      {
+        "epoch": 0.471953825058197,
+        "grad_norm": 0.4079609513282776,
+        "learning_rate": 0.0005,
+        "loss": 1.1519,
+        "step": 14800
+      },
+      {
+        "epoch": 0.47227271277783095,
+        "grad_norm": 0.4308188557624817,
+        "learning_rate": 0.0005,
+        "loss": 1.1414,
+        "step": 14810
+      },
+      {
+        "epoch": 0.47259160049746485,
+        "grad_norm": 0.44308769702911377,
+        "learning_rate": 0.0005,
+        "loss": 1.1535,
+        "step": 14820
+      },
+      {
+        "epoch": 0.47291048821709875,
+        "grad_norm": 0.404365599155426,
+        "learning_rate": 0.0005,
+        "loss": 1.1549,
+        "step": 14830
+      },
+      {
+        "epoch": 0.4732293759367327,
+        "grad_norm": 0.4012412428855896,
+        "learning_rate": 0.0005,
+        "loss": 1.1512,
+        "step": 14840
+      },
+      {
+        "epoch": 0.4735482636563666,
+        "grad_norm": 0.3904338479042053,
+        "learning_rate": 0.0005,
+        "loss": 1.1456,
+        "step": 14850
+      },
+      {
+        "epoch": 0.4738671513760005,
+        "grad_norm": 0.4106074571609497,
+        "learning_rate": 0.0005,
+        "loss": 1.1322,
+        "step": 14860
+      },
+      {
+        "epoch": 0.47418603909563445,
+        "grad_norm": 0.41735175251960754,
+        "learning_rate": 0.0005,
+        "loss": 1.133,
+        "step": 14870
+      },
+      {
+        "epoch": 0.47450492681526835,
+        "grad_norm": 0.40341079235076904,
+        "learning_rate": 0.0005,
+        "loss": 1.1376,
+        "step": 14880
+      },
+      {
+        "epoch": 0.47482381453490224,
+        "grad_norm": 0.42312002182006836,
+        "learning_rate": 0.0005,
+        "loss": 1.1462,
+        "step": 14890
+      },
+      {
+        "epoch": 0.4751427022545362,
+        "grad_norm": 0.4028998613357544,
+        "learning_rate": 0.0005,
+        "loss": 1.1457,
+        "step": 14900
+      },
+      {
+        "epoch": 0.4754615899741701,
+        "grad_norm": 0.41534656286239624,
+        "learning_rate": 0.0005,
+        "loss": 1.1572,
+        "step": 14910
+      },
+      {
+        "epoch": 0.475780477693804,
+        "grad_norm": 0.39137589931488037,
+        "learning_rate": 0.0005,
+        "loss": 1.1413,
+        "step": 14920
+      },
+      {
+        "epoch": 0.47609936541343795,
+        "grad_norm": 0.3952145576477051,
+        "learning_rate": 0.0005,
+        "loss": 1.1538,
+        "step": 14930
+      },
+      {
+        "epoch": 0.47641825313307185,
+        "grad_norm": 0.42215418815612793,
+        "learning_rate": 0.0005,
+        "loss": 1.158,
+        "step": 14940
+      },
+      {
+        "epoch": 0.47673714085270574,
+        "grad_norm": 0.40796101093292236,
+        "learning_rate": 0.0005,
+        "loss": 1.1366,
+        "step": 14950
+      },
+      {
+        "epoch": 0.4770560285723397,
+        "grad_norm": 0.387132465839386,
+        "learning_rate": 0.0005,
+        "loss": 1.1482,
+        "step": 14960
+      },
+      {
+        "epoch": 0.4773749162919736,
+        "grad_norm": 0.4274901747703552,
+        "learning_rate": 0.0005,
+        "loss": 1.151,
+        "step": 14970
+      },
+      {
+        "epoch": 0.4776938040116075,
+        "grad_norm": 0.4018956124782562,
+        "learning_rate": 0.0005,
+        "loss": 1.1365,
+        "step": 14980
+      },
+      {
+        "epoch": 0.47801269173124145,
+        "grad_norm": 0.3979787528514862,
+        "learning_rate": 0.0005,
+        "loss": 1.1255,
+        "step": 14990
+      },
+      {
+        "epoch": 0.47833157945087534,
+        "grad_norm": 0.43101415038108826,
+        "learning_rate": 0.0005,
+        "loss": 1.158,
+        "step": 15000
+      },
+      {
+        "epoch": 0.47865046717050924,
+        "grad_norm": 0.4156745672225952,
+        "learning_rate": 0.0005,
+        "loss": 1.1584,
+        "step": 15010
+      },
+      {
+        "epoch": 0.4789693548901432,
+        "grad_norm": 0.4341460168361664,
+        "learning_rate": 0.0005,
+        "loss": 1.1339,
+        "step": 15020
+      },
+      {
+        "epoch": 0.4792882426097771,
+        "grad_norm": 0.4072680175304413,
+        "learning_rate": 0.0005,
+        "loss": 1.1539,
+        "step": 15030
+      },
+      {
+        "epoch": 0.479607130329411,
+        "grad_norm": 0.3964672386646271,
+        "learning_rate": 0.0005,
+        "loss": 1.1333,
+        "step": 15040
+      },
+      {
+        "epoch": 0.47992601804904494,
+        "grad_norm": 0.41430017352104187,
+        "learning_rate": 0.0005,
+        "loss": 1.1457,
+        "step": 15050
+      },
+      {
+        "epoch": 0.48024490576867884,
+        "grad_norm": 0.3992692828178406,
+        "learning_rate": 0.0005,
+        "loss": 1.1347,
+        "step": 15060
+      },
+      {
+        "epoch": 0.48056379348831274,
+        "grad_norm": 0.4212576448917389,
+        "learning_rate": 0.0005,
+        "loss": 1.1504,
+        "step": 15070
+      },
+      {
+        "epoch": 0.4808826812079467,
+        "grad_norm": 0.4008728861808777,
+        "learning_rate": 0.0005,
+        "loss": 1.1461,
+        "step": 15080
+      },
+      {
+        "epoch": 0.4812015689275806,
+        "grad_norm": 0.4172374904155731,
+        "learning_rate": 0.0005,
+        "loss": 1.1416,
+        "step": 15090
+      },
+      {
+        "epoch": 0.4815204566472145,
+        "grad_norm": 0.4110826849937439,
+        "learning_rate": 0.0005,
+        "loss": 1.1534,
+        "step": 15100
+      },
+      {
+        "epoch": 0.48183934436684844,
+        "grad_norm": 0.4187602698802948,
+        "learning_rate": 0.0005,
+        "loss": 1.1269,
+        "step": 15110
+      },
+      {
+        "epoch": 0.48215823208648234,
+        "grad_norm": 0.39021268486976624,
+        "learning_rate": 0.0005,
+        "loss": 1.1219,
+        "step": 15120
+      },
+      {
+        "epoch": 0.48247711980611624,
+        "grad_norm": 0.4110153317451477,
+        "learning_rate": 0.0005,
+        "loss": 1.1314,
+        "step": 15130
+      },
+      {
+        "epoch": 0.4827960075257502,
+        "grad_norm": 0.4028952121734619,
+        "learning_rate": 0.0005,
+        "loss": 1.1363,
+        "step": 15140
+      },
+      {
+        "epoch": 0.4831148952453841,
+        "grad_norm": 0.39289188385009766,
+        "learning_rate": 0.0005,
+        "loss": 1.1257,
+        "step": 15150
+      },
+      {
+        "epoch": 0.48343378296501804,
+        "grad_norm": 0.4015601873397827,
+        "learning_rate": 0.0005,
+        "loss": 1.132,
+        "step": 15160
+      },
+      {
+        "epoch": 0.48375267068465194,
+        "grad_norm": 0.4219626188278198,
+        "learning_rate": 0.0005,
+        "loss": 1.1365,
+        "step": 15170
+      },
+      {
+        "epoch": 0.48407155840428584,
+        "grad_norm": 0.41227683424949646,
+        "learning_rate": 0.0005,
+        "loss": 1.1337,
+        "step": 15180
+      },
+      {
+        "epoch": 0.4843904461239198,
+        "grad_norm": 0.4118657112121582,
+        "learning_rate": 0.0005,
+        "loss": 1.1277,
+        "step": 15190
+      },
+      {
+        "epoch": 0.4847093338435537,
+        "grad_norm": 0.42577946186065674,
+        "learning_rate": 0.0005,
+        "loss": 1.1668,
+        "step": 15200
+      },
+      {
+        "epoch": 0.4850282215631876,
+        "grad_norm": 0.39030152559280396,
+        "learning_rate": 0.0005,
+        "loss": 1.127,
+        "step": 15210
+      },
+      {
+        "epoch": 0.48534710928282154,
+        "grad_norm": 0.3986493945121765,
+        "learning_rate": 0.0005,
+        "loss": 1.1388,
+        "step": 15220
+      },
+      {
+        "epoch": 0.48566599700245544,
+        "grad_norm": 0.40662774443626404,
+        "learning_rate": 0.0005,
+        "loss": 1.1502,
+        "step": 15230
+      },
+      {
+        "epoch": 0.48598488472208934,
+        "grad_norm": 0.4098156690597534,
+        "learning_rate": 0.0005,
+        "loss": 1.1467,
+        "step": 15240
+      },
+      {
+        "epoch": 0.4863037724417233,
+        "grad_norm": 0.38814619183540344,
+        "learning_rate": 0.0005,
+        "loss": 1.132,
+        "step": 15250
+      },
+      {
+        "epoch": 0.4866226601613572,
+        "grad_norm": 0.394285649061203,
+        "learning_rate": 0.0005,
+        "loss": 1.1384,
+        "step": 15260
+      },
+      {
+        "epoch": 0.4869415478809911,
+        "grad_norm": 0.39842134714126587,
+        "learning_rate": 0.0005,
+        "loss": 1.1355,
+        "step": 15270
+      },
+      {
+        "epoch": 0.48726043560062504,
+        "grad_norm": 0.3929298222064972,
+        "learning_rate": 0.0005,
+        "loss": 1.1205,
+        "step": 15280
+      },
+      {
+        "epoch": 0.48757932332025894,
+        "grad_norm": 0.4007817804813385,
+        "learning_rate": 0.0005,
+        "loss": 1.1399,
+        "step": 15290
+      },
+      {
+        "epoch": 0.48789821103989284,
+        "grad_norm": 0.39471328258514404,
+        "learning_rate": 0.0005,
+        "loss": 1.1271,
+        "step": 15300
+      },
+      {
+        "epoch": 0.4882170987595268,
+        "grad_norm": 0.4145445227622986,
+        "learning_rate": 0.0005,
+        "loss": 1.122,
+        "step": 15310
+      },
+      {
+        "epoch": 0.4885359864791607,
+        "grad_norm": 0.4210784137248993,
+        "learning_rate": 0.0005,
+        "loss": 1.1379,
+        "step": 15320
+      },
+      {
+        "epoch": 0.4888548741987946,
+        "grad_norm": 0.4189470708370209,
+        "learning_rate": 0.0005,
+        "loss": 1.1394,
+        "step": 15330
+      },
+      {
+        "epoch": 0.48917376191842854,
+        "grad_norm": 0.41203248500823975,
+        "learning_rate": 0.0005,
+        "loss": 1.1239,
+        "step": 15340
+      },
+      {
+        "epoch": 0.48949264963806244,
+        "grad_norm": 0.4104352593421936,
+        "learning_rate": 0.0005,
+        "loss": 1.1285,
+        "step": 15350
+      },
+      {
+        "epoch": 0.48981153735769634,
+        "grad_norm": 0.39657625555992126,
+        "learning_rate": 0.0005,
+        "loss": 1.1247,
+        "step": 15360
+      },
+      {
+        "epoch": 0.4901304250773303,
+        "grad_norm": 0.4017868936061859,
+        "learning_rate": 0.0005,
+        "loss": 1.1461,
+        "step": 15370
+      },
+      {
+        "epoch": 0.4904493127969642,
+        "grad_norm": 0.4098726511001587,
+        "learning_rate": 0.0005,
+        "loss": 1.141,
+        "step": 15380
+      },
+      {
+        "epoch": 0.4907682005165981,
+        "grad_norm": 0.4074094891548157,
+        "learning_rate": 0.0005,
+        "loss": 1.1486,
+        "step": 15390
+      },
+      {
+        "epoch": 0.49108708823623204,
+        "grad_norm": 0.42989563941955566,
+        "learning_rate": 0.0005,
+        "loss": 1.1252,
+        "step": 15400
+      },
+      {
+        "epoch": 0.49140597595586594,
+        "grad_norm": 0.41132497787475586,
+        "learning_rate": 0.0005,
+        "loss": 1.1439,
+        "step": 15410
+      },
+      {
+        "epoch": 0.49172486367549983,
+        "grad_norm": 0.41117823123931885,
+        "learning_rate": 0.0005,
+        "loss": 1.145,
+        "step": 15420
+      },
+      {
+        "epoch": 0.4920437513951338,
+        "grad_norm": 0.4057653844356537,
+        "learning_rate": 0.0005,
+        "loss": 1.1377,
+        "step": 15430
+      },
+      {
+        "epoch": 0.4923626391147677,
+        "grad_norm": 0.39292091131210327,
+        "learning_rate": 0.0005,
+        "loss": 1.1527,
+        "step": 15440
+      },
+      {
+        "epoch": 0.4926815268344016,
+        "grad_norm": 0.41839951276779175,
+        "learning_rate": 0.0005,
+        "loss": 1.1633,
+        "step": 15450
+      },
+      {
+        "epoch": 0.49300041455403554,
+        "grad_norm": 0.4018307030200958,
+        "learning_rate": 0.0005,
+        "loss": 1.1584,
+        "step": 15460
+      },
+      {
+        "epoch": 0.49331930227366944,
+        "grad_norm": 0.42629310488700867,
+        "learning_rate": 0.0005,
+        "loss": 1.1557,
+        "step": 15470
+      },
+      {
+        "epoch": 0.49363818999330333,
+        "grad_norm": 0.4066459536552429,
+        "learning_rate": 0.0005,
+        "loss": 1.1268,
+        "step": 15480
+      },
+      {
+        "epoch": 0.4939570777129373,
+        "grad_norm": 0.4150345027446747,
+        "learning_rate": 0.0005,
+        "loss": 1.1552,
+        "step": 15490
+      },
+      {
+        "epoch": 0.4942759654325712,
+        "grad_norm": 0.4128059446811676,
+        "learning_rate": 0.0005,
+        "loss": 1.1308,
+        "step": 15500
+      },
+      {
+        "epoch": 0.4945948531522051,
+        "grad_norm": 0.4207080006599426,
+        "learning_rate": 0.0005,
+        "loss": 1.1431,
+        "step": 15510
+      },
+      {
+        "epoch": 0.49491374087183904,
+        "grad_norm": 0.39305001497268677,
+        "learning_rate": 0.0005,
+        "loss": 1.1484,
+        "step": 15520
+      },
+      {
+        "epoch": 0.49523262859147293,
+        "grad_norm": 0.40596818923950195,
+        "learning_rate": 0.0005,
+        "loss": 1.1378,
+        "step": 15530
+      },
+      {
+        "epoch": 0.49555151631110683,
+        "grad_norm": 0.3958894610404968,
+        "learning_rate": 0.0005,
+        "loss": 1.1372,
+        "step": 15540
+      },
+      {
+        "epoch": 0.4958704040307408,
+        "grad_norm": 0.39964666962623596,
+        "learning_rate": 0.0005,
+        "loss": 1.1452,
+        "step": 15550
+      },
+      {
+        "epoch": 0.4961892917503747,
+        "grad_norm": 0.3986499607563019,
+        "learning_rate": 0.0005,
+        "loss": 1.1357,
+        "step": 15560
+      },
+      {
+        "epoch": 0.49650817947000864,
+        "grad_norm": 0.38713058829307556,
+        "learning_rate": 0.0005,
+        "loss": 1.146,
+        "step": 15570
+      },
+      {
+        "epoch": 0.49682706718964254,
+        "grad_norm": 0.4117546081542969,
+        "learning_rate": 0.0005,
+        "loss": 1.1424,
+        "step": 15580
+      },
+      {
+        "epoch": 0.49714595490927643,
+        "grad_norm": 0.41599640250205994,
+        "learning_rate": 0.0005,
+        "loss": 1.1458,
+        "step": 15590
+      },
+      {
+        "epoch": 0.4974648426289104,
+        "grad_norm": 0.4060603678226471,
+        "learning_rate": 0.0005,
+        "loss": 1.1298,
+        "step": 15600
+      },
+      {
+        "epoch": 0.4977837303485443,
+        "grad_norm": 0.40494588017463684,
+        "learning_rate": 0.0005,
+        "loss": 1.1315,
+        "step": 15610
+      },
+      {
+        "epoch": 0.4981026180681782,
+        "grad_norm": 0.39388027787208557,
+        "learning_rate": 0.0005,
+        "loss": 1.1406,
+        "step": 15620
+      },
+      {
+        "epoch": 0.49842150578781214,
+        "grad_norm": 0.4042576849460602,
+        "learning_rate": 0.0005,
+        "loss": 1.1253,
+        "step": 15630
+      },
+      {
+        "epoch": 0.49874039350744603,
+        "grad_norm": 0.40436872839927673,
+        "learning_rate": 0.0005,
+        "loss": 1.1388,
+        "step": 15640
+      },
+      {
+        "epoch": 0.49905928122707993,
+        "grad_norm": 0.40439584851264954,
+        "learning_rate": 0.0005,
+        "loss": 1.139,
+        "step": 15650
+      },
+      {
+        "epoch": 0.4993781689467139,
+        "grad_norm": 0.3996795117855072,
+        "learning_rate": 0.0005,
+        "loss": 1.1563,
+        "step": 15660
+      },
+      {
+        "epoch": 0.4996970566663478,
+        "grad_norm": 0.415203720331192,
+        "learning_rate": 0.0005,
+        "loss": 1.1362,
+        "step": 15670
+      },
+      {
+        "epoch": 0.5000159443859817,
+        "grad_norm": 0.40682661533355713,
+        "learning_rate": 0.0005,
+        "loss": 1.1387,
+        "step": 15680
+      },
+      {
+        "epoch": 0.5003348321056156,
+        "grad_norm": 0.38640740513801575,
+        "learning_rate": 0.0005,
+        "loss": 1.1269,
+        "step": 15690
+      },
+      {
+        "epoch": 0.5006537198252495,
+        "grad_norm": 0.4006939232349396,
+        "learning_rate": 0.0005,
+        "loss": 1.1346,
+        "step": 15700
+      },
+      {
+        "epoch": 0.5009726075448835,
+        "grad_norm": 0.3977952003479004,
+        "learning_rate": 0.0005,
+        "loss": 1.1391,
+        "step": 15710
+      },
+      {
+        "epoch": 0.5012914952645173,
+        "grad_norm": 0.3980969786643982,
+        "learning_rate": 0.0005,
+        "loss": 1.1486,
+        "step": 15720
+      },
+      {
+        "epoch": 0.5016103829841513,
+        "grad_norm": 0.4002665877342224,
+        "learning_rate": 0.0005,
+        "loss": 1.1279,
+        "step": 15730
+      },
+      {
+        "epoch": 0.5019292707037852,
+        "grad_norm": 0.39593833684921265,
+        "learning_rate": 0.0005,
+        "loss": 1.1584,
+        "step": 15740
+      },
+      {
+        "epoch": 0.5022481584234191,
+        "grad_norm": 0.3968457877635956,
+        "learning_rate": 0.0005,
+        "loss": 1.1389,
+        "step": 15750
+      },
+      {
+        "epoch": 0.502567046143053,
+        "grad_norm": 0.4009703993797302,
+        "learning_rate": 0.0005,
+        "loss": 1.1151,
+        "step": 15760
+      },
+      {
+        "epoch": 0.502885933862687,
+        "grad_norm": 0.40247270464897156,
+        "learning_rate": 0.0005,
+        "loss": 1.1419,
+        "step": 15770
+      },
+      {
+        "epoch": 0.5032048215823208,
+        "grad_norm": 0.4238397181034088,
+        "learning_rate": 0.0005,
+        "loss": 1.1362,
+        "step": 15780
+      },
+      {
+        "epoch": 0.5035237093019548,
+        "grad_norm": 0.3963928520679474,
+        "learning_rate": 0.0005,
+        "loss": 1.1311,
+        "step": 15790
+      },
+      {
+        "epoch": 0.5038425970215887,
+        "grad_norm": 0.38498803973197937,
+        "learning_rate": 0.0005,
+        "loss": 1.1301,
+        "step": 15800
+      },
+      {
+        "epoch": 0.5041614847412226,
+        "grad_norm": 0.40925782918930054,
+        "learning_rate": 0.0005,
+        "loss": 1.1447,
+        "step": 15810
+      },
+      {
+        "epoch": 0.5044803724608565,
+        "grad_norm": 0.3941743075847626,
+        "learning_rate": 0.0005,
+        "loss": 1.1325,
+        "step": 15820
+      },
+      {
+        "epoch": 0.5047992601804905,
+        "grad_norm": 0.3878701627254486,
+        "learning_rate": 0.0005,
+        "loss": 1.1247,
+        "step": 15830
+      },
+      {
+        "epoch": 0.5051181479001243,
+        "grad_norm": 0.4078117609024048,
+        "learning_rate": 0.0005,
+        "loss": 1.1386,
+        "step": 15840
+      },
+      {
+        "epoch": 0.5054370356197583,
+        "grad_norm": 0.40194058418273926,
+        "learning_rate": 0.0005,
+        "loss": 1.1454,
+        "step": 15850
+      },
+      {
+        "epoch": 0.5057559233393922,
+        "grad_norm": 0.3996395170688629,
+        "learning_rate": 0.0005,
+        "loss": 1.1478,
+        "step": 15860
+      },
+      {
+        "epoch": 0.5060748110590261,
+        "grad_norm": 0.41087689995765686,
+        "learning_rate": 0.0005,
+        "loss": 1.126,
+        "step": 15870
+      },
+      {
+        "epoch": 0.50639369877866,
+        "grad_norm": 0.3939761817455292,
+        "learning_rate": 0.0005,
+        "loss": 1.1361,
+        "step": 15880
+      },
+      {
+        "epoch": 0.506712586498294,
+        "grad_norm": 0.4000106453895569,
+        "learning_rate": 0.0005,
+        "loss": 1.1259,
+        "step": 15890
+      },
+      {
+        "epoch": 0.5070314742179278,
+        "grad_norm": 0.4082937240600586,
+        "learning_rate": 0.0005,
+        "loss": 1.1292,
+        "step": 15900
+      },
+      {
+        "epoch": 0.5073503619375618,
+        "grad_norm": 0.3926817774772644,
+        "learning_rate": 0.0005,
+        "loss": 1.1273,
+        "step": 15910
+      },
+      {
+        "epoch": 0.5076692496571957,
+        "grad_norm": 0.41120755672454834,
+        "learning_rate": 0.0005,
+        "loss": 1.1279,
+        "step": 15920
+      },
+      {
+        "epoch": 0.5079881373768296,
+        "grad_norm": 0.3957878351211548,
+        "learning_rate": 0.0005,
+        "loss": 1.1347,
+        "step": 15930
+      },
+      {
+        "epoch": 0.5083070250964635,
+        "grad_norm": 0.37787875533103943,
+        "learning_rate": 0.0005,
+        "loss": 1.1279,
+        "step": 15940
+      },
+      {
+        "epoch": 0.5086259128160975,
+        "grad_norm": 0.40983471274375916,
+        "learning_rate": 0.0005,
+        "loss": 1.1298,
+        "step": 15950
+      },
+      {
+        "epoch": 0.5089448005357313,
+        "grad_norm": 0.417005717754364,
+        "learning_rate": 0.0005,
+        "loss": 1.1374,
+        "step": 15960
+      },
+      {
+        "epoch": 0.5092636882553653,
+        "grad_norm": 0.4130145013332367,
+        "learning_rate": 0.0005,
+        "loss": 1.1416,
+        "step": 15970
+      },
+      {
+        "epoch": 0.5095825759749992,
+        "grad_norm": 0.389676034450531,
+        "learning_rate": 0.0005,
+        "loss": 1.131,
+        "step": 15980
+      },
+      {
+        "epoch": 0.5099014636946331,
+        "grad_norm": 0.4049873352050781,
+        "learning_rate": 0.0005,
+        "loss": 1.1198,
+        "step": 15990
+      },
+      {
+        "epoch": 0.510220351414267,
+        "grad_norm": 0.40844064950942993,
+        "learning_rate": 0.0005,
+        "loss": 1.1209,
+        "step": 16000
+      },
+      {
+        "epoch": 0.510539239133901,
+        "grad_norm": 0.4057811498641968,
+        "learning_rate": 0.0005,
+        "loss": 1.1303,
+        "step": 16010
+      },
+      {
+        "epoch": 0.5108581268535348,
+        "grad_norm": 0.39924463629722595,
+        "learning_rate": 0.0005,
+        "loss": 1.1399,
+        "step": 16020
+      },
+      {
+        "epoch": 0.5111770145731688,
+        "grad_norm": 0.39572224020957947,
+        "learning_rate": 0.0005,
+        "loss": 1.1104,
+        "step": 16030
+      },
+      {
+        "epoch": 0.5114959022928027,
+        "grad_norm": 0.42480066418647766,
+        "learning_rate": 0.0005,
+        "loss": 1.1508,
+        "step": 16040
+      },
+      {
+        "epoch": 0.5118147900124366,
+        "grad_norm": 0.40182217955589294,
+        "learning_rate": 0.0005,
+        "loss": 1.1243,
+        "step": 16050
+      },
+      {
+        "epoch": 0.5121336777320705,
+        "grad_norm": 0.41160985827445984,
+        "learning_rate": 0.0005,
+        "loss": 1.1514,
+        "step": 16060
+      },
+      {
+        "epoch": 0.5124525654517045,
+        "grad_norm": 0.40019845962524414,
+        "learning_rate": 0.0005,
+        "loss": 1.1029,
+        "step": 16070
+      },
+      {
+        "epoch": 0.5127714531713383,
+        "grad_norm": 0.40229541063308716,
+        "learning_rate": 0.0005,
+        "loss": 1.1206,
+        "step": 16080
+      },
+      {
+        "epoch": 0.5130903408909723,
+        "grad_norm": 0.39097946882247925,
+        "learning_rate": 0.0005,
+        "loss": 1.1369,
+        "step": 16090
+      },
+      {
+        "epoch": 0.5134092286106062,
+        "grad_norm": 0.4011414051055908,
+        "learning_rate": 0.0005,
+        "loss": 1.1363,
+        "step": 16100
+      },
+      {
+        "epoch": 0.5137281163302401,
+        "grad_norm": 0.3968360424041748,
+        "learning_rate": 0.0005,
+        "loss": 1.1295,
+        "step": 16110
+      },
+      {
+        "epoch": 0.514047004049874,
+        "grad_norm": 0.39342719316482544,
+        "learning_rate": 0.0005,
+        "loss": 1.1258,
+        "step": 16120
+      },
+      {
+        "epoch": 0.514365891769508,
+        "grad_norm": 0.38369807600975037,
+        "learning_rate": 0.0005,
+        "loss": 1.1257,
+        "step": 16130
+      },
+      {
+        "epoch": 0.5146847794891418,
+        "grad_norm": 0.38821354508399963,
+        "learning_rate": 0.0005,
+        "loss": 1.1276,
+        "step": 16140
+      },
+      {
+        "epoch": 0.5150036672087758,
+        "grad_norm": 0.41388481855392456,
+        "learning_rate": 0.0005,
+        "loss": 1.1349,
+        "step": 16150
+      },
+      {
+        "epoch": 0.5153225549284097,
+        "grad_norm": 0.40181416273117065,
+        "learning_rate": 0.0005,
+        "loss": 1.1299,
+        "step": 16160
+      },
+      {
+        "epoch": 0.5156414426480436,
+        "grad_norm": 0.4062349498271942,
+        "learning_rate": 0.0005,
+        "loss": 1.1163,
+        "step": 16170
+      },
+      {
+        "epoch": 0.5159603303676775,
+        "grad_norm": 0.4037853181362152,
+        "learning_rate": 0.0005,
+        "loss": 1.1189,
+        "step": 16180
+      },
+      {
+        "epoch": 0.5162792180873115,
+        "grad_norm": 0.3858471214771271,
+        "learning_rate": 0.0005,
+        "loss": 1.123,
+        "step": 16190
+      },
+      {
+        "epoch": 0.5165981058069454,
+        "grad_norm": 0.3916098475456238,
+        "learning_rate": 0.0005,
+        "loss": 1.1359,
+        "step": 16200
+      },
+      {
+        "epoch": 0.5169169935265793,
+        "grad_norm": 0.40132418274879456,
+        "learning_rate": 0.0005,
+        "loss": 1.1262,
+        "step": 16210
+      },
+      {
+        "epoch": 0.5172358812462132,
+        "grad_norm": 0.3784230351448059,
+        "learning_rate": 0.0005,
+        "loss": 1.1295,
+        "step": 16220
+      },
+      {
+        "epoch": 0.5175547689658472,
+        "grad_norm": 0.4010845720767975,
+        "learning_rate": 0.0005,
+        "loss": 1.1272,
+        "step": 16230
+      },
+      {
+        "epoch": 0.517873656685481,
+        "grad_norm": 0.4017618000507355,
+        "learning_rate": 0.0005,
+        "loss": 1.128,
+        "step": 16240
+      },
+      {
+        "epoch": 0.518192544405115,
+        "grad_norm": 0.3865583539009094,
+        "learning_rate": 0.0005,
+        "loss": 1.1378,
+        "step": 16250
+      },
+      {
+        "epoch": 0.5185114321247489,
+        "grad_norm": 0.38878291845321655,
+        "learning_rate": 0.0005,
+        "loss": 1.1296,
+        "step": 16260
+      },
+      {
+        "epoch": 0.5188303198443828,
+        "grad_norm": 0.3995644152164459,
+        "learning_rate": 0.0005,
+        "loss": 1.1167,
+        "step": 16270
+      },
+      {
+        "epoch": 0.5191492075640167,
+        "grad_norm": 0.3930690586566925,
+        "learning_rate": 0.0005,
+        "loss": 1.1164,
+        "step": 16280
+      },
+      {
+        "epoch": 0.5194680952836507,
+        "grad_norm": 0.4176512658596039,
+        "learning_rate": 0.0005,
+        "loss": 1.1214,
+        "step": 16290
+      },
+      {
+        "epoch": 0.5197869830032845,
+        "grad_norm": 0.37958794832229614,
+        "learning_rate": 0.0005,
+        "loss": 1.1471,
+        "step": 16300
+      },
+      {
+        "epoch": 0.5201058707229185,
+        "grad_norm": 0.3942332863807678,
+        "learning_rate": 0.0005,
+        "loss": 1.1451,
+        "step": 16310
+      },
+      {
+        "epoch": 0.5204247584425524,
+        "grad_norm": 0.3994196355342865,
+        "learning_rate": 0.0005,
+        "loss": 1.125,
+        "step": 16320
+      },
+      {
+        "epoch": 0.5207436461621863,
+        "grad_norm": 0.40982067584991455,
+        "learning_rate": 0.0005,
+        "loss": 1.1321,
+        "step": 16330
+      },
+      {
+        "epoch": 0.5210625338818202,
+        "grad_norm": 0.4163874387741089,
+        "learning_rate": 0.0005,
+        "loss": 1.1433,
+        "step": 16340
+      },
+      {
+        "epoch": 0.5213814216014542,
+        "grad_norm": 0.4115857183933258,
+        "learning_rate": 0.0005,
+        "loss": 1.1186,
+        "step": 16350
+      },
+      {
+        "epoch": 0.521700309321088,
+        "grad_norm": 0.3939268887042999,
+        "learning_rate": 0.0005,
+        "loss": 1.1093,
+        "step": 16360
+      },
+      {
+        "epoch": 0.522019197040722,
+        "grad_norm": 0.4047539234161377,
+        "learning_rate": 0.0005,
+        "loss": 1.1203,
+        "step": 16370
+      },
+      {
+        "epoch": 0.5223380847603559,
+        "grad_norm": 0.4223722815513611,
+        "learning_rate": 0.0005,
+        "loss": 1.1159,
+        "step": 16380
+      },
+      {
+        "epoch": 0.5226569724799898,
+        "grad_norm": 0.3986281454563141,
+        "learning_rate": 0.0005,
+        "loss": 1.1577,
+        "step": 16390
+      },
+      {
+        "epoch": 0.5229758601996237,
+        "grad_norm": 0.40140584111213684,
+        "learning_rate": 0.0005,
+        "loss": 1.1186,
+        "step": 16400
+      },
+      {
+        "epoch": 0.5232947479192577,
+        "grad_norm": 0.41486597061157227,
+        "learning_rate": 0.0005,
+        "loss": 1.1207,
+        "step": 16410
+      },
+      {
+        "epoch": 0.5236136356388915,
+        "grad_norm": 0.39416739344596863,
+        "learning_rate": 0.0005,
+        "loss": 1.1158,
+        "step": 16420
+      },
+      {
+        "epoch": 0.5239325233585255,
+        "grad_norm": 0.39388272166252136,
+        "learning_rate": 0.0005,
+        "loss": 1.1466,
+        "step": 16430
+      },
+      {
+        "epoch": 0.5242514110781594,
+        "grad_norm": 0.40144652128219604,
+        "learning_rate": 0.0005,
+        "loss": 1.1374,
+        "step": 16440
+      },
+      {
+        "epoch": 0.5245702987977933,
+        "grad_norm": 0.4018017053604126,
+        "learning_rate": 0.0005,
+        "loss": 1.1444,
+        "step": 16450
+      },
+      {
+        "epoch": 0.5248891865174272,
+        "grad_norm": 0.40438687801361084,
+        "learning_rate": 0.0005,
+        "loss": 1.1275,
+        "step": 16460
+      },
+      {
+        "epoch": 0.5252080742370612,
+        "grad_norm": 0.4010547697544098,
+        "learning_rate": 0.0005,
+        "loss": 1.1383,
+        "step": 16470
+      },
+      {
+        "epoch": 0.525526961956695,
+        "grad_norm": 0.41225293278694153,
+        "learning_rate": 0.0005,
+        "loss": 1.1536,
+        "step": 16480
+      },
+      {
+        "epoch": 0.525845849676329,
+        "grad_norm": 0.39326658844947815,
+        "learning_rate": 0.0005,
+        "loss": 1.1366,
+        "step": 16490
+      },
+      {
+        "epoch": 0.5261647373959629,
+        "grad_norm": 0.3904922306537628,
+        "learning_rate": 0.0005,
+        "loss": 1.1103,
+        "step": 16500
+      },
+      {
+        "epoch": 0.5264836251155968,
+        "grad_norm": 0.3908495008945465,
+        "learning_rate": 0.0005,
+        "loss": 1.1172,
+        "step": 16510
+      },
+      {
+        "epoch": 0.5268025128352307,
+        "grad_norm": 0.412609726190567,
+        "learning_rate": 0.0005,
+        "loss": 1.1317,
+        "step": 16520
+      },
+      {
+        "epoch": 0.5271214005548647,
+        "grad_norm": 0.38488397002220154,
+        "learning_rate": 0.0005,
+        "loss": 1.1207,
+        "step": 16530
+      },
+      {
+        "epoch": 0.5274402882744985,
+        "grad_norm": 0.40790820121765137,
+        "learning_rate": 0.0005,
+        "loss": 1.1597,
+        "step": 16540
+      },
+      {
+        "epoch": 0.5277591759941325,
+        "grad_norm": 0.4010877013206482,
+        "learning_rate": 0.0005,
+        "loss": 1.1119,
+        "step": 16550
+      },
+      {
+        "epoch": 0.5280780637137664,
+        "grad_norm": 0.39443594217300415,
+        "learning_rate": 0.0005,
+        "loss": 1.1303,
+        "step": 16560
+      },
+      {
+        "epoch": 0.5283969514334003,
+        "grad_norm": 0.41272708773612976,
+        "learning_rate": 0.0005,
+        "loss": 1.1182,
+        "step": 16570
+      },
+      {
+        "epoch": 0.5287158391530342,
+        "grad_norm": 0.40090036392211914,
+        "learning_rate": 0.0005,
+        "loss": 1.1059,
+        "step": 16580
+      },
+      {
+        "epoch": 0.5290347268726682,
+        "grad_norm": 0.3933262526988983,
+        "learning_rate": 0.0005,
+        "loss": 1.1184,
+        "step": 16590
+      },
+      {
+        "epoch": 0.529353614592302,
+        "grad_norm": 0.3988369405269623,
+        "learning_rate": 0.0005,
+        "loss": 1.1239,
+        "step": 16600
+      },
+      {
+        "epoch": 0.529672502311936,
+        "grad_norm": 0.3974078595638275,
+        "learning_rate": 0.0005,
+        "loss": 1.1132,
+        "step": 16610
+      },
+      {
+        "epoch": 0.5299913900315699,
+        "grad_norm": 0.39606592059135437,
+        "learning_rate": 0.0005,
+        "loss": 1.1491,
+        "step": 16620
+      },
+      {
+        "epoch": 0.5303102777512038,
+        "grad_norm": 0.39793723821640015,
+        "learning_rate": 0.0005,
+        "loss": 1.1312,
+        "step": 16630
+      },
+      {
+        "epoch": 0.5306291654708377,
+        "grad_norm": 0.38791102170944214,
+        "learning_rate": 0.0005,
+        "loss": 1.1274,
+        "step": 16640
+      },
+      {
+        "epoch": 0.5309480531904717,
+        "grad_norm": 0.4097859263420105,
+        "learning_rate": 0.0005,
+        "loss": 1.1315,
+        "step": 16650
+      },
+      {
+        "epoch": 0.5312669409101055,
+        "grad_norm": 0.3862121105194092,
+        "learning_rate": 0.0005,
+        "loss": 1.1332,
+        "step": 16660
+      },
+      {
+        "epoch": 0.5315858286297395,
+        "grad_norm": 0.3790407180786133,
+        "learning_rate": 0.0005,
+        "loss": 1.1102,
+        "step": 16670
+      },
+      {
+        "epoch": 0.5319047163493734,
+        "grad_norm": 0.39420583844184875,
+        "learning_rate": 0.0005,
+        "loss": 1.1593,
+        "step": 16680
+      },
+      {
+        "epoch": 0.5322236040690073,
+        "grad_norm": 0.38086998462677,
+        "learning_rate": 0.0005,
+        "loss": 1.1187,
+        "step": 16690
+      },
+      {
+        "epoch": 0.5325424917886412,
+        "grad_norm": 0.3970024287700653,
+        "learning_rate": 0.0005,
+        "loss": 1.1196,
+        "step": 16700
+      },
+      {
+        "epoch": 0.5328613795082752,
+        "grad_norm": 0.40243950486183167,
+        "learning_rate": 0.0005,
+        "loss": 1.1154,
+        "step": 16710
+      },
+      {
+        "epoch": 0.533180267227909,
+        "grad_norm": 0.392729789018631,
+        "learning_rate": 0.0005,
+        "loss": 1.14,
+        "step": 16720
+      },
+      {
+        "epoch": 0.533499154947543,
+        "grad_norm": 0.4108237028121948,
+        "learning_rate": 0.0005,
+        "loss": 1.1352,
+        "step": 16730
+      },
+      {
+        "epoch": 0.5338180426671769,
+        "grad_norm": 0.3943747580051422,
+        "learning_rate": 0.0005,
+        "loss": 1.1215,
+        "step": 16740
+      },
+      {
+        "epoch": 0.5341369303868108,
+        "grad_norm": 0.4041931927204132,
+        "learning_rate": 0.0005,
+        "loss": 1.1134,
+        "step": 16750
+      },
+      {
+        "epoch": 0.5344558181064447,
+        "grad_norm": 0.3883357346057892,
+        "learning_rate": 0.0005,
+        "loss": 1.1344,
+        "step": 16760
+      },
+      {
+        "epoch": 0.5347747058260787,
+        "grad_norm": 0.38386496901512146,
+        "learning_rate": 0.0005,
+        "loss": 1.1285,
+        "step": 16770
+      },
+      {
+        "epoch": 0.5350935935457125,
+        "grad_norm": 0.3961685597896576,
+        "learning_rate": 0.0005,
+        "loss": 1.1328,
+        "step": 16780
+      },
+      {
+        "epoch": 0.5354124812653465,
+        "grad_norm": 0.4196777939796448,
+        "learning_rate": 0.0005,
+        "loss": 1.1319,
+        "step": 16790
+      },
+      {
+        "epoch": 0.5357313689849804,
+        "grad_norm": 0.39921364188194275,
+        "learning_rate": 0.0005,
+        "loss": 1.1217,
+        "step": 16800
+      },
+      {
+        "epoch": 0.5360502567046143,
+        "grad_norm": 0.40605226159095764,
+        "learning_rate": 0.0005,
+        "loss": 1.1503,
+        "step": 16810
+      },
+      {
+        "epoch": 0.5363691444242482,
+        "grad_norm": 0.4120405614376068,
+        "learning_rate": 0.0005,
+        "loss": 1.1284,
+        "step": 16820
+      },
+      {
+        "epoch": 0.5366880321438822,
+        "grad_norm": 0.3855202794075012,
+        "learning_rate": 0.0005,
+        "loss": 1.1233,
+        "step": 16830
+      },
+      {
+        "epoch": 0.537006919863516,
+        "grad_norm": 0.4094316065311432,
+        "learning_rate": 0.0005,
+        "loss": 1.1495,
+        "step": 16840
+      },
+      {
+        "epoch": 0.53732580758315,
+        "grad_norm": 0.395519495010376,
+        "learning_rate": 0.0005,
+        "loss": 1.131,
+        "step": 16850
+      },
+      {
+        "epoch": 0.5376446953027839,
+        "grad_norm": 0.3976709544658661,
+        "learning_rate": 0.0005,
+        "loss": 1.1316,
+        "step": 16860
+      },
+      {
+        "epoch": 0.5379635830224178,
+        "grad_norm": 0.38913607597351074,
+        "learning_rate": 0.0005,
+        "loss": 1.1177,
+        "step": 16870
+      },
+      {
+        "epoch": 0.5382824707420517,
+        "grad_norm": 0.3925860524177551,
+        "learning_rate": 0.0005,
+        "loss": 1.1052,
+        "step": 16880
+      },
+      {
+        "epoch": 0.5386013584616857,
+        "grad_norm": 0.384029358625412,
+        "learning_rate": 0.0005,
+        "loss": 1.1369,
+        "step": 16890
+      },
+      {
+        "epoch": 0.5389202461813195,
+        "grad_norm": 0.3836941123008728,
+        "learning_rate": 0.0005,
+        "loss": 1.1342,
+        "step": 16900
+      },
+      {
+        "epoch": 0.5392391339009535,
+        "grad_norm": 0.39144232869148254,
+        "learning_rate": 0.0005,
+        "loss": 1.112,
+        "step": 16910
+      },
+      {
+        "epoch": 0.5395580216205874,
+        "grad_norm": 0.3911305069923401,
+        "learning_rate": 0.0005,
+        "loss": 1.14,
+        "step": 16920
+      },
+      {
+        "epoch": 0.5398769093402213,
+        "grad_norm": 0.39506804943084717,
+        "learning_rate": 0.0005,
+        "loss": 1.1223,
+        "step": 16930
+      },
+      {
+        "epoch": 0.5401957970598552,
+        "grad_norm": 0.39567089080810547,
+        "learning_rate": 0.0005,
+        "loss": 1.1295,
+        "step": 16940
+      },
+      {
+        "epoch": 0.5405146847794892,
+        "grad_norm": 0.40075957775115967,
+        "learning_rate": 0.0005,
+        "loss": 1.119,
+        "step": 16950
+      },
+      {
+        "epoch": 0.540833572499123,
+        "grad_norm": 0.42073139548301697,
+        "learning_rate": 0.0005,
+        "loss": 1.1242,
+        "step": 16960
+      },
+      {
+        "epoch": 0.541152460218757,
+        "grad_norm": 0.38740274310112,
+        "learning_rate": 0.0005,
+        "loss": 1.124,
+        "step": 16970
+      },
+      {
+        "epoch": 0.5414713479383909,
+        "grad_norm": 0.39646822214126587,
+        "learning_rate": 0.0005,
+        "loss": 1.1167,
+        "step": 16980
+      },
+      {
+        "epoch": 0.5417902356580248,
+        "grad_norm": 0.3964032232761383,
+        "learning_rate": 0.0005,
+        "loss": 1.1422,
+        "step": 16990
+      },
+      {
+        "epoch": 0.5421091233776587,
+        "grad_norm": 0.38478758931159973,
+        "learning_rate": 0.0005,
+        "loss": 1.1057,
+        "step": 17000
+      },
+      {
+        "epoch": 0.5424280110972927,
+        "grad_norm": 0.3826146423816681,
+        "learning_rate": 0.0005,
+        "loss": 1.1097,
+        "step": 17010
+      },
+      {
+        "epoch": 0.5427468988169266,
+        "grad_norm": 0.3929329514503479,
+        "learning_rate": 0.0005,
+        "loss": 1.0978,
+        "step": 17020
+      },
+      {
+        "epoch": 0.5430657865365605,
+        "grad_norm": 0.3981013000011444,
+        "learning_rate": 0.0005,
+        "loss": 1.1299,
+        "step": 17030
+      },
+      {
+        "epoch": 0.5433846742561944,
+        "grad_norm": 0.3714570701122284,
+        "learning_rate": 0.0005,
+        "loss": 1.1061,
+        "step": 17040
+      },
+      {
+        "epoch": 0.5437035619758284,
+        "grad_norm": 0.38850119709968567,
+        "learning_rate": 0.0005,
+        "loss": 1.1096,
+        "step": 17050
+      },
+      {
+        "epoch": 0.5440224496954622,
+        "grad_norm": 0.3953388035297394,
+        "learning_rate": 0.0005,
+        "loss": 1.1347,
+        "step": 17060
+      },
+      {
+        "epoch": 0.5443413374150962,
+        "grad_norm": 0.3949526846408844,
+        "learning_rate": 0.0005,
+        "loss": 1.1354,
+        "step": 17070
+      },
+      {
+        "epoch": 0.5446602251347301,
+        "grad_norm": 0.39056915044784546,
+        "learning_rate": 0.0005,
+        "loss": 1.1398,
+        "step": 17080
+      },
+      {
+        "epoch": 0.544979112854364,
+        "grad_norm": 0.3963756561279297,
+        "learning_rate": 0.0005,
+        "loss": 1.1321,
+        "step": 17090
+      },
+      {
+        "epoch": 0.5452980005739979,
+        "grad_norm": 0.39296120405197144,
+        "learning_rate": 0.0005,
+        "loss": 1.118,
+        "step": 17100
+      },
+      {
+        "epoch": 0.5456168882936319,
+        "grad_norm": 0.39938679337501526,
+        "learning_rate": 0.0005,
+        "loss": 1.1324,
+        "step": 17110
+      },
+      {
+        "epoch": 0.5459357760132657,
+        "grad_norm": 0.4136159121990204,
+        "learning_rate": 0.0005,
+        "loss": 1.1346,
+        "step": 17120
+      },
+      {
+        "epoch": 0.5462546637328997,
+        "grad_norm": 0.39979586005210876,
+        "learning_rate": 0.0005,
+        "loss": 1.1239,
+        "step": 17130
+      },
+      {
+        "epoch": 0.5465735514525336,
+        "grad_norm": 0.3993443548679352,
+        "learning_rate": 0.0005,
+        "loss": 1.1163,
+        "step": 17140
+      },
+      {
+        "epoch": 0.5468924391721675,
+        "grad_norm": 0.3846668601036072,
+        "learning_rate": 0.0005,
+        "loss": 1.1193,
+        "step": 17150
+      },
+      {
+        "epoch": 0.5472113268918014,
+        "grad_norm": 0.3863378167152405,
+        "learning_rate": 0.0005,
+        "loss": 1.1212,
+        "step": 17160
+      },
+      {
+        "epoch": 0.5475302146114354,
+        "grad_norm": 0.3799877166748047,
+        "learning_rate": 0.0005,
+        "loss": 1.1148,
+        "step": 17170
+      },
+      {
+        "epoch": 0.5478491023310692,
+        "grad_norm": 0.40554550290107727,
+        "learning_rate": 0.0005,
+        "loss": 1.1378,
+        "step": 17180
+      },
+      {
+        "epoch": 0.5481679900507032,
+        "grad_norm": 0.3773590922355652,
+        "learning_rate": 0.0005,
+        "loss": 1.1133,
+        "step": 17190
+      },
+      {
+        "epoch": 0.5484868777703371,
+        "grad_norm": 0.39433354139328003,
+        "learning_rate": 0.0005,
+        "loss": 1.1343,
+        "step": 17200
+      },
+      {
+        "epoch": 0.548805765489971,
+        "grad_norm": 0.3862454891204834,
+        "learning_rate": 0.0005,
+        "loss": 1.1232,
+        "step": 17210
+      },
+      {
+        "epoch": 0.5491246532096049,
+        "grad_norm": 0.38670477271080017,
+        "learning_rate": 0.0005,
+        "loss": 1.1136,
+        "step": 17220
+      },
+      {
+        "epoch": 0.5494435409292389,
+        "grad_norm": 0.39055192470550537,
+        "learning_rate": 0.0005,
+        "loss": 1.1129,
+        "step": 17230
+      },
+      {
+        "epoch": 0.5497624286488727,
+        "grad_norm": 0.40744680166244507,
+        "learning_rate": 0.0005,
+        "loss": 1.13,
+        "step": 17240
+      },
+      {
+        "epoch": 0.5500813163685067,
+        "grad_norm": 0.38350918889045715,
+        "learning_rate": 0.0005,
+        "loss": 1.1296,
+        "step": 17250
+      },
+      {
+        "epoch": 0.5504002040881406,
+        "grad_norm": 0.4025372564792633,
+        "learning_rate": 0.0005,
+        "loss": 1.1229,
+        "step": 17260
+      },
+      {
+        "epoch": 0.5507190918077745,
+        "grad_norm": 0.3817196190357208,
+        "learning_rate": 0.0005,
+        "loss": 1.1481,
+        "step": 17270
+      },
+      {
+        "epoch": 0.5510379795274084,
+        "grad_norm": 0.37860438227653503,
+        "learning_rate": 0.0005,
+        "loss": 1.1058,
+        "step": 17280
+      },
+      {
+        "epoch": 0.5513568672470424,
+        "grad_norm": 0.41163647174835205,
+        "learning_rate": 0.0005,
+        "loss": 1.1423,
+        "step": 17290
+      },
+      {
+        "epoch": 0.5516757549666762,
+        "grad_norm": 0.389542818069458,
+        "learning_rate": 0.0005,
+        "loss": 1.1164,
+        "step": 17300
+      },
+      {
+        "epoch": 0.5519946426863102,
+        "grad_norm": 0.39946138858795166,
+        "learning_rate": 0.0005,
+        "loss": 1.1228,
+        "step": 17310
+      },
+      {
+        "epoch": 0.5523135304059441,
+        "grad_norm": 0.3970085084438324,
+        "learning_rate": 0.0005,
+        "loss": 1.1236,
+        "step": 17320
+      },
+      {
+        "epoch": 0.552632418125578,
+        "grad_norm": 0.3923346996307373,
+        "learning_rate": 0.0005,
+        "loss": 1.1115,
+        "step": 17330
+      },
+      {
+        "epoch": 0.5529513058452119,
+        "grad_norm": 0.3953569531440735,
+        "learning_rate": 0.0005,
+        "loss": 1.1243,
+        "step": 17340
+      },
+      {
+        "epoch": 0.5532701935648459,
+        "grad_norm": 0.4045715034008026,
+        "learning_rate": 0.0005,
+        "loss": 1.1226,
+        "step": 17350
+      },
+      {
+        "epoch": 0.5535890812844797,
+        "grad_norm": 0.38618502020835876,
+        "learning_rate": 0.0005,
+        "loss": 1.1002,
+        "step": 17360
+      },
+      {
+        "epoch": 0.5539079690041137,
+        "grad_norm": 0.38182103633880615,
+        "learning_rate": 0.0005,
+        "loss": 1.1163,
+        "step": 17370
+      },
+      {
+        "epoch": 0.5542268567237476,
+        "grad_norm": 0.3986895978450775,
+        "learning_rate": 0.0005,
+        "loss": 1.1312,
+        "step": 17380
+      },
+      {
+        "epoch": 0.5545457444433814,
+        "grad_norm": 0.4195130169391632,
+        "learning_rate": 0.0005,
+        "loss": 1.1188,
+        "step": 17390
+      },
+      {
+        "epoch": 0.5548646321630154,
+        "grad_norm": 0.3816862404346466,
+        "learning_rate": 0.0005,
+        "loss": 1.1481,
+        "step": 17400
+      },
+      {
+        "epoch": 0.5551835198826494,
+        "grad_norm": 0.3993696868419647,
+        "learning_rate": 0.0005,
+        "loss": 1.1273,
+        "step": 17410
+      },
+      {
+        "epoch": 0.5555024076022832,
+        "grad_norm": 0.38697922229766846,
+        "learning_rate": 0.0005,
+        "loss": 1.1092,
+        "step": 17420
+      },
+      {
+        "epoch": 0.5558212953219172,
+        "grad_norm": 0.40851518511772156,
+        "learning_rate": 0.0005,
+        "loss": 1.1213,
+        "step": 17430
+      },
+      {
+        "epoch": 0.5561401830415511,
+        "grad_norm": 0.3991389870643616,
+        "learning_rate": 0.0005,
+        "loss": 1.1257,
+        "step": 17440
+      },
+      {
+        "epoch": 0.556459070761185,
+        "grad_norm": 0.392604798078537,
+        "learning_rate": 0.0005,
+        "loss": 1.1252,
+        "step": 17450
+      },
+      {
+        "epoch": 0.5567779584808189,
+        "grad_norm": 0.3909078538417816,
+        "learning_rate": 0.0005,
+        "loss": 1.1249,
+        "step": 17460
+      },
+      {
+        "epoch": 0.5570968462004529,
+        "grad_norm": 0.3915567100048065,
+        "learning_rate": 0.0005,
+        "loss": 1.1257,
+        "step": 17470
+      },
+      {
+        "epoch": 0.5574157339200867,
+        "grad_norm": 0.38957881927490234,
+        "learning_rate": 0.0005,
+        "loss": 1.1027,
+        "step": 17480
+      },
+      {
+        "epoch": 0.5577346216397207,
+        "grad_norm": 0.38382086157798767,
+        "learning_rate": 0.0005,
+        "loss": 1.1296,
+        "step": 17490
+      },
+      {
+        "epoch": 0.5580535093593546,
+        "grad_norm": 0.39388954639434814,
+        "learning_rate": 0.0005,
+        "loss": 1.1249,
+        "step": 17500
+      },
+      {
+        "epoch": 0.5583723970789884,
+        "grad_norm": 0.395594984292984,
+        "learning_rate": 0.0005,
+        "loss": 1.1243,
+        "step": 17510
+      },
+      {
+        "epoch": 0.5586912847986224,
+        "grad_norm": 0.4204639792442322,
+        "learning_rate": 0.0005,
+        "loss": 1.1221,
+        "step": 17520
+      },
+      {
+        "epoch": 0.5590101725182564,
+        "grad_norm": 0.41280484199523926,
+        "learning_rate": 0.0005,
+        "loss": 1.1094,
+        "step": 17530
+      },
+      {
+        "epoch": 0.5593290602378902,
+        "grad_norm": 0.37822672724723816,
+        "learning_rate": 0.0005,
+        "loss": 1.1089,
+        "step": 17540
+      },
+      {
+        "epoch": 0.5596479479575242,
+        "grad_norm": 0.38607358932495117,
+        "learning_rate": 0.0005,
+        "loss": 1.1461,
+        "step": 17550
+      },
+      {
+        "epoch": 0.5599668356771581,
+        "grad_norm": 0.3974449038505554,
+        "learning_rate": 0.0005,
+        "loss": 1.1057,
+        "step": 17560
+      },
+      {
+        "epoch": 0.560285723396792,
+        "grad_norm": 0.39886006712913513,
+        "learning_rate": 0.0005,
+        "loss": 1.1197,
+        "step": 17570
+      },
+      {
+        "epoch": 0.5606046111164259,
+        "grad_norm": 0.4036238193511963,
+        "learning_rate": 0.0005,
+        "loss": 1.1223,
+        "step": 17580
+      },
+      {
+        "epoch": 0.5609234988360599,
+        "grad_norm": 0.40054237842559814,
+        "learning_rate": 0.0005,
+        "loss": 1.1176,
+        "step": 17590
+      },
+      {
+        "epoch": 0.5612423865556937,
+        "grad_norm": 0.38854557275772095,
+        "learning_rate": 0.0005,
+        "loss": 1.1162,
+        "step": 17600
+      },
+      {
+        "epoch": 0.5615612742753276,
+        "grad_norm": 0.38694027066230774,
+        "learning_rate": 0.0005,
+        "loss": 1.121,
+        "step": 17610
+      },
+      {
+        "epoch": 0.5618801619949616,
+        "grad_norm": 0.37990602850914,
+        "learning_rate": 0.0005,
+        "loss": 1.1177,
+        "step": 17620
+      },
+      {
+        "epoch": 0.5621990497145954,
+        "grad_norm": 0.39579102396965027,
+        "learning_rate": 0.0005,
+        "loss": 1.1269,
+        "step": 17630
+      },
+      {
+        "epoch": 0.5625179374342294,
+        "grad_norm": 0.39988967776298523,
+        "learning_rate": 0.0005,
+        "loss": 1.1112,
+        "step": 17640
+      },
+      {
+        "epoch": 0.5628368251538634,
+        "grad_norm": 0.3899117410182953,
+        "learning_rate": 0.0005,
+        "loss": 1.0975,
+        "step": 17650
+      },
+      {
+        "epoch": 0.5631557128734972,
+        "grad_norm": 0.391559362411499,
+        "learning_rate": 0.0005,
+        "loss": 1.1202,
+        "step": 17660
+      },
+      {
+        "epoch": 0.5634746005931311,
+        "grad_norm": 0.3928084075450897,
+        "learning_rate": 0.0005,
+        "loss": 1.1191,
+        "step": 17670
+      },
+      {
+        "epoch": 0.5637934883127651,
+        "grad_norm": 0.4055452048778534,
+        "learning_rate": 0.0005,
+        "loss": 1.114,
+        "step": 17680
+      },
+      {
+        "epoch": 0.5641123760323989,
+        "grad_norm": 0.39144572615623474,
+        "learning_rate": 0.0005,
+        "loss": 1.1277,
+        "step": 17690
+      },
+      {
+        "epoch": 0.5644312637520329,
+        "grad_norm": 0.3878569006919861,
+        "learning_rate": 0.0005,
+        "loss": 1.1205,
+        "step": 17700
+      },
+      {
+        "epoch": 0.5647501514716669,
+        "grad_norm": 0.3888893127441406,
+        "learning_rate": 0.0005,
+        "loss": 1.1238,
+        "step": 17710
+      },
+      {
+        "epoch": 0.5650690391913007,
+        "grad_norm": 0.38882070779800415,
+        "learning_rate": 0.0005,
+        "loss": 1.1176,
+        "step": 17720
+      },
+      {
+        "epoch": 0.5653879269109346,
+        "grad_norm": 0.3775336444377899,
+        "learning_rate": 0.0005,
+        "loss": 1.0967,
+        "step": 17730
+      },
+      {
+        "epoch": 0.5657068146305686,
+        "grad_norm": 0.386237233877182,
+        "learning_rate": 0.0005,
+        "loss": 1.1149,
+        "step": 17740
+      },
+      {
+        "epoch": 0.5660257023502024,
+        "grad_norm": 0.38431134819984436,
+        "learning_rate": 0.0005,
+        "loss": 1.1073,
+        "step": 17750
+      },
+      {
+        "epoch": 0.5663445900698364,
+        "grad_norm": 0.4077931344509125,
+        "learning_rate": 0.0005,
+        "loss": 1.1196,
+        "step": 17760
+      },
+      {
+        "epoch": 0.5666634777894703,
+        "grad_norm": 0.397979736328125,
+        "learning_rate": 0.0005,
+        "loss": 1.1078,
+        "step": 17770
+      },
+      {
+        "epoch": 0.5669823655091042,
+        "grad_norm": 0.40055903792381287,
+        "learning_rate": 0.0005,
+        "loss": 1.1186,
+        "step": 17780
+      },
+      {
+        "epoch": 0.5673012532287381,
+        "grad_norm": 0.40928784012794495,
+        "learning_rate": 0.0005,
+        "loss": 1.1277,
+        "step": 17790
+      },
+      {
+        "epoch": 0.5676201409483721,
+        "grad_norm": 0.38073426485061646,
+        "learning_rate": 0.0005,
+        "loss": 1.0975,
+        "step": 17800
+      },
+      {
+        "epoch": 0.5679390286680059,
+        "grad_norm": 0.39049458503723145,
+        "learning_rate": 0.0005,
+        "loss": 1.1193,
+        "step": 17810
+      },
+      {
+        "epoch": 0.5682579163876399,
+        "grad_norm": 0.39856138825416565,
+        "learning_rate": 0.0005,
+        "loss": 1.1151,
+        "step": 17820
+      },
+      {
+        "epoch": 0.5685768041072738,
+        "grad_norm": 0.3846971094608307,
+        "learning_rate": 0.0005,
+        "loss": 1.1088,
+        "step": 17830
+      },
+      {
+        "epoch": 0.5688956918269078,
+        "grad_norm": 0.4002894461154938,
+        "learning_rate": 0.0005,
+        "loss": 1.105,
+        "step": 17840
+      },
+      {
+        "epoch": 0.5692145795465416,
+        "grad_norm": 0.3965441584587097,
+        "learning_rate": 0.0005,
+        "loss": 1.1151,
+        "step": 17850
+      },
+      {
+        "epoch": 0.5695334672661756,
+        "grad_norm": 0.38790929317474365,
+        "learning_rate": 0.0005,
+        "loss": 1.109,
+        "step": 17860
+      },
+      {
+        "epoch": 0.5698523549858096,
+        "grad_norm": 0.3736358880996704,
+        "learning_rate": 0.0005,
+        "loss": 1.1091,
+        "step": 17870
+      },
+      {
+        "epoch": 0.5701712427054434,
+        "grad_norm": 0.3990086317062378,
+        "learning_rate": 0.0005,
+        "loss": 1.1215,
+        "step": 17880
+      },
+      {
+        "epoch": 0.5704901304250773,
+        "grad_norm": 0.3745183050632477,
+        "learning_rate": 0.0005,
+        "loss": 1.1109,
+        "step": 17890
+      },
+      {
+        "epoch": 0.5708090181447113,
+        "grad_norm": 0.39467811584472656,
+        "learning_rate": 0.0005,
+        "loss": 1.1141,
+        "step": 17900
+      },
+      {
+        "epoch": 0.5711279058643451,
+        "grad_norm": 0.3837774395942688,
+        "learning_rate": 0.0005,
+        "loss": 1.1156,
+        "step": 17910
+      },
+      {
+        "epoch": 0.5714467935839791,
+        "grad_norm": 0.3765852153301239,
+        "learning_rate": 0.0005,
+        "loss": 1.1126,
+        "step": 17920
+      },
+      {
+        "epoch": 0.571765681303613,
+        "grad_norm": 0.37944474816322327,
+        "learning_rate": 0.0005,
+        "loss": 1.1152,
+        "step": 17930
+      },
+      {
+        "epoch": 0.5720845690232469,
+        "grad_norm": 0.4042120575904846,
+        "learning_rate": 0.0005,
+        "loss": 1.1293,
+        "step": 17940
+      },
+      {
+        "epoch": 0.5724034567428808,
+        "grad_norm": 0.39046669006347656,
+        "learning_rate": 0.0005,
+        "loss": 1.1078,
+        "step": 17950
+      },
+      {
+        "epoch": 0.5727223444625148,
+        "grad_norm": 0.39052364230155945,
+        "learning_rate": 0.0005,
+        "loss": 1.0919,
+        "step": 17960
+      },
+      {
+        "epoch": 0.5730412321821486,
+        "grad_norm": 0.3719790577888489,
+        "learning_rate": 0.0005,
+        "loss": 1.108,
+        "step": 17970
+      },
+      {
+        "epoch": 0.5733601199017826,
+        "grad_norm": 0.4557461738586426,
+        "learning_rate": 0.0005,
+        "loss": 1.1243,
+        "step": 17980
+      },
+      {
+        "epoch": 0.5736790076214165,
+        "grad_norm": 0.3853502571582794,
+        "learning_rate": 0.0005,
+        "loss": 1.1299,
+        "step": 17990
+      },
+      {
+        "epoch": 0.5739978953410504,
+        "grad_norm": 0.4154195785522461,
+        "learning_rate": 0.0005,
+        "loss": 1.1106,
+        "step": 18000
+      },
+      {
+        "epoch": 0.5743167830606843,
+        "grad_norm": 0.37731653451919556,
+        "learning_rate": 0.0005,
+        "loss": 1.105,
+        "step": 18010
+      },
+      {
+        "epoch": 0.5746356707803183,
+        "grad_norm": 0.40308648347854614,
+        "learning_rate": 0.0005,
+        "loss": 1.1173,
+        "step": 18020
+      },
+      {
+        "epoch": 0.5749545584999521,
+        "grad_norm": 0.3943016231060028,
+        "learning_rate": 0.0005,
+        "loss": 1.1209,
+        "step": 18030
+      },
+      {
+        "epoch": 0.5752734462195861,
+        "grad_norm": 0.3769054114818573,
+        "learning_rate": 0.0005,
+        "loss": 1.0992,
+        "step": 18040
+      },
+      {
+        "epoch": 0.57559233393922,
+        "grad_norm": 0.40331166982650757,
+        "learning_rate": 0.0005,
+        "loss": 1.1115,
+        "step": 18050
+      },
+      {
+        "epoch": 0.5759112216588539,
+        "grad_norm": 0.40439677238464355,
+        "learning_rate": 0.0005,
+        "loss": 1.1066,
+        "step": 18060
+      },
+      {
+        "epoch": 0.5762301093784878,
+        "grad_norm": 0.3903745710849762,
+        "learning_rate": 0.0005,
+        "loss": 1.1162,
+        "step": 18070
+      },
+      {
+        "epoch": 0.5765489970981218,
+        "grad_norm": 0.388918936252594,
+        "learning_rate": 0.0005,
+        "loss": 1.1097,
+        "step": 18080
+      },
+      {
+        "epoch": 0.5768678848177556,
+        "grad_norm": 0.40826553106307983,
+        "learning_rate": 0.0005,
+        "loss": 1.1326,
+        "step": 18090
+      },
+      {
+        "epoch": 0.5771867725373896,
+        "grad_norm": 0.385471910238266,
+        "learning_rate": 0.0005,
+        "loss": 1.1257,
+        "step": 18100
+      },
+      {
+        "epoch": 0.5775056602570235,
+        "grad_norm": 0.3771023750305176,
+        "learning_rate": 0.0005,
+        "loss": 1.1121,
+        "step": 18110
+      },
+      {
+        "epoch": 0.5778245479766574,
+        "grad_norm": 0.40232816338539124,
+        "learning_rate": 0.0005,
+        "loss": 1.1218,
+        "step": 18120
+      },
+      {
+        "epoch": 0.5781434356962913,
+        "grad_norm": 0.3964270353317261,
+        "learning_rate": 0.0005,
+        "loss": 1.0979,
+        "step": 18130
+      },
+      {
+        "epoch": 0.5784623234159253,
+        "grad_norm": 0.3804101049900055,
+        "learning_rate": 0.0005,
+        "loss": 1.114,
+        "step": 18140
+      },
+      {
+        "epoch": 0.5787812111355591,
+        "grad_norm": 0.3876723647117615,
+        "learning_rate": 0.0005,
+        "loss": 1.0882,
+        "step": 18150
+      },
+      {
+        "epoch": 0.5791000988551931,
+        "grad_norm": 0.39598479866981506,
+        "learning_rate": 0.0005,
+        "loss": 1.1089,
+        "step": 18160
+      },
+      {
+        "epoch": 0.579418986574827,
+        "grad_norm": 0.40432387590408325,
+        "learning_rate": 0.0005,
+        "loss": 1.0953,
+        "step": 18170
+      },
+      {
+        "epoch": 0.5797378742944609,
+        "grad_norm": 0.37542155385017395,
+        "learning_rate": 0.0005,
+        "loss": 1.1234,
+        "step": 18180
+      },
+      {
+        "epoch": 0.5800567620140948,
+        "grad_norm": 0.379314124584198,
+        "learning_rate": 0.0005,
+        "loss": 1.1158,
+        "step": 18190
+      },
+      {
+        "epoch": 0.5803756497337288,
+        "grad_norm": 0.3933337330818176,
+        "learning_rate": 0.0005,
+        "loss": 1.1233,
+        "step": 18200
+      },
+      {
+        "epoch": 0.5806945374533626,
+        "grad_norm": 0.3895704448223114,
+        "learning_rate": 0.0005,
+        "loss": 1.1099,
+        "step": 18210
+      },
+      {
+        "epoch": 0.5810134251729966,
+        "grad_norm": 0.37216246128082275,
+        "learning_rate": 0.0005,
+        "loss": 1.1094,
+        "step": 18220
+      },
+      {
+        "epoch": 0.5813323128926305,
+        "grad_norm": 0.37889426946640015,
+        "learning_rate": 0.0005,
+        "loss": 1.1331,
+        "step": 18230
+      },
+      {
+        "epoch": 0.5816512006122644,
+        "grad_norm": 0.3779814839363098,
+        "learning_rate": 0.0005,
+        "loss": 1.1053,
+        "step": 18240
+      },
+      {
+        "epoch": 0.5819700883318983,
+        "grad_norm": 0.384083092212677,
+        "learning_rate": 0.0005,
+        "loss": 1.1169,
+        "step": 18250
+      },
+      {
+        "epoch": 0.5822889760515323,
+        "grad_norm": 0.39483460783958435,
+        "learning_rate": 0.0005,
+        "loss": 1.1075,
+        "step": 18260
+      },
+      {
+        "epoch": 0.5826078637711661,
+        "grad_norm": 0.4070870280265808,
+        "learning_rate": 0.0005,
+        "loss": 1.1181,
+        "step": 18270
+      },
+      {
+        "epoch": 0.5829267514908001,
+        "grad_norm": 0.396951287984848,
+        "learning_rate": 0.0005,
+        "loss": 1.1002,
+        "step": 18280
+      },
+      {
+        "epoch": 0.583245639210434,
+        "grad_norm": 0.3835306763648987,
+        "learning_rate": 0.0005,
+        "loss": 1.1126,
+        "step": 18290
+      },
+      {
+        "epoch": 0.5835645269300679,
+        "grad_norm": 0.3818987309932709,
+        "learning_rate": 0.0005,
+        "loss": 1.1075,
+        "step": 18300
+      },
+      {
+        "epoch": 0.5838834146497018,
+        "grad_norm": 0.3779695928096771,
+        "learning_rate": 0.0005,
+        "loss": 1.0975,
+        "step": 18310
+      },
+      {
+        "epoch": 0.5842023023693358,
+        "grad_norm": 0.3806639611721039,
+        "learning_rate": 0.0005,
+        "loss": 1.1089,
+        "step": 18320
+      },
+      {
+        "epoch": 0.5845211900889696,
+        "grad_norm": 0.3965880870819092,
+        "learning_rate": 0.0005,
+        "loss": 1.1147,
+        "step": 18330
+      },
+      {
+        "epoch": 0.5848400778086036,
+        "grad_norm": 0.38795900344848633,
+        "learning_rate": 0.0005,
+        "loss": 1.106,
+        "step": 18340
+      },
+      {
+        "epoch": 0.5851589655282375,
+        "grad_norm": 0.3765144646167755,
+        "learning_rate": 0.0005,
+        "loss": 1.1132,
+        "step": 18350
+      },
+      {
+        "epoch": 0.5854778532478714,
+        "grad_norm": 0.374661386013031,
+        "learning_rate": 0.0005,
+        "loss": 1.1295,
+        "step": 18360
+      },
+      {
+        "epoch": 0.5857967409675053,
+        "grad_norm": 0.388966828584671,
+        "learning_rate": 0.0005,
+        "loss": 1.1295,
+        "step": 18370
+      },
+      {
+        "epoch": 0.5861156286871393,
+        "grad_norm": 0.4160568416118622,
+        "learning_rate": 0.0005,
+        "loss": 1.1335,
+        "step": 18380
+      },
+      {
+        "epoch": 0.5864345164067731,
+        "grad_norm": 0.39939290285110474,
+        "learning_rate": 0.0005,
+        "loss": 1.1024,
+        "step": 18390
+      },
+      {
+        "epoch": 0.5867534041264071,
+        "grad_norm": 0.3849220275878906,
+        "learning_rate": 0.0005,
+        "loss": 1.1041,
+        "step": 18400
+      },
+      {
+        "epoch": 0.587072291846041,
+        "grad_norm": 0.3832233250141144,
+        "learning_rate": 0.0005,
+        "loss": 1.1084,
+        "step": 18410
+      },
+      {
+        "epoch": 0.5873911795656749,
+        "grad_norm": 0.3949020504951477,
+        "learning_rate": 0.0005,
+        "loss": 1.1269,
+        "step": 18420
+      },
+      {
+        "epoch": 0.5877100672853088,
+        "grad_norm": 0.39243316650390625,
+        "learning_rate": 0.0005,
+        "loss": 1.1221,
+        "step": 18430
+      },
+      {
+        "epoch": 0.5880289550049428,
+        "grad_norm": 0.3923371136188507,
+        "learning_rate": 0.0005,
+        "loss": 1.1198,
+        "step": 18440
+      },
+      {
+        "epoch": 0.5883478427245766,
+        "grad_norm": 0.3877916932106018,
+        "learning_rate": 0.0005,
+        "loss": 1.1132,
+        "step": 18450
+      },
+      {
+        "epoch": 0.5886667304442106,
+        "grad_norm": 0.3907372057437897,
+        "learning_rate": 0.0005,
+        "loss": 1.0928,
+        "step": 18460
+      },
+      {
+        "epoch": 0.5889856181638445,
+        "grad_norm": 0.39084115624427795,
+        "learning_rate": 0.0005,
+        "loss": 1.1069,
+        "step": 18470
+      },
+      {
+        "epoch": 0.5893045058834784,
+        "grad_norm": 0.374729186296463,
+        "learning_rate": 0.0005,
+        "loss": 1.1187,
+        "step": 18480
+      },
+      {
+        "epoch": 0.5896233936031123,
+        "grad_norm": 0.36638641357421875,
+        "learning_rate": 0.0005,
+        "loss": 1.1038,
+        "step": 18490
+      },
+      {
+        "epoch": 0.5899422813227463,
+        "grad_norm": 0.3777466416358948,
+        "learning_rate": 0.0005,
+        "loss": 1.1147,
+        "step": 18500
+      },
+      {
+        "epoch": 0.5902611690423801,
+        "grad_norm": 0.3874850869178772,
+        "learning_rate": 0.0005,
+        "loss": 1.1116,
+        "step": 18510
+      },
+      {
+        "epoch": 0.5905800567620141,
+        "grad_norm": 0.4039003252983093,
+        "learning_rate": 0.0005,
+        "loss": 1.1216,
+        "step": 18520
+      },
+      {
+        "epoch": 0.590898944481648,
+        "grad_norm": 0.4028369188308716,
+        "learning_rate": 0.0005,
+        "loss": 1.1036,
+        "step": 18530
+      },
+      {
+        "epoch": 0.5912178322012819,
+        "grad_norm": 0.39203813672065735,
+        "learning_rate": 0.0005,
+        "loss": 1.1149,
+        "step": 18540
+      },
+      {
+        "epoch": 0.5915367199209158,
+        "grad_norm": 0.3853701651096344,
+        "learning_rate": 0.0005,
+        "loss": 1.1192,
+        "step": 18550
+      },
+      {
+        "epoch": 0.5918556076405498,
+        "grad_norm": 0.389746755361557,
+        "learning_rate": 0.0005,
+        "loss": 1.132,
+        "step": 18560
+      },
+      {
+        "epoch": 0.5921744953601836,
+        "grad_norm": 0.3897966742515564,
+        "learning_rate": 0.0005,
+        "loss": 1.0949,
+        "step": 18570
+      },
+      {
+        "epoch": 0.5924933830798176,
+        "grad_norm": 0.40840062499046326,
+        "learning_rate": 0.0005,
+        "loss": 1.102,
+        "step": 18580
+      },
+      {
+        "epoch": 0.5928122707994515,
+        "grad_norm": 0.3809534013271332,
+        "learning_rate": 0.0005,
+        "loss": 1.1128,
+        "step": 18590
+      },
+      {
+        "epoch": 0.5931311585190854,
+        "grad_norm": 0.38073065876960754,
+        "learning_rate": 0.0005,
+        "loss": 1.1099,
+        "step": 18600
+      },
+      {
+        "epoch": 0.5934500462387193,
+        "grad_norm": 0.4003779888153076,
+        "learning_rate": 0.0005,
+        "loss": 1.1182,
+        "step": 18610
+      },
+      {
+        "epoch": 0.5937689339583533,
+        "grad_norm": 0.38311606645584106,
+        "learning_rate": 0.0005,
+        "loss": 1.1133,
+        "step": 18620
+      },
+      {
+        "epoch": 0.5940878216779871,
+        "grad_norm": 0.3980417549610138,
+        "learning_rate": 0.0005,
+        "loss": 1.1108,
+        "step": 18630
+      },
+      {
+        "epoch": 0.5944067093976211,
+        "grad_norm": 0.39308735728263855,
+        "learning_rate": 0.0005,
+        "loss": 1.112,
+        "step": 18640
+      },
+      {
+        "epoch": 0.594725597117255,
+        "grad_norm": 0.3650916814804077,
+        "learning_rate": 0.0005,
+        "loss": 1.1127,
+        "step": 18650
+      },
+      {
+        "epoch": 0.5950444848368889,
+        "grad_norm": 0.3832719027996063,
+        "learning_rate": 0.0005,
+        "loss": 1.086,
+        "step": 18660
+      },
+      {
+        "epoch": 0.5953633725565228,
+        "grad_norm": 0.39278343319892883,
+        "learning_rate": 0.0005,
+        "loss": 1.1029,
+        "step": 18670
+      },
+      {
+        "epoch": 0.5956822602761568,
+        "grad_norm": 0.3769834041595459,
+        "learning_rate": 0.0005,
+        "loss": 1.1116,
+        "step": 18680
+      },
+      {
+        "epoch": 0.5960011479957907,
+        "grad_norm": 0.381989061832428,
+        "learning_rate": 0.0005,
+        "loss": 1.1047,
+        "step": 18690
+      },
+      {
+        "epoch": 0.5963200357154246,
+        "grad_norm": 0.38305485248565674,
+        "learning_rate": 0.0005,
+        "loss": 1.0906,
+        "step": 18700
+      },
+      {
+        "epoch": 0.5966389234350585,
+        "grad_norm": 0.3805263638496399,
+        "learning_rate": 0.0005,
+        "loss": 1.1158,
+        "step": 18710
+      },
+      {
+        "epoch": 0.5969578111546925,
+        "grad_norm": 0.3875347673892975,
+        "learning_rate": 0.0005,
+        "loss": 1.1095,
+        "step": 18720
+      },
+      {
+        "epoch": 0.5972766988743263,
+        "grad_norm": 0.3871545195579529,
+        "learning_rate": 0.0005,
+        "loss": 1.1108,
+        "step": 18730
+      },
+      {
+        "epoch": 0.5975955865939603,
+        "grad_norm": 0.3692149519920349,
+        "learning_rate": 0.0005,
+        "loss": 1.1256,
+        "step": 18740
+      },
+      {
+        "epoch": 0.5979144743135942,
+        "grad_norm": 0.3667621612548828,
+        "learning_rate": 0.0005,
+        "loss": 1.0938,
+        "step": 18750
+      },
+      {
+        "epoch": 0.5982333620332281,
+        "grad_norm": 0.38246843218803406,
+        "learning_rate": 0.0005,
+        "loss": 1.1166,
+        "step": 18760
+      },
+      {
+        "epoch": 0.598552249752862,
+        "grad_norm": 0.38617420196533203,
+        "learning_rate": 0.0005,
+        "loss": 1.1086,
+        "step": 18770
+      },
+      {
+        "epoch": 0.598871137472496,
+        "grad_norm": 0.3871334195137024,
+        "learning_rate": 0.0005,
+        "loss": 1.1041,
+        "step": 18780
+      },
+      {
+        "epoch": 0.5991900251921298,
+        "grad_norm": 0.38116347789764404,
+        "learning_rate": 0.0005,
+        "loss": 1.1395,
+        "step": 18790
+      },
+      {
+        "epoch": 0.5995089129117638,
+        "grad_norm": 0.38512861728668213,
+        "learning_rate": 0.0005,
+        "loss": 1.0727,
+        "step": 18800
+      },
+      {
+        "epoch": 0.5998278006313977,
+        "grad_norm": 0.3731023669242859,
+        "learning_rate": 0.0005,
+        "loss": 1.096,
+        "step": 18810
+      },
+      {
+        "epoch": 0.6001466883510316,
+        "grad_norm": 0.3894040286540985,
+        "learning_rate": 0.0005,
+        "loss": 1.1203,
+        "step": 18820
+      },
+      {
+        "epoch": 0.6004655760706655,
+        "grad_norm": 0.38910865783691406,
+        "learning_rate": 0.0005,
+        "loss": 1.1099,
+        "step": 18830
+      },
+      {
+        "epoch": 0.6007844637902995,
+        "grad_norm": 0.39119401574134827,
+        "learning_rate": 0.0005,
+        "loss": 1.1067,
+        "step": 18840
+      },
+      {
+        "epoch": 0.6011033515099333,
+        "grad_norm": 0.39401862025260925,
+        "learning_rate": 0.0005,
+        "loss": 1.1443,
+        "step": 18850
+      },
+      {
+        "epoch": 0.6014222392295673,
+        "grad_norm": 0.38900884985923767,
+        "learning_rate": 0.0005,
+        "loss": 1.1052,
+        "step": 18860
+      },
+      {
+        "epoch": 0.6017411269492012,
+        "grad_norm": 0.3946762681007385,
+        "learning_rate": 0.0005,
+        "loss": 1.1095,
+        "step": 18870
+      },
+      {
+        "epoch": 0.6020600146688351,
+        "grad_norm": 0.38308626413345337,
+        "learning_rate": 0.0005,
+        "loss": 1.0929,
+        "step": 18880
+      },
+      {
+        "epoch": 0.602378902388469,
+        "grad_norm": 0.3667078912258148,
+        "learning_rate": 0.0005,
+        "loss": 1.108,
+        "step": 18890
+      },
+      {
+        "epoch": 0.602697790108103,
+        "grad_norm": 0.3811316192150116,
+        "learning_rate": 0.0005,
+        "loss": 1.1106,
+        "step": 18900
+      },
+      {
+        "epoch": 0.6030166778277368,
+        "grad_norm": 0.3822799026966095,
+        "learning_rate": 0.0005,
+        "loss": 1.11,
+        "step": 18910
+      },
+      {
+        "epoch": 0.6033355655473708,
+        "grad_norm": 0.37707287073135376,
+        "learning_rate": 0.0005,
+        "loss": 1.1164,
+        "step": 18920
+      },
+      {
+        "epoch": 0.6036544532670047,
+        "grad_norm": 0.3722653388977051,
+        "learning_rate": 0.0005,
+        "loss": 1.087,
+        "step": 18930
+      },
+      {
+        "epoch": 0.6039733409866386,
+        "grad_norm": 0.3983718454837799,
+        "learning_rate": 0.0005,
+        "loss": 1.0896,
+        "step": 18940
+      },
+      {
+        "epoch": 0.6042922287062725,
+        "grad_norm": 0.3902481496334076,
+        "learning_rate": 0.0005,
+        "loss": 1.115,
+        "step": 18950
+      },
+      {
+        "epoch": 0.6046111164259065,
+        "grad_norm": 0.3895500898361206,
+        "learning_rate": 0.0005,
+        "loss": 1.1194,
+        "step": 18960
+      },
+      {
+        "epoch": 0.6049300041455403,
+        "grad_norm": 0.3960655629634857,
+        "learning_rate": 0.0005,
+        "loss": 1.1132,
+        "step": 18970
+      },
+      {
+        "epoch": 0.6052488918651743,
+        "grad_norm": 0.39928701519966125,
+        "learning_rate": 0.0005,
+        "loss": 1.1145,
+        "step": 18980
+      },
+      {
+        "epoch": 0.6055677795848082,
+        "grad_norm": 0.40216585993766785,
+        "learning_rate": 0.0005,
+        "loss": 1.1151,
+        "step": 18990
+      },
+      {
+        "epoch": 0.6058866673044421,
+        "grad_norm": 0.37362903356552124,
+        "learning_rate": 0.0005,
+        "loss": 1.1073,
+        "step": 19000
+      },
+      {
+        "epoch": 0.606205555024076,
+        "grad_norm": 0.40238964557647705,
+        "learning_rate": 0.0005,
+        "loss": 1.1007,
+        "step": 19010
+      },
+      {
+        "epoch": 0.60652444274371,
+        "grad_norm": 0.38304927945137024,
+        "learning_rate": 0.0005,
+        "loss": 1.1208,
+        "step": 19020
+      },
+      {
+        "epoch": 0.6068433304633438,
+        "grad_norm": 0.38120222091674805,
+        "learning_rate": 0.0005,
+        "loss": 1.1043,
+        "step": 19030
+      },
+      {
+        "epoch": 0.6071622181829778,
+        "grad_norm": 0.3859049081802368,
+        "learning_rate": 0.0005,
+        "loss": 1.0943,
+        "step": 19040
+      },
+      {
+        "epoch": 0.6074811059026117,
+        "grad_norm": 0.37868672609329224,
+        "learning_rate": 0.0005,
+        "loss": 1.0978,
+        "step": 19050
+      },
+      {
+        "epoch": 0.6077999936222456,
+        "grad_norm": 0.3834395110607147,
+        "learning_rate": 0.0005,
+        "loss": 1.1088,
+        "step": 19060
+      },
+      {
+        "epoch": 0.6081188813418795,
+        "grad_norm": 0.3883598744869232,
+        "learning_rate": 0.0005,
+        "loss": 1.1118,
+        "step": 19070
+      },
+      {
+        "epoch": 0.6084377690615135,
+        "grad_norm": 0.3869861364364624,
+        "learning_rate": 0.0005,
+        "loss": 1.103,
+        "step": 19080
+      },
+      {
+        "epoch": 0.6087566567811473,
+        "grad_norm": 0.3876883387565613,
+        "learning_rate": 0.0005,
+        "loss": 1.0852,
+        "step": 19090
+      },
+      {
+        "epoch": 0.6090755445007813,
+        "grad_norm": 0.402549684047699,
+        "learning_rate": 0.0005,
+        "loss": 1.1133,
+        "step": 19100
+      },
+      {
+        "epoch": 0.6093944322204152,
+        "grad_norm": 0.3869187831878662,
+        "learning_rate": 0.0005,
+        "loss": 1.0844,
+        "step": 19110
+      },
+      {
+        "epoch": 0.6097133199400491,
+        "grad_norm": 0.40940797328948975,
+        "learning_rate": 0.0005,
+        "loss": 1.1096,
+        "step": 19120
+      },
+      {
+        "epoch": 0.610032207659683,
+        "grad_norm": 0.3896699547767639,
+        "learning_rate": 0.0005,
+        "loss": 1.1233,
+        "step": 19130
+      },
+      {
+        "epoch": 0.610351095379317,
+        "grad_norm": 0.38502994179725647,
+        "learning_rate": 0.0005,
+        "loss": 1.12,
+        "step": 19140
+      },
+      {
+        "epoch": 0.6106699830989508,
+        "grad_norm": 0.37965127825737,
+        "learning_rate": 0.0005,
+        "loss": 1.1192,
+        "step": 19150
+      },
+      {
+        "epoch": 0.6109888708185848,
+        "grad_norm": 0.3973645269870758,
+        "learning_rate": 0.0005,
+        "loss": 1.0942,
+        "step": 19160
+      },
+      {
+        "epoch": 0.6113077585382187,
+        "grad_norm": 0.3911329209804535,
+        "learning_rate": 0.0005,
+        "loss": 1.121,
+        "step": 19170
+      },
+      {
+        "epoch": 0.6116266462578526,
+        "grad_norm": 0.3806923031806946,
+        "learning_rate": 0.0005,
+        "loss": 1.1192,
+        "step": 19180
+      },
+      {
+        "epoch": 0.6119455339774865,
+        "grad_norm": 0.38102033734321594,
+        "learning_rate": 0.0005,
+        "loss": 1.1168,
+        "step": 19190
+      },
+      {
+        "epoch": 0.6122644216971205,
+        "grad_norm": 0.39193785190582275,
+        "learning_rate": 0.0005,
+        "loss": 1.0999,
+        "step": 19200
+      },
+      {
+        "epoch": 0.6125833094167543,
+        "grad_norm": 0.365323007106781,
+        "learning_rate": 0.0005,
+        "loss": 1.0853,
+        "step": 19210
+      },
+      {
+        "epoch": 0.6129021971363883,
+        "grad_norm": 0.39409881830215454,
+        "learning_rate": 0.0005,
+        "loss": 1.1065,
+        "step": 19220
+      },
+      {
+        "epoch": 0.6132210848560222,
+        "grad_norm": 0.3841363787651062,
+        "learning_rate": 0.0005,
+        "loss": 1.094,
+        "step": 19230
+      },
+      {
+        "epoch": 0.6135399725756561,
+        "grad_norm": 0.3903886079788208,
+        "learning_rate": 0.0005,
+        "loss": 1.1018,
+        "step": 19240
+      },
+      {
+        "epoch": 0.61385886029529,
+        "grad_norm": 0.3933718800544739,
+        "learning_rate": 0.0005,
+        "loss": 1.0862,
+        "step": 19250
+      },
+      {
+        "epoch": 0.614177748014924,
+        "grad_norm": 0.38507020473480225,
+        "learning_rate": 0.0005,
+        "loss": 1.1148,
+        "step": 19260
+      },
+      {
+        "epoch": 0.6144966357345578,
+        "grad_norm": 0.3783412575721741,
+        "learning_rate": 0.0005,
+        "loss": 1.1174,
+        "step": 19270
+      },
+      {
+        "epoch": 0.6148155234541918,
+        "grad_norm": 0.3768440783023834,
+        "learning_rate": 0.0005,
+        "loss": 1.1083,
+        "step": 19280
+      },
+      {
+        "epoch": 0.6151344111738257,
+        "grad_norm": 0.37053754925727844,
+        "learning_rate": 0.0005,
+        "loss": 1.09,
+        "step": 19290
+      },
+      {
+        "epoch": 0.6154532988934596,
+        "grad_norm": 0.3970535397529602,
+        "learning_rate": 0.0005,
+        "loss": 1.1147,
+        "step": 19300
+      },
+      {
+        "epoch": 0.6157721866130935,
+        "grad_norm": 0.3832497298717499,
+        "learning_rate": 0.0005,
+        "loss": 1.1118,
+        "step": 19310
+      },
+      {
+        "epoch": 0.6160910743327275,
+        "grad_norm": 0.41348445415496826,
+        "learning_rate": 0.0005,
+        "loss": 1.1003,
+        "step": 19320
+      },
+      {
+        "epoch": 0.6164099620523613,
+        "grad_norm": 0.3903905749320984,
+        "learning_rate": 0.0005,
+        "loss": 1.1281,
+        "step": 19330
+      },
+      {
+        "epoch": 0.6167288497719953,
+        "grad_norm": 0.3750537931919098,
+        "learning_rate": 0.0005,
+        "loss": 1.1033,
+        "step": 19340
+      },
+      {
+        "epoch": 0.6170477374916292,
+        "grad_norm": 0.38821011781692505,
+        "learning_rate": 0.0005,
+        "loss": 1.1045,
+        "step": 19350
+      },
+      {
+        "epoch": 0.6173666252112631,
+        "grad_norm": 0.385237455368042,
+        "learning_rate": 0.0005,
+        "loss": 1.1076,
+        "step": 19360
+      },
+      {
+        "epoch": 0.617685512930897,
+        "grad_norm": 0.37394216656684875,
+        "learning_rate": 0.0005,
+        "loss": 1.1085,
+        "step": 19370
+      },
+      {
+        "epoch": 0.618004400650531,
+        "grad_norm": 0.38589802384376526,
+        "learning_rate": 0.0005,
+        "loss": 1.114,
+        "step": 19380
+      },
+      {
+        "epoch": 0.6183232883701648,
+        "grad_norm": 0.40107953548431396,
+        "learning_rate": 0.0005,
+        "loss": 1.1248,
+        "step": 19390
+      },
+      {
+        "epoch": 0.6186421760897988,
+        "grad_norm": 0.371114045381546,
+        "learning_rate": 0.0005,
+        "loss": 1.0976,
+        "step": 19400
+      },
+      {
+        "epoch": 0.6189610638094327,
+        "grad_norm": 0.3722735345363617,
+        "learning_rate": 0.0005,
+        "loss": 1.0997,
+        "step": 19410
+      },
+      {
+        "epoch": 0.6192799515290666,
+        "grad_norm": 0.4006235599517822,
+        "learning_rate": 0.0005,
+        "loss": 1.1236,
+        "step": 19420
+      },
+      {
+        "epoch": 0.6195988392487005,
+        "grad_norm": 0.3824571371078491,
+        "learning_rate": 0.0005,
+        "loss": 1.1023,
+        "step": 19430
+      },
+      {
+        "epoch": 0.6199177269683345,
+        "grad_norm": 0.38375917077064514,
+        "learning_rate": 0.0005,
+        "loss": 1.1097,
+        "step": 19440
+      },
+      {
+        "epoch": 0.6202366146879683,
+        "grad_norm": 0.36844027042388916,
+        "learning_rate": 0.0005,
+        "loss": 1.0865,
+        "step": 19450
+      },
+      {
+        "epoch": 0.6205555024076023,
+        "grad_norm": 0.38515403866767883,
+        "learning_rate": 0.0005,
+        "loss": 1.1242,
+        "step": 19460
+      },
+      {
+        "epoch": 0.6208743901272362,
+        "grad_norm": 0.39014798402786255,
+        "learning_rate": 0.0005,
+        "loss": 1.0959,
+        "step": 19470
+      },
+      {
+        "epoch": 0.6211932778468701,
+        "grad_norm": 0.3807680904865265,
+        "learning_rate": 0.0005,
+        "loss": 1.1045,
+        "step": 19480
+      },
+      {
+        "epoch": 0.621512165566504,
+        "grad_norm": 0.39916208386421204,
+        "learning_rate": 0.0005,
+        "loss": 1.1077,
+        "step": 19490
+      },
+      {
+        "epoch": 0.621831053286138,
+        "grad_norm": 0.38831251859664917,
+        "learning_rate": 0.0005,
+        "loss": 1.109,
+        "step": 19500
+      },
+      {
+        "epoch": 0.6221499410057719,
+        "grad_norm": 0.3683846592903137,
+        "learning_rate": 0.0005,
+        "loss": 1.1111,
+        "step": 19510
+      },
+      {
+        "epoch": 0.6224688287254058,
+        "grad_norm": 0.3902853727340698,
+        "learning_rate": 0.0005,
+        "loss": 1.102,
+        "step": 19520
+      },
+      {
+        "epoch": 0.6227877164450397,
+        "grad_norm": 0.39055487513542175,
+        "learning_rate": 0.0005,
+        "loss": 1.1183,
+        "step": 19530
+      },
+      {
+        "epoch": 0.6231066041646737,
+        "grad_norm": 0.380648136138916,
+        "learning_rate": 0.0005,
+        "loss": 1.1002,
+        "step": 19540
+      },
+      {
+        "epoch": 0.6234254918843075,
+        "grad_norm": 0.36444613337516785,
+        "learning_rate": 0.0005,
+        "loss": 1.1011,
+        "step": 19550
+      },
+      {
+        "epoch": 0.6237443796039415,
+        "grad_norm": 0.39328739047050476,
+        "learning_rate": 0.0005,
+        "loss": 1.126,
+        "step": 19560
+      },
+      {
+        "epoch": 0.6240632673235754,
+        "grad_norm": 0.3845846354961395,
+        "learning_rate": 0.0005,
+        "loss": 1.1174,
+        "step": 19570
+      },
+      {
+        "epoch": 0.6243821550432093,
+        "grad_norm": 0.3922548294067383,
+        "learning_rate": 0.0005,
+        "loss": 1.1122,
+        "step": 19580
+      },
+      {
+        "epoch": 0.6247010427628432,
+        "grad_norm": 0.3749217987060547,
+        "learning_rate": 0.0005,
+        "loss": 1.1073,
+        "step": 19590
+      },
+      {
+        "epoch": 0.6250199304824772,
+        "grad_norm": 0.3772096335887909,
+        "learning_rate": 0.0005,
+        "loss": 1.1021,
+        "step": 19600
+      },
+      {
+        "epoch": 0.625338818202111,
+        "grad_norm": 0.3921220600605011,
+        "learning_rate": 0.0005,
+        "loss": 1.1128,
+        "step": 19610
+      },
+      {
+        "epoch": 0.625657705921745,
+        "grad_norm": 0.38062894344329834,
+        "learning_rate": 0.0005,
+        "loss": 1.1215,
+        "step": 19620
+      },
+      {
+        "epoch": 0.6259765936413789,
+        "grad_norm": 0.3739464282989502,
+        "learning_rate": 0.0005,
+        "loss": 1.0983,
+        "step": 19630
+      },
+      {
+        "epoch": 0.6262954813610128,
+        "grad_norm": 0.3996415436267853,
+        "learning_rate": 0.0005,
+        "loss": 1.1093,
+        "step": 19640
+      },
+      {
+        "epoch": 0.6266143690806467,
+        "grad_norm": 0.37616029381752014,
+        "learning_rate": 0.0005,
+        "loss": 1.1107,
+        "step": 19650
+      },
+      {
+        "epoch": 0.6269332568002807,
+        "grad_norm": 0.3891390860080719,
+        "learning_rate": 0.0005,
+        "loss": 1.1001,
+        "step": 19660
+      },
+      {
+        "epoch": 0.6272521445199145,
+        "grad_norm": 0.38982507586479187,
+        "learning_rate": 0.0005,
+        "loss": 1.0998,
+        "step": 19670
+      },
+      {
+        "epoch": 0.6275710322395485,
+        "grad_norm": 0.3861311376094818,
+        "learning_rate": 0.0005,
+        "loss": 1.1155,
+        "step": 19680
+      },
+      {
+        "epoch": 0.6278899199591824,
+        "grad_norm": 0.38030752539634705,
+        "learning_rate": 0.0005,
+        "loss": 1.1024,
+        "step": 19690
+      },
+      {
+        "epoch": 0.6282088076788163,
+        "grad_norm": 0.40073028206825256,
+        "learning_rate": 0.0005,
+        "loss": 1.1032,
+        "step": 19700
+      },
+      {
+        "epoch": 0.6285276953984502,
+        "grad_norm": 0.38168585300445557,
+        "learning_rate": 0.0005,
+        "loss": 1.0865,
+        "step": 19710
+      },
+      {
+        "epoch": 0.6288465831180842,
+        "grad_norm": 0.37104278802871704,
+        "learning_rate": 0.0005,
+        "loss": 1.1062,
+        "step": 19720
+      },
+      {
+        "epoch": 0.629165470837718,
+        "grad_norm": 0.3792668879032135,
+        "learning_rate": 0.0005,
+        "loss": 1.098,
+        "step": 19730
+      },
+      {
+        "epoch": 0.629484358557352,
+        "grad_norm": 0.38486728072166443,
+        "learning_rate": 0.0005,
+        "loss": 1.1056,
+        "step": 19740
+      },
+      {
+        "epoch": 0.6298032462769859,
+        "grad_norm": 0.3664900064468384,
+        "learning_rate": 0.0005,
+        "loss": 1.0944,
+        "step": 19750
+      },
+      {
+        "epoch": 0.6301221339966198,
+        "grad_norm": 0.36216047406196594,
+        "learning_rate": 0.0005,
+        "loss": 1.0908,
+        "step": 19760
+      },
+      {
+        "epoch": 0.6304410217162537,
+        "grad_norm": 0.38465407490730286,
+        "learning_rate": 0.0005,
+        "loss": 1.0963,
+        "step": 19770
+      },
+      {
+        "epoch": 0.6307599094358877,
+        "grad_norm": 0.39357104897499084,
+        "learning_rate": 0.0005,
+        "loss": 1.0946,
+        "step": 19780
+      },
+      {
+        "epoch": 0.6310787971555215,
+        "grad_norm": 0.38242307305336,
+        "learning_rate": 0.0005,
+        "loss": 1.1089,
+        "step": 19790
+      },
+      {
+        "epoch": 0.6313976848751555,
+        "grad_norm": 0.394965261220932,
+        "learning_rate": 0.0005,
+        "loss": 1.0989,
+        "step": 19800
+      },
+      {
+        "epoch": 0.6317165725947894,
+        "grad_norm": 0.3696037530899048,
+        "learning_rate": 0.0005,
+        "loss": 1.1052,
+        "step": 19810
+      },
+      {
+        "epoch": 0.6320354603144233,
+        "grad_norm": 0.39433643221855164,
+        "learning_rate": 0.0005,
+        "loss": 1.1002,
+        "step": 19820
+      },
+      {
+        "epoch": 0.6323543480340572,
+        "grad_norm": 0.39929360151290894,
+        "learning_rate": 0.0005,
+        "loss": 1.1046,
+        "step": 19830
+      },
+      {
+        "epoch": 0.6326732357536912,
+        "grad_norm": 0.385371595621109,
+        "learning_rate": 0.0005,
+        "loss": 1.1151,
+        "step": 19840
+      },
+      {
+        "epoch": 0.632992123473325,
+        "grad_norm": 0.36684107780456543,
+        "learning_rate": 0.0005,
+        "loss": 1.0942,
+        "step": 19850
+      },
+      {
+        "epoch": 0.633311011192959,
+        "grad_norm": 0.3931552469730377,
+        "learning_rate": 0.0005,
+        "loss": 1.104,
+        "step": 19860
+      },
+      {
+        "epoch": 0.6336298989125929,
+        "grad_norm": 0.3704059422016144,
+        "learning_rate": 0.0005,
+        "loss": 1.0976,
+        "step": 19870
+      },
+      {
+        "epoch": 0.6339487866322268,
+        "grad_norm": 0.3817865252494812,
+        "learning_rate": 0.0005,
+        "loss": 1.1033,
+        "step": 19880
+      },
+      {
+        "epoch": 0.6342676743518607,
+        "grad_norm": 0.3732118606567383,
+        "learning_rate": 0.0005,
+        "loss": 1.1094,
+        "step": 19890
+      },
+      {
+        "epoch": 0.6345865620714947,
+        "grad_norm": 0.3995611369609833,
+        "learning_rate": 0.0005,
+        "loss": 1.1045,
+        "step": 19900
+      },
+      {
+        "epoch": 0.6349054497911285,
+        "grad_norm": 0.4252803325653076,
+        "learning_rate": 0.0005,
+        "loss": 1.1128,
+        "step": 19910
+      },
+      {
+        "epoch": 0.6352243375107625,
+        "grad_norm": 0.3813573718070984,
+        "learning_rate": 0.0005,
+        "loss": 1.1065,
+        "step": 19920
+      },
+      {
+        "epoch": 0.6355432252303964,
+        "grad_norm": 0.3913743197917938,
+        "learning_rate": 0.0005,
+        "loss": 1.0895,
+        "step": 19930
+      },
+      {
+        "epoch": 0.6358621129500303,
+        "grad_norm": 0.39196574687957764,
+        "learning_rate": 0.0005,
+        "loss": 1.0954,
+        "step": 19940
+      },
+      {
+        "epoch": 0.6361810006696642,
+        "grad_norm": 0.3680713176727295,
+        "learning_rate": 0.0005,
+        "loss": 1.0881,
+        "step": 19950
+      },
+      {
+        "epoch": 0.6364998883892982,
+        "grad_norm": 0.3810781240463257,
+        "learning_rate": 0.0005,
+        "loss": 1.1045,
+        "step": 19960
+      },
+      {
+        "epoch": 0.636818776108932,
+        "grad_norm": 0.3877175748348236,
+        "learning_rate": 0.0005,
+        "loss": 1.1149,
+        "step": 19970
+      },
+      {
+        "epoch": 0.637137663828566,
+        "grad_norm": 0.370029479265213,
+        "learning_rate": 0.0005,
+        "loss": 1.1207,
+        "step": 19980
+      },
+      {
+        "epoch": 0.6374565515481999,
+        "grad_norm": 0.3744882643222809,
+        "learning_rate": 0.0005,
+        "loss": 1.1102,
+        "step": 19990
+      },
+      {
+        "epoch": 0.6377754392678338,
+        "grad_norm": 0.3977319896221161,
+        "learning_rate": 0.0005,
+        "loss": 1.1017,
+        "step": 20000
+      },
+      {
+        "epoch": 0.6380943269874677,
+        "grad_norm": 0.37014520168304443,
+        "learning_rate": 0.0005,
+        "loss": 1.1124,
+        "step": 20010
+      },
+      {
+        "epoch": 0.6384132147071017,
+        "grad_norm": 0.37795954942703247,
+        "learning_rate": 0.0005,
+        "loss": 1.1125,
+        "step": 20020
+      },
+      {
+        "epoch": 0.6387321024267355,
+        "grad_norm": 0.38005727529525757,
+        "learning_rate": 0.0005,
+        "loss": 1.1233,
+        "step": 20030
+      },
+      {
+        "epoch": 0.6390509901463695,
+        "grad_norm": 0.3620345890522003,
+        "learning_rate": 0.0005,
+        "loss": 1.0987,
+        "step": 20040
+      },
+      {
+        "epoch": 0.6393698778660034,
+        "grad_norm": 0.3555848300457001,
+        "learning_rate": 0.0005,
+        "loss": 1.0906,
+        "step": 20050
+      },
+      {
+        "epoch": 0.6396887655856373,
+        "grad_norm": 0.37240397930145264,
+        "learning_rate": 0.0005,
+        "loss": 1.1008,
+        "step": 20060
+      },
+      {
+        "epoch": 0.6400076533052712,
+        "grad_norm": 0.37049400806427,
+        "learning_rate": 0.0005,
+        "loss": 1.1029,
+        "step": 20070
+      },
+      {
+        "epoch": 0.6403265410249052,
+        "grad_norm": 0.37072935700416565,
+        "learning_rate": 0.0005,
+        "loss": 1.0949,
+        "step": 20080
+      },
+      {
+        "epoch": 0.640645428744539,
+        "grad_norm": 0.3630337417125702,
+        "learning_rate": 0.0005,
+        "loss": 1.1095,
+        "step": 20090
+      },
+      {
+        "epoch": 0.640964316464173,
+        "grad_norm": 0.38237258791923523,
+        "learning_rate": 0.0005,
+        "loss": 1.1151,
+        "step": 20100
+      },
+      {
+        "epoch": 0.6412832041838069,
+        "grad_norm": 0.3834942579269409,
+        "learning_rate": 0.0005,
+        "loss": 1.088,
+        "step": 20110
+      },
+      {
+        "epoch": 0.6416020919034408,
+        "grad_norm": 0.3801743686199188,
+        "learning_rate": 0.0005,
+        "loss": 1.1113,
+        "step": 20120
+      },
+      {
+        "epoch": 0.6419209796230747,
+        "grad_norm": 0.36740273237228394,
+        "learning_rate": 0.0005,
+        "loss": 1.1085,
+        "step": 20130
+      },
+      {
+        "epoch": 0.6422398673427087,
+        "grad_norm": 0.38204872608184814,
+        "learning_rate": 0.0005,
+        "loss": 1.1032,
+        "step": 20140
+      },
+      {
+        "epoch": 0.6425587550623425,
+        "grad_norm": 0.38014906644821167,
+        "learning_rate": 0.0005,
+        "loss": 1.1098,
+        "step": 20150
+      },
+      {
+        "epoch": 0.6428776427819765,
+        "grad_norm": 0.36667174100875854,
+        "learning_rate": 0.0005,
+        "loss": 1.109,
+        "step": 20160
+      },
+      {
+        "epoch": 0.6431965305016104,
+        "grad_norm": 0.3805667459964752,
+        "learning_rate": 0.0005,
+        "loss": 1.1101,
+        "step": 20170
+      },
+      {
+        "epoch": 0.6435154182212443,
+        "grad_norm": 0.36804214119911194,
+        "learning_rate": 0.0005,
+        "loss": 1.1068,
+        "step": 20180
+      },
+      {
+        "epoch": 0.6438343059408782,
+        "grad_norm": 0.37310531735420227,
+        "learning_rate": 0.0005,
+        "loss": 1.0728,
+        "step": 20190
+      },
+      {
+        "epoch": 0.6441531936605122,
+        "grad_norm": 0.3750821053981781,
+        "learning_rate": 0.0005,
+        "loss": 1.0997,
+        "step": 20200
+      },
+      {
+        "epoch": 0.644472081380146,
+        "grad_norm": 0.396422803401947,
+        "learning_rate": 0.0005,
+        "loss": 1.1176,
+        "step": 20210
+      },
+      {
+        "epoch": 0.64479096909978,
+        "grad_norm": 0.39535143971443176,
+        "learning_rate": 0.0005,
+        "loss": 1.0924,
+        "step": 20220
+      },
+      {
+        "epoch": 0.6451098568194139,
+        "grad_norm": 0.3851791322231293,
+        "learning_rate": 0.0005,
+        "loss": 1.1042,
+        "step": 20230
+      },
+      {
+        "epoch": 0.6454287445390477,
+        "grad_norm": 0.3816549777984619,
+        "learning_rate": 0.0005,
+        "loss": 1.0909,
+        "step": 20240
+      },
+      {
+        "epoch": 0.6457476322586817,
+        "grad_norm": 0.3699096739292145,
+        "learning_rate": 0.0005,
+        "loss": 1.0845,
+        "step": 20250
+      },
+      {
+        "epoch": 0.6460665199783157,
+        "grad_norm": 0.3914874196052551,
+        "learning_rate": 0.0005,
+        "loss": 1.098,
+        "step": 20260
+      },
+      {
+        "epoch": 0.6463854076979495,
+        "grad_norm": 0.4025990664958954,
+        "learning_rate": 0.0005,
+        "loss": 1.1098,
+        "step": 20270
+      },
+      {
+        "epoch": 0.6467042954175835,
+        "grad_norm": 0.3710470497608185,
+        "learning_rate": 0.0005,
+        "loss": 1.1193,
+        "step": 20280
+      },
+      {
+        "epoch": 0.6470231831372174,
+        "grad_norm": 0.38040199875831604,
+        "learning_rate": 0.0005,
+        "loss": 1.0883,
+        "step": 20290
+      },
+      {
+        "epoch": 0.6473420708568512,
+        "grad_norm": 0.38826727867126465,
+        "learning_rate": 0.0005,
+        "loss": 1.1073,
+        "step": 20300
+      },
+      {
+        "epoch": 0.6476609585764852,
+        "grad_norm": 0.39304104447364807,
+        "learning_rate": 0.0005,
+        "loss": 1.0937,
+        "step": 20310
+      },
+      {
+        "epoch": 0.6479798462961192,
+        "grad_norm": 0.3646523952484131,
+        "learning_rate": 0.0005,
+        "loss": 1.0911,
+        "step": 20320
+      },
+      {
+        "epoch": 0.6482987340157531,
+        "grad_norm": 0.36644312739372253,
+        "learning_rate": 0.0005,
+        "loss": 1.0735,
+        "step": 20330
+      },
+      {
+        "epoch": 0.648617621735387,
+        "grad_norm": 0.37986788153648376,
+        "learning_rate": 0.0005,
+        "loss": 1.1079,
+        "step": 20340
+      },
+      {
+        "epoch": 0.6489365094550209,
+        "grad_norm": 0.3844556212425232,
+        "learning_rate": 0.0005,
+        "loss": 1.1149,
+        "step": 20350
+      },
+      {
+        "epoch": 0.6492553971746549,
+        "grad_norm": 0.3708367347717285,
+        "learning_rate": 0.0005,
+        "loss": 1.1058,
+        "step": 20360
+      },
+      {
+        "epoch": 0.6495742848942887,
+        "grad_norm": 0.3721873462200165,
+        "learning_rate": 0.0005,
+        "loss": 1.096,
+        "step": 20370
+      },
+      {
+        "epoch": 0.6498931726139227,
+        "grad_norm": 0.36754587292671204,
+        "learning_rate": 0.0005,
+        "loss": 1.115,
+        "step": 20380
+      },
+      {
+        "epoch": 0.6502120603335566,
+        "grad_norm": 0.3798966407775879,
+        "learning_rate": 0.0004987293446490035,
+        "loss": 1.0767,
+        "step": 20390
+      },
+      {
+        "epoch": 0.6505309480531904,
+        "grad_norm": 0.40216562151908875,
+        "learning_rate": 0.000496824661048599,
+        "loss": 1.1075,
+        "step": 20400
+      },
+      {
+        "epoch": 0.6508498357728244,
+        "grad_norm": 0.37482136487960815,
+        "learning_rate": 0.0004949272515732417,
+        "loss": 1.0795,
+        "step": 20410
+      },
+      {
+        "epoch": 0.6511687234924584,
+        "grad_norm": 0.37605640292167664,
+        "learning_rate": 0.0004930370884425194,
+        "loss": 1.1036,
+        "step": 20420
+      },
+      {
+        "epoch": 0.6514876112120922,
+        "grad_norm": 0.37845471501350403,
+        "learning_rate": 0.0004911541439821156,
+        "loss": 1.1062,
+        "step": 20430
+      },
+      {
+        "epoch": 0.6518064989317262,
+        "grad_norm": 0.3742932379245758,
+        "learning_rate": 0.000489278390623404,
+        "loss": 1.1239,
+        "step": 20440
+      },
+      {
+        "epoch": 0.6521253866513601,
+        "grad_norm": 0.3708976209163666,
+        "learning_rate": 0.0004874098009030445,
+        "loss": 1.0989,
+        "step": 20450
+      },
+      {
+        "epoch": 0.652444274370994,
+        "grad_norm": 0.38528645038604736,
+        "learning_rate": 0.0004855483474625822,
+        "loss": 1.0936,
+        "step": 20460
+      },
+      {
+        "epoch": 0.6527631620906279,
+        "grad_norm": 0.39533546566963196,
+        "learning_rate": 0.0004836940030480454,
+        "loss": 1.0938,
+        "step": 20470
+      },
+      {
+        "epoch": 0.6530820498102619,
+        "grad_norm": 0.36083102226257324,
+        "learning_rate": 0.0004818467405095477,
+        "loss": 1.0641,
+        "step": 20480
+      },
+      {
+        "epoch": 0.6534009375298957,
+        "grad_norm": 0.3747830092906952,
+        "learning_rate": 0.00048000653280089,
+        "loss": 1.0919,
+        "step": 20490
+      },
+      {
+        "epoch": 0.6537198252495297,
+        "grad_norm": 0.38523662090301514,
+        "learning_rate": 0.00047817335297916465,
+        "loss": 1.0922,
+        "step": 20500
+      },
+      {
+        "epoch": 0.6540387129691636,
+        "grad_norm": 0.3711490333080292,
+        "learning_rate": 0.0004763471742043608,
+        "loss": 1.0861,
+        "step": 20510
+      },
+      {
+        "epoch": 0.6543576006887974,
+        "grad_norm": 0.37779441475868225,
+        "learning_rate": 0.00047452796973897166,
+        "loss": 1.0979,
+        "step": 20520
+      },
+      {
+        "epoch": 0.6546764884084314,
+        "grad_norm": 0.37991273403167725,
+        "learning_rate": 0.0004727157129476028,
+        "loss": 1.082,
+        "step": 20530
+      },
+      {
+        "epoch": 0.6549953761280654,
+        "grad_norm": 0.38536524772644043,
+        "learning_rate": 0.0004709103772965826,
+        "loss": 1.1007,
+        "step": 20540
+      },
+      {
+        "epoch": 0.6553142638476992,
+        "grad_norm": 0.36973538994789124,
+        "learning_rate": 0.0004691119363535731,
+        "loss": 1.0886,
+        "step": 20550
+      },
+      {
+        "epoch": 0.6556331515673332,
+        "grad_norm": 0.35994893312454224,
+        "learning_rate": 0.0004673203637871834,
+        "loss": 1.0805,
+        "step": 20560
+      },
+      {
+        "epoch": 0.6559520392869671,
+        "grad_norm": 0.3755747973918915,
+        "learning_rate": 0.0004655356333665843,
+        "loss": 1.1021,
+        "step": 20570
+      },
+      {
+        "epoch": 0.656270927006601,
+        "grad_norm": 0.3772231638431549,
+        "learning_rate": 0.000463757718961124,
+        "loss": 1.1012,
+        "step": 20580
+      },
+      {
+        "epoch": 0.6565898147262349,
+        "grad_norm": 0.3790730834007263,
+        "learning_rate": 0.00046198659453994544,
+        "loss": 1.0853,
+        "step": 20590
+      },
+      {
+        "epoch": 0.6569087024458689,
+        "grad_norm": 0.36576566100120544,
+        "learning_rate": 0.0004602222341716053,
+        "loss": 1.0923,
+        "step": 20600
+      },
+      {
+        "epoch": 0.6572275901655027,
+        "grad_norm": 0.3624275326728821,
+        "learning_rate": 0.0004584646120236944,
+        "loss": 1.0946,
+        "step": 20610
+      },
+      {
+        "epoch": 0.6575464778851366,
+        "grad_norm": 0.3723544776439667,
+        "learning_rate": 0.0004567137023624594,
+        "loss": 1.0819,
+        "step": 20620
+      },
+      {
+        "epoch": 0.6578653656047706,
+        "grad_norm": 0.3639419376850128,
+        "learning_rate": 0.0004549694795524261,
+        "loss": 1.0697,
+        "step": 20630
+      },
+      {
+        "epoch": 0.6581842533244044,
+        "grad_norm": 0.3809690475463867,
+        "learning_rate": 0.0004532319180560237,
+        "loss": 1.1277,
+        "step": 20640
+      },
+      {
+        "epoch": 0.6585031410440384,
+        "grad_norm": 0.41316747665405273,
+        "learning_rate": 0.00045150099243321175,
+        "loss": 1.0938,
+        "step": 20650
+      },
+      {
+        "epoch": 0.6588220287636724,
+        "grad_norm": 0.36777839064598083,
+        "learning_rate": 0.00044977667734110685,
+        "loss": 1.0731,
+        "step": 20660
+      },
+      {
+        "epoch": 0.6591409164833062,
+        "grad_norm": 0.3870874047279358,
+        "learning_rate": 0.00044805894753361183,
+        "loss": 1.0922,
+        "step": 20670
+      },
+      {
+        "epoch": 0.6594598042029401,
+        "grad_norm": 0.36223119497299194,
+        "learning_rate": 0.0004463477778610465,
+        "loss": 1.0844,
+        "step": 20680
+      },
+      {
+        "epoch": 0.6597786919225741,
+        "grad_norm": 0.3737870752811432,
+        "learning_rate": 0.0004446431432697789,
+        "loss": 1.0826,
+        "step": 20690
+      },
+      {
+        "epoch": 0.6600975796422079,
+        "grad_norm": 0.35541796684265137,
+        "learning_rate": 0.0004429450188018587,
+        "loss": 1.0889,
+        "step": 20700
+      },
+      {
+        "epoch": 0.6604164673618419,
+        "grad_norm": 0.3769436180591583,
+        "learning_rate": 0.00044125337959465206,
+        "loss": 1.088,
+        "step": 20710
+      },
+      {
+        "epoch": 0.6607353550814759,
+        "grad_norm": 0.3823726177215576,
+        "learning_rate": 0.00043956820088047696,
+        "loss": 1.0515,
+        "step": 20720
+      },
+      {
+        "epoch": 0.6610542428011097,
+        "grad_norm": 0.36033788323402405,
+        "learning_rate": 0.0004378894579862412,
+        "loss": 1.0833,
+        "step": 20730
+      },
+      {
+        "epoch": 0.6613731305207436,
+        "grad_norm": 0.3760935664176941,
+        "learning_rate": 0.00043621712633308096,
+        "loss": 1.0874,
+        "step": 20740
+      },
+      {
+        "epoch": 0.6616920182403776,
+        "grad_norm": 0.3767266571521759,
+        "learning_rate": 0.00043455118143600075,
+        "loss": 1.0913,
+        "step": 20750
+      },
+      {
+        "epoch": 0.6620109059600114,
+        "grad_norm": 0.3505248725414276,
+        "learning_rate": 0.00043289159890351497,
+        "loss": 1.0709,
+        "step": 20760
+      },
+      {
+        "epoch": 0.6623297936796454,
+        "grad_norm": 0.3828575015068054,
+        "learning_rate": 0.00043123835443729117,
+        "loss": 1.0933,
+        "step": 20770
+      },
+      {
+        "epoch": 0.6626486813992793,
+        "grad_norm": 0.38556039333343506,
+        "learning_rate": 0.0004295914238317937,
+        "loss": 1.0806,
+        "step": 20780
+      },
+      {
+        "epoch": 0.6629675691189132,
+        "grad_norm": 0.36559784412384033,
+        "learning_rate": 0.00042795078297392985,
+        "loss": 1.0812,
+        "step": 20790
+      },
+      {
+        "epoch": 0.6632864568385471,
+        "grad_norm": 0.36894142627716064,
+        "learning_rate": 0.0004263164078426964,
+        "loss": 1.1001,
+        "step": 20800
+      },
+      {
+        "epoch": 0.6636053445581811,
+        "grad_norm": 0.37960314750671387,
+        "learning_rate": 0.0004246882745088281,
+        "loss": 1.0672,
+        "step": 20810
+      },
+      {
+        "epoch": 0.6639242322778149,
+        "grad_norm": 0.394055038690567,
+        "learning_rate": 0.0004230663591344474,
+        "loss": 1.0909,
+        "step": 20820
+      },
+      {
+        "epoch": 0.6642431199974489,
+        "grad_norm": 0.37684398889541626,
+        "learning_rate": 0.00042145063797271515,
+        "loss": 1.071,
+        "step": 20830
+      },
+      {
+        "epoch": 0.6645620077170828,
+        "grad_norm": 0.36336904764175415,
+        "learning_rate": 0.00041984108736748337,
+        "loss": 1.0739,
+        "step": 20840
+      },
+      {
+        "epoch": 0.6648808954367167,
+        "grad_norm": 0.3808850347995758,
+        "learning_rate": 0.00041823768375294853,
+        "loss": 1.0814,
+        "step": 20850
+      },
+      {
+        "epoch": 0.6651997831563506,
+        "grad_norm": 0.36502712965011597,
+        "learning_rate": 0.0004166404036533064,
+        "loss": 1.0715,
+        "step": 20860
+      },
+      {
+        "epoch": 0.6655186708759846,
+        "grad_norm": 0.3796021640300751,
+        "learning_rate": 0.0004150492236824088,
+        "loss": 1.0884,
+        "step": 20870
+      },
+      {
+        "epoch": 0.6658375585956184,
+        "grad_norm": 0.3758836090564728,
+        "learning_rate": 0.000413464120543421,
+        "loss": 1.076,
+        "step": 20880
+      },
+      {
+        "epoch": 0.6661564463152524,
+        "grad_norm": 0.35722362995147705,
+        "learning_rate": 0.00041188507102848026,
+        "loss": 1.073,
+        "step": 20890
+      },
+      {
+        "epoch": 0.6664753340348863,
+        "grad_norm": 0.3790309429168701,
+        "learning_rate": 0.00041031205201835666,
+        "loss": 1.0811,
+        "step": 20900
+      },
+      {
+        "epoch": 0.6667942217545202,
+        "grad_norm": 0.36754101514816284,
+        "learning_rate": 0.0004087450404821141,
+        "loss": 1.0704,
+        "step": 20910
+      },
+      {
+        "epoch": 0.6671131094741541,
+        "grad_norm": 0.3516288995742798,
+        "learning_rate": 0.0004071840134767734,
+        "loss": 1.0562,
+        "step": 20920
+      },
+      {
+        "epoch": 0.6674319971937881,
+        "grad_norm": 0.35020914673805237,
+        "learning_rate": 0.0004056289481469763,
+        "loss": 1.0726,
+        "step": 20930
+      },
+      {
+        "epoch": 0.6677508849134219,
+        "grad_norm": 0.37123730778694153,
+        "learning_rate": 0.00040407982172465073,
+        "loss": 1.0619,
+        "step": 20940
+      },
+      {
+        "epoch": 0.6680697726330559,
+        "grad_norm": 0.376804381608963,
+        "learning_rate": 0.00040253661152867774,
+        "loss": 1.0813,
+        "step": 20950
+      },
+      {
+        "epoch": 0.6683886603526898,
+        "grad_norm": 0.36201369762420654,
+        "learning_rate": 0.00040099929496455883,
+        "loss": 1.0784,
+        "step": 20960
+      },
+      {
+        "epoch": 0.6687075480723237,
+        "grad_norm": 0.3564017415046692,
+        "learning_rate": 0.0003994678495240859,
+        "loss": 1.0586,
+        "step": 20970
+      },
+      {
+        "epoch": 0.6690264357919576,
+        "grad_norm": 0.3649623394012451,
+        "learning_rate": 0.0003979422527850112,
+        "loss": 1.0657,
+        "step": 20980
+      },
+      {
+        "epoch": 0.6693453235115916,
+        "grad_norm": 0.3724578022956848,
+        "learning_rate": 0.00039642248241071923,
+        "loss": 1.0699,
+        "step": 20990
+      },
+      {
+        "epoch": 0.6696642112312254,
+        "grad_norm": 0.37061256170272827,
+        "learning_rate": 0.0003949085161498995,
+        "loss": 1.0841,
+        "step": 21000
+      },
+      {
+        "epoch": 0.6699830989508594,
+        "grad_norm": 0.3638942539691925,
+        "learning_rate": 0.00039340033183622096,
+        "loss": 1.067,
+        "step": 21010
+      },
+      {
+        "epoch": 0.6703019866704933,
+        "grad_norm": 0.3802492916584015,
+        "learning_rate": 0.00039189790738800755,
+        "loss": 1.0667,
+        "step": 21020
+      },
+      {
+        "epoch": 0.6706208743901272,
+        "grad_norm": 0.3844676911830902,
+        "learning_rate": 0.00039040122080791445,
+        "loss": 1.0759,
+        "step": 21030
+      },
+      {
+        "epoch": 0.6709397621097611,
+        "grad_norm": 0.36140817403793335,
+        "learning_rate": 0.0003889102501826066,
+        "loss": 1.0672,
+        "step": 21040
+      },
+      {
+        "epoch": 0.6712586498293951,
+        "grad_norm": 0.355047345161438,
+        "learning_rate": 0.00038742497368243723,
+        "loss": 1.0726,
+        "step": 21050
+      },
+      {
+        "epoch": 0.6715775375490289,
+        "grad_norm": 0.39054790139198303,
+        "learning_rate": 0.00038594536956112894,
+        "loss": 1.0715,
+        "step": 21060
+      },
+      {
+        "epoch": 0.6718964252686629,
+        "grad_norm": 0.3946945071220398,
+        "learning_rate": 0.00038447141615545455,
+        "loss": 1.0707,
+        "step": 21070
+      },
+      {
+        "epoch": 0.6722153129882968,
+        "grad_norm": 0.3749053478240967,
+        "learning_rate": 0.00038300309188492066,
+        "loss": 1.0905,
+        "step": 21080
+      },
+      {
+        "epoch": 0.6725342007079307,
+        "grad_norm": 0.36572372913360596,
+        "learning_rate": 0.00038154037525145126,
+        "loss": 1.0643,
+        "step": 21090
+      },
+      {
+        "epoch": 0.6728530884275646,
+        "grad_norm": 0.3694365620613098,
+        "learning_rate": 0.0003800832448390728,
+        "loss": 1.0378,
+        "step": 21100
+      },
+      {
+        "epoch": 0.6731719761471986,
+        "grad_norm": 0.36773478984832764,
+        "learning_rate": 0.00037863167931360134,
+        "loss": 1.0899,
+        "step": 21110
+      },
+      {
+        "epoch": 0.6734908638668324,
+        "grad_norm": 0.3820645809173584,
+        "learning_rate": 0.00037718565742232924,
+        "loss": 1.0651,
+        "step": 21120
+      },
+      {
+        "epoch": 0.6738097515864664,
+        "grad_norm": 0.3728492259979248,
+        "learning_rate": 0.0003757451579937149,
+        "loss": 1.0503,
+        "step": 21130
+      },
+      {
+        "epoch": 0.6741286393061003,
+        "grad_norm": 0.3656805753707886,
+        "learning_rate": 0.00037431015993707214,
+        "loss": 1.0667,
+        "step": 21140
+      },
+      {
+        "epoch": 0.6744475270257342,
+        "grad_norm": 0.3748362958431244,
+        "learning_rate": 0.00037288064224226184,
+        "loss": 1.0628,
+        "step": 21150
+      },
+      {
+        "epoch": 0.6747664147453681,
+        "grad_norm": 0.36744022369384766,
+        "learning_rate": 0.00037145658397938383,
+        "loss": 1.086,
+        "step": 21160
+      },
+      {
+        "epoch": 0.6750853024650021,
+        "grad_norm": 0.3650926351547241,
+        "learning_rate": 0.0003700379642984711,
+        "loss": 1.0481,
+        "step": 21170
+      },
+      {
+        "epoch": 0.675404190184636,
+        "grad_norm": 0.3765900135040283,
+        "learning_rate": 0.0003686247624291839,
+        "loss": 1.0837,
+        "step": 21180
+      },
+      {
+        "epoch": 0.6757230779042699,
+        "grad_norm": 0.37482020258903503,
+        "learning_rate": 0.000367216957680506,
+        "loss": 1.0429,
+        "step": 21190
+      },
+      {
+        "epoch": 0.6760419656239038,
+        "grad_norm": 0.3568904995918274,
+        "learning_rate": 0.00036581452944044174,
+        "loss": 1.0468,
+        "step": 21200
+      },
+      {
+        "epoch": 0.6763608533435378,
+        "grad_norm": 0.35615864396095276,
+        "learning_rate": 0.0003644174571757142,
+        "loss": 1.0304,
+        "step": 21210
+      },
+      {
+        "epoch": 0.6766797410631716,
+        "grad_norm": 0.359070360660553,
+        "learning_rate": 0.00036302572043146427,
+        "loss": 1.0553,
+        "step": 21220
+      },
+      {
+        "epoch": 0.6769986287828056,
+        "grad_norm": 0.35599997639656067,
+        "learning_rate": 0.00036163929883095176,
+        "loss": 1.0406,
+        "step": 21230
+      },
+      {
+        "epoch": 0.6773175165024395,
+        "grad_norm": 0.3729799687862396,
+        "learning_rate": 0.0003602581720752565,
+        "loss": 1.0523,
+        "step": 21240
+      },
+      {
+        "epoch": 0.6776364042220734,
+        "grad_norm": 0.35726866126060486,
+        "learning_rate": 0.00035888231994298154,
+        "loss": 1.0702,
+        "step": 21250
+      },
+      {
+        "epoch": 0.6779552919417073,
+        "grad_norm": 0.36645516753196716,
+        "learning_rate": 0.00035751172228995677,
+        "loss": 1.051,
+        "step": 21260
+      },
+      {
+        "epoch": 0.6782741796613413,
+        "grad_norm": 0.37250199913978577,
+        "learning_rate": 0.00035614635904894417,
+        "loss": 1.0621,
+        "step": 21270
+      },
+      {
+        "epoch": 0.6785930673809751,
+        "grad_norm": 0.3667907118797302,
+        "learning_rate": 0.000354786210229344,
+        "loss": 1.0707,
+        "step": 21280
+      },
+      {
+        "epoch": 0.6789119551006091,
+        "grad_norm": 0.35865122079849243,
+        "learning_rate": 0.0003534312559169021,
+        "loss": 1.0424,
+        "step": 21290
+      },
+      {
+        "epoch": 0.679230842820243,
+        "grad_norm": 0.36637625098228455,
+        "learning_rate": 0.00035208147627341824,
+        "loss": 1.0598,
+        "step": 21300
+      },
+      {
+        "epoch": 0.6795497305398769,
+        "grad_norm": 0.36743950843811035,
+        "learning_rate": 0.0003507368515364557,
+        "loss": 1.0601,
+        "step": 21310
+      },
+      {
+        "epoch": 0.6798686182595108,
+        "grad_norm": 0.3567495048046112,
+        "learning_rate": 0.00034939736201905193,
+        "loss": 1.05,
+        "step": 21320
+      },
+      {
+        "epoch": 0.6801875059791448,
+        "grad_norm": 0.3582441806793213,
+        "learning_rate": 0.0003480629881094304,
+        "loss": 1.0754,
+        "step": 21330
+      },
+      {
+        "epoch": 0.6805063936987786,
+        "grad_norm": 0.36522796750068665,
+        "learning_rate": 0.0003467337102707133,
+        "loss": 1.0412,
+        "step": 21340
+      },
+      {
+        "epoch": 0.6808252814184126,
+        "grad_norm": 0.3645346462726593,
+        "learning_rate": 0.0003454095090406356,
+        "loss": 1.0676,
+        "step": 21350
+      },
+      {
+        "epoch": 0.6811441691380465,
+        "grad_norm": 0.36625224351882935,
+        "learning_rate": 0.00034409036503126023,
+        "loss": 1.0462,
+        "step": 21360
+      },
+      {
+        "epoch": 0.6814630568576804,
+        "grad_norm": 0.3716904819011688,
+        "learning_rate": 0.00034277625892869374,
+        "loss": 1.0536,
+        "step": 21370
+      },
+      {
+        "epoch": 0.6817819445773143,
+        "grad_norm": 0.36496251821517944,
+        "learning_rate": 0.0003414671714928041,
+        "loss": 1.0798,
+        "step": 21380
+      },
+      {
+        "epoch": 0.6821008322969483,
+        "grad_norm": 0.386317640542984,
+        "learning_rate": 0.00034016308355693865,
+        "loss": 1.0535,
+        "step": 21390
+      },
+      {
+        "epoch": 0.6824197200165821,
+        "grad_norm": 0.3683474361896515,
+        "learning_rate": 0.0003388639760276436,
+        "loss": 1.0552,
+        "step": 21400
+      },
+      {
+        "epoch": 0.6827386077362161,
+        "grad_norm": 0.3529580533504486,
+        "learning_rate": 0.0003375698298843843,
+        "loss": 1.0574,
+        "step": 21410
+      },
+      {
+        "epoch": 0.68305749545585,
+        "grad_norm": 0.3774348795413971,
+        "learning_rate": 0.00033628062617926724,
+        "loss": 1.0483,
+        "step": 21420
+      },
+      {
+        "epoch": 0.6833763831754839,
+        "grad_norm": 0.3695205748081207,
+        "learning_rate": 0.0003349963460367619,
+        "loss": 1.0497,
+        "step": 21430
+      },
+      {
+        "epoch": 0.6836952708951178,
+        "grad_norm": 0.3666650950908661,
+        "learning_rate": 0.0003337169706534251,
+        "loss": 1.0553,
+        "step": 21440
+      },
+      {
+        "epoch": 0.6840141586147518,
+        "grad_norm": 0.3757452666759491,
+        "learning_rate": 0.00033244248129762514,
+        "loss": 1.0438,
+        "step": 21450
+      },
+      {
+        "epoch": 0.6843330463343856,
+        "grad_norm": 0.3607637882232666,
+        "learning_rate": 0.00033117285930926786,
+        "loss": 1.0489,
+        "step": 21460
+      },
+      {
+        "epoch": 0.6846519340540196,
+        "grad_norm": 0.3701781928539276,
+        "learning_rate": 0.0003299080860995236,
+        "loss": 1.0499,
+        "step": 21470
+      },
+      {
+        "epoch": 0.6849708217736535,
+        "grad_norm": 0.3538482189178467,
+        "learning_rate": 0.00032864814315055425,
+        "loss": 1.0465,
+        "step": 21480
+      },
+      {
+        "epoch": 0.6852897094932874,
+        "grad_norm": 0.3631685972213745,
+        "learning_rate": 0.00032739301201524317,
+        "loss": 1.0559,
+        "step": 21490
+      },
+      {
+        "epoch": 0.6856085972129213,
+        "grad_norm": 0.3718867301940918,
+        "learning_rate": 0.0003261426743169244,
+        "loss": 1.0677,
+        "step": 21500
+      },
+      {
+        "epoch": 0.6859274849325553,
+        "grad_norm": 0.35852405428886414,
+        "learning_rate": 0.0003248971117491138,
+        "loss": 1.0414,
+        "step": 21510
+      },
+      {
+        "epoch": 0.6862463726521891,
+        "grad_norm": 0.3558112680912018,
+        "learning_rate": 0.00032365630607524107,
+        "loss": 1.0516,
+        "step": 21520
+      },
+      {
+        "epoch": 0.6865652603718231,
+        "grad_norm": 0.3524588644504547,
+        "learning_rate": 0.00032242023912838264,
+        "loss": 1.0414,
+        "step": 21530
+      },
+      {
+        "epoch": 0.686884148091457,
+        "grad_norm": 0.3678439259529114,
+        "learning_rate": 0.00032118889281099574,
+        "loss": 1.054,
+        "step": 21540
+      },
+      {
+        "epoch": 0.6872030358110909,
+        "grad_norm": 0.34993577003479004,
+        "learning_rate": 0.0003199622490946535,
+        "loss": 1.0423,
+        "step": 21550
+      },
+      {
+        "epoch": 0.6875219235307248,
+        "grad_norm": 0.36512652039527893,
+        "learning_rate": 0.0003187402900197809,
+        "loss": 1.0519,
+        "step": 21560
+      },
+      {
+        "epoch": 0.6878408112503588,
+        "grad_norm": 0.37657028436660767,
+        "learning_rate": 0.0003175229976953918,
+        "loss": 1.0358,
+        "step": 21570
+      },
+      {
+        "epoch": 0.6881596989699926,
+        "grad_norm": 0.3696797788143158,
+        "learning_rate": 0.000316310354298827,
+        "loss": 1.0247,
+        "step": 21580
+      },
+      {
+        "epoch": 0.6884785866896266,
+        "grad_norm": 0.3626907467842102,
+        "learning_rate": 0.00031510234207549347,
+        "loss": 1.0446,
+        "step": 21590
+      },
+      {
+        "epoch": 0.6887974744092605,
+        "grad_norm": 0.36002790927886963,
+        "learning_rate": 0.0003138989433386042,
+        "loss": 1.0376,
+        "step": 21600
+      },
+      {
+        "epoch": 0.6891163621288944,
+        "grad_norm": 0.3639136552810669,
+        "learning_rate": 0.0003127001404689193,
+        "loss": 1.0464,
+        "step": 21610
+      },
+      {
+        "epoch": 0.6894352498485283,
+        "grad_norm": 0.37046611309051514,
+        "learning_rate": 0.0003115059159144878,
+        "loss": 1.05,
+        "step": 21620
+      },
+      {
+        "epoch": 0.6897541375681623,
+        "grad_norm": 0.3616935610771179,
+        "learning_rate": 0.0003103162521903914,
+        "loss": 1.0301,
+        "step": 21630
+      },
+      {
+        "epoch": 0.6900730252877961,
+        "grad_norm": 0.37073081731796265,
+        "learning_rate": 0.00030913113187848755,
+        "loss": 1.0426,
+        "step": 21640
+      },
+      {
+        "epoch": 0.6903919130074301,
+        "grad_norm": 0.36260053515434265,
+        "learning_rate": 0.00030795053762715507,
+        "loss": 1.0484,
+        "step": 21650
+      },
+      {
+        "epoch": 0.690710800727064,
+        "grad_norm": 0.35945361852645874,
+        "learning_rate": 0.00030677445215103975,
+        "loss": 1.0465,
+        "step": 21660
+      },
+      {
+        "epoch": 0.6910296884466979,
+        "grad_norm": 0.3656920790672302,
+        "learning_rate": 0.00030560285823080156,
+        "loss": 1.0339,
+        "step": 21670
+      },
+      {
+        "epoch": 0.6913485761663318,
+        "grad_norm": 0.3640262186527252,
+        "learning_rate": 0.00030443573871286224,
+        "loss": 1.0495,
+        "step": 21680
+      },
+      {
+        "epoch": 0.6916674638859658,
+        "grad_norm": 0.3602169454097748,
+        "learning_rate": 0.0003032730765091543,
+        "loss": 1.0446,
+        "step": 21690
+      },
+      {
+        "epoch": 0.6919863516055996,
+        "grad_norm": 0.36435046792030334,
+        "learning_rate": 0.000302114854596871,
+        "loss": 1.0479,
+        "step": 21700
+      },
+      {
+        "epoch": 0.6923052393252336,
+        "grad_norm": 0.3639476001262665,
+        "learning_rate": 0.00030096105601821663,
+        "loss": 1.0413,
+        "step": 21710
+      },
+      {
+        "epoch": 0.6926241270448675,
+        "grad_norm": 0.37392833828926086,
+        "learning_rate": 0.0002998116638801587,
+        "loss": 1.0367,
+        "step": 21720
+      },
+      {
+        "epoch": 0.6929430147645014,
+        "grad_norm": 0.3621789216995239,
+        "learning_rate": 0.0002986666613541806,
+        "loss": 1.0523,
+        "step": 21730
+      },
+      {
+        "epoch": 0.6932619024841353,
+        "grad_norm": 0.36604568362236023,
+        "learning_rate": 0.00029752603167603484,
+        "loss": 1.0348,
+        "step": 21740
+      },
+      {
+        "epoch": 0.6935807902037693,
+        "grad_norm": 0.351910799741745,
+        "learning_rate": 0.00029638975814549775,
+        "loss": 1.0403,
+        "step": 21750
+      },
+      {
+        "epoch": 0.6938996779234031,
+        "grad_norm": 0.35878685116767883,
+        "learning_rate": 0.0002952578241261252,
+        "loss": 1.0458,
+        "step": 21760
+      },
+      {
+        "epoch": 0.6942185656430371,
+        "grad_norm": 0.36928388476371765,
+        "learning_rate": 0.00029413021304500875,
+        "loss": 1.0391,
+        "step": 21770
+      },
+      {
+        "epoch": 0.694537453362671,
+        "grad_norm": 0.3780742883682251,
+        "learning_rate": 0.0002930069083925332,
+        "loss": 1.0471,
+        "step": 21780
+      },
+      {
+        "epoch": 0.6948563410823049,
+        "grad_norm": 0.36558210849761963,
+        "learning_rate": 0.0002918878937221347,
+        "loss": 1.0406,
+        "step": 21790
+      },
+      {
+        "epoch": 0.6951752288019388,
+        "grad_norm": 0.36797404289245605,
+        "learning_rate": 0.00029077315265005994,
+        "loss": 1.0435,
+        "step": 21800
+      },
+      {
+        "epoch": 0.6954941165215728,
+        "grad_norm": 0.363120436668396,
+        "learning_rate": 0.00028966266885512655,
+        "loss": 1.0339,
+        "step": 21810
+      },
+      {
+        "epoch": 0.6958130042412066,
+        "grad_norm": 0.36143240332603455,
+        "learning_rate": 0.0002885564260784838,
+        "loss": 1.0464,
+        "step": 21820
+      },
+      {
+        "epoch": 0.6961318919608406,
+        "grad_norm": 0.36355921626091003,
+        "learning_rate": 0.00028745440812337464,
+        "loss": 1.0251,
+        "step": 21830
+      },
+      {
+        "epoch": 0.6964507796804745,
+        "grad_norm": 0.3523947596549988,
+        "learning_rate": 0.00028635659885489876,
+        "loss": 1.0296,
+        "step": 21840
+      },
+      {
+        "epoch": 0.6967696674001084,
+        "grad_norm": 0.36901727318763733,
+        "learning_rate": 0.00028526298219977615,
+        "loss": 1.0425,
+        "step": 21850
+      },
+      {
+        "epoch": 0.6970885551197423,
+        "grad_norm": 0.371360182762146,
+        "learning_rate": 0.0002841735421461118,
+        "loss": 1.0334,
+        "step": 21860
+      },
+      {
+        "epoch": 0.6974074428393763,
+        "grad_norm": 0.35016298294067383,
+        "learning_rate": 0.00028308826274316135,
+        "loss": 1.0492,
+        "step": 21870
+      },
+      {
+        "epoch": 0.6977263305590101,
+        "grad_norm": 0.3725981116294861,
+        "learning_rate": 0.00028200712810109736,
+        "loss": 1.0343,
+        "step": 21880
+      },
+      {
+        "epoch": 0.6980452182786441,
+        "grad_norm": 0.36036965250968933,
+        "learning_rate": 0.00028093012239077697,
+        "loss": 1.0412,
+        "step": 21890
+      },
+      {
+        "epoch": 0.698364105998278,
+        "grad_norm": 0.3615533411502838,
+        "learning_rate": 0.0002798572298435098,
+        "loss": 1.0385,
+        "step": 21900
+      },
+      {
+        "epoch": 0.6986829937179119,
+        "grad_norm": 0.34284889698028564,
+        "learning_rate": 0.00027878843475082743,
+        "loss": 1.0215,
+        "step": 21910
+      },
+      {
+        "epoch": 0.6990018814375458,
+        "grad_norm": 0.35769203305244446,
+        "learning_rate": 0.00027772372146425304,
+        "loss": 1.0215,
+        "step": 21920
+      },
+      {
+        "epoch": 0.6993207691571798,
+        "grad_norm": 0.3554302155971527,
+        "learning_rate": 0.00027666307439507253,
+        "loss": 1.0346,
+        "step": 21930
+      },
+      {
+        "epoch": 0.6996396568768136,
+        "grad_norm": 0.35582250356674194,
+        "learning_rate": 0.0002756064780141064,
+        "loss": 1.0164,
+        "step": 21940
+      },
+      {
+        "epoch": 0.6999585445964476,
+        "grad_norm": 0.35530009865760803,
+        "learning_rate": 0.00027455391685148186,
+        "loss": 1.0241,
+        "step": 21950
+      },
+      {
+        "epoch": 0.7002774323160815,
+        "grad_norm": 0.3660754859447479,
+        "learning_rate": 0.0002735053754964071,
+        "loss": 1.0366,
+        "step": 21960
+      },
+      {
+        "epoch": 0.7005963200357154,
+        "grad_norm": 0.36314699053764343,
+        "learning_rate": 0.00027246083859694485,
+        "loss": 1.0118,
+        "step": 21970
+      },
+      {
+        "epoch": 0.7009152077553493,
+        "grad_norm": 0.3499937355518341,
+        "learning_rate": 0.0002714202908597884,
+        "loss": 1.0371,
+        "step": 21980
+      },
+      {
+        "epoch": 0.7012340954749833,
+        "grad_norm": 0.3634529411792755,
+        "learning_rate": 0.00027038371705003693,
+        "loss": 1.0425,
+        "step": 21990
+      },
+      {
+        "epoch": 0.7015529831946172,
+        "grad_norm": 0.3665272295475006,
+        "learning_rate": 0.0002693511019909731,
+        "loss": 1.0308,
+        "step": 22000
+      },
+      {
+        "epoch": 0.7018718709142511,
+        "grad_norm": 0.3651745617389679,
+        "learning_rate": 0.0002683224305638403,
+        "loss": 1.0397,
+        "step": 22010
+      },
+      {
+        "epoch": 0.702190758633885,
+        "grad_norm": 0.371563196182251,
+        "learning_rate": 0.0002672976877076218,
+        "loss": 1.0231,
+        "step": 22020
+      },
+      {
+        "epoch": 0.702509646353519,
+        "grad_norm": 0.3721901476383209,
+        "learning_rate": 0.0002662768584188198,
+        "loss": 1.0377,
+        "step": 22030
+      },
+      {
+        "epoch": 0.7028285340731528,
+        "grad_norm": 0.36019212007522583,
+        "learning_rate": 0.0002652599277512359,
+        "loss": 1.0302,
+        "step": 22040
+      },
+      {
+        "epoch": 0.7031474217927868,
+        "grad_norm": 0.3766479194164276,
+        "learning_rate": 0.00026424688081575263,
+        "loss": 1.0314,
+        "step": 22050
+      },
+      {
+        "epoch": 0.7034663095124207,
+        "grad_norm": 0.34915393590927124,
+        "learning_rate": 0.0002632377027801149,
+        "loss": 1.0087,
+        "step": 22060
+      },
+      {
+        "epoch": 0.7037851972320546,
+        "grad_norm": 0.36800262331962585,
+        "learning_rate": 0.000262232378868713,
+        "loss": 1.0457,
+        "step": 22070
+      },
+      {
+        "epoch": 0.7041040849516885,
+        "grad_norm": 0.3524855673313141,
+        "learning_rate": 0.00026123089436236655,
+        "loss": 1.0279,
+        "step": 22080
+      },
+      {
+        "epoch": 0.7044229726713225,
+        "grad_norm": 0.3520108461380005,
+        "learning_rate": 0.0002602332345981087,
+        "loss": 1.0168,
+        "step": 22090
+      },
+      {
+        "epoch": 0.7047418603909563,
+        "grad_norm": 0.3598593771457672,
+        "learning_rate": 0.00025923938496897155,
+        "loss": 1.0387,
+        "step": 22100
+      },
+      {
+        "epoch": 0.7050607481105903,
+        "grad_norm": 0.37009191513061523,
+        "learning_rate": 0.0002582493309237722,
+        "loss": 1.0246,
+        "step": 22110
+      },
+      {
+        "epoch": 0.7053796358302242,
+        "grad_norm": 0.35536834597587585,
+        "learning_rate": 0.0002572630579669,
+        "loss": 1.0162,
+        "step": 22120
+      },
+      {
+        "epoch": 0.7056985235498581,
+        "grad_norm": 0.3501632511615753,
+        "learning_rate": 0.0002562805516581037,
+        "loss": 1.0323,
+        "step": 22130
+      },
+      {
+        "epoch": 0.706017411269492,
+        "grad_norm": 0.3461703658103943,
+        "learning_rate": 0.00025530179761228084,
+        "loss": 1.0198,
+        "step": 22140
+      },
+      {
+        "epoch": 0.706336298989126,
+        "grad_norm": 0.34733515977859497,
+        "learning_rate": 0.00025432678149926643,
+        "loss": 1.0226,
+        "step": 22150
+      },
+      {
+        "epoch": 0.7066551867087598,
+        "grad_norm": 0.3601519465446472,
+        "learning_rate": 0.00025335548904362355,
+        "loss": 1.0376,
+        "step": 22160
+      },
+      {
+        "epoch": 0.7069740744283938,
+        "grad_norm": 0.3628826141357422,
+        "learning_rate": 0.0002523879060244341,
+        "loss": 1.0034,
+        "step": 22170
+      },
+      {
+        "epoch": 0.7072929621480277,
+        "grad_norm": 0.3661800026893616,
+        "learning_rate": 0.00025142401827509084,
+        "loss": 1.0231,
+        "step": 22180
+      },
+      {
+        "epoch": 0.7076118498676616,
+        "grad_norm": 0.36240875720977783,
+        "learning_rate": 0.00025046381168308975,
+        "loss": 1.032,
+        "step": 22190
+      },
+      {
+        "epoch": 0.7079307375872955,
+        "grad_norm": 0.35451745986938477,
+        "learning_rate": 0.00024950727218982346,
+        "loss": 1.0213,
+        "step": 22200
+      },
+      {
+        "epoch": 0.7082496253069295,
+        "grad_norm": 0.3512251079082489,
+        "learning_rate": 0.00024855438579037556,
+        "loss": 1.024,
+        "step": 22210
+      },
+      {
+        "epoch": 0.7085685130265633,
+        "grad_norm": 0.349729984998703,
+        "learning_rate": 0.0002476051385333152,
+        "loss": 1.0192,
+        "step": 22220
+      },
+      {
+        "epoch": 0.7088874007461973,
+        "grad_norm": 0.3506690561771393,
+        "learning_rate": 0.0002466595165204934,
+        "loss": 1.0379,
+        "step": 22230
+      },
+      {
+        "epoch": 0.7092062884658312,
+        "grad_norm": 0.368215411901474,
+        "learning_rate": 0.000245717505906839,
+        "loss": 1.0402,
+        "step": 22240
+      },
+      {
+        "epoch": 0.7095251761854651,
+        "grad_norm": 0.36005720496177673,
+        "learning_rate": 0.00024477909290015614,
+        "loss": 1.014,
+        "step": 22250
+      },
+      {
+        "epoch": 0.709844063905099,
+        "grad_norm": 0.34494173526763916,
+        "learning_rate": 0.00024384426376092245,
+        "loss": 1.0154,
+        "step": 22260
+      },
+      {
+        "epoch": 0.710162951624733,
+        "grad_norm": 0.3667897582054138,
+        "learning_rate": 0.00024291300480208775,
+        "loss": 1.0186,
+        "step": 22270
+      },
+      {
+        "epoch": 0.7104818393443668,
+        "grad_norm": 0.36466264724731445,
+        "learning_rate": 0.00024198530238887377,
+        "loss": 1.0204,
+        "step": 22280
+      },
+      {
+        "epoch": 0.7108007270640008,
+        "grad_norm": 0.35865655541419983,
+        "learning_rate": 0.00024106114293857438,
+        "loss": 1.0302,
+        "step": 22290
+      },
+      {
+        "epoch": 0.7111196147836347,
+        "grad_norm": 0.3657973110675812,
+        "learning_rate": 0.00024014051292035676,
+        "loss": 1.0283,
+        "step": 22300
+      },
+      {
+        "epoch": 0.7114385025032686,
+        "grad_norm": 0.36389386653900146,
+        "learning_rate": 0.00023922339885506342,
+        "loss": 1.0234,
+        "step": 22310
+      },
+      {
+        "epoch": 0.7117573902229025,
+        "grad_norm": 0.3584628105163574,
+        "learning_rate": 0.00023830978731501472,
+        "loss": 1.0254,
+        "step": 22320
+      },
+      {
+        "epoch": 0.7120762779425365,
+        "grad_norm": 0.36497706174850464,
+        "learning_rate": 0.00023739966492381225,
+        "loss": 1.0239,
+        "step": 22330
+      },
+      {
+        "epoch": 0.7123951656621703,
+        "grad_norm": 0.36257028579711914,
+        "learning_rate": 0.0002364930183561431,
+        "loss": 1.0161,
+        "step": 22340
+      },
+      {
+        "epoch": 0.7127140533818043,
+        "grad_norm": 0.3534430265426636,
+        "learning_rate": 0.00023558983433758467,
+        "loss": 1.0032,
+        "step": 22350
+      },
+      {
+        "epoch": 0.7130329411014382,
+        "grad_norm": 0.36840540170669556,
+        "learning_rate": 0.0002346900996444103,
+        "loss": 1.0212,
+        "step": 22360
+      },
+      {
+        "epoch": 0.7133518288210721,
+        "grad_norm": 0.35739049315452576,
+        "learning_rate": 0.00023379380110339585,
+        "loss": 1.0179,
+        "step": 22370
+      },
+      {
+        "epoch": 0.713670716540706,
+        "grad_norm": 0.3666969835758209,
+        "learning_rate": 0.00023290092559162653,
+        "loss": 1.0044,
+        "step": 22380
+      },
+      {
+        "epoch": 0.71398960426034,
+        "grad_norm": 0.35255569219589233,
+        "learning_rate": 0.0002320114600363049,
+        "loss": 1.01,
+        "step": 22390
+      },
+      {
+        "epoch": 0.7143084919799738,
+        "grad_norm": 0.3584805130958557,
+        "learning_rate": 0.0002311253914145597,
+        "loss": 0.9974,
+        "step": 22400
+      },
+      {
+        "epoch": 0.7146273796996078,
+        "grad_norm": 0.3672718405723572,
+        "learning_rate": 0.00023024270675325468,
+        "loss": 1.033,
+        "step": 22410
+      },
+      {
+        "epoch": 0.7149462674192417,
+        "grad_norm": 0.3601375222206116,
+        "learning_rate": 0.0002293633931287991,
+        "loss": 1.0273,
+        "step": 22420
+      },
+      {
+        "epoch": 0.7152651551388756,
+        "grad_norm": 0.35628193616867065,
+        "learning_rate": 0.00022848743766695835,
+        "loss": 1.0063,
+        "step": 22430
+      },
+      {
+        "epoch": 0.7155840428585095,
+        "grad_norm": 0.3655938506126404,
+        "learning_rate": 0.00022761482754266545,
+        "loss": 1.0111,
+        "step": 22440
+      },
+      {
+        "epoch": 0.7159029305781435,
+        "grad_norm": 0.37356406450271606,
+        "learning_rate": 0.00022674554997983317,
+        "loss": 1.0128,
+        "step": 22450
+      },
+      {
+        "epoch": 0.7162218182977773,
+        "grad_norm": 0.3504394590854645,
+        "learning_rate": 0.0002258795922511673,
+        "loss": 1.0223,
+        "step": 22460
+      },
+      {
+        "epoch": 0.7165407060174113,
+        "grad_norm": 0.359099417924881,
+        "learning_rate": 0.00022501694167797987,
+        "loss": 1.0093,
+        "step": 22470
+      },
+      {
+        "epoch": 0.7168595937370452,
+        "grad_norm": 0.3635914921760559,
+        "learning_rate": 0.00022415758563000388,
+        "loss": 1.0067,
+        "step": 22480
+      },
+      {
+        "epoch": 0.7171784814566791,
+        "grad_norm": 0.3575493097305298,
+        "learning_rate": 0.0002233015115252082,
+        "loss": 1.0181,
+        "step": 22490
+      },
+      {
+        "epoch": 0.717497369176313,
+        "grad_norm": 0.3696117401123047,
+        "learning_rate": 0.00022244870682961338,
+        "loss": 1.0185,
+        "step": 22500
+      },
+      {
+        "epoch": 0.717816256895947,
+        "grad_norm": 0.3546154201030731,
+        "learning_rate": 0.00022159915905710817,
+        "loss": 1.0165,
+        "step": 22510
+      },
+      {
+        "epoch": 0.7181351446155808,
+        "grad_norm": 0.35239526629447937,
+        "learning_rate": 0.00022075285576926662,
+        "loss": 1.005,
+        "step": 22520
+      },
+      {
+        "epoch": 0.7184540323352148,
+        "grad_norm": 0.3775861859321594,
+        "learning_rate": 0.00021990978457516616,
+        "loss": 1.0082,
+        "step": 22530
+      },
+      {
+        "epoch": 0.7187729200548487,
+        "grad_norm": 0.3526906371116638,
+        "learning_rate": 0.00021906993313120594,
+        "loss": 1.0081,
+        "step": 22540
+      },
+      {
+        "epoch": 0.7190918077744826,
+        "grad_norm": 0.36010169982910156,
+        "learning_rate": 0.00021823328914092629,
+        "loss": 1.0089,
+        "step": 22550
+      },
+      {
+        "epoch": 0.7194106954941165,
+        "grad_norm": 0.3513311445713043,
+        "learning_rate": 0.00021739984035482846,
+        "loss": 1.0087,
+        "step": 22560
+      },
+      {
+        "epoch": 0.7197295832137505,
+        "grad_norm": 0.36615613102912903,
+        "learning_rate": 0.00021656957457019568,
+        "loss": 1.0124,
+        "step": 22570
+      },
+      {
+        "epoch": 0.7200484709333843,
+        "grad_norm": 0.36499765515327454,
+        "learning_rate": 0.00021574247963091408,
+        "loss": 1.0119,
+        "step": 22580
+      },
+      {
+        "epoch": 0.7203673586530183,
+        "grad_norm": 0.37087002396583557,
+        "learning_rate": 0.00021491854342729487,
+        "loss": 1.0047,
+        "step": 22590
+      },
+      {
+        "epoch": 0.7206862463726522,
+        "grad_norm": 0.3546287417411804,
+        "learning_rate": 0.00021409775389589705,
+        "loss": 1.0131,
+        "step": 22600
+      },
+      {
+        "epoch": 0.7210051340922861,
+        "grad_norm": 0.3611092269420624,
+        "learning_rate": 0.00021328009901935087,
+        "loss": 0.9961,
+        "step": 22610
+      },
+      {
+        "epoch": 0.72132402181192,
+        "grad_norm": 0.3478728234767914,
+        "learning_rate": 0.00021246556682618162,
+        "loss": 1.0134,
+        "step": 22620
+      },
+      {
+        "epoch": 0.721642909531554,
+        "grad_norm": 0.3665063977241516,
+        "learning_rate": 0.0002116541453906347,
+        "loss": 1.0069,
+        "step": 22630
+      },
+      {
+        "epoch": 0.7219617972511878,
+        "grad_norm": 0.35230499505996704,
+        "learning_rate": 0.0002108458228325007,
+        "loss": 0.9957,
+        "step": 22640
+      },
+      {
+        "epoch": 0.7222806849708218,
+        "grad_norm": 0.3609510362148285,
+        "learning_rate": 0.0002100405873169417,
+        "loss": 1.0128,
+        "step": 22650
+      },
+      {
+        "epoch": 0.7225995726904557,
+        "grad_norm": 0.38953086733818054,
+        "learning_rate": 0.0002092384270543178,
+        "loss": 1.0046,
+        "step": 22660
+      },
+      {
+        "epoch": 0.7229184604100896,
+        "grad_norm": 0.36705484986305237,
+        "learning_rate": 0.00020843933030001454,
+        "loss": 0.9997,
+        "step": 22670
+      },
+      {
+        "epoch": 0.7232373481297235,
+        "grad_norm": 0.35538023710250854,
+        "learning_rate": 0.00020764328535427122,
+        "loss": 1.005,
+        "step": 22680
+      },
+      {
+        "epoch": 0.7235562358493575,
+        "grad_norm": 0.3577936589717865,
+        "learning_rate": 0.0002068502805620091,
+        "loss": 1.0213,
+        "step": 22690
+      },
+      {
+        "epoch": 0.7238751235689913,
+        "grad_norm": 0.3556760549545288,
+        "learning_rate": 0.0002060603043126613,
+        "loss": 1.0144,
+        "step": 22700
+      },
+      {
+        "epoch": 0.7241940112886253,
+        "grad_norm": 0.37856200337409973,
+        "learning_rate": 0.00020527334504000228,
+        "loss": 1.0109,
+        "step": 22710
+      },
+      {
+        "epoch": 0.7245128990082592,
+        "grad_norm": 0.3712582290172577,
+        "learning_rate": 0.00020448939122197894,
+        "loss": 0.997,
+        "step": 22720
+      },
+      {
+        "epoch": 0.7248317867278931,
+        "grad_norm": 0.3549712598323822,
+        "learning_rate": 0.00020370843138054165,
+        "loss": 1.0027,
+        "step": 22730
+      },
+      {
+        "epoch": 0.725150674447527,
+        "grad_norm": 0.37729141116142273,
+        "learning_rate": 0.00020293045408147637,
+        "loss": 0.9918,
+        "step": 22740
+      },
+      {
+        "epoch": 0.725469562167161,
+        "grad_norm": 0.36283665895462036,
+        "learning_rate": 0.00020215544793423702,
+        "loss": 0.9999,
+        "step": 22750
+      },
+      {
+        "epoch": 0.7257884498867948,
+        "grad_norm": 0.37182876467704773,
+        "learning_rate": 0.00020138340159177898,
+        "loss": 1.0097,
+        "step": 22760
+      },
+      {
+        "epoch": 0.7261073376064288,
+        "grad_norm": 0.36891797184944153,
+        "learning_rate": 0.00020061430375039268,
+        "loss": 1.0067,
+        "step": 22770
+      },
+      {
+        "epoch": 0.7264262253260627,
+        "grad_norm": 0.36455395817756653,
+        "learning_rate": 0.00019984814314953845,
+        "loss": 0.9968,
+        "step": 22780
+      },
+      {
+        "epoch": 0.7267451130456966,
+        "grad_norm": 0.36560770869255066,
+        "learning_rate": 0.0001990849085716812,
+        "loss": 1.0218,
+        "step": 22790
+      },
+      {
+        "epoch": 0.7270640007653305,
+        "grad_norm": 0.3545674681663513,
+        "learning_rate": 0.0001983245888421265,
+        "loss": 1.001,
+        "step": 22800
+      },
+      {
+        "epoch": 0.7273828884849645,
+        "grad_norm": 0.3545055687427521,
+        "learning_rate": 0.000197567172828857,
+        "loss": 1.0128,
+        "step": 22810
+      },
+      {
+        "epoch": 0.7277017762045984,
+        "grad_norm": 0.36697012186050415,
+        "learning_rate": 0.0001968126494423692,
+        "loss": 1.0175,
+        "step": 22820
+      },
+      {
+        "epoch": 0.7280206639242323,
+        "grad_norm": 0.35822004079818726,
+        "learning_rate": 0.0001960610076355113,
+        "loss": 1.0116,
+        "step": 22830
+      },
+      {
+        "epoch": 0.7283395516438662,
+        "grad_norm": 0.3711821436882019,
+        "learning_rate": 0.00019531223640332132,
+        "loss": 1.0206,
+        "step": 22840
+      },
+      {
+        "epoch": 0.7286584393635002,
+        "grad_norm": 0.3546949028968811,
+        "learning_rate": 0.00019456632478286607,
+        "loss": 1.0219,
+        "step": 22850
+      },
+      {
+        "epoch": 0.728977327083134,
+        "grad_norm": 0.36350348591804504,
+        "learning_rate": 0.00019382326185308054,
+        "loss": 0.9975,
+        "step": 22860
+      },
+      {
+        "epoch": 0.729296214802768,
+        "grad_norm": 0.3645276725292206,
+        "learning_rate": 0.00019308303673460815,
+        "loss": 1.0059,
+        "step": 22870
+      },
+      {
+        "epoch": 0.7296151025224019,
+        "grad_norm": 0.3612558841705322,
+        "learning_rate": 0.0001923456385896413,
+        "loss": 0.9885,
+        "step": 22880
+      },
+      {
+        "epoch": 0.7299339902420358,
+        "grad_norm": 0.35814574360847473,
+        "learning_rate": 0.0001916110566217628,
+        "loss": 0.9994,
+        "step": 22890
+      },
+      {
+        "epoch": 0.7302528779616697,
+        "grad_norm": 0.36161062121391296,
+        "learning_rate": 0.00019087928007578776,
+        "loss": 1.0177,
+        "step": 22900
+      },
+      {
+        "epoch": 0.7305717656813037,
+        "grad_norm": 0.350792795419693,
+        "learning_rate": 0.0001901502982376061,
+        "loss": 0.9979,
+        "step": 22910
+      },
+      {
+        "epoch": 0.7308906534009375,
+        "grad_norm": 0.35586076974868774,
+        "learning_rate": 0.00018942410043402573,
+        "loss": 0.9949,
+        "step": 22920
+      },
+      {
+        "epoch": 0.7312095411205715,
+        "grad_norm": 0.3612673282623291,
+        "learning_rate": 0.0001887006760326162,
+        "loss": 1.0101,
+        "step": 22930
+      },
+      {
+        "epoch": 0.7315284288402054,
+        "grad_norm": 0.35979345440864563,
+        "learning_rate": 0.00018798001444155318,
+        "loss": 0.9986,
+        "step": 22940
+      },
+      {
+        "epoch": 0.7318473165598393,
+        "grad_norm": 0.3546121418476105,
+        "learning_rate": 0.00018726210510946323,
+        "loss": 0.9955,
+        "step": 22950
+      },
+      {
+        "epoch": 0.7321662042794732,
+        "grad_norm": 0.35720688104629517,
+        "learning_rate": 0.00018654693752526939,
+        "loss": 0.9991,
+        "step": 22960
+      },
+      {
+        "epoch": 0.7324850919991072,
+        "grad_norm": 0.36118242144584656,
+        "learning_rate": 0.00018583450121803722,
+        "loss": 0.986,
+        "step": 22970
+      },
+      {
+        "epoch": 0.732803979718741,
+        "grad_norm": 0.36114799976348877,
+        "learning_rate": 0.00018512478575682153,
+        "loss": 0.9957,
+        "step": 22980
+      },
+      {
+        "epoch": 0.733122867438375,
+        "grad_norm": 0.35737699270248413,
+        "learning_rate": 0.00018441778075051386,
+        "loss": 0.9852,
+        "step": 22990
+      },
+      {
+        "epoch": 0.7334417551580089,
+        "grad_norm": 0.3749620020389557,
+        "learning_rate": 0.00018371347584768985,
+        "loss": 1.0069,
+        "step": 23000
+      },
+      {
+        "epoch": 0.7337606428776428,
+        "grad_norm": 0.35069161653518677,
+        "learning_rate": 0.0001830118607364582,
+        "loss": 0.9929,
+        "step": 23010
+      },
+      {
+        "epoch": 0.7340795305972767,
+        "grad_norm": 0.36476099491119385,
+        "learning_rate": 0.00018231292514430937,
+        "loss": 1.0061,
+        "step": 23020
+      },
+      {
+        "epoch": 0.7343984183169107,
+        "grad_norm": 0.35989847779273987,
+        "learning_rate": 0.00018161665883796536,
+        "loss": 1.0136,
+        "step": 23030
+      },
+      {
+        "epoch": 0.7347173060365445,
+        "grad_norm": 0.353773295879364,
+        "learning_rate": 0.0001809230516232297,
+        "loss": 0.9859,
+        "step": 23040
+      },
+      {
+        "epoch": 0.7350361937561785,
+        "grad_norm": 0.3702070713043213,
+        "learning_rate": 0.0001802320933448384,
+        "loss": 1.0134,
+        "step": 23050
+      },
+      {
+        "epoch": 0.7353550814758124,
+        "grad_norm": 0.37036627531051636,
+        "learning_rate": 0.0001795437738863111,
+        "loss": 1.001,
+        "step": 23060
+      },
+      {
+        "epoch": 0.7356739691954463,
+        "grad_norm": 0.355307936668396,
+        "learning_rate": 0.000178858083169803,
+        "loss": 0.9962,
+        "step": 23070
+      },
+      {
+        "epoch": 0.7359928569150802,
+        "grad_norm": 0.35053226351737976,
+        "learning_rate": 0.00017817501115595735,
+        "loss": 1.0033,
+        "step": 23080
+      },
+      {
+        "epoch": 0.7363117446347142,
+        "grad_norm": 0.3521139919757843,
+        "learning_rate": 0.0001774945478437585,
+        "loss": 0.9785,
+        "step": 23090
+      },
+      {
+        "epoch": 0.736630632354348,
+        "grad_norm": 0.36836716532707214,
+        "learning_rate": 0.00017681668327038525,
+        "loss": 0.9966,
+        "step": 23100
+      },
+      {
+        "epoch": 0.736949520073982,
+        "grad_norm": 0.3591693937778473,
+        "learning_rate": 0.00017614140751106525,
+        "loss": 0.9818,
+        "step": 23110
+      },
+      {
+        "epoch": 0.7372684077936159,
+        "grad_norm": 0.36045387387275696,
+        "learning_rate": 0.00017546871067892955,
+        "loss": 0.9973,
+        "step": 23120
+      },
+      {
+        "epoch": 0.7375872955132498,
+        "grad_norm": 0.3539069592952728,
+        "learning_rate": 0.00017479858292486792,
+        "loss": 0.99,
+        "step": 23130
+      },
+      {
+        "epoch": 0.7379061832328837,
+        "grad_norm": 0.37071090936660767,
+        "learning_rate": 0.0001741310144373845,
+        "loss": 1.0186,
+        "step": 23140
+      },
+      {
+        "epoch": 0.7382250709525177,
+        "grad_norm": 0.34401729702949524,
+        "learning_rate": 0.00017346599544245428,
+        "loss": 1.0128,
+        "step": 23150
+      },
+      {
+        "epoch": 0.7385439586721515,
+        "grad_norm": 0.362400621175766,
+        "learning_rate": 0.00017280351620338,
+        "loss": 0.9964,
+        "step": 23160
+      },
+      {
+        "epoch": 0.7388628463917855,
+        "grad_norm": 0.3564687669277191,
+        "learning_rate": 0.00017214356702064952,
+        "loss": 1.0032,
+        "step": 23170
+      },
+      {
+        "epoch": 0.7391817341114194,
+        "grad_norm": 0.35708966851234436,
+        "learning_rate": 0.00017148613823179387,
+        "loss": 0.9913,
+        "step": 23180
+      },
+      {
+        "epoch": 0.7395006218310533,
+        "grad_norm": 0.3605559170246124,
+        "learning_rate": 0.00017083122021124575,
+        "loss": 0.9906,
+        "step": 23190
+      },
+      {
+        "epoch": 0.7398195095506872,
+        "grad_norm": 0.36472660303115845,
+        "learning_rate": 0.00017017880337019854,
+        "loss": 0.9896,
+        "step": 23200
+      },
+      {
+        "epoch": 0.7401383972703212,
+        "grad_norm": 0.3542068302631378,
+        "learning_rate": 0.00016952887815646604,
+        "loss": 0.9998,
+        "step": 23210
+      },
+      {
+        "epoch": 0.740457284989955,
+        "grad_norm": 0.3632705807685852,
+        "learning_rate": 0.0001688814350543425,
+        "loss": 0.9837,
+        "step": 23220
+      },
+      {
+        "epoch": 0.740776172709589,
+        "grad_norm": 0.3601367473602295,
+        "learning_rate": 0.00016823646458446337,
+        "loss": 0.9773,
+        "step": 23230
+      },
+      {
+        "epoch": 0.7410950604292229,
+        "grad_norm": 0.3728727102279663,
+        "learning_rate": 0.00016759395730366647,
+        "loss": 0.9863,
+        "step": 23240
+      },
+      {
+        "epoch": 0.7414139481488567,
+        "grad_norm": 0.3813025951385498,
+        "learning_rate": 0.00016695390380485372,
+        "loss": 0.9903,
+        "step": 23250
+      },
+      {
+        "epoch": 0.7417328358684907,
+        "grad_norm": 0.3660871088504791,
+        "learning_rate": 0.00016631629471685345,
+        "loss": 0.9877,
+        "step": 23260
+      },
+      {
+        "epoch": 0.7420517235881247,
+        "grad_norm": 0.35876861214637756,
+        "learning_rate": 0.0001656811207042832,
+        "loss": 0.9885,
+        "step": 23270
+      },
+      {
+        "epoch": 0.7423706113077585,
+        "grad_norm": 0.3620452284812927,
+        "learning_rate": 0.00016504837246741303,
+        "loss": 0.981,
+        "step": 23280
+      },
+      {
+        "epoch": 0.7426894990273925,
+        "grad_norm": 0.3587225675582886,
+        "learning_rate": 0.00016441804074202924,
+        "loss": 0.9838,
+        "step": 23290
+      },
+      {
+        "epoch": 0.7430083867470264,
+        "grad_norm": 0.3670322895050049,
+        "learning_rate": 0.00016379011629929894,
+        "loss": 1.0061,
+        "step": 23300
+      },
+      {
+        "epoch": 0.7433272744666602,
+        "grad_norm": 0.36198362708091736,
+        "learning_rate": 0.00016316458994563475,
+        "loss": 0.9907,
+        "step": 23310
+      },
+      {
+        "epoch": 0.7436461621862942,
+        "grad_norm": 0.3827613890171051,
+        "learning_rate": 0.00016254145252256035,
+        "loss": 1.0024,
+        "step": 23320
+      },
+      {
+        "epoch": 0.7439650499059282,
+        "grad_norm": 0.3546290993690491,
+        "learning_rate": 0.0001619206949065762,
+        "loss": 0.999,
+        "step": 23330
+      },
+      {
+        "epoch": 0.744283937625562,
+        "grad_norm": 0.36129066348075867,
+        "learning_rate": 0.00016130230800902616,
+        "loss": 0.9967,
+        "step": 23340
+      },
+      {
+        "epoch": 0.744602825345196,
+        "grad_norm": 0.356289803981781,
+        "learning_rate": 0.00016068628277596432,
+        "loss": 0.9941,
+        "step": 23350
+      },
+      {
+        "epoch": 0.7449217130648299,
+        "grad_norm": 0.3629865348339081,
+        "learning_rate": 0.00016007261018802238,
+        "loss": 0.9751,
+        "step": 23360
+      },
+      {
+        "epoch": 0.7452406007844637,
+        "grad_norm": 0.34594714641571045,
+        "learning_rate": 0.00015946128126027772,
+        "loss": 0.9881,
+        "step": 23370
+      },
+      {
+        "epoch": 0.7455594885040977,
+        "grad_norm": 0.3698471784591675,
+        "learning_rate": 0.0001588522870421218,
+        "loss": 0.9857,
+        "step": 23380
+      },
+      {
+        "epoch": 0.7458783762237317,
+        "grad_norm": 0.37548011541366577,
+        "learning_rate": 0.00015824561861712907,
+        "loss": 0.9918,
+        "step": 23390
+      },
+      {
+        "epoch": 0.7461972639433655,
+        "grad_norm": 0.3651987612247467,
+        "learning_rate": 0.00015764126710292645,
+        "loss": 0.9896,
+        "step": 23400
+      },
+      {
+        "epoch": 0.7465161516629994,
+        "grad_norm": 0.34970924258232117,
+        "learning_rate": 0.0001570392236510633,
+        "loss": 1.003,
+        "step": 23410
+      },
+      {
+        "epoch": 0.7468350393826334,
+        "grad_norm": 0.3621990978717804,
+        "learning_rate": 0.00015643947944688188,
+        "loss": 0.9941,
+        "step": 23420
+      },
+      {
+        "epoch": 0.7471539271022672,
+        "grad_norm": 0.35395199060440063,
+        "learning_rate": 0.00015584202570938826,
+        "loss": 0.9834,
+        "step": 23430
+      },
+      {
+        "epoch": 0.7474728148219012,
+        "grad_norm": 0.37184348702430725,
+        "learning_rate": 0.00015524685369112376,
+        "loss": 0.9715,
+        "step": 23440
+      },
+      {
+        "epoch": 0.7477917025415352,
+        "grad_norm": 0.349755197763443,
+        "learning_rate": 0.00015465395467803683,
+        "loss": 0.976,
+        "step": 23450
+      },
+      {
+        "epoch": 0.748110590261169,
+        "grad_norm": 0.37444648146629333,
+        "learning_rate": 0.00015406331998935564,
+        "loss": 1.0032,
+        "step": 23460
+      },
+      {
+        "epoch": 0.748429477980803,
+        "grad_norm": 0.36452704668045044,
+        "learning_rate": 0.00015347494097746068,
+        "loss": 0.9924,
+        "step": 23470
+      },
+      {
+        "epoch": 0.7487483657004369,
+        "grad_norm": 0.3637312352657318,
+        "learning_rate": 0.0001528888090277585,
+        "loss": 0.9791,
+        "step": 23480
+      },
+      {
+        "epoch": 0.7490672534200707,
+        "grad_norm": 0.36109039187431335,
+        "learning_rate": 0.0001523049155585553,
+        "loss": 0.9841,
+        "step": 23490
+      },
+      {
+        "epoch": 0.7493861411397047,
+        "grad_norm": 0.36957335472106934,
+        "learning_rate": 0.00015172325202093142,
+        "loss": 0.9906,
+        "step": 23500
+      },
+      {
+        "epoch": 0.7497050288593387,
+        "grad_norm": 0.3457266688346863,
+        "learning_rate": 0.00015114380989861618,
+        "loss": 0.9777,
+        "step": 23510
+      },
+      {
+        "epoch": 0.7500239165789725,
+        "grad_norm": 0.359375,
+        "learning_rate": 0.0001505665807078631,
+        "loss": 0.9763,
+        "step": 23520
+      },
+      {
+        "epoch": 0.7503428042986064,
+        "grad_norm": 0.36092039942741394,
+        "learning_rate": 0.00014999155599732582,
+        "loss": 0.9858,
+        "step": 23530
+      },
+      {
+        "epoch": 0.7506616920182404,
+        "grad_norm": 0.34542810916900635,
+        "learning_rate": 0.00014941872734793423,
+        "loss": 0.9811,
+        "step": 23540
+      },
+      {
+        "epoch": 0.7509805797378742,
+        "grad_norm": 0.36623185873031616,
+        "learning_rate": 0.0001488480863727712,
+        "loss": 0.9933,
+        "step": 23550
+      },
+      {
+        "epoch": 0.7512994674575082,
+        "grad_norm": 0.34604063630104065,
+        "learning_rate": 0.00014827962471694999,
+        "loss": 0.9802,
+        "step": 23560
+      },
+      {
+        "epoch": 0.7516183551771422,
+        "grad_norm": 0.36921989917755127,
+        "learning_rate": 0.00014771333405749166,
+        "loss": 0.993,
+        "step": 23570
+      },
+      {
+        "epoch": 0.751937242896776,
+        "grad_norm": 0.3590368330478668,
+        "learning_rate": 0.00014714920610320336,
+        "loss": 0.9773,
+        "step": 23580
+      },
+      {
+        "epoch": 0.75225613061641,
+        "grad_norm": 0.36252063512802124,
+        "learning_rate": 0.00014658723259455698,
+        "loss": 0.9775,
+        "step": 23590
+      },
+      {
+        "epoch": 0.7525750183360439,
+        "grad_norm": 0.36665475368499756,
+        "learning_rate": 0.00014602740530356807,
+        "loss": 0.9747,
+        "step": 23600
+      },
+      {
+        "epoch": 0.7528939060556777,
+        "grad_norm": 0.36198699474334717,
+        "learning_rate": 0.00014546971603367545,
+        "loss": 0.9912,
+        "step": 23610
+      },
+      {
+        "epoch": 0.7532127937753117,
+        "grad_norm": 0.3599792718887329,
+        "learning_rate": 0.00014491415661962125,
+        "loss": 0.9923,
+        "step": 23620
+      },
+      {
+        "epoch": 0.7535316814949456,
+        "grad_norm": 0.36203503608703613,
+        "learning_rate": 0.00014436071892733128,
+        "loss": 0.9944,
+        "step": 23630
+      },
+      {
+        "epoch": 0.7538505692145795,
+        "grad_norm": 0.34409478306770325,
+        "learning_rate": 0.00014380939485379596,
+        "loss": 0.9663,
+        "step": 23640
+      },
+      {
+        "epoch": 0.7541694569342134,
+        "grad_norm": 0.38052651286125183,
+        "learning_rate": 0.00014326017632695172,
+        "loss": 0.9878,
+        "step": 23650
+      },
+      {
+        "epoch": 0.7544883446538474,
+        "grad_norm": 0.3748937249183655,
+        "learning_rate": 0.00014271305530556277,
+        "loss": 0.9844,
+        "step": 23660
+      },
+      {
+        "epoch": 0.7548072323734814,
+        "grad_norm": 0.3574012517929077,
+        "learning_rate": 0.00014216802377910344,
+        "loss": 0.9798,
+        "step": 23670
+      },
+      {
+        "epoch": 0.7551261200931152,
+        "grad_norm": 0.3533320128917694,
+        "learning_rate": 0.00014162507376764066,
+        "loss": 0.9785,
+        "step": 23680
+      },
+      {
+        "epoch": 0.7554450078127491,
+        "grad_norm": 0.3751009702682495,
+        "learning_rate": 0.0001410841973217176,
+        "loss": 0.9992,
+        "step": 23690
+      },
+      {
+        "epoch": 0.7557638955323831,
+        "grad_norm": 0.35940220952033997,
+        "learning_rate": 0.0001405453865222367,
+        "loss": 0.9833,
+        "step": 23700
+      },
+      {
+        "epoch": 0.7560827832520169,
+        "grad_norm": 0.350617378950119,
+        "learning_rate": 0.0001400086334803442,
+        "loss": 0.9887,
+        "step": 23710
+      },
+      {
+        "epoch": 0.7564016709716509,
+        "grad_norm": 0.37829482555389404,
+        "learning_rate": 0.0001394739303373143,
+        "loss": 0.9735,
+        "step": 23720
+      },
+      {
+        "epoch": 0.7567205586912849,
+        "grad_norm": 0.3666497468948364,
+        "learning_rate": 0.0001389412692644344,
+        "loss": 0.9778,
+        "step": 23730
+      },
+      {
+        "epoch": 0.7570394464109187,
+        "grad_norm": 0.3569285273551941,
+        "learning_rate": 0.00013841064246289018,
+        "loss": 0.9804,
+        "step": 23740
+      },
+      {
+        "epoch": 0.7573583341305526,
+        "grad_norm": 0.3598553240299225,
+        "learning_rate": 0.00013788204216365162,
+        "loss": 0.9758,
+        "step": 23750
+      },
+      {
+        "epoch": 0.7576772218501866,
+        "grad_norm": 0.3719578683376312,
+        "learning_rate": 0.00013735546062735915,
+        "loss": 0.9679,
+        "step": 23760
+      },
+      {
+        "epoch": 0.7579961095698204,
+        "grad_norm": 0.37967368960380554,
+        "learning_rate": 0.00013683089014421051,
+        "loss": 0.9899,
+        "step": 23770
+      },
+      {
+        "epoch": 0.7583149972894544,
+        "grad_norm": 0.3560192584991455,
+        "learning_rate": 0.0001363083230338476,
+        "loss": 0.9694,
+        "step": 23780
+      },
+      {
+        "epoch": 0.7586338850090883,
+        "grad_norm": 0.36281514167785645,
+        "learning_rate": 0.00013578775164524418,
+        "loss": 0.9864,
+        "step": 23790
+      },
+      {
+        "epoch": 0.7589527727287222,
+        "grad_norm": 0.36002716422080994,
+        "learning_rate": 0.00013526916835659387,
+        "loss": 1.0041,
+        "step": 23800
+      },
+      {
+        "epoch": 0.7592716604483561,
+        "grad_norm": 0.3768812119960785,
+        "learning_rate": 0.00013475256557519852,
+        "loss": 0.9687,
+        "step": 23810
+      },
+      {
+        "epoch": 0.7595905481679901,
+        "grad_norm": 0.36057835817337036,
+        "learning_rate": 0.00013423793573735703,
+        "loss": 0.9824,
+        "step": 23820
+      },
+      {
+        "epoch": 0.7599094358876239,
+        "grad_norm": 0.3561815619468689,
+        "learning_rate": 0.00013372527130825463,
+        "loss": 0.9826,
+        "step": 23830
+      },
+      {
+        "epoch": 0.7602283236072579,
+        "grad_norm": 0.3604488968849182,
+        "learning_rate": 0.00013321456478185254,
+        "loss": 0.9808,
+        "step": 23840
+      },
+      {
+        "epoch": 0.7605472113268918,
+        "grad_norm": 0.3659672141075134,
+        "learning_rate": 0.00013270580868077813,
+        "loss": 0.9812,
+        "step": 23850
+      },
+      {
+        "epoch": 0.7608660990465257,
+        "grad_norm": 0.3699098527431488,
+        "learning_rate": 0.00013219899555621538,
+        "loss": 0.9706,
+        "step": 23860
+      },
+      {
+        "epoch": 0.7611849867661596,
+        "grad_norm": 0.3599204123020172,
+        "learning_rate": 0.00013169411798779585,
+        "loss": 0.9992,
+        "step": 23870
+      },
+      {
+        "epoch": 0.7615038744857936,
+        "grad_norm": 0.3598966896533966,
+        "learning_rate": 0.00013119116858348995,
+        "loss": 0.9848,
+        "step": 23880
+      },
+      {
+        "epoch": 0.7618227622054274,
+        "grad_norm": 0.3556804358959198,
+        "learning_rate": 0.0001306901399794989,
+        "loss": 0.9872,
+        "step": 23890
+      },
+      {
+        "epoch": 0.7621416499250614,
+        "grad_norm": 0.3712446689605713,
+        "learning_rate": 0.0001301910248401467,
+        "loss": 0.9711,
+        "step": 23900
+      },
+      {
+        "epoch": 0.7624605376446953,
+        "grad_norm": 0.3657471835613251,
+        "learning_rate": 0.0001296938158577729,
+        "loss": 0.9854,
+        "step": 23910
+      },
+      {
+        "epoch": 0.7627794253643292,
+        "grad_norm": 0.3565327227115631,
+        "learning_rate": 0.0001291985057526255,
+        "loss": 0.9698,
+        "step": 23920
+      },
+      {
+        "epoch": 0.7630983130839631,
+        "grad_norm": 0.35236185789108276,
+        "learning_rate": 0.00012870508727275446,
+        "loss": 0.9737,
+        "step": 23930
+      },
+      {
+        "epoch": 0.7634172008035971,
+        "grad_norm": 0.3755548894405365,
+        "learning_rate": 0.0001282135531939054,
+        "loss": 0.9882,
+        "step": 23940
+      },
+      {
+        "epoch": 0.7637360885232309,
+        "grad_norm": 0.36770176887512207,
+        "learning_rate": 0.00012772389631941392,
+        "loss": 0.9807,
+        "step": 23950
+      },
+      {
+        "epoch": 0.7640549762428649,
+        "grad_norm": 0.3637702763080597,
+        "learning_rate": 0.00012723610948010017,
+        "loss": 0.977,
+        "step": 23960
+      },
+      {
+        "epoch": 0.7643738639624988,
+        "grad_norm": 0.36087551712989807,
+        "learning_rate": 0.00012675018553416398,
+        "loss": 0.9674,
+        "step": 23970
+      },
+      {
+        "epoch": 0.7646927516821327,
+        "grad_norm": 0.3561502993106842,
+        "learning_rate": 0.00012626611736708018,
+        "loss": 0.9812,
+        "step": 23980
+      },
+      {
+        "epoch": 0.7650116394017666,
+        "grad_norm": 0.37487906217575073,
+        "learning_rate": 0.00012578389789149453,
+        "loss": 0.9791,
+        "step": 23990
+      },
+      {
+        "epoch": 0.7653305271214006,
+        "grad_norm": 0.3568199872970581,
+        "learning_rate": 0.00012530352004711987,
+        "loss": 0.9932,
+        "step": 24000
+      },
+      {
+        "epoch": 0.7656494148410344,
+        "grad_norm": 0.3541073799133301,
+        "learning_rate": 0.00012482497680063275,
+        "loss": 0.9596,
+        "step": 24010
+      },
+      {
+        "epoch": 0.7659683025606684,
+        "grad_norm": 0.3517587184906006,
+        "learning_rate": 0.00012434826114557067,
+        "loss": 0.9755,
+        "step": 24020
+      },
+      {
+        "epoch": 0.7662871902803023,
+        "grad_norm": 0.37279677391052246,
+        "learning_rate": 0.00012387336610222914,
+        "loss": 0.9945,
+        "step": 24030
+      },
+      {
+        "epoch": 0.7666060779999362,
+        "grad_norm": 0.34747180342674255,
+        "learning_rate": 0.00012340028471755974,
+        "loss": 0.9832,
+        "step": 24040
+      },
+      {
+        "epoch": 0.7669249657195701,
+        "grad_norm": 0.36472681164741516,
+        "learning_rate": 0.00012292901006506822,
+        "loss": 0.9802,
+        "step": 24050
+      },
+      {
+        "epoch": 0.7672438534392041,
+        "grad_norm": 0.3583838939666748,
+        "learning_rate": 0.00012245953524471316,
+        "loss": 0.9837,
+        "step": 24060
+      },
+      {
+        "epoch": 0.7675627411588379,
+        "grad_norm": 0.3646237254142761,
+        "learning_rate": 0.00012199185338280484,
+        "loss": 1.0001,
+        "step": 24070
+      },
+      {
+        "epoch": 0.7678816288784719,
+        "grad_norm": 0.3578408360481262,
+        "learning_rate": 0.00012152595763190471,
+        "loss": 0.9867,
+        "step": 24080
+      },
+      {
+        "epoch": 0.7682005165981058,
+        "grad_norm": 0.358577162027359,
+        "learning_rate": 0.00012106184117072507,
+        "loss": 0.9757,
+        "step": 24090
+      },
+      {
+        "epoch": 0.7685194043177397,
+        "grad_norm": 0.3546871840953827,
+        "learning_rate": 0.00012059949720402918,
+        "loss": 0.9805,
+        "step": 24100
+      },
+      {
+        "epoch": 0.7688382920373736,
+        "grad_norm": 0.3737013638019562,
+        "learning_rate": 0.00012013891896253189,
+        "loss": 0.9851,
+        "step": 24110
+      },
+      {
+        "epoch": 0.7691571797570076,
+        "grad_norm": 0.35779833793640137,
+        "learning_rate": 0.00011968009970280033,
+        "loss": 0.972,
+        "step": 24120
+      },
+      {
+        "epoch": 0.7694760674766414,
+        "grad_norm": 0.3553289473056793,
+        "learning_rate": 0.00011922303270715539,
+        "loss": 0.9745,
+        "step": 24130
+      },
+      {
+        "epoch": 0.7697949551962754,
+        "grad_norm": 0.3517867922782898,
+        "learning_rate": 0.00011876771128357317,
+        "loss": 0.9777,
+        "step": 24140
+      },
+      {
+        "epoch": 0.7701138429159093,
+        "grad_norm": 0.36188703775405884,
+        "learning_rate": 0.00011831412876558719,
+        "loss": 0.9814,
+        "step": 24150
+      },
+      {
+        "epoch": 0.7704327306355432,
+        "grad_norm": 0.3496958911418915,
+        "learning_rate": 0.00011786227851219062,
+        "loss": 0.9648,
+        "step": 24160
+      },
+      {
+        "epoch": 0.7707516183551771,
+        "grad_norm": 0.3597332835197449,
+        "learning_rate": 0.00011741215390773915,
+        "loss": 0.9686,
+        "step": 24170
+      },
+      {
+        "epoch": 0.7710705060748111,
+        "grad_norm": 0.35886675119400024,
+        "learning_rate": 0.00011696374836185405,
+        "loss": 0.9839,
+        "step": 24180
+      },
+      {
+        "epoch": 0.7713893937944449,
+        "grad_norm": 0.3740817904472351,
+        "learning_rate": 0.00011651705530932584,
+        "loss": 0.9757,
+        "step": 24190
+      },
+      {
+        "epoch": 0.7717082815140789,
+        "grad_norm": 0.35636112093925476,
+        "learning_rate": 0.00011607206821001794,
+        "loss": 0.9643,
+        "step": 24200
+      },
+      {
+        "epoch": 0.7720271692337128,
+        "grad_norm": 0.34713831543922424,
+        "learning_rate": 0.00011562878054877104,
+        "loss": 0.9722,
+        "step": 24210
+      },
+      {
+        "epoch": 0.7723460569533467,
+        "grad_norm": 0.3590894043445587,
+        "learning_rate": 0.00011518718583530775,
+        "loss": 0.9621,
+        "step": 24220
+      },
+      {
+        "epoch": 0.7726649446729806,
+        "grad_norm": 0.3767780065536499,
+        "learning_rate": 0.00011474727760413749,
+        "loss": 0.9457,
+        "step": 24230
+      },
+      {
+        "epoch": 0.7729838323926146,
+        "grad_norm": 0.35947349667549133,
+        "learning_rate": 0.00011430904941446183,
+        "loss": 0.9726,
+        "step": 24240
+      },
+      {
+        "epoch": 0.7733027201122484,
+        "grad_norm": 0.3609798848628998,
+        "learning_rate": 0.00011387249485008026,
+        "loss": 0.9639,
+        "step": 24250
+      },
+      {
+        "epoch": 0.7736216078318824,
+        "grad_norm": 0.3725574016571045,
+        "learning_rate": 0.0001134376075192962,
+        "loss": 0.9655,
+        "step": 24260
+      },
+      {
+        "epoch": 0.7739404955515163,
+        "grad_norm": 0.35596516728401184,
+        "learning_rate": 0.00011300438105482337,
+        "loss": 0.9748,
+        "step": 24270
+      },
+      {
+        "epoch": 0.7742593832711502,
+        "grad_norm": 0.36799517273902893,
+        "learning_rate": 0.00011257280911369273,
+        "loss": 0.983,
+        "step": 24280
+      },
+      {
+        "epoch": 0.7745782709907841,
+        "grad_norm": 0.38049396872520447,
+        "learning_rate": 0.00011214288537715942,
+        "loss": 0.982,
+        "step": 24290
+      },
+      {
+        "epoch": 0.7748971587104181,
+        "grad_norm": 0.35354548692703247,
+        "learning_rate": 0.00011171460355061031,
+        "loss": 0.9748,
+        "step": 24300
+      },
+      {
+        "epoch": 0.7752160464300519,
+        "grad_norm": 0.36302658915519714,
+        "learning_rate": 0.0001112879573634719,
+        "loss": 0.978,
+        "step": 24310
+      },
+      {
+        "epoch": 0.7755349341496859,
+        "grad_norm": 0.37108802795410156,
+        "learning_rate": 0.00011086294056911845,
+        "loss": 0.9826,
+        "step": 24320
+      },
+      {
+        "epoch": 0.7758538218693198,
+        "grad_norm": 0.35210826992988586,
+        "learning_rate": 0.00011043954694478053,
+        "loss": 0.9664,
+        "step": 24330
+      },
+      {
+        "epoch": 0.7761727095889537,
+        "grad_norm": 0.3648816645145416,
+        "learning_rate": 0.00011001777029145393,
+        "loss": 0.9785,
+        "step": 24340
+      },
+      {
+        "epoch": 0.7764915973085876,
+        "grad_norm": 0.354042112827301,
+        "learning_rate": 0.00010959760443380887,
+        "loss": 0.9784,
+        "step": 24350
+      },
+      {
+        "epoch": 0.7768104850282216,
+        "grad_norm": 0.3622848093509674,
+        "learning_rate": 0.00010917904322009965,
+        "loss": 0.9731,
+        "step": 24360
+      },
+      {
+        "epoch": 0.7771293727478554,
+        "grad_norm": 0.3674158751964569,
+        "learning_rate": 0.00010876208052207448,
+        "loss": 0.9596,
+        "step": 24370
+      },
+      {
+        "epoch": 0.7774482604674894,
+        "grad_norm": 0.35103335976600647,
+        "learning_rate": 0.00010834671023488582,
+        "loss": 0.9525,
+        "step": 24380
+      },
+      {
+        "epoch": 0.7777671481871233,
+        "grad_norm": 0.3551388084888458,
+        "learning_rate": 0.000107932926277001,
+        "loss": 0.9719,
+        "step": 24390
+      },
+      {
+        "epoch": 0.7780860359067572,
+        "grad_norm": 0.36495012044906616,
+        "learning_rate": 0.00010752072259011318,
+        "loss": 0.9728,
+        "step": 24400
+      },
+      {
+        "epoch": 0.7784049236263911,
+        "grad_norm": 0.3599463999271393,
+        "learning_rate": 0.00010711009313905259,
+        "loss": 0.9761,
+        "step": 24410
+      },
+      {
+        "epoch": 0.7787238113460251,
+        "grad_norm": 0.36352306604385376,
+        "learning_rate": 0.00010670103191169825,
+        "loss": 0.981,
+        "step": 24420
+      },
+      {
+        "epoch": 0.7790426990656589,
+        "grad_norm": 0.3560295104980469,
+        "learning_rate": 0.00010629353291888986,
+        "loss": 0.9718,
+        "step": 24430
+      },
+      {
+        "epoch": 0.7793615867852929,
+        "grad_norm": 0.3523396849632263,
+        "learning_rate": 0.00010588759019434024,
+        "loss": 0.9728,
+        "step": 24440
+      },
+      {
+        "epoch": 0.7796804745049268,
+        "grad_norm": 0.35102060437202454,
+        "learning_rate": 0.00010548319779454782,
+        "loss": 0.961,
+        "step": 24450
+      },
+      {
+        "epoch": 0.7799993622245607,
+        "grad_norm": 0.3619765639305115,
+        "learning_rate": 0.00010508034979870972,
+        "loss": 0.9715,
+        "step": 24460
+      },
+      {
+        "epoch": 0.7803182499441946,
+        "grad_norm": 0.35871872305870056,
+        "learning_rate": 0.00010467904030863507,
+        "loss": 0.9693,
+        "step": 24470
+      },
+      {
+        "epoch": 0.7806371376638286,
+        "grad_norm": 0.3562319278717041,
+        "learning_rate": 0.00010427926344865853,
+        "loss": 0.9565,
+        "step": 24480
+      },
+      {
+        "epoch": 0.7809560253834625,
+        "grad_norm": 0.35921573638916016,
+        "learning_rate": 0.00010388101336555442,
+        "loss": 0.9703,
+        "step": 24490
+      },
+      {
+        "epoch": 0.7812749131030964,
+        "grad_norm": 0.3715077340602875,
+        "learning_rate": 0.00010348428422845097,
+        "loss": 0.9746,
+        "step": 24500
+      },
+      {
+        "epoch": 0.7815938008227303,
+        "grad_norm": 0.3662187457084656,
+        "learning_rate": 0.00010308907022874489,
+        "loss": 0.9526,
+        "step": 24510
+      },
+      {
+        "epoch": 0.7819126885423643,
+        "grad_norm": 0.3650236427783966,
+        "learning_rate": 0.00010269536558001635,
+        "loss": 0.9763,
+        "step": 24520
+      },
+      {
+        "epoch": 0.7822315762619981,
+        "grad_norm": 0.358048677444458,
+        "learning_rate": 0.00010230316451794439,
+        "loss": 0.9842,
+        "step": 24530
+      },
+      {
+        "epoch": 0.7825504639816321,
+        "grad_norm": 0.3543560802936554,
+        "learning_rate": 0.00010191246130022226,
+        "loss": 0.959,
+        "step": 24540
+      },
+      {
+        "epoch": 0.782869351701266,
+        "grad_norm": 0.36472126841545105,
+        "learning_rate": 0.00010152325020647362,
+        "loss": 0.9696,
+        "step": 24550
+      },
+      {
+        "epoch": 0.7831882394208999,
+        "grad_norm": 0.3696049153804779,
+        "learning_rate": 0.00010113552553816855,
+        "loss": 0.9638,
+        "step": 24560
+      },
+      {
+        "epoch": 0.7835071271405338,
+        "grad_norm": 0.36587175726890564,
+        "learning_rate": 0.00010074928161854036,
+        "loss": 0.9732,
+        "step": 24570
+      },
+      {
+        "epoch": 0.7838260148601678,
+        "grad_norm": 0.3524242043495178,
+        "learning_rate": 0.00010036451279250222,
+        "loss": 0.9643,
+        "step": 24580
+      },
+      {
+        "epoch": 0.7841449025798016,
+        "grad_norm": 0.3479771018028259,
+        "learning_rate": 9.998121342656457e-05,
+        "loss": 0.9707,
+        "step": 24590
+      },
+      {
+        "epoch": 0.7844637902994356,
+        "grad_norm": 0.35497039556503296,
+        "learning_rate": 9.959937790875249e-05,
+        "loss": 0.9698,
+        "step": 24600
+      },
+      {
+        "epoch": 0.7847826780190695,
+        "grad_norm": 0.35893678665161133,
+        "learning_rate": 9.92190006485237e-05,
+        "loss": 0.9627,
+        "step": 24610
+      },
+      {
+        "epoch": 0.7851015657387034,
+        "grad_norm": 0.3627142906188965,
+        "learning_rate": 9.884007607668652e-05,
+        "loss": 0.9529,
+        "step": 24620
+      },
+      {
+        "epoch": 0.7854204534583373,
+        "grad_norm": 0.3647702634334564,
+        "learning_rate": 9.846259864531842e-05,
+        "loss": 0.9648,
+        "step": 24630
+      },
+      {
+        "epoch": 0.7857393411779713,
+        "grad_norm": 0.36695319414138794,
+        "learning_rate": 9.808656282768486e-05,
+        "loss": 0.9627,
+        "step": 24640
+      },
+      {
+        "epoch": 0.7860582288976051,
+        "grad_norm": 0.36442694067955017,
+        "learning_rate": 9.77119631181582e-05,
+        "loss": 0.9624,
+        "step": 24650
+      },
+      {
+        "epoch": 0.7863771166172391,
+        "grad_norm": 0.36140966415405273,
+        "learning_rate": 9.733879403213728e-05,
+        "loss": 0.9696,
+        "step": 24660
+      },
+      {
+        "epoch": 0.786696004336873,
+        "grad_norm": 0.35129281878471375,
+        "learning_rate": 9.696705010596698e-05,
+        "loss": 0.9566,
+        "step": 24670
+      },
+      {
+        "epoch": 0.7870148920565069,
+        "grad_norm": 0.36112743616104126,
+        "learning_rate": 9.65967258968583e-05,
+        "loss": 0.9759,
+        "step": 24680
+      },
+      {
+        "epoch": 0.7873337797761408,
+        "grad_norm": 0.36856353282928467,
+        "learning_rate": 9.622781598280861e-05,
+        "loss": 0.9725,
+        "step": 24690
+      },
+      {
+        "epoch": 0.7876526674957748,
+        "grad_norm": 0.3470911979675293,
+        "learning_rate": 9.586031496252237e-05,
+        "loss": 0.953,
+        "step": 24700
+      },
+      {
+        "epoch": 0.7879715552154086,
+        "grad_norm": 0.3622485101222992,
+        "learning_rate": 9.549421745533192e-05,
+        "loss": 0.9723,
+        "step": 24710
+      },
+      {
+        "epoch": 0.7882904429350426,
+        "grad_norm": 0.36138638854026794,
+        "learning_rate": 9.512951810111879e-05,
+        "loss": 0.9514,
+        "step": 24720
+      },
+      {
+        "epoch": 0.7886093306546765,
+        "grad_norm": 0.36271795630455017,
+        "learning_rate": 9.476621156023519e-05,
+        "loss": 0.9726,
+        "step": 24730
+      },
+      {
+        "epoch": 0.7889282183743104,
+        "grad_norm": 0.3721517324447632,
+        "learning_rate": 9.44042925134258e-05,
+        "loss": 0.9702,
+        "step": 24740
+      },
+      {
+        "epoch": 0.7892471060939443,
+        "grad_norm": 0.3487117290496826,
+        "learning_rate": 9.404375566174994e-05,
+        "loss": 0.9836,
+        "step": 24750
+      },
+      {
+        "epoch": 0.7895659938135783,
+        "grad_norm": 0.3537936210632324,
+        "learning_rate": 9.368459572650401e-05,
+        "loss": 0.9605,
+        "step": 24760
+      },
+      {
+        "epoch": 0.7898848815332121,
+        "grad_norm": 0.359280526638031,
+        "learning_rate": 9.33268074491441e-05,
+        "loss": 0.9518,
+        "step": 24770
+      },
+      {
+        "epoch": 0.7902037692528461,
+        "grad_norm": 0.34653404355049133,
+        "learning_rate": 9.297038559120912e-05,
+        "loss": 0.9743,
+        "step": 24780
+      },
+      {
+        "epoch": 0.79052265697248,
+        "grad_norm": 0.360307902097702,
+        "learning_rate": 9.2615324934244e-05,
+        "loss": 0.9869,
+        "step": 24790
+      },
+      {
+        "epoch": 0.7908415446921139,
+        "grad_norm": 0.3570634126663208,
+        "learning_rate": 9.226162027972337e-05,
+        "loss": 0.9536,
+        "step": 24800
+      },
+      {
+        "epoch": 0.7911604324117478,
+        "grad_norm": 0.3584236204624176,
+        "learning_rate": 9.190926644897531e-05,
+        "loss": 0.9724,
+        "step": 24810
+      },
+      {
+        "epoch": 0.7914793201313818,
+        "grad_norm": 0.36408424377441406,
+        "learning_rate": 9.155825828310578e-05,
+        "loss": 0.9641,
+        "step": 24820
+      },
+      {
+        "epoch": 0.7917982078510156,
+        "grad_norm": 0.3730533719062805,
+        "learning_rate": 9.120859064292278e-05,
+        "loss": 0.9673,
+        "step": 24830
+      },
+      {
+        "epoch": 0.7921170955706496,
+        "grad_norm": 0.35709601640701294,
+        "learning_rate": 9.086025840886135e-05,
+        "loss": 0.9588,
+        "step": 24840
+      },
+      {
+        "epoch": 0.7924359832902835,
+        "grad_norm": 0.36663979291915894,
+        "learning_rate": 9.05132564809085e-05,
+        "loss": 0.9696,
+        "step": 24850
+      },
+      {
+        "epoch": 0.7927548710099174,
+        "grad_norm": 0.3638729453086853,
+        "learning_rate": 9.016757977852851e-05,
+        "loss": 0.9577,
+        "step": 24860
+      },
+      {
+        "epoch": 0.7930737587295513,
+        "grad_norm": 0.35785070061683655,
+        "learning_rate": 8.982322324058868e-05,
+        "loss": 0.9525,
+        "step": 24870
+      },
+      {
+        "epoch": 0.7933926464491853,
+        "grad_norm": 0.36241087317466736,
+        "learning_rate": 8.948018182528511e-05,
+        "loss": 0.9621,
+        "step": 24880
+      },
+      {
+        "epoch": 0.7937115341688191,
+        "grad_norm": 0.35988399386405945,
+        "learning_rate": 8.913845051006889e-05,
+        "loss": 0.9541,
+        "step": 24890
+      },
+      {
+        "epoch": 0.7940304218884531,
+        "grad_norm": 0.3533669710159302,
+        "learning_rate": 8.879802429157259e-05,
+        "loss": 0.9546,
+        "step": 24900
+      },
+      {
+        "epoch": 0.794349309608087,
+        "grad_norm": 0.36953115463256836,
+        "learning_rate": 8.845889818553704e-05,
+        "loss": 0.9684,
+        "step": 24910
+      },
+      {
+        "epoch": 0.7946681973277209,
+        "grad_norm": 0.3601612150669098,
+        "learning_rate": 8.81210672267383e-05,
+        "loss": 0.9636,
+        "step": 24920
+      },
+      {
+        "epoch": 0.7949870850473548,
+        "grad_norm": 0.3613516688346863,
+        "learning_rate": 8.778452646891497e-05,
+        "loss": 0.957,
+        "step": 24930
+      },
+      {
+        "epoch": 0.7953059727669888,
+        "grad_norm": 0.3623802065849304,
+        "learning_rate": 8.744927098469577e-05,
+        "loss": 0.9593,
+        "step": 24940
+      },
+      {
+        "epoch": 0.7956248604866226,
+        "grad_norm": 0.3671809732913971,
+        "learning_rate": 8.711529586552748e-05,
+        "loss": 0.9708,
+        "step": 24950
+      },
+      {
+        "epoch": 0.7959437482062566,
+        "grad_norm": 0.3662140667438507,
+        "learning_rate": 8.678259622160293e-05,
+        "loss": 0.9586,
+        "step": 24960
+      },
+      {
+        "epoch": 0.7962626359258905,
+        "grad_norm": 0.35915324091911316,
+        "learning_rate": 8.645116718178947e-05,
+        "loss": 0.9563,
+        "step": 24970
+      },
+      {
+        "epoch": 0.7965815236455244,
+        "grad_norm": 0.3539397716522217,
+        "learning_rate": 8.612100389355772e-05,
+        "loss": 0.9631,
+        "step": 24980
+      },
+      {
+        "epoch": 0.7969004113651583,
+        "grad_norm": 0.36351755261421204,
+        "learning_rate": 8.579210152291042e-05,
+        "loss": 0.9371,
+        "step": 24990
+      },
+      {
+        "epoch": 0.7972192990847923,
+        "grad_norm": 0.36191827058792114,
+        "learning_rate": 8.546445525431173e-05,
+        "loss": 0.9566,
+        "step": 25000
+      },
+      {
+        "epoch": 0.7975381868044261,
+        "grad_norm": 0.37281543016433716,
+        "learning_rate": 8.513806029061663e-05,
+        "loss": 0.975,
+        "step": 25010
+      },
+      {
+        "epoch": 0.7978570745240601,
+        "grad_norm": 0.3713143467903137,
+        "learning_rate": 8.481291185300078e-05,
+        "loss": 0.956,
+        "step": 25020
+      },
+      {
+        "epoch": 0.798175962243694,
+        "grad_norm": 0.37444189190864563,
+        "learning_rate": 8.448900518089058e-05,
+        "loss": 0.9518,
+        "step": 25030
+      },
+      {
+        "epoch": 0.7984948499633279,
+        "grad_norm": 0.3919963538646698,
+        "learning_rate": 8.416633553189332e-05,
+        "loss": 0.9714,
+        "step": 25040
+      },
+      {
+        "epoch": 0.7988137376829618,
+        "grad_norm": 0.3554207384586334,
+        "learning_rate": 8.384489818172784e-05,
+        "loss": 0.9544,
+        "step": 25050
+      },
+      {
+        "epoch": 0.7991326254025958,
+        "grad_norm": 0.3600543439388275,
+        "learning_rate": 8.35246884241554e-05,
+        "loss": 0.9526,
+        "step": 25060
+      },
+      {
+        "epoch": 0.7994515131222296,
+        "grad_norm": 0.36619463562965393,
+        "learning_rate": 8.32057015709107e-05,
+        "loss": 0.9734,
+        "step": 25070
+      },
+      {
+        "epoch": 0.7997704008418636,
+        "grad_norm": 0.3708662986755371,
+        "learning_rate": 8.288793295163325e-05,
+        "loss": 0.9637,
+        "step": 25080
+      },
+      {
+        "epoch": 0.8000892885614975,
+        "grad_norm": 0.35670143365859985,
+        "learning_rate": 8.257137791379903e-05,
+        "loss": 0.9712,
+        "step": 25090
+      },
+      {
+        "epoch": 0.8004081762811314,
+        "grad_norm": 0.3521883189678192,
+        "learning_rate": 8.225603182265234e-05,
+        "loss": 0.9477,
+        "step": 25100
+      },
+      {
+        "epoch": 0.8007270640007653,
+        "grad_norm": 0.352546364068985,
+        "learning_rate": 8.194189006113793e-05,
+        "loss": 0.9609,
+        "step": 25110
+      },
+      {
+        "epoch": 0.8010459517203993,
+        "grad_norm": 0.36682891845703125,
+        "learning_rate": 8.162894802983348e-05,
+        "loss": 0.961,
+        "step": 25120
+      },
+      {
+        "epoch": 0.8013648394400331,
+        "grad_norm": 0.37029045820236206,
+        "learning_rate": 8.131720114688214e-05,
+        "loss": 0.9521,
+        "step": 25130
+      },
+      {
+        "epoch": 0.8016837271596671,
+        "grad_norm": 0.36519482731819153,
+        "learning_rate": 8.100664484792551e-05,
+        "loss": 0.9574,
+        "step": 25140
+      },
+      {
+        "epoch": 0.802002614879301,
+        "grad_norm": 0.36591729521751404,
+        "learning_rate": 8.069727458603683e-05,
+        "loss": 0.9464,
+        "step": 25150
+      },
+      {
+        "epoch": 0.8023215025989349,
+        "grad_norm": 0.3642500340938568,
+        "learning_rate": 8.038908583165436e-05,
+        "loss": 0.9746,
+        "step": 25160
+      },
+      {
+        "epoch": 0.8026403903185688,
+        "grad_norm": 0.36521655321121216,
+        "learning_rate": 8.008207407251511e-05,
+        "loss": 0.9394,
+        "step": 25170
+      },
+      {
+        "epoch": 0.8029592780382028,
+        "grad_norm": 0.3553067445755005,
+        "learning_rate": 7.977623481358872e-05,
+        "loss": 0.9494,
+        "step": 25180
+      },
+      {
+        "epoch": 0.8032781657578366,
+        "grad_norm": 0.3684810400009155,
+        "learning_rate": 7.947156357701168e-05,
+        "loss": 0.9455,
+        "step": 25190
+      },
+      {
+        "epoch": 0.8035970534774706,
+        "grad_norm": 0.3643917739391327,
+        "learning_rate": 7.916805590202182e-05,
+        "loss": 0.9567,
+        "step": 25200
+      },
+      {
+        "epoch": 0.8039159411971045,
+        "grad_norm": 0.3606899380683899,
+        "learning_rate": 7.886570734489289e-05,
+        "loss": 0.9681,
+        "step": 25210
+      },
+      {
+        "epoch": 0.8042348289167384,
+        "grad_norm": 0.36119991540908813,
+        "learning_rate": 7.856451347886955e-05,
+        "loss": 0.9555,
+        "step": 25220
+      },
+      {
+        "epoch": 0.8045537166363723,
+        "grad_norm": 0.35547545552253723,
+        "learning_rate": 7.82644698941026e-05,
+        "loss": 0.944,
+        "step": 25230
+      },
+      {
+        "epoch": 0.8048726043560063,
+        "grad_norm": 0.3734472990036011,
+        "learning_rate": 7.796557219758433e-05,
+        "loss": 0.97,
+        "step": 25240
+      },
+      {
+        "epoch": 0.8051914920756401,
+        "grad_norm": 0.36550241708755493,
+        "learning_rate": 7.766781601308432e-05,
+        "loss": 0.9647,
+        "step": 25250
+      },
+      {
+        "epoch": 0.8055103797952741,
+        "grad_norm": 0.3795914947986603,
+        "learning_rate": 7.737119698108516e-05,
+        "loss": 0.9377,
+        "step": 25260
+      },
+      {
+        "epoch": 0.805829267514908,
+        "grad_norm": 0.34035786986351013,
+        "learning_rate": 7.70757107587189e-05,
+        "loss": 0.9327,
+        "step": 25270
+      },
+      {
+        "epoch": 0.8061481552345419,
+        "grad_norm": 0.3551400601863861,
+        "learning_rate": 7.678135301970321e-05,
+        "loss": 0.9606,
+        "step": 25280
+      },
+      {
+        "epoch": 0.8064670429541758,
+        "grad_norm": 0.35824182629585266,
+        "learning_rate": 7.648811945427822e-05,
+        "loss": 0.9599,
+        "step": 25290
+      },
+      {
+        "epoch": 0.8067859306738098,
+        "grad_norm": 0.3525572419166565,
+        "learning_rate": 7.61960057691433e-05,
+        "loss": 0.9565,
+        "step": 25300
+      },
+      {
+        "epoch": 0.8071048183934437,
+        "grad_norm": 0.36219167709350586,
+        "learning_rate": 7.590500768739427e-05,
+        "loss": 0.9413,
+        "step": 25310
+      },
+      {
+        "epoch": 0.8074237061130776,
+        "grad_norm": 0.36099758744239807,
+        "learning_rate": 7.561512094846075e-05,
+        "loss": 0.9746,
+        "step": 25320
+      },
+      {
+        "epoch": 0.8077425938327115,
+        "grad_norm": 0.35688111186027527,
+        "learning_rate": 7.532634130804376e-05,
+        "loss": 0.958,
+        "step": 25330
+      },
+      {
+        "epoch": 0.8080614815523455,
+        "grad_norm": 0.36312273144721985,
+        "learning_rate": 7.503866453805368e-05,
+        "loss": 0.9557,
+        "step": 25340
+      },
+      {
+        "epoch": 0.8083803692719793,
+        "grad_norm": 0.3614059090614319,
+        "learning_rate": 7.475208642654819e-05,
+        "loss": 0.9591,
+        "step": 25350
+      },
+      {
+        "epoch": 0.8086992569916133,
+        "grad_norm": 0.3655224144458771,
+        "learning_rate": 7.44666027776707e-05,
+        "loss": 0.9518,
+        "step": 25360
+      },
+      {
+        "epoch": 0.8090181447112472,
+        "grad_norm": 0.3513615131378174,
+        "learning_rate": 7.418220941158897e-05,
+        "loss": 0.9465,
+        "step": 25370
+      },
+      {
+        "epoch": 0.8093370324308811,
+        "grad_norm": 0.3549514710903168,
+        "learning_rate": 7.389890216443378e-05,
+        "loss": 0.9628,
+        "step": 25380
+      },
+      {
+        "epoch": 0.809655920150515,
+        "grad_norm": 0.3620496690273285,
+        "learning_rate": 7.361667688823803e-05,
+        "loss": 0.952,
+        "step": 25390
+      },
+      {
+        "epoch": 0.809974807870149,
+        "grad_norm": 0.356606662273407,
+        "learning_rate": 7.333552945087602e-05,
+        "loss": 0.9508,
+        "step": 25400
+      },
+      {
+        "epoch": 0.8102936955897828,
+        "grad_norm": 0.3640885055065155,
+        "learning_rate": 7.305545573600293e-05,
+        "loss": 0.9496,
+        "step": 25410
+      },
+      {
+        "epoch": 0.8106125833094168,
+        "grad_norm": 0.3723457455635071,
+        "learning_rate": 7.277645164299459e-05,
+        "loss": 0.9541,
+        "step": 25420
+      },
+      {
+        "epoch": 0.8109314710290507,
+        "grad_norm": 0.3665878474712372,
+        "learning_rate": 7.249851308688739e-05,
+        "loss": 0.9561,
+        "step": 25430
+      },
+      {
+        "epoch": 0.8112503587486846,
+        "grad_norm": 0.3625575602054596,
+        "learning_rate": 7.222163599831843e-05,
+        "loss": 0.9649,
+        "step": 25440
+      },
+      {
+        "epoch": 0.8115692464683185,
+        "grad_norm": 0.3609040677547455,
+        "learning_rate": 7.194581632346617e-05,
+        "loss": 0.9548,
+        "step": 25450
+      },
+      {
+        "epoch": 0.8118881341879525,
+        "grad_norm": 0.34985458850860596,
+        "learning_rate": 7.167105002399073e-05,
+        "loss": 0.9395,
+        "step": 25460
+      },
+      {
+        "epoch": 0.8122070219075863,
+        "grad_norm": 0.3752239942550659,
+        "learning_rate": 7.139733307697503e-05,
+        "loss": 0.9666,
+        "step": 25470
+      },
+      {
+        "epoch": 0.8125259096272203,
+        "grad_norm": 0.34349602460861206,
+        "learning_rate": 7.112466147486579e-05,
+        "loss": 0.9695,
+        "step": 25480
+      },
+      {
+        "epoch": 0.8128447973468542,
+        "grad_norm": 0.37290871143341064,
+        "learning_rate": 7.08530312254149e-05,
+        "loss": 0.9622,
+        "step": 25490
+      },
+      {
+        "epoch": 0.8131636850664881,
+        "grad_norm": 0.3771199584007263,
+        "learning_rate": 7.058243835162084e-05,
+        "loss": 0.9633,
+        "step": 25500
+      },
+      {
+        "epoch": 0.813482572786122,
+        "grad_norm": 0.3719727396965027,
+        "learning_rate": 7.031287889167066e-05,
+        "loss": 0.9571,
+        "step": 25510
+      },
+      {
+        "epoch": 0.813801460505756,
+        "grad_norm": 0.37617650628089905,
+        "learning_rate": 7.004434889888181e-05,
+        "loss": 0.9532,
+        "step": 25520
+      },
+      {
+        "epoch": 0.8141203482253898,
+        "grad_norm": 0.3691602349281311,
+        "learning_rate": 6.977684444164437e-05,
+        "loss": 0.9604,
+        "step": 25530
+      },
+      {
+        "epoch": 0.8144392359450238,
+        "grad_norm": 0.35761821269989014,
+        "learning_rate": 6.951036160336359e-05,
+        "loss": 0.961,
+        "step": 25540
+      },
+      {
+        "epoch": 0.8147581236646577,
+        "grad_norm": 0.3590810000896454,
+        "learning_rate": 6.924489648240246e-05,
+        "loss": 0.9575,
+        "step": 25550
+      },
+      {
+        "epoch": 0.8150770113842916,
+        "grad_norm": 0.3555402159690857,
+        "learning_rate": 6.89804451920246e-05,
+        "loss": 0.9462,
+        "step": 25560
+      },
+      {
+        "epoch": 0.8153958991039255,
+        "grad_norm": 0.35681602358818054,
+        "learning_rate": 6.871700386033734e-05,
+        "loss": 0.9583,
+        "step": 25570
+      },
+      {
+        "epoch": 0.8157147868235595,
+        "grad_norm": 0.3650819957256317,
+        "learning_rate": 6.845456863023507e-05,
+        "loss": 0.9551,
+        "step": 25580
+      },
+      {
+        "epoch": 0.8160336745431933,
+        "grad_norm": 0.36943039298057556,
+        "learning_rate": 6.819313565934276e-05,
+        "loss": 0.9673,
+        "step": 25590
+      },
+      {
+        "epoch": 0.8163525622628273,
+        "grad_norm": 0.3585793673992157,
+        "learning_rate": 6.793270111995963e-05,
+        "loss": 0.9552,
+        "step": 25600
+      },
+      {
+        "epoch": 0.8166714499824612,
+        "grad_norm": 0.3606432378292084,
+        "learning_rate": 6.767326119900324e-05,
+        "loss": 0.9538,
+        "step": 25610
+      },
+      {
+        "epoch": 0.8169903377020951,
+        "grad_norm": 0.36778682470321655,
+        "learning_rate": 6.741481209795356e-05,
+        "loss": 0.9552,
+        "step": 25620
+      },
+      {
+        "epoch": 0.817309225421729,
+        "grad_norm": 0.362591415643692,
+        "learning_rate": 6.715735003279736e-05,
+        "loss": 0.9656,
+        "step": 25630
+      },
+      {
+        "epoch": 0.817628113141363,
+        "grad_norm": 0.3498934805393219,
+        "learning_rate": 6.69008712339729e-05,
+        "loss": 0.9598,
+        "step": 25640
+      },
+      {
+        "epoch": 0.8179470008609968,
+        "grad_norm": 0.361508846282959,
+        "learning_rate": 6.664537194631458e-05,
+        "loss": 0.9403,
+        "step": 25650
+      },
+      {
+        "epoch": 0.8182658885806308,
+        "grad_norm": 0.36598604917526245,
+        "learning_rate": 6.639084842899812e-05,
+        "loss": 0.9568,
+        "step": 25660
+      },
+      {
+        "epoch": 0.8185847763002647,
+        "grad_norm": 0.3673774302005768,
+        "learning_rate": 6.613729695548567e-05,
+        "loss": 0.9531,
+        "step": 25670
+      },
+      {
+        "epoch": 0.8189036640198986,
+        "grad_norm": 0.3787930905818939,
+        "learning_rate": 6.588471381347133e-05,
+        "loss": 0.9603,
+        "step": 25680
+      },
+      {
+        "epoch": 0.8192225517395325,
+        "grad_norm": 0.36597883701324463,
+        "learning_rate": 6.563309530482677e-05,
+        "loss": 0.9596,
+        "step": 25690
+      },
+      {
+        "epoch": 0.8195414394591665,
+        "grad_norm": 0.3917224705219269,
+        "learning_rate": 6.538243774554706e-05,
+        "loss": 0.9636,
+        "step": 25700
+      },
+      {
+        "epoch": 0.8198603271788003,
+        "grad_norm": 0.3685980439186096,
+        "learning_rate": 6.513273746569676e-05,
+        "loss": 0.9542,
+        "step": 25710
+      },
+      {
+        "epoch": 0.8201792148984343,
+        "grad_norm": 0.35182175040245056,
+        "learning_rate": 6.48839908093562e-05,
+        "loss": 0.9366,
+        "step": 25720
+      },
+      {
+        "epoch": 0.8204981026180682,
+        "grad_norm": 0.36965203285217285,
+        "learning_rate": 6.463619413456787e-05,
+        "loss": 0.9558,
+        "step": 25730
+      },
+      {
+        "epoch": 0.8208169903377021,
+        "grad_norm": 0.36861446499824524,
+        "learning_rate": 6.438934381328326e-05,
+        "loss": 0.9519,
+        "step": 25740
+      },
+      {
+        "epoch": 0.821135878057336,
+        "grad_norm": 0.3652392625808716,
+        "learning_rate": 6.414343623130956e-05,
+        "loss": 0.9546,
+        "step": 25750
+      },
+      {
+        "epoch": 0.82145476577697,
+        "grad_norm": 0.3519989550113678,
+        "learning_rate": 6.389846778825685e-05,
+        "loss": 0.928,
+        "step": 25760
+      },
+      {
+        "epoch": 0.8217736534966038,
+        "grad_norm": 0.36737608909606934,
+        "learning_rate": 6.365443489748536e-05,
+        "loss": 0.9531,
+        "step": 25770
+      },
+      {
+        "epoch": 0.8220925412162378,
+        "grad_norm": 0.36519476771354675,
+        "learning_rate": 6.341133398605295e-05,
+        "loss": 0.9467,
+        "step": 25780
+      },
+      {
+        "epoch": 0.8224114289358717,
+        "grad_norm": 0.3782409727573395,
+        "learning_rate": 6.316916149466283e-05,
+        "loss": 0.9584,
+        "step": 25790
+      },
+      {
+        "epoch": 0.8227303166555056,
+        "grad_norm": 0.36387908458709717,
+        "learning_rate": 6.292791387761143e-05,
+        "loss": 0.9313,
+        "step": 25800
+      },
+      {
+        "epoch": 0.8230492043751395,
+        "grad_norm": 0.3558565080165863,
+        "learning_rate": 6.268758760273642e-05,
+        "loss": 0.9525,
+        "step": 25810
+      },
+      {
+        "epoch": 0.8233680920947735,
+        "grad_norm": 0.3587815463542938,
+        "learning_rate": 6.24481791513651e-05,
+        "loss": 0.9593,
+        "step": 25820
+      },
+      {
+        "epoch": 0.8236869798144073,
+        "grad_norm": 0.37237703800201416,
+        "learning_rate": 6.220968501826283e-05,
+        "loss": 0.9481,
+        "step": 25830
+      },
+      {
+        "epoch": 0.8240058675340413,
+        "grad_norm": 0.35571229457855225,
+        "learning_rate": 6.197210171158173e-05,
+        "loss": 0.9452,
+        "step": 25840
+      },
+      {
+        "epoch": 0.8243247552536752,
+        "grad_norm": 0.37391969561576843,
+        "learning_rate": 6.173542575280949e-05,
+        "loss": 0.9587,
+        "step": 25850
+      },
+      {
+        "epoch": 0.824643642973309,
+        "grad_norm": 0.37451764941215515,
+        "learning_rate": 6.149965367671856e-05,
+        "loss": 0.9464,
+        "step": 25860
+      },
+      {
+        "epoch": 0.824962530692943,
+        "grad_norm": 0.3647792339324951,
+        "learning_rate": 6.126478203131529e-05,
+        "loss": 0.9339,
+        "step": 25870
+      },
+      {
+        "epoch": 0.825281418412577,
+        "grad_norm": 0.36442533135414124,
+        "learning_rate": 6.1030807377789486e-05,
+        "loss": 0.9413,
+        "step": 25880
+      },
+      {
+        "epoch": 0.8256003061322108,
+        "grad_norm": 0.36027491092681885,
+        "learning_rate": 6.0797726290463996e-05,
+        "loss": 0.9562,
+        "step": 25890
+      },
+      {
+        "epoch": 0.8259191938518448,
+        "grad_norm": 0.3716151714324951,
+        "learning_rate": 6.056553535674458e-05,
+        "loss": 0.9368,
+        "step": 25900
+      },
+      {
+        "epoch": 0.8262380815714787,
+        "grad_norm": 0.382212370634079,
+        "learning_rate": 6.033423117706994e-05,
+        "loss": 0.9551,
+        "step": 25910
+      },
+      {
+        "epoch": 0.8265569692911126,
+        "grad_norm": 0.3520452082157135,
+        "learning_rate": 6.0103810364861955e-05,
+        "loss": 0.9539,
+        "step": 25920
+      },
+      {
+        "epoch": 0.8268758570107465,
+        "grad_norm": 0.36965930461883545,
+        "learning_rate": 5.9874269546476105e-05,
+        "loss": 0.9474,
+        "step": 25930
+      },
+      {
+        "epoch": 0.8271947447303805,
+        "grad_norm": 0.35486385226249695,
+        "learning_rate": 5.964560536115204e-05,
+        "loss": 0.9547,
+        "step": 25940
+      },
+      {
+        "epoch": 0.8275136324500143,
+        "grad_norm": 0.3544781506061554,
+        "learning_rate": 5.941781446096441e-05,
+        "loss": 0.9271,
+        "step": 25950
+      },
+      {
+        "epoch": 0.8278325201696483,
+        "grad_norm": 0.3677733540534973,
+        "learning_rate": 5.9190893510773834e-05,
+        "loss": 0.955,
+        "step": 25960
+      },
+      {
+        "epoch": 0.8281514078892822,
+        "grad_norm": 0.3709496855735779,
+        "learning_rate": 5.896483918817807e-05,
+        "loss": 0.9493,
+        "step": 25970
+      },
+      {
+        "epoch": 0.828470295608916,
+        "grad_norm": 0.36802056431770325,
+        "learning_rate": 5.873964818346338e-05,
+        "loss": 0.9589,
+        "step": 25980
+      },
+      {
+        "epoch": 0.82878918332855,
+        "grad_norm": 0.3739546239376068,
+        "learning_rate": 5.8515317199556014e-05,
+        "loss": 0.9445,
+        "step": 25990
+      },
+      {
+        "epoch": 0.829108071048184,
+        "grad_norm": 0.3755469024181366,
+        "learning_rate": 5.829184295197409e-05,
+        "loss": 0.9434,
+        "step": 26000
+      },
+      {
+        "epoch": 0.8294269587678178,
+        "grad_norm": 0.3417413532733917,
+        "learning_rate": 5.806922216877932e-05,
+        "loss": 0.9374,
+        "step": 26010
+      },
+      {
+        "epoch": 0.8297458464874518,
+        "grad_norm": 0.36555731296539307,
+        "learning_rate": 5.784745159052919e-05,
+        "loss": 0.9373,
+        "step": 26020
+      },
+      {
+        "epoch": 0.8300647342070857,
+        "grad_norm": 0.3578369617462158,
+        "learning_rate": 5.7626527970229256e-05,
+        "loss": 0.9585,
+        "step": 26030
+      },
+      {
+        "epoch": 0.8303836219267196,
+        "grad_norm": 0.37149980664253235,
+        "learning_rate": 5.7406448073285566e-05,
+        "loss": 0.9487,
+        "step": 26040
+      },
+      {
+        "epoch": 0.8307025096463535,
+        "grad_norm": 0.3658088147640228,
+        "learning_rate": 5.718720867745734e-05,
+        "loss": 0.9591,
+        "step": 26050
+      },
+      {
+        "epoch": 0.8310213973659875,
+        "grad_norm": 0.3635800778865814,
+        "learning_rate": 5.6968806572809736e-05,
+        "loss": 0.946,
+        "step": 26060
+      },
+      {
+        "epoch": 0.8313402850856213,
+        "grad_norm": 0.3664982318878174,
+        "learning_rate": 5.675123856166692e-05,
+        "loss": 0.9351,
+        "step": 26070
+      },
+      {
+        "epoch": 0.8316591728052553,
+        "grad_norm": 0.36701416969299316,
+        "learning_rate": 5.653450145856519e-05,
+        "loss": 0.946,
+        "step": 26080
+      },
+      {
+        "epoch": 0.8319780605248892,
+        "grad_norm": 0.3673885762691498,
+        "learning_rate": 5.631859209020637e-05,
+        "loss": 0.9445,
+        "step": 26090
+      },
+      {
+        "epoch": 0.832296948244523,
+        "grad_norm": 0.3711278438568115,
+        "learning_rate": 5.6103507295411355e-05,
+        "loss": 0.9504,
+        "step": 26100
+      },
+      {
+        "epoch": 0.832615835964157,
+        "grad_norm": 0.36567771434783936,
+        "learning_rate": 5.5889243925073783e-05,
+        "loss": 0.9828,
+        "step": 26110
+      },
+      {
+        "epoch": 0.832934723683791,
+        "grad_norm": 0.3659859001636505,
+        "learning_rate": 5.5675798842113984e-05,
+        "loss": 0.9513,
+        "step": 26120
+      },
+      {
+        "epoch": 0.8332536114034248,
+        "grad_norm": 0.36194849014282227,
+        "learning_rate": 5.546316892143301e-05,
+        "loss": 0.9443,
+        "step": 26130
+      },
+      {
+        "epoch": 0.8335724991230588,
+        "grad_norm": 0.36451077461242676,
+        "learning_rate": 5.525135104986689e-05,
+        "loss": 0.938,
+        "step": 26140
+      },
+      {
+        "epoch": 0.8338913868426927,
+        "grad_norm": 0.3581928610801697,
+        "learning_rate": 5.5040342126141065e-05,
+        "loss": 0.9473,
+        "step": 26150
+      },
+      {
+        "epoch": 0.8342102745623267,
+        "grad_norm": 0.3767908811569214,
+        "learning_rate": 5.483013906082494e-05,
+        "loss": 0.9584,
+        "step": 26160
+      },
+      {
+        "epoch": 0.8345291622819605,
+        "grad_norm": 0.3569718301296234,
+        "learning_rate": 5.462073877628674e-05,
+        "loss": 0.9194,
+        "step": 26170
+      },
+      {
+        "epoch": 0.8348480500015945,
+        "grad_norm": 0.37712088227272034,
+        "learning_rate": 5.4412138206648335e-05,
+        "loss": 0.9696,
+        "step": 26180
+      },
+      {
+        "epoch": 0.8351669377212284,
+        "grad_norm": 0.3611338436603546,
+        "learning_rate": 5.420433429774042e-05,
+        "loss": 0.9433,
+        "step": 26190
+      },
+      {
+        "epoch": 0.8354858254408623,
+        "grad_norm": 0.365477055311203,
+        "learning_rate": 5.3997324007057795e-05,
+        "loss": 0.9307,
+        "step": 26200
+      },
+      {
+        "epoch": 0.8358047131604962,
+        "grad_norm": 0.3720219135284424,
+        "learning_rate": 5.379110430371478e-05,
+        "loss": 0.9351,
+        "step": 26210
+      },
+      {
+        "epoch": 0.8361236008801302,
+        "grad_norm": 0.35642099380493164,
+        "learning_rate": 5.358567216840091e-05,
+        "loss": 0.9441,
+        "step": 26220
+      },
+      {
+        "epoch": 0.836442488599764,
+        "grad_norm": 0.3763499855995178,
+        "learning_rate": 5.338102459333666e-05,
+        "loss": 0.9426,
+        "step": 26230
+      },
+      {
+        "epoch": 0.836761376319398,
+        "grad_norm": 0.3568282127380371,
+        "learning_rate": 5.317715858222943e-05,
+        "loss": 0.9431,
+        "step": 26240
+      },
+      {
+        "epoch": 0.8370802640390319,
+        "grad_norm": 0.3555208444595337,
+        "learning_rate": 5.297407115022969e-05,
+        "loss": 0.9416,
+        "step": 26250
+      },
+      {
+        "epoch": 0.8373991517586657,
+        "grad_norm": 0.3644401729106903,
+        "learning_rate": 5.277175932388726e-05,
+        "loss": 0.919,
+        "step": 26260
+      },
+      {
+        "epoch": 0.8377180394782997,
+        "grad_norm": 0.35538750886917114,
+        "learning_rate": 5.2570220141107796e-05,
+        "loss": 0.9656,
+        "step": 26270
+      },
+      {
+        "epoch": 0.8380369271979337,
+        "grad_norm": 0.36755502223968506,
+        "learning_rate": 5.236945065110938e-05,
+        "loss": 0.9645,
+        "step": 26280
+      },
+      {
+        "epoch": 0.8383558149175675,
+        "grad_norm": 0.35435423254966736,
+        "learning_rate": 5.216944791437939e-05,
+        "loss": 0.9412,
+        "step": 26290
+      },
+      {
+        "epoch": 0.8386747026372015,
+        "grad_norm": 0.3524496555328369,
+        "learning_rate": 5.1970209002631356e-05,
+        "loss": 0.9513,
+        "step": 26300
+      },
+      {
+        "epoch": 0.8389935903568354,
+        "grad_norm": 0.36997079849243164,
+        "learning_rate": 5.177173099876221e-05,
+        "loss": 0.9451,
+        "step": 26310
+      },
+      {
+        "epoch": 0.8393124780764692,
+        "grad_norm": 0.3610512912273407,
+        "learning_rate": 5.157401099680946e-05,
+        "loss": 0.94,
+        "step": 26320
+      },
+      {
+        "epoch": 0.8396313657961032,
+        "grad_norm": 0.36900800466537476,
+        "learning_rate": 5.13770461019087e-05,
+        "loss": 0.9453,
+        "step": 26330
+      },
+      {
+        "epoch": 0.8399502535157372,
+        "grad_norm": 0.3552963435649872,
+        "learning_rate": 5.118083343025129e-05,
+        "loss": 0.9421,
+        "step": 26340
+      },
+      {
+        "epoch": 0.840269141235371,
+        "grad_norm": 0.3669191002845764,
+        "learning_rate": 5.0985370109041986e-05,
+        "loss": 0.9286,
+        "step": 26350
+      },
+      {
+        "epoch": 0.840588028955005,
+        "grad_norm": 0.3605242371559143,
+        "learning_rate": 5.079065327645699e-05,
+        "loss": 0.9564,
+        "step": 26360
+      },
+      {
+        "epoch": 0.8409069166746389,
+        "grad_norm": 0.3721179962158203,
+        "learning_rate": 5.059668008160201e-05,
+        "loss": 0.951,
+        "step": 26370
+      },
+      {
+        "epoch": 0.8412258043942727,
+        "grad_norm": 0.37630388140678406,
+        "learning_rate": 5.0403447684470534e-05,
+        "loss": 0.9437,
+        "step": 26380
+      },
+      {
+        "epoch": 0.8415446921139067,
+        "grad_norm": 0.3522168695926666,
+        "learning_rate": 5.021095325590223e-05,
+        "loss": 0.9353,
+        "step": 26390
+      },
+      {
+        "epoch": 0.8418635798335407,
+        "grad_norm": 0.3843633532524109,
+        "learning_rate": 5.001919397754154e-05,
+        "loss": 0.961,
+        "step": 26400
+      },
+      {
+        "epoch": 0.8421824675531745,
+        "grad_norm": 0.35673603415489197,
+        "learning_rate": 4.9828167041796415e-05,
+        "loss": 0.9534,
+        "step": 26410
+      },
+      {
+        "epoch": 0.8425013552728084,
+        "grad_norm": 0.3700405955314636,
+        "learning_rate": 4.9637869651797196e-05,
+        "loss": 0.9264,
+        "step": 26420
+      },
+      {
+        "epoch": 0.8428202429924424,
+        "grad_norm": 0.371220201253891,
+        "learning_rate": 4.944829902135568e-05,
+        "loss": 0.947,
+        "step": 26430
+      },
+      {
+        "epoch": 0.8431391307120762,
+        "grad_norm": 0.35385560989379883,
+        "learning_rate": 4.925945237492433e-05,
+        "loss": 0.949,
+        "step": 26440
+      },
+      {
+        "epoch": 0.8434580184317102,
+        "grad_norm": 0.36943143606185913,
+        "learning_rate": 4.9071326947555595e-05,
+        "loss": 0.9618,
+        "step": 26450
+      },
+      {
+        "epoch": 0.8437769061513442,
+        "grad_norm": 0.3573361933231354,
+        "learning_rate": 4.8883919984861484e-05,
+        "loss": 0.9538,
+        "step": 26460
+      },
+      {
+        "epoch": 0.844095793870978,
+        "grad_norm": 0.37482595443725586,
+        "learning_rate": 4.86972287429732e-05,
+        "loss": 0.933,
+        "step": 26470
+      },
+      {
+        "epoch": 0.844414681590612,
+        "grad_norm": 0.3452557623386383,
+        "learning_rate": 4.8511250488501005e-05,
+        "loss": 0.9393,
+        "step": 26480
+      },
+      {
+        "epoch": 0.8447335693102459,
+        "grad_norm": 0.35981285572052,
+        "learning_rate": 4.832598249849414e-05,
+        "loss": 0.9365,
+        "step": 26490
+      },
+      {
+        "epoch": 0.8450524570298797,
+        "grad_norm": 0.34943288564682007,
+        "learning_rate": 4.814142206040099e-05,
+        "loss": 0.9419,
+        "step": 26500
+      },
+      {
+        "epoch": 0.8453713447495137,
+        "grad_norm": 0.35167521238327026,
+        "learning_rate": 4.795756647202945e-05,
+        "loss": 0.9393,
+        "step": 26510
+      },
+      {
+        "epoch": 0.8456902324691477,
+        "grad_norm": 0.35662704706192017,
+        "learning_rate": 4.777441304150719e-05,
+        "loss": 0.9325,
+        "step": 26520
+      },
+      {
+        "epoch": 0.8460091201887815,
+        "grad_norm": 0.3618640601634979,
+        "learning_rate": 4.759195908724238e-05,
+        "loss": 0.9379,
+        "step": 26530
+      },
+      {
+        "epoch": 0.8463280079084154,
+        "grad_norm": 0.3570464551448822,
+        "learning_rate": 4.741020193788437e-05,
+        "loss": 0.9363,
+        "step": 26540
+      },
+      {
+        "epoch": 0.8466468956280494,
+        "grad_norm": 0.37565356492996216,
+        "learning_rate": 4.72291389322846e-05,
+        "loss": 0.9437,
+        "step": 26550
+      },
+      {
+        "epoch": 0.8469657833476832,
+        "grad_norm": 0.36229774355888367,
+        "learning_rate": 4.7048767419457626e-05,
+        "loss": 0.9409,
+        "step": 26560
+      },
+      {
+        "epoch": 0.8472846710673172,
+        "grad_norm": 0.3552214205265045,
+        "learning_rate": 4.686908475854231e-05,
+        "loss": 0.9561,
+        "step": 26570
+      },
+      {
+        "epoch": 0.8476035587869512,
+        "grad_norm": 0.36420536041259766,
+        "learning_rate": 4.669008831876315e-05,
+        "loss": 0.9338,
+        "step": 26580
+      },
+      {
+        "epoch": 0.847922446506585,
+        "grad_norm": 0.3625452518463135,
+        "learning_rate": 4.651177547939179e-05,
+        "loss": 0.9357,
+        "step": 26590
+      },
+      {
+        "epoch": 0.848241334226219,
+        "grad_norm": 0.3741529881954193,
+        "learning_rate": 4.633414362970859e-05,
+        "loss": 0.9488,
+        "step": 26600
+      },
+      {
+        "epoch": 0.8485602219458529,
+        "grad_norm": 0.3811874985694885,
+        "learning_rate": 4.6157190168964464e-05,
+        "loss": 0.9377,
+        "step": 26610
+      },
+      {
+        "epoch": 0.8488791096654867,
+        "grad_norm": 0.3749126195907593,
+        "learning_rate": 4.598091250634277e-05,
+        "loss": 0.945,
+        "step": 26620
+      },
+      {
+        "epoch": 0.8491979973851207,
+        "grad_norm": 0.3762841820716858,
+        "learning_rate": 4.580530806092137e-05,
+        "loss": 0.946,
+        "step": 26630
+      },
+      {
+        "epoch": 0.8495168851047546,
+        "grad_norm": 0.3608279824256897,
+        "learning_rate": 4.563037426163488e-05,
+        "loss": 0.9599,
+        "step": 26640
+      },
+      {
+        "epoch": 0.8498357728243885,
+        "grad_norm": 0.35772505402565,
+        "learning_rate": 4.545610854723698e-05,
+        "loss": 0.9429,
+        "step": 26650
+      },
+      {
+        "epoch": 0.8501546605440224,
+        "grad_norm": 0.3615608811378479,
+        "learning_rate": 4.528250836626295e-05,
+        "loss": 0.9347,
+        "step": 26660
+      },
+      {
+        "epoch": 0.8504735482636564,
+        "grad_norm": 0.35870596766471863,
+        "learning_rate": 4.5109571176992264e-05,
+        "loss": 0.9432,
+        "step": 26670
+      },
+      {
+        "epoch": 0.8507924359832902,
+        "grad_norm": 0.3548199534416199,
+        "learning_rate": 4.493729444741149e-05,
+        "loss": 0.949,
+        "step": 26680
+      },
+      {
+        "epoch": 0.8511113237029242,
+        "grad_norm": 0.34394973516464233,
+        "learning_rate": 4.476567565517706e-05,
+        "loss": 0.9346,
+        "step": 26690
+      },
+      {
+        "epoch": 0.8514302114225581,
+        "grad_norm": 0.35724857449531555,
+        "learning_rate": 4.459471228757844e-05,
+        "loss": 0.9341,
+        "step": 26700
+      },
+      {
+        "epoch": 0.851749099142192,
+        "grad_norm": 0.3644699454307556,
+        "learning_rate": 4.442440184150135e-05,
+        "loss": 0.9465,
+        "step": 26710
+      },
+      {
+        "epoch": 0.8520679868618259,
+        "grad_norm": 0.36522284150123596,
+        "learning_rate": 4.425474182339106e-05,
+        "loss": 0.9398,
+        "step": 26720
+      },
+      {
+        "epoch": 0.8523868745814599,
+        "grad_norm": 0.3725425899028778,
+        "learning_rate": 4.40857297492159e-05,
+        "loss": 0.9481,
+        "step": 26730
+      },
+      {
+        "epoch": 0.8527057623010937,
+        "grad_norm": 0.36529541015625,
+        "learning_rate": 4.391736314443091e-05,
+        "loss": 0.9542,
+        "step": 26740
+      },
+      {
+        "epoch": 0.8530246500207277,
+        "grad_norm": 0.3558250665664673,
+        "learning_rate": 4.37496395439416e-05,
+        "loss": 0.9518,
+        "step": 26750
+      },
+      {
+        "epoch": 0.8533435377403616,
+        "grad_norm": 0.3752404749393463,
+        "learning_rate": 4.3582556492067844e-05,
+        "loss": 0.936,
+        "step": 26760
+      },
+      {
+        "epoch": 0.8536624254599955,
+        "grad_norm": 0.37050309777259827,
+        "learning_rate": 4.341611154250795e-05,
+        "loss": 0.9322,
+        "step": 26770
+      },
+      {
+        "epoch": 0.8539813131796294,
+        "grad_norm": 0.36518236994743347,
+        "learning_rate": 4.325030225830281e-05,
+        "loss": 0.9567,
+        "step": 26780
+      },
+      {
+        "epoch": 0.8543002008992634,
+        "grad_norm": 0.3709019720554352,
+        "learning_rate": 4.308512621180027e-05,
+        "loss": 0.9464,
+        "step": 26790
+      },
+      {
+        "epoch": 0.8546190886188972,
+        "grad_norm": 0.3707132637500763,
+        "learning_rate": 4.2920580984619533e-05,
+        "loss": 0.9245,
+        "step": 26800
+      },
+      {
+        "epoch": 0.8549379763385312,
+        "grad_norm": 0.35442495346069336,
+        "learning_rate": 4.275666416761577e-05,
+        "loss": 0.9439,
+        "step": 26810
+      },
+      {
+        "epoch": 0.8552568640581651,
+        "grad_norm": 0.3763255178928375,
+        "learning_rate": 4.259337336084486e-05,
+        "loss": 0.9557,
+        "step": 26820
+      },
+      {
+        "epoch": 0.855575751777799,
+        "grad_norm": 0.3710927367210388,
+        "learning_rate": 4.243070617352825e-05,
+        "loss": 0.9561,
+        "step": 26830
+      },
+      {
+        "epoch": 0.8558946394974329,
+        "grad_norm": 0.36720240116119385,
+        "learning_rate": 4.226866022401794e-05,
+        "loss": 0.9419,
+        "step": 26840
+      },
+      {
+        "epoch": 0.8562135272170669,
+        "grad_norm": 0.37940865755081177,
+        "learning_rate": 4.2107233139761615e-05,
+        "loss": 0.9378,
+        "step": 26850
+      },
+      {
+        "epoch": 0.8565324149367007,
+        "grad_norm": 0.35762712359428406,
+        "learning_rate": 4.194642255726791e-05,
+        "loss": 0.9367,
+        "step": 26860
+      },
+      {
+        "epoch": 0.8568513026563347,
+        "grad_norm": 0.3704802095890045,
+        "learning_rate": 4.1786226122071794e-05,
+        "loss": 0.9501,
+        "step": 26870
+      },
+      {
+        "epoch": 0.8571701903759686,
+        "grad_norm": 0.3747062385082245,
+        "learning_rate": 4.162664148870013e-05,
+        "loss": 0.9374,
+        "step": 26880
+      },
+      {
+        "epoch": 0.8574890780956025,
+        "grad_norm": 0.3695331811904907,
+        "learning_rate": 4.146766632063729e-05,
+        "loss": 0.9485,
+        "step": 26890
+      },
+      {
+        "epoch": 0.8578079658152364,
+        "grad_norm": 0.3804053068161011,
+        "learning_rate": 4.1309298290290994e-05,
+        "loss": 0.9504,
+        "step": 26900
+      },
+      {
+        "epoch": 0.8581268535348704,
+        "grad_norm": 0.3694726228713989,
+        "learning_rate": 4.1151535078958185e-05,
+        "loss": 0.9515,
+        "step": 26910
+      },
+      {
+        "epoch": 0.8584457412545042,
+        "grad_norm": 0.37324094772338867,
+        "learning_rate": 4.099437437679111e-05,
+        "loss": 0.9379,
+        "step": 26920
+      },
+      {
+        "epoch": 0.8587646289741382,
+        "grad_norm": 0.35455772280693054,
+        "learning_rate": 4.08378138827635e-05,
+        "loss": 0.9565,
+        "step": 26930
+      },
+      {
+        "epoch": 0.8590835166937721,
+        "grad_norm": 0.3658784329891205,
+        "learning_rate": 4.0681851304636857e-05,
+        "loss": 0.9395,
+        "step": 26940
+      },
+      {
+        "epoch": 0.859402404413406,
+        "grad_norm": 0.36860978603363037,
+        "learning_rate": 4.052648435892692e-05,
+        "loss": 0.9333,
+        "step": 26950
+      },
+      {
+        "epoch": 0.8597212921330399,
+        "grad_norm": 0.3664551377296448,
+        "learning_rate": 4.037171077087022e-05,
+        "loss": 0.9441,
+        "step": 26960
+      },
+      {
+        "epoch": 0.8600401798526739,
+        "grad_norm": 0.36110377311706543,
+        "learning_rate": 4.021752827439075e-05,
+        "loss": 0.9171,
+        "step": 26970
+      },
+      {
+        "epoch": 0.8603590675723078,
+        "grad_norm": 0.3666580617427826,
+        "learning_rate": 4.0063934612066855e-05,
+        "loss": 0.9306,
+        "step": 26980
+      },
+      {
+        "epoch": 0.8606779552919417,
+        "grad_norm": 0.3739660978317261,
+        "learning_rate": 3.991092753509812e-05,
+        "loss": 0.9396,
+        "step": 26990
+      },
+      {
+        "epoch": 0.8609968430115756,
+        "grad_norm": 0.36052122712135315,
+        "learning_rate": 3.975850480327241e-05,
+        "loss": 0.9236,
+        "step": 27000
+      },
+      {
+        "epoch": 0.8613157307312096,
+        "grad_norm": 0.3642176389694214,
+        "learning_rate": 3.960666418493324e-05,
+        "loss": 0.9218,
+        "step": 27010
+      },
+      {
+        "epoch": 0.8616346184508434,
+        "grad_norm": 0.36625078320503235,
+        "learning_rate": 3.9455403456946875e-05,
+        "loss": 0.9433,
+        "step": 27020
+      },
+      {
+        "epoch": 0.8619535061704774,
+        "grad_norm": 0.3641470670700073,
+        "learning_rate": 3.930472040466995e-05,
+        "loss": 0.9443,
+        "step": 27030
+      },
+      {
+        "epoch": 0.8622723938901113,
+        "grad_norm": 0.3567174971103668,
+        "learning_rate": 3.915461282191693e-05,
+        "loss": 0.9546,
+        "step": 27040
+      },
+      {
+        "epoch": 0.8625912816097452,
+        "grad_norm": 0.3644573390483856,
+        "learning_rate": 3.900507851092791e-05,
+        "loss": 0.9586,
+        "step": 27050
+      },
+      {
+        "epoch": 0.8629101693293791,
+        "grad_norm": 0.369090735912323,
+        "learning_rate": 3.885611528233638e-05,
+        "loss": 0.9501,
+        "step": 27060
+      },
+      {
+        "epoch": 0.8632290570490131,
+        "grad_norm": 0.3611607849597931,
+        "learning_rate": 3.870772095513717e-05,
+        "loss": 0.9329,
+        "step": 27070
+      },
+      {
+        "epoch": 0.8635479447686469,
+        "grad_norm": 0.3491160571575165,
+        "learning_rate": 3.855989335665453e-05,
+        "loss": 0.952,
+        "step": 27080
+      },
+      {
+        "epoch": 0.8638668324882809,
+        "grad_norm": 0.368991881608963,
+        "learning_rate": 3.841263032251032e-05,
+        "loss": 0.95,
+        "step": 27090
+      },
+      {
+        "epoch": 0.8641857202079148,
+        "grad_norm": 0.351935476064682,
+        "learning_rate": 3.8265929696592315e-05,
+        "loss": 0.9426,
+        "step": 27100
+      },
+      {
+        "epoch": 0.8645046079275487,
+        "grad_norm": 0.3519881069660187,
+        "learning_rate": 3.811978933102264e-05,
+        "loss": 0.9313,
+        "step": 27110
+      },
+      {
+        "epoch": 0.8648234956471826,
+        "grad_norm": 0.3692045509815216,
+        "learning_rate": 3.797420708612632e-05,
+        "loss": 0.9482,
+        "step": 27120
+      },
+      {
+        "epoch": 0.8651423833668166,
+        "grad_norm": 0.36540305614471436,
+        "learning_rate": 3.7829180830399963e-05,
+        "loss": 0.9294,
+        "step": 27130
+      },
+      {
+        "epoch": 0.8654612710864504,
+        "grad_norm": 0.35637855529785156,
+        "learning_rate": 3.768470844048052e-05,
+        "loss": 0.9244,
+        "step": 27140
+      },
+      {
+        "epoch": 0.8657801588060844,
+        "grad_norm": 0.3536495268344879,
+        "learning_rate": 3.7540787801114243e-05,
+        "loss": 0.9418,
+        "step": 27150
+      },
+      {
+        "epoch": 0.8660990465257183,
+        "grad_norm": 0.36629825830459595,
+        "learning_rate": 3.739741680512569e-05,
+        "loss": 0.9484,
+        "step": 27160
+      },
+      {
+        "epoch": 0.8664179342453522,
+        "grad_norm": 0.3662286698818207,
+        "learning_rate": 3.725459335338685e-05,
+        "loss": 0.9328,
+        "step": 27170
+      },
+      {
+        "epoch": 0.8667368219649861,
+        "grad_norm": 0.36647507548332214,
+        "learning_rate": 3.711231535478648e-05,
+        "loss": 0.9442,
+        "step": 27180
+      },
+      {
+        "epoch": 0.8670557096846201,
+        "grad_norm": 0.3591477572917938,
+        "learning_rate": 3.697058072619941e-05,
+        "loss": 0.9401,
+        "step": 27190
+      },
+      {
+        "epoch": 0.8673745974042539,
+        "grad_norm": 0.35763055086135864,
+        "learning_rate": 3.6829387392456075e-05,
+        "loss": 0.9369,
+        "step": 27200
+      },
+      {
+        "epoch": 0.8676934851238879,
+        "grad_norm": 0.3626702129840851,
+        "learning_rate": 3.668873328631214e-05,
+        "loss": 0.937,
+        "step": 27210
+      },
+      {
+        "epoch": 0.8680123728435218,
+        "grad_norm": 0.3550717234611511,
+        "learning_rate": 3.6548616348418236e-05,
+        "loss": 0.9322,
+        "step": 27220
+      },
+      {
+        "epoch": 0.8683312605631557,
+        "grad_norm": 0.36245885491371155,
+        "learning_rate": 3.640903452728978e-05,
+        "loss": 0.9334,
+        "step": 27230
+      },
+      {
+        "epoch": 0.8686501482827896,
+        "grad_norm": 0.36782872676849365,
+        "learning_rate": 3.626998577927698e-05,
+        "loss": 0.9322,
+        "step": 27240
+      },
+      {
+        "epoch": 0.8689690360024236,
+        "grad_norm": 0.38292792439460754,
+        "learning_rate": 3.6131468068534876e-05,
+        "loss": 0.9441,
+        "step": 27250
+      },
+      {
+        "epoch": 0.8692879237220574,
+        "grad_norm": 0.37048473954200745,
+        "learning_rate": 3.599347936699354e-05,
+        "loss": 0.9355,
+        "step": 27260
+      },
+      {
+        "epoch": 0.8696068114416914,
+        "grad_norm": 0.3561244010925293,
+        "learning_rate": 3.585601765432841e-05,
+        "loss": 0.9362,
+        "step": 27270
+      },
+      {
+        "epoch": 0.8699256991613253,
+        "grad_norm": 0.3700783848762512,
+        "learning_rate": 3.571908091793068e-05,
+        "loss": 0.9346,
+        "step": 27280
+      },
+      {
+        "epoch": 0.8702445868809592,
+        "grad_norm": 0.358857661485672,
+        "learning_rate": 3.558266715287785e-05,
+        "loss": 0.9398,
+        "step": 27290
+      },
+      {
+        "epoch": 0.8705634746005931,
+        "grad_norm": 0.3824344277381897,
+        "learning_rate": 3.544677436190435e-05,
+        "loss": 0.9489,
+        "step": 27300
+      },
+      {
+        "epoch": 0.8708823623202271,
+        "grad_norm": 0.376953125,
+        "learning_rate": 3.5311400555372326e-05,
+        "loss": 0.9566,
+        "step": 27310
+      },
+      {
+        "epoch": 0.8712012500398609,
+        "grad_norm": 0.35705527663230896,
+        "learning_rate": 3.517654375124249e-05,
+        "loss": 0.9313,
+        "step": 27320
+      },
+      {
+        "epoch": 0.8715201377594949,
+        "grad_norm": 0.3527824878692627,
+        "learning_rate": 3.5042201975045115e-05,
+        "loss": 0.9491,
+        "step": 27330
+      },
+      {
+        "epoch": 0.8718390254791288,
+        "grad_norm": 0.361565500497818,
+        "learning_rate": 3.490837325985108e-05,
+        "loss": 0.9515,
+        "step": 27340
+      },
+      {
+        "epoch": 0.8721579131987627,
+        "grad_norm": 0.3611353039741516,
+        "learning_rate": 3.4775055646243186e-05,
+        "loss": 0.933,
+        "step": 27350
+      },
+      {
+        "epoch": 0.8724768009183966,
+        "grad_norm": 0.36242130398750305,
+        "learning_rate": 3.464224718228731e-05,
+        "loss": 0.9474,
+        "step": 27360
+      },
+      {
+        "epoch": 0.8727956886380306,
+        "grad_norm": 0.36209696531295776,
+        "learning_rate": 3.450994592350395e-05,
+        "loss": 0.9306,
+        "step": 27370
+      },
+      {
+        "epoch": 0.8731145763576644,
+        "grad_norm": 0.3568836748600006,
+        "learning_rate": 3.437814993283972e-05,
+        "loss": 0.9371,
+        "step": 27380
+      },
+      {
+        "epoch": 0.8734334640772984,
+        "grad_norm": 0.3655485212802887,
+        "learning_rate": 3.424685728063894e-05,
+        "loss": 0.9409,
+        "step": 27390
+      },
+      {
+        "epoch": 0.8737523517969323,
+        "grad_norm": 0.36494147777557373,
+        "learning_rate": 3.411606604461545e-05,
+        "loss": 0.9339,
+        "step": 27400
+      },
+      {
+        "epoch": 0.8740712395165662,
+        "grad_norm": 0.36795875430107117,
+        "learning_rate": 3.398577430982446e-05,
+        "loss": 0.9249,
+        "step": 27410
+      },
+      {
+        "epoch": 0.8743901272362001,
+        "grad_norm": 0.36251020431518555,
+        "learning_rate": 3.385598016863445e-05,
+        "loss": 0.9301,
+        "step": 27420
+      },
+      {
+        "epoch": 0.8747090149558341,
+        "grad_norm": 0.35069212317466736,
+        "learning_rate": 3.372668172069933e-05,
+        "loss": 0.9388,
+        "step": 27430
+      },
+      {
+        "epoch": 0.8750279026754679,
+        "grad_norm": 0.3661258816719055,
+        "learning_rate": 3.3597877072930536e-05,
+        "loss": 0.9326,
+        "step": 27440
+      },
+      {
+        "epoch": 0.8753467903951019,
+        "grad_norm": 0.37232705950737,
+        "learning_rate": 3.3469564339469355e-05,
+        "loss": 0.9388,
+        "step": 27450
+      },
+      {
+        "epoch": 0.8756656781147358,
+        "grad_norm": 0.3602105379104614,
+        "learning_rate": 3.334174164165931e-05,
+        "loss": 0.9348,
+        "step": 27460
+      },
+      {
+        "epoch": 0.8759845658343697,
+        "grad_norm": 0.3558805286884308,
+        "learning_rate": 3.321440710801865e-05,
+        "loss": 0.9433,
+        "step": 27470
+      },
+      {
+        "epoch": 0.8763034535540036,
+        "grad_norm": 0.38075533509254456,
+        "learning_rate": 3.308755887421296e-05,
+        "loss": 0.9363,
+        "step": 27480
+      },
+      {
+        "epoch": 0.8766223412736376,
+        "grad_norm": 0.35899147391319275,
+        "learning_rate": 3.296119508302781e-05,
+        "loss": 0.9301,
+        "step": 27490
+      },
+      {
+        "epoch": 0.8769412289932714,
+        "grad_norm": 0.3670910894870758,
+        "learning_rate": 3.2835313884341655e-05,
+        "loss": 0.9487,
+        "step": 27500
+      },
+      {
+        "epoch": 0.8772601167129054,
+        "grad_norm": 0.35804957151412964,
+        "learning_rate": 3.2709913435098666e-05,
+        "loss": 0.933,
+        "step": 27510
+      },
+      {
+        "epoch": 0.8775790044325393,
+        "grad_norm": 0.37163132429122925,
+        "learning_rate": 3.2584991899281825e-05,
+        "loss": 0.9448,
+        "step": 27520
+      },
+      {
+        "epoch": 0.8778978921521732,
+        "grad_norm": 0.35967984795570374,
+        "learning_rate": 3.246054744788594e-05,
+        "loss": 0.9144,
+        "step": 27530
+      },
+      {
+        "epoch": 0.8782167798718071,
+        "grad_norm": 0.36492854356765747,
+        "learning_rate": 3.233657825889095e-05,
+        "loss": 0.9395,
+        "step": 27540
+      },
+      {
+        "epoch": 0.8785356675914411,
+        "grad_norm": 0.3674900233745575,
+        "learning_rate": 3.221308251723522e-05,
+        "loss": 0.9434,
+        "step": 27550
+      },
+      {
+        "epoch": 0.8788545553110749,
+        "grad_norm": 0.3588939607143402,
+        "learning_rate": 3.2090058414788956e-05,
+        "loss": 0.9357,
+        "step": 27560
+      },
+      {
+        "epoch": 0.8791734430307089,
+        "grad_norm": 0.35960447788238525,
+        "learning_rate": 3.196750415032777e-05,
+        "loss": 0.9354,
+        "step": 27570
+      },
+      {
+        "epoch": 0.8794923307503428,
+        "grad_norm": 0.3683167099952698,
+        "learning_rate": 3.1845417929506246e-05,
+        "loss": 0.9415,
+        "step": 27580
+      },
+      {
+        "epoch": 0.8798112184699767,
+        "grad_norm": 0.36023595929145813,
+        "learning_rate": 3.1723797964831725e-05,
+        "loss": 0.9382,
+        "step": 27590
+      },
+      {
+        "epoch": 0.8801301061896106,
+        "grad_norm": 0.36763185262680054,
+        "learning_rate": 3.160264247563812e-05,
+        "loss": 0.936,
+        "step": 27600
+      },
+      {
+        "epoch": 0.8804489939092446,
+        "grad_norm": 0.362166166305542,
+        "learning_rate": 3.1481949688059806e-05,
+        "loss": 0.9333,
+        "step": 27610
+      },
+      {
+        "epoch": 0.8807678816288784,
+        "grad_norm": 0.36778682470321655,
+        "learning_rate": 3.1361717835005704e-05,
+        "loss": 0.9392,
+        "step": 27620
+      },
+      {
+        "epoch": 0.8810867693485124,
+        "grad_norm": 0.3676985502243042,
+        "learning_rate": 3.1241945156133386e-05,
+        "loss": 0.9399,
+        "step": 27630
+      },
+      {
+        "epoch": 0.8814056570681463,
+        "grad_norm": 0.3742704391479492,
+        "learning_rate": 3.1122629897823284e-05,
+        "loss": 0.9337,
+        "step": 27640
+      },
+      {
+        "epoch": 0.8817245447877802,
+        "grad_norm": 0.36281928420066833,
+        "learning_rate": 3.100377031315304e-05,
+        "loss": 0.9327,
+        "step": 27650
+      },
+      {
+        "epoch": 0.8820434325074141,
+        "grad_norm": 0.3667699992656708,
+        "learning_rate": 3.088536466187193e-05,
+        "loss": 0.9452,
+        "step": 27660
+      },
+      {
+        "epoch": 0.8823623202270481,
+        "grad_norm": 0.36892062425613403,
+        "learning_rate": 3.076741121037534e-05,
+        "loss": 0.9261,
+        "step": 27670
+      },
+      {
+        "epoch": 0.8826812079466819,
+        "grad_norm": 0.3637031614780426,
+        "learning_rate": 3.064990823167945e-05,
+        "loss": 0.9376,
+        "step": 27680
+      },
+      {
+        "epoch": 0.8830000956663159,
+        "grad_norm": 0.3612978458404541,
+        "learning_rate": 3.053285400539591e-05,
+        "loss": 0.9573,
+        "step": 27690
+      },
+      {
+        "epoch": 0.8833189833859498,
+        "grad_norm": 0.3765890300273895,
+        "learning_rate": 3.041624681770667e-05,
+        "loss": 0.9367,
+        "step": 27700
+      },
+      {
+        "epoch": 0.8836378711055837,
+        "grad_norm": 0.36441487073898315,
+        "learning_rate": 3.030008496133884e-05,
+        "loss": 0.9228,
+        "step": 27710
+      },
+      {
+        "epoch": 0.8839567588252176,
+        "grad_norm": 0.361269474029541,
+        "learning_rate": 3.0184366735539748e-05,
+        "loss": 0.9288,
+        "step": 27720
+      },
+      {
+        "epoch": 0.8842756465448516,
+        "grad_norm": 0.37258458137512207,
+        "learning_rate": 3.006909044605203e-05,
+        "loss": 0.9353,
+        "step": 27730
+      },
+      {
+        "epoch": 0.8845945342644854,
+        "grad_norm": 0.364519864320755,
+        "learning_rate": 2.995425440508881e-05,
+        "loss": 0.9169,
+        "step": 27740
+      },
+      {
+        "epoch": 0.8849134219841194,
+        "grad_norm": 0.35370734333992004,
+        "learning_rate": 2.983985693130898e-05,
+        "loss": 0.9495,
+        "step": 27750
+      },
+      {
+        "epoch": 0.8852323097037533,
+        "grad_norm": 0.38416677713394165,
+        "learning_rate": 2.9725896349792608e-05,
+        "loss": 0.9371,
+        "step": 27760
+      },
+      {
+        "epoch": 0.8855511974233872,
+        "grad_norm": 0.3630027174949646,
+        "learning_rate": 2.9612370992016398e-05,
+        "loss": 0.9366,
+        "step": 27770
+      },
+      {
+        "epoch": 0.8858700851430211,
+        "grad_norm": 0.36047670245170593,
+        "learning_rate": 2.9499279195829267e-05,
+        "loss": 0.9219,
+        "step": 27780
+      },
+      {
+        "epoch": 0.8861889728626551,
+        "grad_norm": 0.35632213950157166,
+        "learning_rate": 2.9386619305428005e-05,
+        "loss": 0.9305,
+        "step": 27790
+      },
+      {
+        "epoch": 0.886507860582289,
+        "grad_norm": 0.3719460368156433,
+        "learning_rate": 2.9274389671333044e-05,
+        "loss": 0.9455,
+        "step": 27800
+      },
+      {
+        "epoch": 0.8868267483019229,
+        "grad_norm": 0.37200984358787537,
+        "learning_rate": 2.9162588650364285e-05,
+        "loss": 0.9365,
+        "step": 27810
+      },
+      {
+        "epoch": 0.8871456360215568,
+        "grad_norm": 0.3699634373188019,
+        "learning_rate": 2.9051214605617067e-05,
+        "loss": 0.9182,
+        "step": 27820
+      },
+      {
+        "epoch": 0.8874645237411908,
+        "grad_norm": 0.36919641494750977,
+        "learning_rate": 2.8940265906438172e-05,
+        "loss": 0.9275,
+        "step": 27830
+      },
+      {
+        "epoch": 0.8877834114608246,
+        "grad_norm": 0.3644123673439026,
+        "learning_rate": 2.882974092840196e-05,
+        "loss": 0.9334,
+        "step": 27840
+      },
+      {
+        "epoch": 0.8881022991804586,
+        "grad_norm": 0.36178484559059143,
+        "learning_rate": 2.871963805328664e-05,
+        "loss": 0.9486,
+        "step": 27850
+      },
+      {
+        "epoch": 0.8884211869000925,
+        "grad_norm": 0.36924728751182556,
+        "learning_rate": 2.860995566905046e-05,
+        "loss": 0.9361,
+        "step": 27860
+      },
+      {
+        "epoch": 0.8887400746197264,
+        "grad_norm": 0.349809467792511,
+        "learning_rate": 2.850069216980822e-05,
+        "loss": 0.9321,
+        "step": 27870
+      },
+      {
+        "epoch": 0.8890589623393603,
+        "grad_norm": 0.37572547793388367,
+        "learning_rate": 2.8391845955807693e-05,
+        "loss": 0.9389,
+        "step": 27880
+      },
+      {
+        "epoch": 0.8893778500589943,
+        "grad_norm": 0.36704355478286743,
+        "learning_rate": 2.8283415433406215e-05,
+        "loss": 0.919,
+        "step": 27890
+      },
+      {
+        "epoch": 0.8896967377786281,
+        "grad_norm": 0.35531923174858093,
+        "learning_rate": 2.8175399015047376e-05,
+        "loss": 0.9412,
+        "step": 27900
+      },
+      {
+        "epoch": 0.8900156254982621,
+        "grad_norm": 0.35668015480041504,
+        "learning_rate": 2.8067795119237755e-05,
+        "loss": 0.9412,
+        "step": 27910
+      },
+      {
+        "epoch": 0.890334513217896,
+        "grad_norm": 0.3683586120605469,
+        "learning_rate": 2.796060217052376e-05,
+        "loss": 0.9495,
+        "step": 27920
+      },
+      {
+        "epoch": 0.8906534009375299,
+        "grad_norm": 0.3708488345146179,
+        "learning_rate": 2.785381859946858e-05,
+        "loss": 0.9303,
+        "step": 27930
+      },
+      {
+        "epoch": 0.8909722886571638,
+        "grad_norm": 0.3647977113723755,
+        "learning_rate": 2.7747442842629192e-05,
+        "loss": 0.9326,
+        "step": 27940
+      },
+      {
+        "epoch": 0.8912911763767978,
+        "grad_norm": 0.3660475015640259,
+        "learning_rate": 2.7641473342533486e-05,
+        "loss": 0.9385,
+        "step": 27950
+      },
+      {
+        "epoch": 0.8916100640964316,
+        "grad_norm": 0.373983770608902,
+        "learning_rate": 2.753590854765743e-05,
+        "loss": 0.9424,
+        "step": 27960
+      },
+      {
+        "epoch": 0.8919289518160656,
+        "grad_norm": 0.36659717559814453,
+        "learning_rate": 2.7430746912402397e-05,
+        "loss": 0.9463,
+        "step": 27970
+      },
+      {
+        "epoch": 0.8922478395356995,
+        "grad_norm": 0.3667260706424713,
+        "learning_rate": 2.73259868970725e-05,
+        "loss": 0.9566,
+        "step": 27980
+      },
+      {
+        "epoch": 0.8925667272553334,
+        "grad_norm": 0.3601289689540863,
+        "learning_rate": 2.722162696785207e-05,
+        "loss": 0.9342,
+        "step": 27990
+      },
+      {
+        "epoch": 0.8928856149749673,
+        "grad_norm": 0.3438984453678131,
+        "learning_rate": 2.711766559678319e-05,
+        "loss": 0.9327,
+        "step": 28000
+      },
+      {
+        "epoch": 0.8932045026946013,
+        "grad_norm": 0.37399184703826904,
+        "learning_rate": 2.7014101261743303e-05,
+        "loss": 0.9334,
+        "step": 28010
+      },
+      {
+        "epoch": 0.8935233904142351,
+        "grad_norm": 0.35472747683525085,
+        "learning_rate": 2.6910932446423003e-05,
+        "loss": 0.9376,
+        "step": 28020
+      },
+      {
+        "epoch": 0.8938422781338691,
+        "grad_norm": 0.3564119338989258,
+        "learning_rate": 2.6808157640303726e-05,
+        "loss": 0.9407,
+        "step": 28030
+      },
+      {
+        "epoch": 0.894161165853503,
+        "grad_norm": 0.3564133644104004,
+        "learning_rate": 2.670577533863571e-05,
+        "loss": 0.9287,
+        "step": 28040
+      },
+      {
+        "epoch": 0.8944800535731369,
+        "grad_norm": 0.3710532784461975,
+        "learning_rate": 2.6603784042415948e-05,
+        "loss": 0.9168,
+        "step": 28050
+      },
+      {
+        "epoch": 0.8947989412927708,
+        "grad_norm": 0.37403547763824463,
+        "learning_rate": 2.6502182258366217e-05,
+        "loss": 0.9314,
+        "step": 28060
+      },
+      {
+        "epoch": 0.8951178290124048,
+        "grad_norm": 0.38338151574134827,
+        "learning_rate": 2.640096849891124e-05,
+        "loss": 0.9309,
+        "step": 28070
+      },
+      {
+        "epoch": 0.8954367167320386,
+        "grad_norm": 0.3532122075557709,
+        "learning_rate": 2.630014128215691e-05,
+        "loss": 0.9309,
+        "step": 28080
+      },
+      {
+        "epoch": 0.8957556044516726,
+        "grad_norm": 0.37361615896224976,
+        "learning_rate": 2.6199699131868562e-05,
+        "loss": 0.9278,
+        "step": 28090
+      },
+      {
+        "epoch": 0.8960744921713065,
+        "grad_norm": 0.36820292472839355,
+        "learning_rate": 2.6099640577449387e-05,
+        "loss": 0.9379,
+        "step": 28100
+      },
+      {
+        "epoch": 0.8963933798909404,
+        "grad_norm": 0.3665463626384735,
+        "learning_rate": 2.5999964153918897e-05,
+        "loss": 0.9271,
+        "step": 28110
+      },
+      {
+        "epoch": 0.8967122676105743,
+        "grad_norm": 0.3633195161819458,
+        "learning_rate": 2.5900668401891458e-05,
+        "loss": 0.9474,
+        "step": 28120
+      },
+      {
+        "epoch": 0.8970311553302083,
+        "grad_norm": 0.36331549286842346,
+        "learning_rate": 2.5801751867554953e-05,
+        "loss": 0.9264,
+        "step": 28130
+      },
+      {
+        "epoch": 0.8973500430498421,
+        "grad_norm": 0.3619482219219208,
+        "learning_rate": 2.570321310264946e-05,
+        "loss": 0.9533,
+        "step": 28140
+      },
+      {
+        "epoch": 0.8976689307694761,
+        "grad_norm": 0.3781132996082306,
+        "learning_rate": 2.5605050664446088e-05,
+        "loss": 0.9373,
+        "step": 28150
+      },
+      {
+        "epoch": 0.89798781848911,
+        "grad_norm": 0.36209985613822937,
+        "learning_rate": 2.550726311572581e-05,
+        "loss": 0.9373,
+        "step": 28160
+      },
+      {
+        "epoch": 0.8983067062087439,
+        "grad_norm": 0.3673580288887024,
+        "learning_rate": 2.540984902475846e-05,
+        "loss": 0.9245,
+        "step": 28170
+      },
+      {
+        "epoch": 0.8986255939283778,
+        "grad_norm": 0.3630748391151428,
+        "learning_rate": 2.5312806965281734e-05,
+        "loss": 0.9319,
+        "step": 28180
+      },
+      {
+        "epoch": 0.8989444816480118,
+        "grad_norm": 0.36070698499679565,
+        "learning_rate": 2.5216135516480365e-05,
+        "loss": 0.9287,
+        "step": 28190
+      },
+      {
+        "epoch": 0.8992633693676456,
+        "grad_norm": 0.3583659529685974,
+        "learning_rate": 2.5119833262965248e-05,
+        "loss": 0.9333,
+        "step": 28200
+      },
+      {
+        "epoch": 0.8995822570872796,
+        "grad_norm": 0.36007121205329895,
+        "learning_rate": 2.502389879475276e-05,
+        "loss": 0.9368,
+        "step": 28210
+      },
+      {
+        "epoch": 0.8999011448069135,
+        "grad_norm": 0.37289687991142273,
+        "learning_rate": 2.4928330707244114e-05,
+        "loss": 0.949,
+        "step": 28220
+      },
+      {
+        "epoch": 0.9002200325265474,
+        "grad_norm": 0.3503265678882599,
+        "learning_rate": 2.4833127601204777e-05,
+        "loss": 0.9214,
+        "step": 28230
+      },
+      {
+        "epoch": 0.9005389202461813,
+        "grad_norm": 0.35933589935302734,
+        "learning_rate": 2.4738288082744006e-05,
+        "loss": 0.9406,
+        "step": 28240
+      },
+      {
+        "epoch": 0.9008578079658153,
+        "grad_norm": 0.36435064673423767,
+        "learning_rate": 2.4643810763294427e-05,
+        "loss": 0.94,
+        "step": 28250
+      },
+      {
+        "epoch": 0.9011766956854491,
+        "grad_norm": 0.3533856272697449,
+        "learning_rate": 2.4549694259591695e-05,
+        "loss": 0.9161,
+        "step": 28260
+      },
+      {
+        "epoch": 0.9014955834050831,
+        "grad_norm": 0.3634324073791504,
+        "learning_rate": 2.4455937193654268e-05,
+        "loss": 0.9391,
+        "step": 28270
+      },
+      {
+        "epoch": 0.901814471124717,
+        "grad_norm": 0.3563332259654999,
+        "learning_rate": 2.4362538192763198e-05,
+        "loss": 0.9248,
+        "step": 28280
+      },
+      {
+        "epoch": 0.9021333588443509,
+        "grad_norm": 0.3568084239959717,
+        "learning_rate": 2.4269495889442062e-05,
+        "loss": 0.9422,
+        "step": 28290
+      },
+      {
+        "epoch": 0.9024522465639848,
+        "grad_norm": 0.36897915601730347,
+        "learning_rate": 2.417680892143693e-05,
+        "loss": 0.9609,
+        "step": 28300
+      },
+      {
+        "epoch": 0.9027711342836188,
+        "grad_norm": 0.3564818799495697,
+        "learning_rate": 2.4084475931696406e-05,
+        "loss": 0.9402,
+        "step": 28310
+      },
+      {
+        "epoch": 0.9030900220032526,
+        "grad_norm": 0.36032170057296753,
+        "learning_rate": 2.399249556835179e-05,
+        "loss": 0.928,
+        "step": 28320
+      },
+      {
+        "epoch": 0.9034089097228866,
+        "grad_norm": 0.362189382314682,
+        "learning_rate": 2.390086648469725e-05,
+        "loss": 0.9331,
+        "step": 28330
+      },
+      {
+        "epoch": 0.9037277974425205,
+        "grad_norm": 0.36130815744400024,
+        "learning_rate": 2.3809587339170133e-05,
+        "loss": 0.9158,
+        "step": 28340
+      },
+      {
+        "epoch": 0.9040466851621544,
+        "grad_norm": 0.3681578040122986,
+        "learning_rate": 2.3718656795331296e-05,
+        "loss": 0.9383,
+        "step": 28350
+      },
+      {
+        "epoch": 0.9043655728817883,
+        "grad_norm": 0.359544038772583,
+        "learning_rate": 2.362807352184559e-05,
+        "loss": 0.9365,
+        "step": 28360
+      },
+      {
+        "epoch": 0.9046844606014223,
+        "grad_norm": 0.36564376950263977,
+        "learning_rate": 2.3537836192462286e-05,
+        "loss": 0.9365,
+        "step": 28370
+      },
+      {
+        "epoch": 0.9050033483210561,
+        "grad_norm": 0.36053574085235596,
+        "learning_rate": 2.344794348599573e-05,
+        "loss": 0.9249,
+        "step": 28380
+      },
+      {
+        "epoch": 0.9053222360406901,
+        "grad_norm": 0.3556235134601593,
+        "learning_rate": 2.3358394086305966e-05,
+        "loss": 0.9233,
+        "step": 28390
+      },
+      {
+        "epoch": 0.905641123760324,
+        "grad_norm": 0.3614828884601593,
+        "learning_rate": 2.3269186682279475e-05,
+        "loss": 0.9318,
+        "step": 28400
+      },
+      {
+        "epoch": 0.9059600114799579,
+        "grad_norm": 0.3779839277267456,
+        "learning_rate": 2.3180319967809967e-05,
+        "loss": 0.9378,
+        "step": 28410
+      },
+      {
+        "epoch": 0.9062788991995918,
+        "grad_norm": 0.3644469678401947,
+        "learning_rate": 2.3091792641779272e-05,
+        "loss": 0.9291,
+        "step": 28420
+      },
+      {
+        "epoch": 0.9065977869192258,
+        "grad_norm": 0.35818055272102356,
+        "learning_rate": 2.300360340803829e-05,
+        "loss": 0.9551,
+        "step": 28430
+      },
+      {
+        "epoch": 0.9069166746388596,
+        "grad_norm": 0.3682006299495697,
+        "learning_rate": 2.2915750975388005e-05,
+        "loss": 0.9419,
+        "step": 28440
+      },
+      {
+        "epoch": 0.9072355623584936,
+        "grad_norm": 0.3534643352031708,
+        "learning_rate": 2.2828234057560574e-05,
+        "loss": 0.9411,
+        "step": 28450
+      },
+      {
+        "epoch": 0.9075544500781275,
+        "grad_norm": 0.37065860629081726,
+        "learning_rate": 2.2741051373200522e-05,
+        "loss": 0.9274,
+        "step": 28460
+      },
+      {
+        "epoch": 0.9078733377977614,
+        "grad_norm": 0.3683619797229767,
+        "learning_rate": 2.265420164584595e-05,
+        "loss": 0.9381,
+        "step": 28470
+      },
+      {
+        "epoch": 0.9081922255173953,
+        "grad_norm": 0.3642658293247223,
+        "learning_rate": 2.2567683603909864e-05,
+        "loss": 0.9379,
+        "step": 28480
+      },
+      {
+        "epoch": 0.9085111132370293,
+        "grad_norm": 0.38277313113212585,
+        "learning_rate": 2.2481495980661557e-05,
+        "loss": 0.9378,
+        "step": 28490
+      },
+      {
+        "epoch": 0.9088300009566631,
+        "grad_norm": 0.3668515682220459,
+        "learning_rate": 2.239563751420805e-05,
+        "loss": 0.9369,
+        "step": 28500
+      },
+      {
+        "epoch": 0.9091488886762971,
+        "grad_norm": 0.36150866746902466,
+        "learning_rate": 2.2310106947475637e-05,
+        "loss": 0.9302,
+        "step": 28510
+      },
+      {
+        "epoch": 0.909467776395931,
+        "grad_norm": 0.3532048463821411,
+        "learning_rate": 2.2224903028191445e-05,
+        "loss": 0.9422,
+        "step": 28520
+      },
+      {
+        "epoch": 0.9097866641155649,
+        "grad_norm": 0.3641529083251953,
+        "learning_rate": 2.2140024508865157e-05,
+        "loss": 0.906,
+        "step": 28530
+      },
+      {
+        "epoch": 0.9101055518351988,
+        "grad_norm": 0.3548017740249634,
+        "learning_rate": 2.205547014677069e-05,
+        "loss": 0.9347,
+        "step": 28540
+      },
+      {
+        "epoch": 0.9104244395548328,
+        "grad_norm": 0.35412734746932983,
+        "learning_rate": 2.197123870392802e-05,
+        "loss": 0.9245,
+        "step": 28550
+      },
+      {
+        "epoch": 0.9107433272744666,
+        "grad_norm": 0.36349058151245117,
+        "learning_rate": 2.1887328947085065e-05,
+        "loss": 0.9468,
+        "step": 28560
+      },
+      {
+        "epoch": 0.9110622149941006,
+        "grad_norm": 0.3631904125213623,
+        "learning_rate": 2.1803739647699623e-05,
+        "loss": 0.9366,
+        "step": 28570
+      },
+      {
+        "epoch": 0.9113811027137345,
+        "grad_norm": 0.35928815603256226,
+        "learning_rate": 2.172046958192138e-05,
+        "loss": 0.9385,
+        "step": 28580
+      },
+      {
+        "epoch": 0.9116999904333684,
+        "grad_norm": 0.3604944944381714,
+        "learning_rate": 2.1637517530574002e-05,
+        "loss": 0.9448,
+        "step": 28590
+      },
+      {
+        "epoch": 0.9120188781530023,
+        "grad_norm": 0.38209038972854614,
+        "learning_rate": 2.155488227913727e-05,
+        "loss": 0.9502,
+        "step": 28600
+      },
+      {
+        "epoch": 0.9123377658726363,
+        "grad_norm": 0.3778879642486572,
+        "learning_rate": 2.14725626177293e-05,
+        "loss": 0.9441,
+        "step": 28610
+      },
+      {
+        "epoch": 0.9126566535922701,
+        "grad_norm": 0.36064228415489197,
+        "learning_rate": 2.139055734108886e-05,
+        "loss": 0.9304,
+        "step": 28620
+      },
+      {
+        "epoch": 0.9129755413119041,
+        "grad_norm": 0.3631671071052551,
+        "learning_rate": 2.1308865248557674e-05,
+        "loss": 0.9226,
+        "step": 28630
+      },
+      {
+        "epoch": 0.913294429031538,
+        "grad_norm": 0.35460859537124634,
+        "learning_rate": 2.122748514406288e-05,
+        "loss": 0.9311,
+        "step": 28640
+      },
+      {
+        "epoch": 0.913613316751172,
+        "grad_norm": 0.38259729743003845,
+        "learning_rate": 2.1146415836099496e-05,
+        "loss": 0.9268,
+        "step": 28650
+      },
+      {
+        "epoch": 0.9139322044708058,
+        "grad_norm": 0.37809115648269653,
+        "learning_rate": 2.1065656137712995e-05,
+        "loss": 0.9334,
+        "step": 28660
+      },
+      {
+        "epoch": 0.9142510921904398,
+        "grad_norm": 0.36083880066871643,
+        "learning_rate": 2.0985204866481902e-05,
+        "loss": 0.9367,
+        "step": 28670
+      },
+      {
+        "epoch": 0.9145699799100737,
+        "grad_norm": 0.36416465044021606,
+        "learning_rate": 2.0905060844500506e-05,
+        "loss": 0.9435,
+        "step": 28680
+      },
+      {
+        "epoch": 0.9148888676297076,
+        "grad_norm": 0.359147846698761,
+        "learning_rate": 2.0825222898361594e-05,
+        "loss": 0.9449,
+        "step": 28690
+      },
+      {
+        "epoch": 0.9152077553493415,
+        "grad_norm": 0.36608290672302246,
+        "learning_rate": 2.07456898591393e-05,
+        "loss": 0.9321,
+        "step": 28700
+      },
+      {
+        "epoch": 0.9155266430689755,
+        "grad_norm": 0.3677798807621002,
+        "learning_rate": 2.0666460562371957e-05,
+        "loss": 0.9344,
+        "step": 28710
+      },
+      {
+        "epoch": 0.9158455307886093,
+        "grad_norm": 0.3638545572757721,
+        "learning_rate": 2.0587533848045053e-05,
+        "loss": 0.9397,
+        "step": 28720
+      },
+      {
+        "epoch": 0.9161644185082433,
+        "grad_norm": 0.37101566791534424,
+        "learning_rate": 2.050890856057427e-05,
+        "loss": 0.9513,
+        "step": 28730
+      },
+      {
+        "epoch": 0.9164833062278772,
+        "grad_norm": 0.3840157091617584,
+        "learning_rate": 2.0430583548788544e-05,
+        "loss": 0.9536,
+        "step": 28740
+      },
+      {
+        "epoch": 0.9168021939475111,
+        "grad_norm": 0.3592657446861267,
+        "learning_rate": 2.0352557665913218e-05,
+        "loss": 0.9424,
+        "step": 28750
+      },
+      {
+        "epoch": 0.917121081667145,
+        "grad_norm": 0.3559933602809906,
+        "learning_rate": 2.027482976955325e-05,
+        "loss": 0.9398,
+        "step": 28760
+      },
+      {
+        "epoch": 0.917439969386779,
+        "grad_norm": 0.38258349895477295,
+        "learning_rate": 2.0197398721676495e-05,
+        "loss": 0.9341,
+        "step": 28770
+      },
+      {
+        "epoch": 0.9177588571064128,
+        "grad_norm": 0.3753907084465027,
+        "learning_rate": 2.0120263388597025e-05,
+        "loss": 0.9365,
+        "step": 28780
+      },
+      {
+        "epoch": 0.9180777448260468,
+        "grad_norm": 0.351815789937973,
+        "learning_rate": 2.004342264095854e-05,
+        "loss": 0.932,
+        "step": 28790
+      },
+      {
+        "epoch": 0.9183966325456807,
+        "grad_norm": 0.35386383533477783,
+        "learning_rate": 1.9966875353717854e-05,
+        "loss": 0.9312,
+        "step": 28800
+      },
+      {
+        "epoch": 0.9187155202653146,
+        "grad_norm": 0.36701905727386475,
+        "learning_rate": 1.9890620406128368e-05,
+        "loss": 0.9287,
+        "step": 28810
+      },
+      {
+        "epoch": 0.9190344079849485,
+        "grad_norm": 0.369667112827301,
+        "learning_rate": 1.981465668172373e-05,
+        "loss": 0.9143,
+        "step": 28820
+      },
+      {
+        "epoch": 0.9193532957045825,
+        "grad_norm": 0.36041995882987976,
+        "learning_rate": 1.9738983068301432e-05,
+        "loss": 0.928,
+        "step": 28830
+      },
+      {
+        "epoch": 0.9196721834242163,
+        "grad_norm": 0.36003410816192627,
+        "learning_rate": 1.966359845790656e-05,
+        "loss": 0.935,
+        "step": 28840
+      },
+      {
+        "epoch": 0.9199910711438503,
+        "grad_norm": 0.3556961715221405,
+        "learning_rate": 1.9588501746815556e-05,
+        "loss": 0.9373,
+        "step": 28850
+      },
+      {
+        "epoch": 0.9203099588634842,
+        "grad_norm": 0.36700302362442017,
+        "learning_rate": 1.9513691835520046e-05,
+        "loss": 0.9397,
+        "step": 28860
+      },
+      {
+        "epoch": 0.920628846583118,
+        "grad_norm": 0.36221426725387573,
+        "learning_rate": 1.9439167628710803e-05,
+        "loss": 0.9221,
+        "step": 28870
+      },
+      {
+        "epoch": 0.920947734302752,
+        "grad_norm": 0.36436378955841064,
+        "learning_rate": 1.936492803526162e-05,
+        "loss": 0.9292,
+        "step": 28880
+      },
+      {
+        "epoch": 0.921266622022386,
+        "grad_norm": 0.3724479079246521,
+        "learning_rate": 1.9290971968213404e-05,
+        "loss": 0.9406,
+        "step": 28890
+      },
+      {
+        "epoch": 0.9215855097420198,
+        "grad_norm": 0.3676651418209076,
+        "learning_rate": 1.9217298344758223e-05,
+        "loss": 0.9428,
+        "step": 28900
+      },
+      {
+        "epoch": 0.9219043974616538,
+        "grad_norm": 0.36604923009872437,
+        "learning_rate": 1.9143906086223483e-05,
+        "loss": 0.9308,
+        "step": 28910
+      },
+      {
+        "epoch": 0.9222232851812877,
+        "grad_norm": 0.35990041494369507,
+        "learning_rate": 1.907079411805611e-05,
+        "loss": 0.9362,
+        "step": 28920
+      },
+      {
+        "epoch": 0.9225421729009216,
+        "grad_norm": 0.3726346492767334,
+        "learning_rate": 1.8997961369806826e-05,
+        "loss": 0.9438,
+        "step": 28930
+      },
+      {
+        "epoch": 0.9228610606205555,
+        "grad_norm": 0.3746441900730133,
+        "learning_rate": 1.8925406775114476e-05,
+        "loss": 0.9323,
+        "step": 28940
+      },
+      {
+        "epoch": 0.9231799483401895,
+        "grad_norm": 0.3630741834640503,
+        "learning_rate": 1.8853129271690422e-05,
+        "loss": 0.9359,
+        "step": 28950
+      },
+      {
+        "epoch": 0.9234988360598233,
+        "grad_norm": 0.3569645583629608,
+        "learning_rate": 1.878112780130298e-05,
+        "loss": 0.9336,
+        "step": 28960
+      },
+      {
+        "epoch": 0.9238177237794573,
+        "grad_norm": 0.37686362862586975,
+        "learning_rate": 1.8709401309761924e-05,
+        "loss": 0.9282,
+        "step": 28970
+      },
+      {
+        "epoch": 0.9241366114990912,
+        "grad_norm": 0.37148985266685486,
+        "learning_rate": 1.863794874690307e-05,
+        "loss": 0.9229,
+        "step": 28980
+      },
+      {
+        "epoch": 0.924455499218725,
+        "grad_norm": 0.3649923503398895,
+        "learning_rate": 1.8566769066572868e-05,
+        "loss": 0.9283,
+        "step": 28990
+      },
+      {
+        "epoch": 0.924774386938359,
+        "grad_norm": 0.3782605528831482,
+        "learning_rate": 1.849586122661313e-05,
+        "loss": 0.9235,
+        "step": 29000
+      },
+      {
+        "epoch": 0.925093274657993,
+        "grad_norm": 0.3647046387195587,
+        "learning_rate": 1.842522418884572e-05,
+        "loss": 0.9334,
+        "step": 29010
+      },
+      {
+        "epoch": 0.9254121623776268,
+        "grad_norm": 0.3661080300807953,
+        "learning_rate": 1.8354856919057388e-05,
+        "loss": 0.934,
+        "step": 29020
+      },
+      {
+        "epoch": 0.9257310500972608,
+        "grad_norm": 0.36003896594047546,
+        "learning_rate": 1.8284758386984637e-05,
+        "loss": 0.9305,
+        "step": 29030
+      },
+      {
+        "epoch": 0.9260499378168947,
+        "grad_norm": 0.3618946373462677,
+        "learning_rate": 1.8214927566298603e-05,
+        "loss": 0.9215,
+        "step": 29040
+      },
+      {
+        "epoch": 0.9263688255365286,
+        "grad_norm": 0.35207855701446533,
+        "learning_rate": 1.814536343459005e-05,
+        "loss": 0.9259,
+        "step": 29050
+      },
+      {
+        "epoch": 0.9266877132561625,
+        "grad_norm": 0.3590606153011322,
+        "learning_rate": 1.8076064973354396e-05,
+        "loss": 0.9411,
+        "step": 29060
+      },
+      {
+        "epoch": 0.9270066009757965,
+        "grad_norm": 0.36371833086013794,
+        "learning_rate": 1.80070311679768e-05,
+        "loss": 0.9291,
+        "step": 29070
+      },
+      {
+        "epoch": 0.9273254886954303,
+        "grad_norm": 0.35812294483184814,
+        "learning_rate": 1.793826100771732e-05,
+        "loss": 0.9124,
+        "step": 29080
+      },
+      {
+        "epoch": 0.9276443764150643,
+        "grad_norm": 0.36259791254997253,
+        "learning_rate": 1.7869753485696093e-05,
+        "loss": 0.9206,
+        "step": 29090
+      },
+      {
+        "epoch": 0.9279632641346982,
+        "grad_norm": 0.3575153648853302,
+        "learning_rate": 1.7801507598878604e-05,
+        "loss": 0.94,
+        "step": 29100
+      },
+      {
+        "epoch": 0.928282151854332,
+        "grad_norm": 0.37916064262390137,
+        "learning_rate": 1.7733522348061006e-05,
+        "loss": 0.9586,
+        "step": 29110
+      },
+      {
+        "epoch": 0.928601039573966,
+        "grad_norm": 0.3657781779766083,
+        "learning_rate": 1.7665796737855474e-05,
+        "loss": 0.9447,
+        "step": 29120
+      },
+      {
+        "epoch": 0.9289199272936,
+        "grad_norm": 0.3646107614040375,
+        "learning_rate": 1.7598329776675653e-05,
+        "loss": 0.926,
+        "step": 29130
+      },
+      {
+        "epoch": 0.9292388150132338,
+        "grad_norm": 0.3722059428691864,
+        "learning_rate": 1.753112047672212e-05,
+        "loss": 0.9316,
+        "step": 29140
+      },
+      {
+        "epoch": 0.9295577027328678,
+        "grad_norm": 0.3507155776023865,
+        "learning_rate": 1.7464167853967936e-05,
+        "loss": 0.9156,
+        "step": 29150
+      },
+      {
+        "epoch": 0.9298765904525017,
+        "grad_norm": 0.36297523975372314,
+        "learning_rate": 1.739747092814422e-05,
+        "loss": 0.9153,
+        "step": 29160
+      },
+      {
+        "epoch": 0.9301954781721355,
+        "grad_norm": 0.37203118205070496,
+        "learning_rate": 1.7331028722725825e-05,
+        "loss": 0.9411,
+        "step": 29170
+      },
+      {
+        "epoch": 0.9305143658917695,
+        "grad_norm": 0.3708671033382416,
+        "learning_rate": 1.726484026491702e-05,
+        "loss": 0.9293,
+        "step": 29180
+      },
+      {
+        "epoch": 0.9308332536114035,
+        "grad_norm": 0.3578820824623108,
+        "learning_rate": 1.7198904585637235e-05,
+        "loss": 0.9084,
+        "step": 29190
+      },
+      {
+        "epoch": 0.9311521413310373,
+        "grad_norm": 0.37341031432151794,
+        "learning_rate": 1.713322071950692e-05,
+        "loss": 0.9341,
+        "step": 29200
+      },
+      {
+        "epoch": 0.9314710290506713,
+        "grad_norm": 0.3641612231731415,
+        "learning_rate": 1.7067787704833355e-05,
+        "loss": 0.9479,
+        "step": 29210
+      },
+      {
+        "epoch": 0.9317899167703052,
+        "grad_norm": 0.3764784634113312,
+        "learning_rate": 1.7002604583596607e-05,
+        "loss": 0.9345,
+        "step": 29220
+      },
+      {
+        "epoch": 0.932108804489939,
+        "grad_norm": 0.3550204336643219,
+        "learning_rate": 1.6937670401435477e-05,
+        "loss": 0.9257,
+        "step": 29230
+      },
+      {
+        "epoch": 0.932427692209573,
+        "grad_norm": 0.36775845289230347,
+        "learning_rate": 1.6872984207633558e-05,
+        "loss": 0.9263,
+        "step": 29240
+      },
+      {
+        "epoch": 0.932746579929207,
+        "grad_norm": 0.3655807673931122,
+        "learning_rate": 1.6808545055105283e-05,
+        "loss": 0.9296,
+        "step": 29250
+      },
+      {
+        "epoch": 0.9330654676488408,
+        "grad_norm": 0.3733557164669037,
+        "learning_rate": 1.6744352000382084e-05,
+        "loss": 0.9325,
+        "step": 29260
+      },
+      {
+        "epoch": 0.9333843553684747,
+        "grad_norm": 0.37416717410087585,
+        "learning_rate": 1.6680404103598565e-05,
+        "loss": 0.9246,
+        "step": 29270
+      },
+      {
+        "epoch": 0.9337032430881087,
+        "grad_norm": 0.3751501739025116,
+        "learning_rate": 1.6616700428478738e-05,
+        "loss": 0.9244,
+        "step": 29280
+      },
+      {
+        "epoch": 0.9340221308077425,
+        "grad_norm": 0.36550331115722656,
+        "learning_rate": 1.6553240042322332e-05,
+        "loss": 0.9258,
+        "step": 29290
+      },
+      {
+        "epoch": 0.9343410185273765,
+        "grad_norm": 0.36365216970443726,
+        "learning_rate": 1.6490022015991115e-05,
+        "loss": 0.9206,
+        "step": 29300
+      },
+      {
+        "epoch": 0.9346599062470105,
+        "grad_norm": 0.36747580766677856,
+        "learning_rate": 1.6427045423895318e-05,
+        "loss": 0.9354,
+        "step": 29310
+      },
+      {
+        "epoch": 0.9349787939666443,
+        "grad_norm": 0.3531034588813782,
+        "learning_rate": 1.636430934398004e-05,
+        "loss": 0.9283,
+        "step": 29320
+      },
+      {
+        "epoch": 0.9352976816862782,
+        "grad_norm": 0.36130332946777344,
+        "learning_rate": 1.63018128577118e-05,
+        "loss": 0.9481,
+        "step": 29330
+      },
+      {
+        "epoch": 0.9356165694059122,
+        "grad_norm": 0.3671252727508545,
+        "learning_rate": 1.623955505006505e-05,
+        "loss": 0.9184,
+        "step": 29340
+      },
+      {
+        "epoch": 0.935935457125546,
+        "grad_norm": 0.36904624104499817,
+        "learning_rate": 1.61775350095088e-05,
+        "loss": 0.9209,
+        "step": 29350
+      },
+      {
+        "epoch": 0.93625434484518,
+        "grad_norm": 0.3691551685333252,
+        "learning_rate": 1.611575182799324e-05,
+        "loss": 0.9094,
+        "step": 29360
+      },
+      {
+        "epoch": 0.936573232564814,
+        "grad_norm": 0.3752463161945343,
+        "learning_rate": 1.605420460093651e-05,
+        "loss": 0.9387,
+        "step": 29370
+      },
+      {
+        "epoch": 0.9368921202844478,
+        "grad_norm": 0.36518752574920654,
+        "learning_rate": 1.5992892427211376e-05,
+        "loss": 0.9332,
+        "step": 29380
+      },
+      {
+        "epoch": 0.9372110080040817,
+        "grad_norm": 0.37571388483047485,
+        "learning_rate": 1.5931814409132093e-05,
+        "loss": 0.9314,
+        "step": 29390
+      },
+      {
+        "epoch": 0.9375298957237157,
+        "grad_norm": 0.3575666844844818,
+        "learning_rate": 1.587096965244123e-05,
+        "loss": 0.9375,
+        "step": 29400
+      },
+      {
+        "epoch": 0.9378487834433495,
+        "grad_norm": 0.3595915138721466,
+        "learning_rate": 1.5810357266296604e-05,
+        "loss": 0.9316,
+        "step": 29410
+      },
+      {
+        "epoch": 0.9381676711629835,
+        "grad_norm": 0.36832576990127563,
+        "learning_rate": 1.5749976363258215e-05,
+        "loss": 0.936,
+        "step": 29420
+      },
+      {
+        "epoch": 0.9384865588826174,
+        "grad_norm": 0.38170093297958374,
+        "learning_rate": 1.5689826059275262e-05,
+        "loss": 0.9243,
+        "step": 29430
+      },
+      {
+        "epoch": 0.9388054466022513,
+        "grad_norm": 0.36499080061912537,
+        "learning_rate": 1.5629905473673202e-05,
+        "loss": 0.9185,
+        "step": 29440
+      },
+      {
+        "epoch": 0.9391243343218852,
+        "grad_norm": 0.3608885407447815,
+        "learning_rate": 1.5570213729140846e-05,
+        "loss": 0.9256,
+        "step": 29450
+      },
+      {
+        "epoch": 0.9394432220415192,
+        "grad_norm": 0.3616710901260376,
+        "learning_rate": 1.5510749951717523e-05,
+        "loss": 0.9342,
+        "step": 29460
+      },
+      {
+        "epoch": 0.9397621097611532,
+        "grad_norm": 0.3602309226989746,
+        "learning_rate": 1.5451513270780287e-05,
+        "loss": 0.9291,
+        "step": 29470
+      },
+      {
+        "epoch": 0.940080997480787,
+        "grad_norm": 0.3594343662261963,
+        "learning_rate": 1.539250281903115e-05,
+        "loss": 0.9269,
+        "step": 29480
+      },
+      {
+        "epoch": 0.940399885200421,
+        "grad_norm": 0.37072819471359253,
+        "learning_rate": 1.5333717732484415e-05,
+        "loss": 0.928,
+        "step": 29490
+      },
+      {
+        "epoch": 0.9407187729200549,
+        "grad_norm": 0.3712194263935089,
+        "learning_rate": 1.5275157150454e-05,
+        "loss": 0.9367,
+        "step": 29500
+      },
+      {
+        "epoch": 0.9410376606396887,
+        "grad_norm": 0.3576986789703369,
+        "learning_rate": 1.5216820215540841e-05,
+        "loss": 0.9242,
+        "step": 29510
+      },
+      {
+        "epoch": 0.9413565483593227,
+        "grad_norm": 0.3759133517742157,
+        "learning_rate": 1.5158706073620354e-05,
+        "loss": 0.9216,
+        "step": 29520
+      },
+      {
+        "epoch": 0.9416754360789567,
+        "grad_norm": 0.37520280480384827,
+        "learning_rate": 1.5100813873829904e-05,
+        "loss": 0.9113,
+        "step": 29530
+      },
+      {
+        "epoch": 0.9419943237985905,
+        "grad_norm": 0.36566489934921265,
+        "learning_rate": 1.5043142768556388e-05,
+        "loss": 0.9398,
+        "step": 29540
+      },
+      {
+        "epoch": 0.9423132115182244,
+        "grad_norm": 0.38121917843818665,
+        "learning_rate": 1.4985691913423778e-05,
+        "loss": 0.9336,
+        "step": 29550
+      },
+      {
+        "epoch": 0.9426320992378584,
+        "grad_norm": 0.3673277795314789,
+        "learning_rate": 1.4928460467280777e-05,
+        "loss": 0.9235,
+        "step": 29560
+      },
+      {
+        "epoch": 0.9429509869574922,
+        "grad_norm": 0.36724069714546204,
+        "learning_rate": 1.4871447592188524e-05,
+        "loss": 0.938,
+        "step": 29570
+      },
+      {
+        "epoch": 0.9432698746771262,
+        "grad_norm": 0.3662523925304413,
+        "learning_rate": 1.4814652453408288e-05,
+        "loss": 0.9178,
+        "step": 29580
+      },
+      {
+        "epoch": 0.9435887623967602,
+        "grad_norm": 0.37952548265457153,
+        "learning_rate": 1.4758074219389278e-05,
+        "loss": 0.9195,
+        "step": 29590
+      },
+      {
+        "epoch": 0.943907650116394,
+        "grad_norm": 0.3726538121700287,
+        "learning_rate": 1.4701712061756454e-05,
+        "loss": 0.9359,
+        "step": 29600
+      },
+      {
+        "epoch": 0.944226537836028,
+        "grad_norm": 0.36995553970336914,
+        "learning_rate": 1.4645565155298395e-05,
+        "loss": 0.9376,
+        "step": 29610
+      },
+      {
+        "epoch": 0.9445454255556619,
+        "grad_norm": 0.3546082675457001,
+        "learning_rate": 1.458963267795523e-05,
+        "loss": 0.9383,
+        "step": 29620
+      },
+      {
+        "epoch": 0.9448643132752957,
+        "grad_norm": 0.38009509444236755,
+        "learning_rate": 1.4533913810806589e-05,
+        "loss": 0.9405,
+        "step": 29630
+      },
+      {
+        "epoch": 0.9451832009949297,
+        "grad_norm": 0.37401118874549866,
+        "learning_rate": 1.4478407738059622e-05,
+        "loss": 0.9234,
+        "step": 29640
+      },
+      {
+        "epoch": 0.9455020887145636,
+        "grad_norm": 0.36730608344078064,
+        "learning_rate": 1.4423113647037045e-05,
+        "loss": 0.9182,
+        "step": 29650
+      },
+      {
+        "epoch": 0.9458209764341975,
+        "grad_norm": 0.3604319393634796,
+        "learning_rate": 1.4368030728165257e-05,
+        "loss": 0.928,
+        "step": 29660
+      },
+      {
+        "epoch": 0.9461398641538314,
+        "grad_norm": 0.3672143518924713,
+        "learning_rate": 1.4313158174962467e-05,
+        "loss": 0.9416,
+        "step": 29670
+      },
+      {
+        "epoch": 0.9464587518734654,
+        "grad_norm": 0.36120569705963135,
+        "learning_rate": 1.4258495184026909e-05,
+        "loss": 0.9331,
+        "step": 29680
+      },
+      {
+        "epoch": 0.9467776395930992,
+        "grad_norm": 0.3656829297542572,
+        "learning_rate": 1.4204040955025053e-05,
+        "loss": 0.9283,
+        "step": 29690
+      },
+      {
+        "epoch": 0.9470965273127332,
+        "grad_norm": 0.3733628988265991,
+        "learning_rate": 1.4149794690679904e-05,
+        "loss": 0.9242,
+        "step": 29700
+      },
+      {
+        "epoch": 0.9474154150323671,
+        "grad_norm": 0.35865920782089233,
+        "learning_rate": 1.409575559675934e-05,
+        "loss": 0.9078,
+        "step": 29710
+      },
+      {
+        "epoch": 0.947734302752001,
+        "grad_norm": 0.3627816438674927,
+        "learning_rate": 1.4041922882064455e-05,
+        "loss": 0.9204,
+        "step": 29720
+      },
+      {
+        "epoch": 0.9480531904716349,
+        "grad_norm": 0.37665456533432007,
+        "learning_rate": 1.398829575841799e-05,
+        "loss": 0.9425,
+        "step": 29730
+      },
+      {
+        "epoch": 0.9483720781912689,
+        "grad_norm": 0.36541375517845154,
+        "learning_rate": 1.3934873440652796e-05,
+        "loss": 0.9157,
+        "step": 29740
+      },
+      {
+        "epoch": 0.9486909659109027,
+        "grad_norm": 0.3650599420070648,
+        "learning_rate": 1.3881655146600332e-05,
+        "loss": 0.9322,
+        "step": 29750
+      },
+      {
+        "epoch": 0.9490098536305367,
+        "grad_norm": 0.35816624760627747,
+        "learning_rate": 1.3828640097079218e-05,
+        "loss": 0.9298,
+        "step": 29760
+      },
+      {
+        "epoch": 0.9493287413501706,
+        "grad_norm": 0.3787805140018463,
+        "learning_rate": 1.377582751588382e-05,
+        "loss": 0.926,
+        "step": 29770
+      },
+      {
+        "epoch": 0.9496476290698045,
+        "grad_norm": 0.3747621476650238,
+        "learning_rate": 1.3723216629772897e-05,
+        "loss": 0.9266,
+        "step": 29780
+      },
+      {
+        "epoch": 0.9499665167894384,
+        "grad_norm": 0.3703758716583252,
+        "learning_rate": 1.3670806668458264e-05,
+        "loss": 0.9331,
+        "step": 29790
+      },
+      {
+        "epoch": 0.9502854045090724,
+        "grad_norm": 0.3748500347137451,
+        "learning_rate": 1.3618596864593529e-05,
+        "loss": 0.9279,
+        "step": 29800
+      },
+      {
+        "epoch": 0.9506042922287062,
+        "grad_norm": 0.36413055658340454,
+        "learning_rate": 1.3566586453762848e-05,
+        "loss": 0.9146,
+        "step": 29810
+      },
+      {
+        "epoch": 0.9509231799483402,
+        "grad_norm": 0.3669073283672333,
+        "learning_rate": 1.3514774674469737e-05,
+        "loss": 0.9376,
+        "step": 29820
+      },
+      {
+        "epoch": 0.9512420676679741,
+        "grad_norm": 0.3571595847606659,
+        "learning_rate": 1.3463160768125927e-05,
+        "loss": 0.9411,
+        "step": 29830
+      },
+      {
+        "epoch": 0.951560955387608,
+        "grad_norm": 0.3696027398109436,
+        "learning_rate": 1.3411743979040244e-05,
+        "loss": 0.9342,
+        "step": 29840
+      },
+      {
+        "epoch": 0.9518798431072419,
+        "grad_norm": 0.3621768057346344,
+        "learning_rate": 1.3360523554407562e-05,
+        "loss": 0.9339,
+        "step": 29850
+      },
+      {
+        "epoch": 0.9521987308268759,
+        "grad_norm": 0.3710097372531891,
+        "learning_rate": 1.3309498744297768e-05,
+        "loss": 0.922,
+        "step": 29860
+      },
+      {
+        "epoch": 0.9525176185465097,
+        "grad_norm": 0.3726564347743988,
+        "learning_rate": 1.3258668801644778e-05,
+        "loss": 0.9394,
+        "step": 29870
+      },
+      {
+        "epoch": 0.9528365062661437,
+        "grad_norm": 0.35227444767951965,
+        "learning_rate": 1.3208032982235637e-05,
+        "loss": 0.9212,
+        "step": 29880
+      },
+      {
+        "epoch": 0.9531553939857776,
+        "grad_norm": 0.3657701313495636,
+        "learning_rate": 1.315759054469956e-05,
+        "loss": 0.9183,
+        "step": 29890
+      },
+      {
+        "epoch": 0.9534742817054115,
+        "grad_norm": 0.36237528920173645,
+        "learning_rate": 1.3107340750497128e-05,
+        "loss": 0.9242,
+        "step": 29900
+      },
+      {
+        "epoch": 0.9537931694250454,
+        "grad_norm": 0.35844841599464417,
+        "learning_rate": 1.3057282863909452e-05,
+        "loss": 0.9318,
+        "step": 29910
+      },
+      {
+        "epoch": 0.9541120571446794,
+        "grad_norm": 0.36002638936042786,
+        "learning_rate": 1.3007416152027412e-05,
+        "loss": 0.9337,
+        "step": 29920
+      },
+      {
+        "epoch": 0.9544309448643132,
+        "grad_norm": 0.3549569547176361,
+        "learning_rate": 1.2957739884740917e-05,
+        "loss": 0.9228,
+        "step": 29930
+      },
+      {
+        "epoch": 0.9547498325839472,
+        "grad_norm": 0.3807508945465088,
+        "learning_rate": 1.290825333472822e-05,
+        "loss": 0.9276,
+        "step": 29940
+      },
+      {
+        "epoch": 0.9550687203035811,
+        "grad_norm": 0.37096288800239563,
+        "learning_rate": 1.2858955777445266e-05,
+        "loss": 0.9555,
+        "step": 29950
+      },
+      {
+        "epoch": 0.955387608023215,
+        "grad_norm": 0.36706507205963135,
+        "learning_rate": 1.2809846491115096e-05,
+        "loss": 0.9261,
+        "step": 29960
+      },
+      {
+        "epoch": 0.9557064957428489,
+        "grad_norm": 0.3656713366508484,
+        "learning_rate": 1.276092475671726e-05,
+        "loss": 0.937,
+        "step": 29970
+      },
+      {
+        "epoch": 0.9560253834624829,
+        "grad_norm": 0.36048075556755066,
+        "learning_rate": 1.27121898579773e-05,
+        "loss": 0.9251,
+        "step": 29980
+      },
+      {
+        "epoch": 0.9563442711821167,
+        "grad_norm": 0.3719075620174408,
+        "learning_rate": 1.266364108135627e-05,
+        "loss": 0.9472,
+        "step": 29990
+      },
+      {
+        "epoch": 0.9566631589017507,
+        "grad_norm": 0.3709539771080017,
+        "learning_rate": 1.2615277716040274e-05,
+        "loss": 0.9418,
+        "step": 30000
+      },
+      {
+        "epoch": 0.9569820466213846,
+        "grad_norm": 0.3729771673679352,
+        "learning_rate": 1.2567099053930065e-05,
+        "loss": 0.943,
+        "step": 30010
+      },
+      {
+        "epoch": 0.9573009343410185,
+        "grad_norm": 0.37757793068885803,
+        "learning_rate": 1.2519104389630684e-05,
+        "loss": 0.9169,
+        "step": 30020
+      },
+      {
+        "epoch": 0.9576198220606524,
+        "grad_norm": 0.3567389249801636,
+        "learning_rate": 1.2471293020441117e-05,
+        "loss": 0.9265,
+        "step": 30030
+      },
+      {
+        "epoch": 0.9579387097802864,
+        "grad_norm": 0.36267825961112976,
+        "learning_rate": 1.2423664246344036e-05,
+        "loss": 0.9259,
+        "step": 30040
+      },
+      {
+        "epoch": 0.9582575974999202,
+        "grad_norm": 0.3645220398902893,
+        "learning_rate": 1.2376217369995511e-05,
+        "loss": 0.9184,
+        "step": 30050
+      },
+      {
+        "epoch": 0.9585764852195542,
+        "grad_norm": 0.37258005142211914,
+        "learning_rate": 1.2328951696714822e-05,
+        "loss": 0.946,
+        "step": 30060
+      },
+      {
+        "epoch": 0.9588953729391881,
+        "grad_norm": 0.3623313903808594,
+        "learning_rate": 1.2281866534474292e-05,
+        "loss": 0.9436,
+        "step": 30070
+      },
+      {
+        "epoch": 0.959214260658822,
+        "grad_norm": 0.36983102560043335,
+        "learning_rate": 1.2234961193889144e-05,
+        "loss": 0.9276,
+        "step": 30080
+      },
+      {
+        "epoch": 0.9595331483784559,
+        "grad_norm": 0.37520304322242737,
+        "learning_rate": 1.218823498820741e-05,
+        "loss": 0.9289,
+        "step": 30090
+      },
+      {
+        "epoch": 0.9598520360980899,
+        "grad_norm": 0.37535175681114197,
+        "learning_rate": 1.214168723329988e-05,
+        "loss": 0.9358,
+        "step": 30100
+      },
+      {
+        "epoch": 0.9601709238177237,
+        "grad_norm": 0.3650328516960144,
+        "learning_rate": 1.2095317247650083e-05,
+        "loss": 0.9339,
+        "step": 30110
+      },
+      {
+        "epoch": 0.9604898115373577,
+        "grad_norm": 0.36983540654182434,
+        "learning_rate": 1.204912435234431e-05,
+        "loss": 0.9241,
+        "step": 30120
+      },
+      {
+        "epoch": 0.9608086992569916,
+        "grad_norm": 0.3635700047016144,
+        "learning_rate": 1.200310787106167e-05,
+        "loss": 0.9177,
+        "step": 30130
+      },
+      {
+        "epoch": 0.9611275869766255,
+        "grad_norm": 0.37635406851768494,
+        "learning_rate": 1.195726713006419e-05,
+        "loss": 0.9353,
+        "step": 30140
+      },
+      {
+        "epoch": 0.9614464746962594,
+        "grad_norm": 0.36136430501937866,
+        "learning_rate": 1.1911601458186958e-05,
+        "loss": 0.9326,
+        "step": 30150
+      },
+      {
+        "epoch": 0.9617653624158934,
+        "grad_norm": 0.36432045698165894,
+        "learning_rate": 1.186611018682828e-05,
+        "loss": 0.9207,
+        "step": 30160
+      },
+      {
+        "epoch": 0.9620842501355272,
+        "grad_norm": 0.3685867488384247,
+        "learning_rate": 1.1820792649939912e-05,
+        "loss": 0.9199,
+        "step": 30170
+      },
+      {
+        "epoch": 0.9624031378551612,
+        "grad_norm": 0.3692206144332886,
+        "learning_rate": 1.1775648184017282e-05,
+        "loss": 0.9319,
+        "step": 30180
+      },
+      {
+        "epoch": 0.9627220255747951,
+        "grad_norm": 0.37184929847717285,
+        "learning_rate": 1.1730676128089802e-05,
+        "loss": 0.9388,
+        "step": 30190
+      },
+      {
+        "epoch": 0.963040913294429,
+        "grad_norm": 0.3655369281768799,
+        "learning_rate": 1.1685875823711168e-05,
+        "loss": 0.9248,
+        "step": 30200
+      },
+      {
+        "epoch": 0.9633598010140629,
+        "grad_norm": 0.3684496283531189,
+        "learning_rate": 1.164124661494975e-05,
+        "loss": 0.9341,
+        "step": 30210
+      },
+      {
+        "epoch": 0.9636786887336969,
+        "grad_norm": 0.36519452929496765,
+        "learning_rate": 1.1596787848378949e-05,
+        "loss": 0.9407,
+        "step": 30220
+      },
+      {
+        "epoch": 0.9639975764533307,
+        "grad_norm": 0.3807597756385803,
+        "learning_rate": 1.1552498873067655e-05,
+        "loss": 0.9397,
+        "step": 30230
+      },
+      {
+        "epoch": 0.9643164641729647,
+        "grad_norm": 0.36671775579452515,
+        "learning_rate": 1.1508379040570714e-05,
+        "loss": 0.9339,
+        "step": 30240
+      },
+      {
+        "epoch": 0.9646353518925986,
+        "grad_norm": 0.3577711582183838,
+        "learning_rate": 1.146442770491943e-05,
+        "loss": 0.9075,
+        "step": 30250
+      },
+      {
+        "epoch": 0.9649542396122325,
+        "grad_norm": 0.3761603832244873,
+        "learning_rate": 1.1420644222612106e-05,
+        "loss": 0.9202,
+        "step": 30260
+      },
+      {
+        "epoch": 0.9652731273318664,
+        "grad_norm": 0.36671191453933716,
+        "learning_rate": 1.1377027952604628e-05,
+        "loss": 0.9265,
+        "step": 30270
+      },
+      {
+        "epoch": 0.9655920150515004,
+        "grad_norm": 0.3775249421596527,
+        "learning_rate": 1.1333578256301075e-05,
+        "loss": 0.9136,
+        "step": 30280
+      },
+      {
+        "epoch": 0.9659109027711343,
+        "grad_norm": 0.3641510307788849,
+        "learning_rate": 1.129029449754437e-05,
+        "loss": 0.9238,
+        "step": 30290
+      },
+      {
+        "epoch": 0.9662297904907682,
+        "grad_norm": 0.3587871491909027,
+        "learning_rate": 1.1247176042606964e-05,
+        "loss": 0.9374,
+        "step": 30300
+      },
+      {
+        "epoch": 0.9665486782104021,
+        "grad_norm": 0.3657750189304352,
+        "learning_rate": 1.1204222260181564e-05,
+        "loss": 0.9234,
+        "step": 30310
+      },
+      {
+        "epoch": 0.9668675659300361,
+        "grad_norm": 0.367485374212265,
+        "learning_rate": 1.1161432521371883e-05,
+        "loss": 0.9302,
+        "step": 30320
+      },
+      {
+        "epoch": 0.9671864536496699,
+        "grad_norm": 0.3634987771511078,
+        "learning_rate": 1.1118806199683434e-05,
+        "loss": 0.9283,
+        "step": 30330
+      },
+      {
+        "epoch": 0.9675053413693039,
+        "grad_norm": 0.3690277338027954,
+        "learning_rate": 1.1076342671014357e-05,
+        "loss": 0.9502,
+        "step": 30340
+      },
+      {
+        "epoch": 0.9678242290889378,
+        "grad_norm": 0.39022156596183777,
+        "learning_rate": 1.1034041313646285e-05,
+        "loss": 0.944,
+        "step": 30350
+      },
+      {
+        "epoch": 0.9681431168085717,
+        "grad_norm": 0.3651580214500427,
+        "learning_rate": 1.099190150823524e-05,
+        "loss": 0.9317,
+        "step": 30360
+      },
+      {
+        "epoch": 0.9684620045282056,
+        "grad_norm": 0.37234464287757874,
+        "learning_rate": 1.0949922637802553e-05,
+        "loss": 0.9482,
+        "step": 30370
+      },
+      {
+        "epoch": 0.9687808922478396,
+        "grad_norm": 0.36375316977500916,
+        "learning_rate": 1.0908104087725861e-05,
+        "loss": 0.9191,
+        "step": 30380
+      },
+      {
+        "epoch": 0.9690997799674734,
+        "grad_norm": 0.36848706007003784,
+        "learning_rate": 1.0866445245730072e-05,
+        "loss": 0.927,
+        "step": 30390
+      },
+      {
+        "epoch": 0.9694186676871074,
+        "grad_norm": 0.3552470803260803,
+        "learning_rate": 1.0824945501878422e-05,
+        "loss": 0.9389,
+        "step": 30400
+      },
+      {
+        "epoch": 0.9697375554067413,
+        "grad_norm": 0.3650756776332855,
+        "learning_rate": 1.0783604248563538e-05,
+        "loss": 0.9215,
+        "step": 30410
+      },
+      {
+        "epoch": 0.9700564431263752,
+        "grad_norm": 0.36942705512046814,
+        "learning_rate": 1.0742420880498549e-05,
+        "loss": 0.9352,
+        "step": 30420
+      },
+      {
+        "epoch": 0.9703753308460091,
+        "grad_norm": 0.3720155656337738,
+        "learning_rate": 1.0701394794708213e-05,
+        "loss": 0.9347,
+        "step": 30430
+      },
+      {
+        "epoch": 0.9706942185656431,
+        "grad_norm": 0.365630567073822,
+        "learning_rate": 1.0660525390520096e-05,
+        "loss": 0.9179,
+        "step": 30440
+      },
+      {
+        "epoch": 0.9710131062852769,
+        "grad_norm": 0.3534587621688843,
+        "learning_rate": 1.0619812069555778e-05,
+        "loss": 0.933,
+        "step": 30450
+      },
+      {
+        "epoch": 0.9713319940049109,
+        "grad_norm": 0.38552242517471313,
+        "learning_rate": 1.0579254235722086e-05,
+        "loss": 0.9374,
+        "step": 30460
+      },
+      {
+        "epoch": 0.9716508817245448,
+        "grad_norm": 0.35881030559539795,
+        "learning_rate": 1.0538851295202372e-05,
+        "loss": 0.9409,
+        "step": 30470
+      },
+      {
+        "epoch": 0.9719697694441787,
+        "grad_norm": 0.3738090693950653,
+        "learning_rate": 1.0498602656447817e-05,
+        "loss": 0.9397,
+        "step": 30480
+      },
+      {
+        "epoch": 0.9722886571638126,
+        "grad_norm": 0.3673436939716339,
+        "learning_rate": 1.0458507730168771e-05,
+        "loss": 0.9377,
+        "step": 30490
+      },
+      {
+        "epoch": 0.9726075448834466,
+        "grad_norm": 0.35112425684928894,
+        "learning_rate": 1.0418565929326121e-05,
+        "loss": 0.9298,
+        "step": 30500
+      },
+      {
+        "epoch": 0.9729264326030804,
+        "grad_norm": 0.36641091108322144,
+        "learning_rate": 1.0378776669122702e-05,
+        "loss": 0.934,
+        "step": 30510
+      },
+      {
+        "epoch": 0.9732453203227144,
+        "grad_norm": 0.36989322304725647,
+        "learning_rate": 1.0339139366994728e-05,
+        "loss": 0.9395,
+        "step": 30520
+      },
+      {
+        "epoch": 0.9735642080423483,
+        "grad_norm": 0.36697059869766235,
+        "learning_rate": 1.0299653442603272e-05,
+        "loss": 0.9317,
+        "step": 30530
+      },
+      {
+        "epoch": 0.9738830957619822,
+        "grad_norm": 0.36741960048675537,
+        "learning_rate": 1.0260318317825752e-05,
+        "loss": 0.9241,
+        "step": 30540
+      },
+      {
+        "epoch": 0.9742019834816161,
+        "grad_norm": 0.36927348375320435,
+        "learning_rate": 1.0221133416747503e-05,
+        "loss": 0.9392,
+        "step": 30550
+      },
+      {
+        "epoch": 0.9745208712012501,
+        "grad_norm": 0.37633517384529114,
+        "learning_rate": 1.0182098165653291e-05,
+        "loss": 0.9382,
+        "step": 30560
+      },
+      {
+        "epoch": 0.9748397589208839,
+        "grad_norm": 0.3640388548374176,
+        "learning_rate": 1.014321199301896e-05,
+        "loss": 0.9333,
+        "step": 30570
+      },
+      {
+        "epoch": 0.9751586466405179,
+        "grad_norm": 0.3635682165622711,
+        "learning_rate": 1.0104474329503038e-05,
+        "loss": 0.9291,
+        "step": 30580
+      },
+      {
+        "epoch": 0.9754775343601518,
+        "grad_norm": 0.3714246153831482,
+        "learning_rate": 1.0065884607938414e-05,
+        "loss": 0.9402,
+        "step": 30590
+      },
+      {
+        "epoch": 0.9757964220797857,
+        "grad_norm": 0.37632328271865845,
+        "learning_rate": 1.0027442263324029e-05,
+        "loss": 0.9568,
+        "step": 30600
+      },
+      {
+        "epoch": 0.9761153097994196,
+        "grad_norm": 0.3554982542991638,
+        "learning_rate": 9.989146732816599e-06,
+        "loss": 0.9331,
+        "step": 30610
+      },
+      {
+        "epoch": 0.9764341975190536,
+        "grad_norm": 0.36082297563552856,
+        "learning_rate": 9.950997455722386e-06,
+        "loss": 0.9289,
+        "step": 30620
+      },
+      {
+        "epoch": 0.9767530852386874,
+        "grad_norm": 0.36781561374664307,
+        "learning_rate": 9.912993873488982e-06,
+        "loss": 0.9246,
+        "step": 30630
+      },
+      {
+        "epoch": 0.9770719729583214,
+        "grad_norm": 0.376779705286026,
+        "learning_rate": 9.875135429697123e-06,
+        "loss": 0.9382,
+        "step": 30640
+      },
+      {
+        "epoch": 0.9773908606779553,
+        "grad_norm": 0.3738979399204254,
+        "learning_rate": 9.83742157005256e-06,
+        "loss": 0.938,
+        "step": 30650
+      },
+      {
+        "epoch": 0.9777097483975892,
+        "grad_norm": 0.3823069930076599,
+        "learning_rate": 9.79985174237793e-06,
+        "loss": 0.9239,
+        "step": 30660
+      },
+      {
+        "epoch": 0.9780286361172231,
+        "grad_norm": 0.3665352165699005,
+        "learning_rate": 9.762425396604675e-06,
+        "loss": 0.9287,
+        "step": 30670
+      },
+      {
+        "epoch": 0.9783475238368571,
+        "grad_norm": 0.367678165435791,
+        "learning_rate": 9.72514198476499e-06,
+        "loss": 0.9297,
+        "step": 30680
+      },
+      {
+        "epoch": 0.9786664115564909,
+        "grad_norm": 0.3722160756587982,
+        "learning_rate": 9.688000960983798e-06,
+        "loss": 0.9356,
+        "step": 30690
+      },
+      {
+        "epoch": 0.9789852992761249,
+        "grad_norm": 0.3580091595649719,
+        "learning_rate": 9.65100178147076e-06,
+        "loss": 0.9336,
+        "step": 30700
+      },
+      {
+        "epoch": 0.9793041869957588,
+        "grad_norm": 0.36442312598228455,
+        "learning_rate": 9.614143904512304e-06,
+        "loss": 0.9423,
+        "step": 30710
+      },
+      {
+        "epoch": 0.9796230747153927,
+        "grad_norm": 0.3623570501804352,
+        "learning_rate": 9.577426790463718e-06,
+        "loss": 0.9267,
+        "step": 30720
+      },
+      {
+        "epoch": 0.9799419624350266,
+        "grad_norm": 0.3609250485897064,
+        "learning_rate": 9.540849901741222e-06,
+        "loss": 0.9132,
+        "step": 30730
+      },
+      {
+        "epoch": 0.9802608501546606,
+        "grad_norm": 0.3623647093772888,
+        "learning_rate": 9.504412702814105e-06,
+        "loss": 0.9124,
+        "step": 30740
+      },
+      {
+        "epoch": 0.9805797378742944,
+        "grad_norm": 0.354446142911911,
+        "learning_rate": 9.468114660196888e-06,
+        "loss": 0.9293,
+        "step": 30750
+      },
+      {
+        "epoch": 0.9808986255939284,
+        "grad_norm": 0.3737424910068512,
+        "learning_rate": 9.431955242441515e-06,
+        "loss": 0.9363,
+        "step": 30760
+      },
+      {
+        "epoch": 0.9812175133135623,
+        "grad_norm": 0.3702537715435028,
+        "learning_rate": 9.395933920129571e-06,
+        "loss": 0.9238,
+        "step": 30770
+      },
+      {
+        "epoch": 0.9815364010331962,
+        "grad_norm": 0.3564508855342865,
+        "learning_rate": 9.360050165864518e-06,
+        "loss": 0.9313,
+        "step": 30780
+      },
+      {
+        "epoch": 0.9818552887528301,
+        "grad_norm": 0.38061270117759705,
+        "learning_rate": 9.324303454263998e-06,
+        "loss": 0.9226,
+        "step": 30790
+      },
+      {
+        "epoch": 0.9821741764724641,
+        "grad_norm": 0.3773099482059479,
+        "learning_rate": 9.288693261952112e-06,
+        "loss": 0.9213,
+        "step": 30800
+      },
+      {
+        "epoch": 0.9824930641920979,
+        "grad_norm": 0.3710877597332001,
+        "learning_rate": 9.253219067551782e-06,
+        "loss": 0.9158,
+        "step": 30810
+      },
+      {
+        "epoch": 0.9828119519117319,
+        "grad_norm": 0.35554859042167664,
+        "learning_rate": 9.217880351677104e-06,
+        "loss": 0.9177,
+        "step": 30820
+      },
+      {
+        "epoch": 0.9831308396313658,
+        "grad_norm": 0.36319994926452637,
+        "learning_rate": 9.182676596925743e-06,
+        "loss": 0.9222,
+        "step": 30830
+      },
+      {
+        "epoch": 0.9834497273509997,
+        "grad_norm": 0.3607344329357147,
+        "learning_rate": 9.147607287871367e-06,
+        "loss": 0.9144,
+        "step": 30840
+      },
+      {
+        "epoch": 0.9837686150706336,
+        "grad_norm": 0.3698658049106598,
+        "learning_rate": 9.112671911056089e-06,
+        "loss": 0.9272,
+        "step": 30850
+      },
+      {
+        "epoch": 0.9840875027902676,
+        "grad_norm": 0.36637309193611145,
+        "learning_rate": 9.077869954982961e-06,
+        "loss": 0.9225,
+        "step": 30860
+      },
+      {
+        "epoch": 0.9844063905099014,
+        "grad_norm": 0.35287508368492126,
+        "learning_rate": 9.043200910108472e-06,
+        "loss": 0.9217,
+        "step": 30870
+      },
+      {
+        "epoch": 0.9847252782295354,
+        "grad_norm": 0.3669984042644501,
+        "learning_rate": 9.008664268835097e-06,
+        "loss": 0.9206,
+        "step": 30880
+      },
+      {
+        "epoch": 0.9850441659491693,
+        "grad_norm": 0.3705748915672302,
+        "learning_rate": 8.97425952550387e-06,
+        "loss": 0.9284,
+        "step": 30890
+      },
+      {
+        "epoch": 0.9853630536688032,
+        "grad_norm": 0.36321109533309937,
+        "learning_rate": 8.939986176386965e-06,
+        "loss": 0.9291,
+        "step": 30900
+      },
+      {
+        "epoch": 0.9856819413884371,
+        "grad_norm": 0.36526602506637573,
+        "learning_rate": 8.90584371968033e-06,
+        "loss": 0.9173,
+        "step": 30910
+      },
+      {
+        "epoch": 0.9860008291080711,
+        "grad_norm": 0.3631839454174042,
+        "learning_rate": 8.871831655496341e-06,
+        "loss": 0.932,
+        "step": 30920
+      },
+      {
+        "epoch": 0.9863197168277049,
+        "grad_norm": 0.3734815716743469,
+        "learning_rate": 8.837949485856484e-06,
+        "loss": 0.9255,
+        "step": 30930
+      },
+      {
+        "epoch": 0.9866386045473389,
+        "grad_norm": 0.35923781991004944,
+        "learning_rate": 8.804196714684051e-06,
+        "loss": 0.9157,
+        "step": 30940
+      },
+      {
+        "epoch": 0.9869574922669728,
+        "grad_norm": 0.375,
+        "learning_rate": 8.770572847796902e-06,
+        "loss": 0.9278,
+        "step": 30950
+      },
+      {
+        "epoch": 0.9872763799866067,
+        "grad_norm": 0.3728904128074646,
+        "learning_rate": 8.737077392900202e-06,
+        "loss": 0.9347,
+        "step": 30960
+      },
+      {
+        "epoch": 0.9875952677062406,
+        "grad_norm": 0.37896835803985596,
+        "learning_rate": 8.703709859579232e-06,
+        "loss": 0.9285,
+        "step": 30970
+      },
+      {
+        "epoch": 0.9879141554258746,
+        "grad_norm": 0.3802531659603119,
+        "learning_rate": 8.670469759292197e-06,
+        "loss": 0.9317,
+        "step": 30980
+      },
+      {
+        "epoch": 0.9882330431455084,
+        "grad_norm": 0.37327489256858826,
+        "learning_rate": 8.637356605363086e-06,
+        "loss": 0.9468,
+        "step": 30990
+      },
+      {
+        "epoch": 0.9885519308651424,
+        "grad_norm": 0.364010751247406,
+        "learning_rate": 8.604369912974537e-06,
+        "loss": 0.9214,
+        "step": 31000
+      },
+      {
+        "epoch": 0.9888708185847763,
+        "grad_norm": 0.3760271370410919,
+        "learning_rate": 8.571509199160733e-06,
+        "loss": 0.9575,
+        "step": 31010
+      },
+      {
+        "epoch": 0.9891897063044102,
+        "grad_norm": 0.3785746097564697,
+        "learning_rate": 8.53877398280035e-06,
+        "loss": 0.9258,
+        "step": 31020
+      },
+      {
+        "epoch": 0.9895085940240441,
+        "grad_norm": 0.3546142876148224,
+        "learning_rate": 8.506163784609493e-06,
+        "loss": 0.9279,
+        "step": 31030
+      },
+      {
+        "epoch": 0.9898274817436781,
+        "grad_norm": 0.3674685060977936,
+        "learning_rate": 8.47367812713469e-06,
+        "loss": 0.9417,
+        "step": 31040
+      },
+      {
+        "epoch": 0.9901463694633119,
+        "grad_norm": 0.3799147307872772,
+        "learning_rate": 8.441316534745896e-06,
+        "loss": 0.9308,
+        "step": 31050
+      },
+      {
+        "epoch": 0.9904652571829459,
+        "grad_norm": 0.36110377311706543,
+        "learning_rate": 8.409078533629542e-06,
+        "loss": 0.9188,
+        "step": 31060
+      },
+      {
+        "epoch": 0.9907841449025798,
+        "grad_norm": 0.3820250928401947,
+        "learning_rate": 8.376963651781573e-06,
+        "loss": 0.9259,
+        "step": 31070
+      },
+      {
+        "epoch": 0.9911030326222137,
+        "grad_norm": 0.3705957531929016,
+        "learning_rate": 8.34497141900056e-06,
+        "loss": 0.9231,
+        "step": 31080
+      },
+      {
+        "epoch": 0.9914219203418476,
+        "grad_norm": 0.3767108917236328,
+        "learning_rate": 8.313101366880808e-06,
+        "loss": 0.9255,
+        "step": 31090
+      },
+      {
+        "epoch": 0.9917408080614816,
+        "grad_norm": 0.36577731370925903,
+        "learning_rate": 8.28135302880549e-06,
+        "loss": 0.926,
+        "step": 31100
+      },
+      {
+        "epoch": 0.9920596957811154,
+        "grad_norm": 0.3568488359451294,
+        "learning_rate": 8.24972593993983e-06,
+        "loss": 0.908,
+        "step": 31110
+      },
+      {
+        "epoch": 0.9923785835007494,
+        "grad_norm": 0.36216244101524353,
+        "learning_rate": 8.21821963722429e-06,
+        "loss": 0.9278,
+        "step": 31120
+      },
+      {
+        "epoch": 0.9926974712203833,
+        "grad_norm": 0.37096574902534485,
+        "learning_rate": 8.186833659367786e-06,
+        "loss": 0.9247,
+        "step": 31130
+      },
+      {
+        "epoch": 0.9930163589400173,
+        "grad_norm": 0.36498740315437317,
+        "learning_rate": 8.155567546840937e-06,
+        "loss": 0.9356,
+        "step": 31140
+      },
+      {
+        "epoch": 0.9933352466596511,
+        "grad_norm": 0.3782957196235657,
+        "learning_rate": 8.124420841869342e-06,
+        "loss": 0.9269,
+        "step": 31150
+      },
+      {
+        "epoch": 0.9936541343792851,
+        "grad_norm": 0.36367443203926086,
+        "learning_rate": 8.093393088426874e-06,
+        "loss": 0.9232,
+        "step": 31160
+      },
+      {
+        "epoch": 0.993973022098919,
+        "grad_norm": 0.35848551988601685,
+        "learning_rate": 8.062483832229001e-06,
+        "loss": 0.9172,
+        "step": 31170
+      },
+      {
+        "epoch": 0.9942919098185529,
+        "grad_norm": 0.3594017028808594,
+        "learning_rate": 8.031692620726136e-06,
+        "loss": 0.9167,
+        "step": 31180
+      },
+      {
+        "epoch": 0.9946107975381868,
+        "grad_norm": 0.3806443214416504,
+        "learning_rate": 8.001019003097014e-06,
+        "loss": 0.9188,
+        "step": 31190
+      },
+      {
+        "epoch": 0.9949296852578208,
+        "grad_norm": 0.3811398148536682,
+        "learning_rate": 7.970462530242083e-06,
+        "loss": 0.9247,
+        "step": 31200
+      },
+      {
+        "epoch": 0.9952485729774546,
+        "grad_norm": 0.37172773480415344,
+        "learning_rate": 7.940022754776945e-06,
+        "loss": 0.9483,
+        "step": 31210
+      },
+      {
+        "epoch": 0.9955674606970886,
+        "grad_norm": 0.3615531921386719,
+        "learning_rate": 7.90969923102579e-06,
+        "loss": 0.9141,
+        "step": 31220
+      },
+      {
+        "epoch": 0.9958863484167225,
+        "grad_norm": 0.3517124354839325,
+        "learning_rate": 7.879491515014875e-06,
+        "loss": 0.9185,
+        "step": 31230
+      },
+      {
+        "epoch": 0.9962052361363564,
+        "grad_norm": 0.37004709243774414,
+        "learning_rate": 7.849399164466028e-06,
+        "loss": 0.9141,
+        "step": 31240
+      },
+      {
+        "epoch": 0.9965241238559903,
+        "grad_norm": 0.3684113919734955,
+        "learning_rate": 7.819421738790164e-06,
+        "loss": 0.934,
+        "step": 31250
+      },
+      {
+        "epoch": 0.9968430115756243,
+        "grad_norm": 0.364093154668808,
+        "learning_rate": 7.78955879908085e-06,
+        "loss": 0.9143,
+        "step": 31260
+      },
+      {
+        "epoch": 0.9971618992952581,
+        "grad_norm": 0.3692590892314911,
+        "learning_rate": 7.759809908107856e-06,
+        "loss": 0.9348,
+        "step": 31270
+      },
+      {
+        "epoch": 0.9974807870148921,
+        "grad_norm": 0.37234070897102356,
+        "learning_rate": 7.730174630310774e-06,
+        "loss": 0.9188,
+        "step": 31280
+      },
+      {
+        "epoch": 0.997799674734526,
+        "grad_norm": 0.3636535406112671,
+        "learning_rate": 7.700652531792634e-06,
+        "loss": 0.9163,
+        "step": 31290
+      },
+      {
+        "epoch": 0.9981185624541599,
+        "grad_norm": 0.37809842824935913,
+        "learning_rate": 7.671243180313544e-06,
+        "loss": 0.9327,
+        "step": 31300
+      },
+      {
+        "epoch": 0.9984374501737938,
+        "grad_norm": 0.36570221185684204,
+        "learning_rate": 7.641946145284373e-06,
+        "loss": 0.9096,
+        "step": 31310
+      },
+      {
+        "epoch": 0.9987563378934278,
+        "grad_norm": 0.36101430654525757,
+        "learning_rate": 7.612760997760436e-06,
+        "loss": 0.927,
+        "step": 31320
+      },
+      {
+        "epoch": 0.9990752256130616,
+        "grad_norm": 0.3641717731952667,
+        "learning_rate": 7.583687310435224e-06,
+        "loss": 0.9353,
+        "step": 31330
+      },
+      {
+        "epoch": 0.9993941133326956,
+        "grad_norm": 0.36067697405815125,
+        "learning_rate": 7.5547246576341395e-06,
+        "loss": 0.911,
+        "step": 31340
+      },
+      {
+        "epoch": 0.9997130010523295,
+        "grad_norm": 0.371115118265152,
+        "learning_rate": 7.525872615308268e-06,
+        "loss": 0.9357,
+        "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": 5.46535369518422e+17,
+    "train_batch_size": 512,
+    "trial_name": null,
+    "trial_params": null
+  }
+  
\ No newline at end of file