diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21978 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 31359,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003188877196339169,
+      "grad_norm": 3.036648750305176,
+      "learning_rate": 3.1888771963391685e-05,
+      "loss": 9.533,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006377754392678338,
+      "grad_norm": 1.8982125520706177,
+      "learning_rate": 6.377754392678337e-05,
+      "loss": 8.4311,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009566631589017507,
+      "grad_norm": 1.0077489614486694,
+      "learning_rate": 9.566631589017505e-05,
+      "loss": 7.5347,
+      "step": 30
+    },
+    {
+      "epoch": 0.0012755508785356677,
+      "grad_norm": 0.7245733737945557,
+      "learning_rate": 0.00012755508785356674,
+      "loss": 6.8032,
+      "step": 40
+    },
+    {
+      "epoch": 0.0015944385981695845,
+      "grad_norm": 0.6618061065673828,
+      "learning_rate": 0.00015944385981695843,
+      "loss": 6.3497,
+      "step": 50
+    },
+    {
+      "epoch": 0.0019133263178035013,
+      "grad_norm": 0.9375441074371338,
+      "learning_rate": 0.0001913326317803501,
+      "loss": 6.1371,
+      "step": 60
+    },
+    {
+      "epoch": 0.0022322140374374183,
+      "grad_norm": 0.8749001622200012,
+      "learning_rate": 0.00022322140374374182,
+      "loss": 5.9688,
+      "step": 70
+    },
+    {
+      "epoch": 0.0025511017570713354,
+      "grad_norm": 0.7610182762145996,
+      "learning_rate": 0.0002551101757071335,
+      "loss": 5.802,
+      "step": 80
+    },
+    {
+      "epoch": 0.002869989476705252,
+      "grad_norm": 0.6044188737869263,
+      "learning_rate": 0.0002869989476705252,
+      "loss": 5.6593,
+      "step": 90
+    },
+    {
+      "epoch": 0.003188877196339169,
+      "grad_norm": 0.8063629865646362,
+      "learning_rate": 0.00031888771963391687,
+      "loss": 5.5041,
+      "step": 100
+    },
+    {
+      "epoch": 0.003507764915973086,
+      "grad_norm": 0.776926577091217,
+      "learning_rate": 0.00035077649159730856,
+      "loss": 5.3573,
+      "step": 110
+    },
+    {
+      "epoch": 0.0038266526356070026,
+      "grad_norm": 0.7012314200401306,
+      "learning_rate": 0.0003826652635607002,
+      "loss": 5.1943,
+      "step": 120
+    },
+    {
+      "epoch": 0.00414554035524092,
+      "grad_norm": 0.8802277445793152,
+      "learning_rate": 0.00041455403552409194,
+      "loss": 5.0659,
+      "step": 130
+    },
+    {
+      "epoch": 0.004464428074874837,
+      "grad_norm": 1.1483536958694458,
+      "learning_rate": 0.00044644280748748364,
+      "loss": 4.8957,
+      "step": 140
+    },
+    {
+      "epoch": 0.004783315794508754,
+      "grad_norm": 0.7169669270515442,
+      "learning_rate": 0.0004783315794508753,
+      "loss": 4.7524,
+      "step": 150
+    },
+    {
+      "epoch": 0.005102203514142671,
+      "grad_norm": 0.6984596848487854,
+      "learning_rate": 0.000510220351414267,
+      "loss": 4.6166,
+      "step": 160
+    },
+    {
+      "epoch": 0.005421091233776587,
+      "grad_norm": 0.8619917035102844,
+      "learning_rate": 0.0005421091233776587,
+      "loss": 4.5516,
+      "step": 170
+    },
+    {
+      "epoch": 0.005739978953410504,
+      "grad_norm": 0.9251404404640198,
+      "learning_rate": 0.0005739978953410504,
+      "loss": 4.4415,
+      "step": 180
+    },
+    {
+      "epoch": 0.006058866673044421,
+      "grad_norm": 0.9755675792694092,
+      "learning_rate": 0.0006058866673044421,
+      "loss": 4.3531,
+      "step": 190
+    },
+    {
+      "epoch": 0.006377754392678338,
+      "grad_norm": 0.7827635407447815,
+      "learning_rate": 0.0006377754392678337,
+      "loss": 4.2329,
+      "step": 200
+    },
+    {
+      "epoch": 0.006696642112312255,
+      "grad_norm": 1.1084445714950562,
+      "learning_rate": 0.0006696642112312254,
+      "loss": 4.165,
+      "step": 210
+    },
+    {
+      "epoch": 0.007015529831946172,
+      "grad_norm": 0.7396318912506104,
+      "learning_rate": 0.0007015529831946171,
+      "loss": 4.0607,
+      "step": 220
+    },
+    {
+      "epoch": 0.007334417551580088,
+      "grad_norm": 0.8335429430007935,
+      "learning_rate": 0.0007334417551580089,
+      "loss": 3.9912,
+      "step": 230
+    },
+    {
+      "epoch": 0.007653305271214005,
+      "grad_norm": 0.7607409954071045,
+      "learning_rate": 0.0007653305271214004,
+      "loss": 3.9232,
+      "step": 240
+    },
+    {
+      "epoch": 0.007972192990847922,
+      "grad_norm": 0.7046091556549072,
+      "learning_rate": 0.0007972192990847921,
+      "loss": 3.8883,
+      "step": 250
+    },
+    {
+      "epoch": 0.00829108071048184,
+      "grad_norm": 0.858257532119751,
+      "learning_rate": 0.0008291080710481839,
+      "loss": 3.8052,
+      "step": 260
+    },
+    {
+      "epoch": 0.008609968430115756,
+      "grad_norm": 0.7207130789756775,
+      "learning_rate": 0.0008609968430115756,
+      "loss": 3.7916,
+      "step": 270
+    },
+    {
+      "epoch": 0.008928856149749673,
+      "grad_norm": 0.8144423961639404,
+      "learning_rate": 0.0008928856149749673,
+      "loss": 3.7006,
+      "step": 280
+    },
+    {
+      "epoch": 0.00924774386938359,
+      "grad_norm": 0.824480414390564,
+      "learning_rate": 0.0009247743869383588,
+      "loss": 3.6646,
+      "step": 290
+    },
+    {
+      "epoch": 0.009566631589017507,
+      "grad_norm": 0.8660809397697449,
+      "learning_rate": 0.0009566631589017505,
+      "loss": 3.6161,
+      "step": 300
+    },
+    {
+      "epoch": 0.009885519308651424,
+      "grad_norm": 0.7533829808235168,
+      "learning_rate": 0.0009885519308651422,
+      "loss": 3.5742,
+      "step": 310
+    },
+    {
+      "epoch": 0.010204407028285341,
+      "grad_norm": 0.7015078067779541,
+      "learning_rate": 0.001,
+      "loss": 3.5328,
+      "step": 320
+    },
+    {
+      "epoch": 0.010523294747919257,
+      "grad_norm": 0.7448273301124573,
+      "learning_rate": 0.001,
+      "loss": 3.4694,
+      "step": 330
+    },
+    {
+      "epoch": 0.010842182467553174,
+      "grad_norm": 0.7898711562156677,
+      "learning_rate": 0.001,
+      "loss": 3.4442,
+      "step": 340
+    },
+    {
+      "epoch": 0.01116107018718709,
+      "grad_norm": 0.9917442202568054,
+      "learning_rate": 0.001,
+      "loss": 3.3712,
+      "step": 350
+    },
+    {
+      "epoch": 0.011479957906821008,
+      "grad_norm": 0.8575316071510315,
+      "learning_rate": 0.001,
+      "loss": 3.3555,
+      "step": 360
+    },
+    {
+      "epoch": 0.011798845626454925,
+      "grad_norm": 0.7570849061012268,
+      "learning_rate": 0.001,
+      "loss": 3.2849,
+      "step": 370
+    },
+    {
+      "epoch": 0.012117733346088842,
+      "grad_norm": 0.7412615418434143,
+      "learning_rate": 0.001,
+      "loss": 3.2602,
+      "step": 380
+    },
+    {
+      "epoch": 0.012436621065722759,
+      "grad_norm": 0.6936795711517334,
+      "learning_rate": 0.001,
+      "loss": 3.2217,
+      "step": 390
+    },
+    {
+      "epoch": 0.012755508785356676,
+      "grad_norm": 0.8034941554069519,
+      "learning_rate": 0.001,
+      "loss": 3.1868,
+      "step": 400
+    },
+    {
+      "epoch": 0.013074396504990593,
+      "grad_norm": 0.7071384787559509,
+      "learning_rate": 0.001,
+      "loss": 3.1318,
+      "step": 410
+    },
+    {
+      "epoch": 0.01339328422462451,
+      "grad_norm": 0.7512100338935852,
+      "learning_rate": 0.001,
+      "loss": 3.1087,
+      "step": 420
+    },
+    {
+      "epoch": 0.013712171944258427,
+      "grad_norm": 0.8786042928695679,
+      "learning_rate": 0.001,
+      "loss": 3.0601,
+      "step": 430
+    },
+    {
+      "epoch": 0.014031059663892344,
+      "grad_norm": 0.7732911705970764,
+      "learning_rate": 0.001,
+      "loss": 3.0246,
+      "step": 440
+    },
+    {
+      "epoch": 0.014349947383526261,
+      "grad_norm": 0.8299092054367065,
+      "learning_rate": 0.001,
+      "loss": 2.9614,
+      "step": 450
+    },
+    {
+      "epoch": 0.014668835103160176,
+      "grad_norm": 0.6985538005828857,
+      "learning_rate": 0.001,
+      "loss": 2.9618,
+      "step": 460
+    },
+    {
+      "epoch": 0.014987722822794093,
+      "grad_norm": 0.8050011396408081,
+      "learning_rate": 0.001,
+      "loss": 2.9392,
+      "step": 470
+    },
+    {
+      "epoch": 0.01530661054242801,
+      "grad_norm": 0.7188740372657776,
+      "learning_rate": 0.001,
+      "loss": 2.908,
+      "step": 480
+    },
+    {
+      "epoch": 0.01562549826206193,
+      "grad_norm": 0.9172284603118896,
+      "learning_rate": 0.001,
+      "loss": 2.8388,
+      "step": 490
+    },
+    {
+      "epoch": 0.015944385981695845,
+      "grad_norm": 0.7237478494644165,
+      "learning_rate": 0.001,
+      "loss": 2.841,
+      "step": 500
+    },
+    {
+      "epoch": 0.016263273701329763,
+      "grad_norm": 0.8718260526657104,
+      "learning_rate": 0.001,
+      "loss": 2.7981,
+      "step": 510
+    },
+    {
+      "epoch": 0.01658216142096368,
+      "grad_norm": 0.7130722999572754,
+      "learning_rate": 0.001,
+      "loss": 2.7755,
+      "step": 520
+    },
+    {
+      "epoch": 0.016901049140597594,
+      "grad_norm": 0.6788746118545532,
+      "learning_rate": 0.001,
+      "loss": 2.7742,
+      "step": 530
+    },
+    {
+      "epoch": 0.017219936860231513,
+      "grad_norm": 0.7213516235351562,
+      "learning_rate": 0.001,
+      "loss": 2.7421,
+      "step": 540
+    },
+    {
+      "epoch": 0.017538824579865428,
+      "grad_norm": 0.7722727060317993,
+      "learning_rate": 0.001,
+      "loss": 2.6919,
+      "step": 550
+    },
+    {
+      "epoch": 0.017857712299499347,
+      "grad_norm": 0.8629282712936401,
+      "learning_rate": 0.001,
+      "loss": 2.6804,
+      "step": 560
+    },
+    {
+      "epoch": 0.018176600019133262,
+      "grad_norm": 0.7852650880813599,
+      "learning_rate": 0.001,
+      "loss": 2.6518,
+      "step": 570
+    },
+    {
+      "epoch": 0.01849548773876718,
+      "grad_norm": 0.6630326509475708,
+      "learning_rate": 0.001,
+      "loss": 2.6447,
+      "step": 580
+    },
+    {
+      "epoch": 0.018814375458401096,
+      "grad_norm": 0.6685476899147034,
+      "learning_rate": 0.001,
+      "loss": 2.599,
+      "step": 590
+    },
+    {
+      "epoch": 0.019133263178035015,
+      "grad_norm": 0.7120856642723083,
+      "learning_rate": 0.001,
+      "loss": 2.5967,
+      "step": 600
+    },
+    {
+      "epoch": 0.01945215089766893,
+      "grad_norm": 0.6978277564048767,
+      "learning_rate": 0.001,
+      "loss": 2.5731,
+      "step": 610
+    },
+    {
+      "epoch": 0.01977103861730285,
+      "grad_norm": 0.6799705624580383,
+      "learning_rate": 0.001,
+      "loss": 2.5141,
+      "step": 620
+    },
+    {
+      "epoch": 0.020089926336936764,
+      "grad_norm": 0.625628650188446,
+      "learning_rate": 0.001,
+      "loss": 2.5176,
+      "step": 630
+    },
+    {
+      "epoch": 0.020408814056570683,
+      "grad_norm": 0.6856659054756165,
+      "learning_rate": 0.001,
+      "loss": 2.5204,
+      "step": 640
+    },
+    {
+      "epoch": 0.0207277017762046,
+      "grad_norm": 0.6310821175575256,
+      "learning_rate": 0.001,
+      "loss": 2.4801,
+      "step": 650
+    },
+    {
+      "epoch": 0.021046589495838514,
+      "grad_norm": 0.6268860697746277,
+      "learning_rate": 0.001,
+      "loss": 2.4809,
+      "step": 660
+    },
+    {
+      "epoch": 0.021365477215472432,
+      "grad_norm": 0.6450563073158264,
+      "learning_rate": 0.001,
+      "loss": 2.4476,
+      "step": 670
+    },
+    {
+      "epoch": 0.021684364935106348,
+      "grad_norm": 0.650497317314148,
+      "learning_rate": 0.001,
+      "loss": 2.4374,
+      "step": 680
+    },
+    {
+      "epoch": 0.022003252654740266,
+      "grad_norm": 0.6631786227226257,
+      "learning_rate": 0.001,
+      "loss": 2.4252,
+      "step": 690
+    },
+    {
+      "epoch": 0.02232214037437418,
+      "grad_norm": 0.6338786482810974,
+      "learning_rate": 0.001,
+      "loss": 2.3957,
+      "step": 700
+    },
+    {
+      "epoch": 0.0226410280940081,
+      "grad_norm": 0.6474817991256714,
+      "learning_rate": 0.001,
+      "loss": 2.424,
+      "step": 710
+    },
+    {
+      "epoch": 0.022959915813642016,
+      "grad_norm": 0.6207141280174255,
+      "learning_rate": 0.001,
+      "loss": 2.3646,
+      "step": 720
+    },
+    {
+      "epoch": 0.023278803533275935,
+      "grad_norm": 0.708824098110199,
+      "learning_rate": 0.001,
+      "loss": 2.3651,
+      "step": 730
+    },
+    {
+      "epoch": 0.02359769125290985,
+      "grad_norm": 0.6238510608673096,
+      "learning_rate": 0.001,
+      "loss": 2.3275,
+      "step": 740
+    },
+    {
+      "epoch": 0.02391657897254377,
+      "grad_norm": 0.6039829254150391,
+      "learning_rate": 0.001,
+      "loss": 2.3622,
+      "step": 750
+    },
+    {
+      "epoch": 0.024235466692177684,
+      "grad_norm": 0.6189019083976746,
+      "learning_rate": 0.001,
+      "loss": 2.3433,
+      "step": 760
+    },
+    {
+      "epoch": 0.024554354411811603,
+      "grad_norm": 0.5963674187660217,
+      "learning_rate": 0.001,
+      "loss": 2.3492,
+      "step": 770
+    },
+    {
+      "epoch": 0.024873242131445518,
+      "grad_norm": 0.6473168730735779,
+      "learning_rate": 0.001,
+      "loss": 2.2995,
+      "step": 780
+    },
+    {
+      "epoch": 0.025192129851079433,
+      "grad_norm": 0.590446412563324,
+      "learning_rate": 0.001,
+      "loss": 2.3136,
+      "step": 790
+    },
+    {
+      "epoch": 0.025511017570713352,
+      "grad_norm": 0.6440556049346924,
+      "learning_rate": 0.001,
+      "loss": 2.333,
+      "step": 800
+    },
+    {
+      "epoch": 0.025829905290347267,
+      "grad_norm": 0.5967952013015747,
+      "learning_rate": 0.001,
+      "loss": 2.2718,
+      "step": 810
+    },
+    {
+      "epoch": 0.026148793009981186,
+      "grad_norm": 0.6391497850418091,
+      "learning_rate": 0.001,
+      "loss": 2.3109,
+      "step": 820
+    },
+    {
+      "epoch": 0.0264676807296151,
+      "grad_norm": 0.5951425433158875,
+      "learning_rate": 0.001,
+      "loss": 2.2615,
+      "step": 830
+    },
+    {
+      "epoch": 0.02678656844924902,
+      "grad_norm": 0.7214036583900452,
+      "learning_rate": 0.001,
+      "loss": 2.2651,
+      "step": 840
+    },
+    {
+      "epoch": 0.027105456168882935,
+      "grad_norm": 0.5951161980628967,
+      "learning_rate": 0.001,
+      "loss": 2.2469,
+      "step": 850
+    },
+    {
+      "epoch": 0.027424343888516854,
+      "grad_norm": 0.643730103969574,
+      "learning_rate": 0.001,
+      "loss": 2.2456,
+      "step": 860
+    },
+    {
+      "epoch": 0.02774323160815077,
+      "grad_norm": 0.6100723147392273,
+      "learning_rate": 0.001,
+      "loss": 2.2127,
+      "step": 870
+    },
+    {
+      "epoch": 0.028062119327784688,
+      "grad_norm": 0.6163387894630432,
+      "learning_rate": 0.001,
+      "loss": 2.215,
+      "step": 880
+    },
+    {
+      "epoch": 0.028381007047418604,
+      "grad_norm": 0.6332610845565796,
+      "learning_rate": 0.001,
+      "loss": 2.1957,
+      "step": 890
+    },
+    {
+      "epoch": 0.028699894767052522,
+      "grad_norm": 0.7613614201545715,
+      "learning_rate": 0.001,
+      "loss": 2.2013,
+      "step": 900
+    },
+    {
+      "epoch": 0.029018782486686438,
+      "grad_norm": 0.6165406703948975,
+      "learning_rate": 0.001,
+      "loss": 2.1597,
+      "step": 910
+    },
+    {
+      "epoch": 0.029337670206320353,
+      "grad_norm": 0.5806939005851746,
+      "learning_rate": 0.001,
+      "loss": 2.1633,
+      "step": 920
+    },
+    {
+      "epoch": 0.02965655792595427,
+      "grad_norm": 0.5569039583206177,
+      "learning_rate": 0.001,
+      "loss": 2.1748,
+      "step": 930
+    },
+    {
+      "epoch": 0.029975445645588187,
+      "grad_norm": 0.5546286106109619,
+      "learning_rate": 0.001,
+      "loss": 2.1717,
+      "step": 940
+    },
+    {
+      "epoch": 0.030294333365222106,
+      "grad_norm": 0.5814953446388245,
+      "learning_rate": 0.001,
+      "loss": 2.1502,
+      "step": 950
+    },
+    {
+      "epoch": 0.03061322108485602,
+      "grad_norm": 0.5773137807846069,
+      "learning_rate": 0.001,
+      "loss": 2.1499,
+      "step": 960
+    },
+    {
+      "epoch": 0.03093210880448994,
+      "grad_norm": 0.5882891416549683,
+      "learning_rate": 0.001,
+      "loss": 2.1352,
+      "step": 970
+    },
+    {
+      "epoch": 0.03125099652412386,
+      "grad_norm": 0.597629964351654,
+      "learning_rate": 0.001,
+      "loss": 2.1569,
+      "step": 980
+    },
+    {
+      "epoch": 0.03156988424375777,
+      "grad_norm": 0.5862744450569153,
+      "learning_rate": 0.001,
+      "loss": 2.1362,
+      "step": 990
+    },
+    {
+      "epoch": 0.03188877196339169,
+      "grad_norm": 0.5462473630905151,
+      "learning_rate": 0.001,
+      "loss": 2.115,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03220765968302561,
+      "grad_norm": 0.5664170384407043,
+      "learning_rate": 0.001,
+      "loss": 2.114,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03252654740265953,
+      "grad_norm": 0.6027910113334656,
+      "learning_rate": 0.001,
+      "loss": 2.1258,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03284543512229344,
+      "grad_norm": 0.6072728633880615,
+      "learning_rate": 0.001,
+      "loss": 2.0963,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03316432284192736,
+      "grad_norm": 0.5642842650413513,
+      "learning_rate": 0.001,
+      "loss": 2.0512,
+      "step": 1040
+    },
+    {
+      "epoch": 0.033483210561561276,
+      "grad_norm": 0.5518870949745178,
+      "learning_rate": 0.001,
+      "loss": 2.0815,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03380209828119519,
+      "grad_norm": 0.5724993348121643,
+      "learning_rate": 0.001,
+      "loss": 2.0478,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03412098600082911,
+      "grad_norm": 0.576343834400177,
+      "learning_rate": 0.001,
+      "loss": 2.0811,
+      "step": 1070
+    },
+    {
+      "epoch": 0.034439873720463025,
+      "grad_norm": 0.5714786052703857,
+      "learning_rate": 0.001,
+      "loss": 2.0707,
+      "step": 1080
+    },
+    {
+      "epoch": 0.034758761440096944,
+      "grad_norm": 0.5584716796875,
+      "learning_rate": 0.001,
+      "loss": 2.0623,
+      "step": 1090
+    },
+    {
+      "epoch": 0.035077649159730856,
+      "grad_norm": 0.6107367277145386,
+      "learning_rate": 0.001,
+      "loss": 2.0665,
+      "step": 1100
+    },
+    {
+      "epoch": 0.035396536879364775,
+      "grad_norm": 0.5658422112464905,
+      "learning_rate": 0.001,
+      "loss": 2.0287,
+      "step": 1110
+    },
+    {
+      "epoch": 0.035715424598998693,
+      "grad_norm": 0.6898009777069092,
+      "learning_rate": 0.001,
+      "loss": 2.0313,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03603431231863261,
+      "grad_norm": 0.5509852766990662,
+      "learning_rate": 0.001,
+      "loss": 2.0321,
+      "step": 1130
+    },
+    {
+      "epoch": 0.036353200038266524,
+      "grad_norm": 0.5971260070800781,
+      "learning_rate": 0.001,
+      "loss": 2.049,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03667208775790044,
+      "grad_norm": 0.5642650127410889,
+      "learning_rate": 0.001,
+      "loss": 2.0325,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03699097547753436,
+      "grad_norm": 0.5502352118492126,
+      "learning_rate": 0.001,
+      "loss": 2.0189,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03730986319716828,
+      "grad_norm": 0.5586209297180176,
+      "learning_rate": 0.001,
+      "loss": 1.9837,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03762875091680219,
+      "grad_norm": 0.5707424879074097,
+      "learning_rate": 0.001,
+      "loss": 1.9626,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03794763863643611,
+      "grad_norm": 0.513467013835907,
+      "learning_rate": 0.001,
+      "loss": 1.9747,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03826652635607003,
+      "grad_norm": 0.5255717039108276,
+      "learning_rate": 0.001,
+      "loss": 1.9826,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03858541407570394,
+      "grad_norm": 0.5995370745658875,
+      "learning_rate": 0.001,
+      "loss": 1.9606,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03890430179533786,
+      "grad_norm": 0.5771000385284424,
+      "learning_rate": 0.001,
+      "loss": 1.9829,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03922318951497178,
+      "grad_norm": 0.5424252152442932,
+      "learning_rate": 0.001,
+      "loss": 1.9634,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0395420772346057,
+      "grad_norm": 0.6467852592468262,
+      "learning_rate": 0.001,
+      "loss": 1.975,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03986096495423961,
+      "grad_norm": 0.5047982931137085,
+      "learning_rate": 0.001,
+      "loss": 1.9443,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04017985267387353,
+      "grad_norm": 0.5370773077011108,
+      "learning_rate": 0.001,
+      "loss": 1.9455,
+      "step": 1260
+    },
+    {
+      "epoch": 0.04049874039350745,
+      "grad_norm": 0.5400688648223877,
+      "learning_rate": 0.001,
+      "loss": 1.9559,
+      "step": 1270
+    },
+    {
+      "epoch": 0.040817628113141366,
+      "grad_norm": 0.49548593163490295,
+      "learning_rate": 0.001,
+      "loss": 1.9353,
+      "step": 1280
+    },
+    {
+      "epoch": 0.04113651583277528,
+      "grad_norm": 0.5320776104927063,
+      "learning_rate": 0.001,
+      "loss": 1.9386,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0414554035524092,
+      "grad_norm": 0.516173243522644,
+      "learning_rate": 0.001,
+      "loss": 1.9209,
+      "step": 1300
+    },
+    {
+      "epoch": 0.041774291272043115,
+      "grad_norm": 0.5142052173614502,
+      "learning_rate": 0.001,
+      "loss": 1.9009,
+      "step": 1310
+    },
+    {
+      "epoch": 0.04209317899167703,
+      "grad_norm": 0.5320322513580322,
+      "learning_rate": 0.001,
+      "loss": 1.94,
+      "step": 1320
+    },
+    {
+      "epoch": 0.042412066711310946,
+      "grad_norm": 0.5104788541793823,
+      "learning_rate": 0.001,
+      "loss": 1.9197,
+      "step": 1330
+    },
+    {
+      "epoch": 0.042730954430944865,
+      "grad_norm": 0.5254496932029724,
+      "learning_rate": 0.001,
+      "loss": 1.9007,
+      "step": 1340
+    },
+    {
+      "epoch": 0.04304984215057878,
+      "grad_norm": 0.5628265738487244,
+      "learning_rate": 0.001,
+      "loss": 1.9229,
+      "step": 1350
+    },
+    {
+      "epoch": 0.043368729870212695,
+      "grad_norm": 0.5209566950798035,
+      "learning_rate": 0.001,
+      "loss": 1.9035,
+      "step": 1360
+    },
+    {
+      "epoch": 0.043687617589846614,
+      "grad_norm": 0.5213086605072021,
+      "learning_rate": 0.001,
+      "loss": 1.9135,
+      "step": 1370
+    },
+    {
+      "epoch": 0.04400650530948053,
+      "grad_norm": 0.5099833011627197,
+      "learning_rate": 0.001,
+      "loss": 1.9162,
+      "step": 1380
+    },
+    {
+      "epoch": 0.04432539302911445,
+      "grad_norm": 0.5214992761611938,
+      "learning_rate": 0.001,
+      "loss": 1.8639,
+      "step": 1390
+    },
+    {
+      "epoch": 0.04464428074874836,
+      "grad_norm": 0.53340744972229,
+      "learning_rate": 0.001,
+      "loss": 1.8961,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04496316846838228,
+      "grad_norm": 0.4767109453678131,
+      "learning_rate": 0.001,
+      "loss": 1.8945,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0452820561880162,
+      "grad_norm": 0.5631620287895203,
+      "learning_rate": 0.001,
+      "loss": 1.8566,
+      "step": 1420
+    },
+    {
+      "epoch": 0.04560094390765012,
+      "grad_norm": 0.5479603409767151,
+      "learning_rate": 0.001,
+      "loss": 1.8817,
+      "step": 1430
+    },
+    {
+      "epoch": 0.04591983162728403,
+      "grad_norm": 0.49137023091316223,
+      "learning_rate": 0.001,
+      "loss": 1.8552,
+      "step": 1440
+    },
+    {
+      "epoch": 0.04623871934691795,
+      "grad_norm": 0.5349081158638,
+      "learning_rate": 0.001,
+      "loss": 1.8952,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04655760706655187,
+      "grad_norm": 0.4876307249069214,
+      "learning_rate": 0.001,
+      "loss": 1.8497,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04687649478618578,
+      "grad_norm": 0.4965214729309082,
+      "learning_rate": 0.001,
+      "loss": 1.8491,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0471953825058197,
+      "grad_norm": 0.5440576672554016,
+      "learning_rate": 0.001,
+      "loss": 1.8561,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04751427022545362,
+      "grad_norm": 0.48483118414878845,
+      "learning_rate": 0.001,
+      "loss": 1.8507,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04783315794508754,
+      "grad_norm": 0.4601515233516693,
+      "learning_rate": 0.001,
+      "loss": 1.8341,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04815204566472145,
+      "grad_norm": 0.5070135593414307,
+      "learning_rate": 0.001,
+      "loss": 1.8136,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04847093338435537,
+      "grad_norm": 0.4945167899131775,
+      "learning_rate": 0.001,
+      "loss": 1.8423,
+      "step": 1520
+    },
+    {
+      "epoch": 0.048789821103989287,
+      "grad_norm": 0.4869058132171631,
+      "learning_rate": 0.001,
+      "loss": 1.8465,
+      "step": 1530
+    },
+    {
+      "epoch": 0.049108708823623205,
+      "grad_norm": 0.5155792832374573,
+      "learning_rate": 0.001,
+      "loss": 1.8085,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04942759654325712,
+      "grad_norm": 0.48565489053726196,
+      "learning_rate": 0.001,
+      "loss": 1.8332,
+      "step": 1550
+    },
+    {
+      "epoch": 0.049746484262891036,
+      "grad_norm": 0.5483407974243164,
+      "learning_rate": 0.001,
+      "loss": 1.8151,
+      "step": 1560
+    },
+    {
+      "epoch": 0.050065371982524955,
+      "grad_norm": 0.473037451505661,
+      "learning_rate": 0.001,
+      "loss": 1.8127,
+      "step": 1570
+    },
+    {
+      "epoch": 0.050384259702158866,
+      "grad_norm": 0.48309481143951416,
+      "learning_rate": 0.001,
+      "loss": 1.792,
+      "step": 1580
+    },
+    {
+      "epoch": 0.050703147421792785,
+      "grad_norm": 0.4604003131389618,
+      "learning_rate": 0.001,
+      "loss": 1.7976,
+      "step": 1590
+    },
+    {
+      "epoch": 0.051022035141426704,
+      "grad_norm": 0.5088821649551392,
+      "learning_rate": 0.001,
+      "loss": 1.7901,
+      "step": 1600
+    },
+    {
+      "epoch": 0.05134092286106062,
+      "grad_norm": 0.46633222699165344,
+      "learning_rate": 0.001,
+      "loss": 1.7978,
+      "step": 1610
+    },
+    {
+      "epoch": 0.051659810580694535,
+      "grad_norm": 0.48440852761268616,
+      "learning_rate": 0.001,
+      "loss": 1.8161,
+      "step": 1620
+    },
+    {
+      "epoch": 0.05197869830032845,
+      "grad_norm": 0.46742019057273865,
+      "learning_rate": 0.001,
+      "loss": 1.7755,
+      "step": 1630
+    },
+    {
+      "epoch": 0.05229758601996237,
+      "grad_norm": 0.4699628949165344,
+      "learning_rate": 0.001,
+      "loss": 1.7872,
+      "step": 1640
+    },
+    {
+      "epoch": 0.05261647373959629,
+      "grad_norm": 0.46953830122947693,
+      "learning_rate": 0.001,
+      "loss": 1.8104,
+      "step": 1650
+    },
+    {
+      "epoch": 0.0529353614592302,
+      "grad_norm": 0.4661422073841095,
+      "learning_rate": 0.001,
+      "loss": 1.7835,
+      "step": 1660
+    },
+    {
+      "epoch": 0.05325424917886412,
+      "grad_norm": 0.4410492479801178,
+      "learning_rate": 0.001,
+      "loss": 1.7827,
+      "step": 1670
+    },
+    {
+      "epoch": 0.05357313689849804,
+      "grad_norm": 0.4911010265350342,
+      "learning_rate": 0.001,
+      "loss": 1.7439,
+      "step": 1680
+    },
+    {
+      "epoch": 0.05389202461813196,
+      "grad_norm": 0.48298680782318115,
+      "learning_rate": 0.001,
+      "loss": 1.7601,
+      "step": 1690
+    },
+    {
+      "epoch": 0.05421091233776587,
+      "grad_norm": 0.482377827167511,
+      "learning_rate": 0.001,
+      "loss": 1.769,
+      "step": 1700
+    },
+    {
+      "epoch": 0.05452980005739979,
+      "grad_norm": 0.45547738671302795,
+      "learning_rate": 0.001,
+      "loss": 1.7547,
+      "step": 1710
+    },
+    {
+      "epoch": 0.05484868777703371,
+      "grad_norm": 0.4433269798755646,
+      "learning_rate": 0.001,
+      "loss": 1.7642,
+      "step": 1720
+    },
+    {
+      "epoch": 0.05516757549666762,
+      "grad_norm": 0.4870554804801941,
+      "learning_rate": 0.001,
+      "loss": 1.756,
+      "step": 1730
+    },
+    {
+      "epoch": 0.05548646321630154,
+      "grad_norm": 0.5187873840332031,
+      "learning_rate": 0.001,
+      "loss": 1.7672,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05580535093593546,
+      "grad_norm": 0.4401336908340454,
+      "learning_rate": 0.001,
+      "loss": 1.7468,
+      "step": 1750
+    },
+    {
+      "epoch": 0.056124238655569376,
+      "grad_norm": 0.45578229427337646,
+      "learning_rate": 0.001,
+      "loss": 1.7545,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05644312637520329,
+      "grad_norm": 0.46454453468322754,
+      "learning_rate": 0.001,
+      "loss": 1.7371,
+      "step": 1770
+    },
+    {
+      "epoch": 0.05676201409483721,
+      "grad_norm": 0.4536410868167877,
+      "learning_rate": 0.001,
+      "loss": 1.7634,
+      "step": 1780
+    },
+    {
+      "epoch": 0.057080901814471126,
+      "grad_norm": 0.4952872395515442,
+      "learning_rate": 0.001,
+      "loss": 1.7373,
+      "step": 1790
+    },
+    {
+      "epoch": 0.057399789534105045,
+      "grad_norm": 0.47220513224601746,
+      "learning_rate": 0.001,
+      "loss": 1.7348,
+      "step": 1800
+    },
+    {
+      "epoch": 0.057718677253738956,
+      "grad_norm": 0.44693368673324585,
+      "learning_rate": 0.001,
+      "loss": 1.7569,
+      "step": 1810
+    },
+    {
+      "epoch": 0.058037564973372875,
+      "grad_norm": 0.4495336711406708,
+      "learning_rate": 0.001,
+      "loss": 1.7362,
+      "step": 1820
+    },
+    {
+      "epoch": 0.058356452693006794,
+      "grad_norm": 0.44833892583847046,
+      "learning_rate": 0.001,
+      "loss": 1.7259,
+      "step": 1830
+    },
+    {
+      "epoch": 0.058675340412640706,
+      "grad_norm": 0.44210535287857056,
+      "learning_rate": 0.001,
+      "loss": 1.7251,
+      "step": 1840
+    },
+    {
+      "epoch": 0.058994228132274625,
+      "grad_norm": 0.45084482431411743,
+      "learning_rate": 0.001,
+      "loss": 1.7018,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05931311585190854,
+      "grad_norm": 0.45676374435424805,
+      "learning_rate": 0.001,
+      "loss": 1.7216,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05963200357154246,
+      "grad_norm": 0.45795685052871704,
+      "learning_rate": 0.001,
+      "loss": 1.723,
+      "step": 1870
+    },
+    {
+      "epoch": 0.059950891291176374,
+      "grad_norm": 0.42781588435173035,
+      "learning_rate": 0.001,
+      "loss": 1.7298,
+      "step": 1880
+    },
+    {
+      "epoch": 0.06026977901081029,
+      "grad_norm": 0.44484594464302063,
+      "learning_rate": 0.001,
+      "loss": 1.7131,
+      "step": 1890
+    },
+    {
+      "epoch": 0.06058866673044421,
+      "grad_norm": 0.49411898851394653,
+      "learning_rate": 0.001,
+      "loss": 1.7175,
+      "step": 1900
+    },
+    {
+      "epoch": 0.06090755445007813,
+      "grad_norm": 0.43948128819465637,
+      "learning_rate": 0.001,
+      "loss": 1.709,
+      "step": 1910
+    },
+    {
+      "epoch": 0.06122644216971204,
+      "grad_norm": 0.44391316175460815,
+      "learning_rate": 0.001,
+      "loss": 1.6897,
+      "step": 1920
+    },
+    {
+      "epoch": 0.06154532988934596,
+      "grad_norm": 0.4589037001132965,
+      "learning_rate": 0.001,
+      "loss": 1.7213,
+      "step": 1930
+    },
+    {
+      "epoch": 0.06186421760897988,
+      "grad_norm": 0.41747188568115234,
+      "learning_rate": 0.001,
+      "loss": 1.6885,
+      "step": 1940
+    },
+    {
+      "epoch": 0.0621831053286138,
+      "grad_norm": 0.41558709740638733,
+      "learning_rate": 0.001,
+      "loss": 1.6909,
+      "step": 1950
+    },
+    {
+      "epoch": 0.06250199304824772,
+      "grad_norm": 0.43651485443115234,
+      "learning_rate": 0.001,
+      "loss": 1.6924,
+      "step": 1960
+    },
+    {
+      "epoch": 0.06282088076788163,
+      "grad_norm": 0.44822901487350464,
+      "learning_rate": 0.001,
+      "loss": 1.7026,
+      "step": 1970
+    },
+    {
+      "epoch": 0.06313976848751554,
+      "grad_norm": 0.43817171454429626,
+      "learning_rate": 0.001,
+      "loss": 1.683,
+      "step": 1980
+    },
+    {
+      "epoch": 0.06345865620714947,
+      "grad_norm": 0.45964375138282776,
+      "learning_rate": 0.001,
+      "loss": 1.6755,
+      "step": 1990
+    },
+    {
+      "epoch": 0.06377754392678338,
+      "grad_norm": 0.4256100356578827,
+      "learning_rate": 0.001,
+      "loss": 1.6814,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06409643164641729,
+      "grad_norm": 0.4580231010913849,
+      "learning_rate": 0.001,
+      "loss": 1.6761,
+      "step": 2010
+    },
+    {
+      "epoch": 0.06441531936605122,
+      "grad_norm": 0.43515661358833313,
+      "learning_rate": 0.001,
+      "loss": 1.6977,
+      "step": 2020
+    },
+    {
+      "epoch": 0.06473420708568513,
+      "grad_norm": 0.5032089352607727,
+      "learning_rate": 0.001,
+      "loss": 1.6764,
+      "step": 2030
+    },
+    {
+      "epoch": 0.06505309480531905,
+      "grad_norm": 0.45531144738197327,
+      "learning_rate": 0.001,
+      "loss": 1.6734,
+      "step": 2040
+    },
+    {
+      "epoch": 0.06537198252495297,
+      "grad_norm": 0.42781633138656616,
+      "learning_rate": 0.001,
+      "loss": 1.6501,
+      "step": 2050
+    },
+    {
+      "epoch": 0.06569087024458688,
+      "grad_norm": 0.4123460352420807,
+      "learning_rate": 0.001,
+      "loss": 1.6385,
+      "step": 2060
+    },
+    {
+      "epoch": 0.0660097579642208,
+      "grad_norm": 0.44210684299468994,
+      "learning_rate": 0.001,
+      "loss": 1.6733,
+      "step": 2070
+    },
+    {
+      "epoch": 0.06632864568385471,
+      "grad_norm": 0.4108719229698181,
+      "learning_rate": 0.001,
+      "loss": 1.658,
+      "step": 2080
+    },
+    {
+      "epoch": 0.06664753340348863,
+      "grad_norm": 0.4030722677707672,
+      "learning_rate": 0.001,
+      "loss": 1.6742,
+      "step": 2090
+    },
+    {
+      "epoch": 0.06696642112312255,
+      "grad_norm": 0.4500997066497803,
+      "learning_rate": 0.001,
+      "loss": 1.6714,
+      "step": 2100
+    },
+    {
+      "epoch": 0.06728530884275646,
+      "grad_norm": 0.4074576795101166,
+      "learning_rate": 0.001,
+      "loss": 1.6419,
+      "step": 2110
+    },
+    {
+      "epoch": 0.06760419656239038,
+      "grad_norm": 0.40393441915512085,
+      "learning_rate": 0.001,
+      "loss": 1.6462,
+      "step": 2120
+    },
+    {
+      "epoch": 0.0679230842820243,
+      "grad_norm": 0.4103441536426544,
+      "learning_rate": 0.001,
+      "loss": 1.6531,
+      "step": 2130
+    },
+    {
+      "epoch": 0.06824197200165821,
+      "grad_norm": 0.4321202337741852,
+      "learning_rate": 0.001,
+      "loss": 1.6399,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06856085972129214,
+      "grad_norm": 0.41198042035102844,
+      "learning_rate": 0.001,
+      "loss": 1.6556,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06887974744092605,
+      "grad_norm": 0.4039364457130432,
+      "learning_rate": 0.001,
+      "loss": 1.654,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06919863516055996,
+      "grad_norm": 0.4295644462108612,
+      "learning_rate": 0.001,
+      "loss": 1.6432,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06951752288019389,
+      "grad_norm": 0.4189327657222748,
+      "learning_rate": 0.001,
+      "loss": 1.6393,
+      "step": 2180
+    },
+    {
+      "epoch": 0.0698364105998278,
+      "grad_norm": 0.4218682050704956,
+      "learning_rate": 0.001,
+      "loss": 1.6204,
+      "step": 2190
+    },
+    {
+      "epoch": 0.07015529831946171,
+      "grad_norm": 0.40319690108299255,
+      "learning_rate": 0.001,
+      "loss": 1.6381,
+      "step": 2200
+    },
+    {
+      "epoch": 0.07047418603909564,
+      "grad_norm": 0.40946927666664124,
+      "learning_rate": 0.001,
+      "loss": 1.6121,
+      "step": 2210
+    },
+    {
+      "epoch": 0.07079307375872955,
+      "grad_norm": 0.397491455078125,
+      "learning_rate": 0.001,
+      "loss": 1.6261,
+      "step": 2220
+    },
+    {
+      "epoch": 0.07111196147836348,
+      "grad_norm": 0.4268970191478729,
+      "learning_rate": 0.001,
+      "loss": 1.6144,
+      "step": 2230
+    },
+    {
+      "epoch": 0.07143084919799739,
+      "grad_norm": 0.46040934324264526,
+      "learning_rate": 0.001,
+      "loss": 1.6208,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0717497369176313,
+      "grad_norm": 0.41573449969291687,
+      "learning_rate": 0.001,
+      "loss": 1.6021,
+      "step": 2250
+    },
+    {
+      "epoch": 0.07206862463726522,
+      "grad_norm": 0.40803709626197815,
+      "learning_rate": 0.001,
+      "loss": 1.5945,
+      "step": 2260
+    },
+    {
+      "epoch": 0.07238751235689914,
+      "grad_norm": 0.4061513841152191,
+      "learning_rate": 0.001,
+      "loss": 1.6189,
+      "step": 2270
+    },
+    {
+      "epoch": 0.07270640007653305,
+      "grad_norm": 0.40856054425239563,
+      "learning_rate": 0.001,
+      "loss": 1.6073,
+      "step": 2280
+    },
+    {
+      "epoch": 0.07302528779616697,
+      "grad_norm": 0.39824768900871277,
+      "learning_rate": 0.001,
+      "loss": 1.5878,
+      "step": 2290
+    },
+    {
+      "epoch": 0.07334417551580089,
+      "grad_norm": 0.4029414653778076,
+      "learning_rate": 0.001,
+      "loss": 1.5974,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0736630632354348,
+      "grad_norm": 0.3928816318511963,
+      "learning_rate": 0.001,
+      "loss": 1.5884,
+      "step": 2310
+    },
+    {
+      "epoch": 0.07398195095506872,
+      "grad_norm": 0.403522253036499,
+      "learning_rate": 0.001,
+      "loss": 1.6077,
+      "step": 2320
+    },
+    {
+      "epoch": 0.07430083867470264,
+      "grad_norm": 0.39391088485717773,
+      "learning_rate": 0.001,
+      "loss": 1.6057,
+      "step": 2330
+    },
+    {
+      "epoch": 0.07461972639433656,
+      "grad_norm": 0.40311235189437866,
+      "learning_rate": 0.001,
+      "loss": 1.5865,
+      "step": 2340
+    },
+    {
+      "epoch": 0.07493861411397047,
+      "grad_norm": 0.4081204831600189,
+      "learning_rate": 0.001,
+      "loss": 1.6074,
+      "step": 2350
+    },
+    {
+      "epoch": 0.07525750183360438,
+      "grad_norm": 0.3867422044277191,
+      "learning_rate": 0.001,
+      "loss": 1.5975,
+      "step": 2360
+    },
+    {
+      "epoch": 0.07557638955323831,
+      "grad_norm": 0.3857383131980896,
+      "learning_rate": 0.001,
+      "loss": 1.5611,
+      "step": 2370
+    },
+    {
+      "epoch": 0.07589527727287222,
+      "grad_norm": 0.3840004801750183,
+      "learning_rate": 0.001,
+      "loss": 1.5612,
+      "step": 2380
+    },
+    {
+      "epoch": 0.07621416499250613,
+      "grad_norm": 0.42938390374183655,
+      "learning_rate": 0.001,
+      "loss": 1.6057,
+      "step": 2390
+    },
+    {
+      "epoch": 0.07653305271214006,
+      "grad_norm": 0.3979109227657318,
+      "learning_rate": 0.001,
+      "loss": 1.5863,
+      "step": 2400
+    },
+    {
+      "epoch": 0.07685194043177397,
+      "grad_norm": 0.38790082931518555,
+      "learning_rate": 0.001,
+      "loss": 1.5973,
+      "step": 2410
+    },
+    {
+      "epoch": 0.07717082815140788,
+      "grad_norm": 0.3939755856990814,
+      "learning_rate": 0.001,
+      "loss": 1.5764,
+      "step": 2420
+    },
+    {
+      "epoch": 0.07748971587104181,
+      "grad_norm": 0.3828144371509552,
+      "learning_rate": 0.001,
+      "loss": 1.5704,
+      "step": 2430
+    },
+    {
+      "epoch": 0.07780860359067572,
+      "grad_norm": 0.42103809118270874,
+      "learning_rate": 0.001,
+      "loss": 1.5667,
+      "step": 2440
+    },
+    {
+      "epoch": 0.07812749131030965,
+      "grad_norm": 0.43680015206336975,
+      "learning_rate": 0.001,
+      "loss": 1.5534,
+      "step": 2450
+    },
+    {
+      "epoch": 0.07844637902994356,
+      "grad_norm": 0.38762131333351135,
+      "learning_rate": 0.001,
+      "loss": 1.5724,
+      "step": 2460
+    },
+    {
+      "epoch": 0.07876526674957747,
+      "grad_norm": 0.38347816467285156,
+      "learning_rate": 0.001,
+      "loss": 1.5616,
+      "step": 2470
+    },
+    {
+      "epoch": 0.0790841544692114,
+      "grad_norm": 0.3741878271102905,
+      "learning_rate": 0.001,
+      "loss": 1.5503,
+      "step": 2480
+    },
+    {
+      "epoch": 0.07940304218884531,
+      "grad_norm": 0.36821624636650085,
+      "learning_rate": 0.001,
+      "loss": 1.5792,
+      "step": 2490
+    },
+    {
+      "epoch": 0.07972192990847922,
+      "grad_norm": 0.39044198393821716,
+      "learning_rate": 0.001,
+      "loss": 1.5657,
+      "step": 2500
+    },
+    {
+      "epoch": 0.08004081762811315,
+      "grad_norm": 0.3792630434036255,
+      "learning_rate": 0.001,
+      "loss": 1.5419,
+      "step": 2510
+    },
+    {
+      "epoch": 0.08035970534774706,
+      "grad_norm": 0.40021875500679016,
+      "learning_rate": 0.001,
+      "loss": 1.565,
+      "step": 2520
+    },
+    {
+      "epoch": 0.08067859306738097,
+      "grad_norm": 0.38357385993003845,
+      "learning_rate": 0.001,
+      "loss": 1.5643,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0809974807870149,
+      "grad_norm": 0.3646472096443176,
+      "learning_rate": 0.001,
+      "loss": 1.5621,
+      "step": 2540
+    },
+    {
+      "epoch": 0.0813163685066488,
+      "grad_norm": 0.37614285945892334,
+      "learning_rate": 0.001,
+      "loss": 1.5516,
+      "step": 2550
+    },
+    {
+      "epoch": 0.08163525622628273,
+      "grad_norm": 0.37552952766418457,
+      "learning_rate": 0.001,
+      "loss": 1.5506,
+      "step": 2560
+    },
+    {
+      "epoch": 0.08195414394591664,
+      "grad_norm": 0.3884067237377167,
+      "learning_rate": 0.001,
+      "loss": 1.5232,
+      "step": 2570
+    },
+    {
+      "epoch": 0.08227303166555056,
+      "grad_norm": 0.393189936876297,
+      "learning_rate": 0.001,
+      "loss": 1.5588,
+      "step": 2580
+    },
+    {
+      "epoch": 0.08259191938518448,
+      "grad_norm": 0.3752188980579376,
+      "learning_rate": 0.001,
+      "loss": 1.5298,
+      "step": 2590
+    },
+    {
+      "epoch": 0.0829108071048184,
+      "grad_norm": 0.37879225611686707,
+      "learning_rate": 0.001,
+      "loss": 1.5462,
+      "step": 2600
+    },
+    {
+      "epoch": 0.0832296948244523,
+      "grad_norm": 0.37227118015289307,
+      "learning_rate": 0.001,
+      "loss": 1.5473,
+      "step": 2610
+    },
+    {
+      "epoch": 0.08354858254408623,
+      "grad_norm": 0.3972522020339966,
+      "learning_rate": 0.001,
+      "loss": 1.5261,
+      "step": 2620
+    },
+    {
+      "epoch": 0.08386747026372014,
+      "grad_norm": 0.37113243341445923,
+      "learning_rate": 0.001,
+      "loss": 1.5337,
+      "step": 2630
+    },
+    {
+      "epoch": 0.08418635798335405,
+      "grad_norm": 0.364286333322525,
+      "learning_rate": 0.001,
+      "loss": 1.5293,
+      "step": 2640
+    },
+    {
+      "epoch": 0.08450524570298798,
+      "grad_norm": 0.362486869096756,
+      "learning_rate": 0.001,
+      "loss": 1.5333,
+      "step": 2650
+    },
+    {
+      "epoch": 0.08482413342262189,
+      "grad_norm": 0.3880142271518707,
+      "learning_rate": 0.001,
+      "loss": 1.5459,
+      "step": 2660
+    },
+    {
+      "epoch": 0.08514302114225582,
+      "grad_norm": 0.3843633532524109,
+      "learning_rate": 0.001,
+      "loss": 1.5181,
+      "step": 2670
+    },
+    {
+      "epoch": 0.08546190886188973,
+      "grad_norm": 0.36711886525154114,
+      "learning_rate": 0.001,
+      "loss": 1.5501,
+      "step": 2680
+    },
+    {
+      "epoch": 0.08578079658152364,
+      "grad_norm": 0.3663885295391083,
+      "learning_rate": 0.001,
+      "loss": 1.5248,
+      "step": 2690
+    },
+    {
+      "epoch": 0.08609968430115757,
+      "grad_norm": 0.3654477298259735,
+      "learning_rate": 0.001,
+      "loss": 1.5316,
+      "step": 2700
+    },
+    {
+      "epoch": 0.08641857202079148,
+      "grad_norm": 0.39503544569015503,
+      "learning_rate": 0.001,
+      "loss": 1.5214,
+      "step": 2710
+    },
+    {
+      "epoch": 0.08673745974042539,
+      "grad_norm": 0.39475134015083313,
+      "learning_rate": 0.001,
+      "loss": 1.5385,
+      "step": 2720
+    },
+    {
+      "epoch": 0.08705634746005932,
+      "grad_norm": 0.3583775758743286,
+      "learning_rate": 0.001,
+      "loss": 1.539,
+      "step": 2730
+    },
+    {
+      "epoch": 0.08737523517969323,
+      "grad_norm": 0.3696213364601135,
+      "learning_rate": 0.001,
+      "loss": 1.5548,
+      "step": 2740
+    },
+    {
+      "epoch": 0.08769412289932715,
+      "grad_norm": 0.3889085352420807,
+      "learning_rate": 0.001,
+      "loss": 1.5136,
+      "step": 2750
+    },
+    {
+      "epoch": 0.08801301061896107,
+      "grad_norm": 0.3819277882575989,
+      "learning_rate": 0.001,
+      "loss": 1.5187,
+      "step": 2760
+    },
+    {
+      "epoch": 0.08833189833859498,
+      "grad_norm": 0.3591541647911072,
+      "learning_rate": 0.001,
+      "loss": 1.5146,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0886507860582289,
+      "grad_norm": 0.37153083086013794,
+      "learning_rate": 0.001,
+      "loss": 1.539,
+      "step": 2780
+    },
+    {
+      "epoch": 0.08896967377786281,
+      "grad_norm": 0.3519269526004791,
+      "learning_rate": 0.001,
+      "loss": 1.4952,
+      "step": 2790
+    },
+    {
+      "epoch": 0.08928856149749673,
+      "grad_norm": 0.3622041344642639,
+      "learning_rate": 0.001,
+      "loss": 1.5245,
+      "step": 2800
+    },
+    {
+      "epoch": 0.08960744921713065,
+      "grad_norm": 0.3769047260284424,
+      "learning_rate": 0.001,
+      "loss": 1.4881,
+      "step": 2810
+    },
+    {
+      "epoch": 0.08992633693676456,
+      "grad_norm": 0.3731653392314911,
+      "learning_rate": 0.001,
+      "loss": 1.5169,
+      "step": 2820
+    },
+    {
+      "epoch": 0.09024522465639848,
+      "grad_norm": 0.36590540409088135,
+      "learning_rate": 0.001,
+      "loss": 1.511,
+      "step": 2830
+    },
+    {
+      "epoch": 0.0905641123760324,
+      "grad_norm": 0.35569992661476135,
+      "learning_rate": 0.001,
+      "loss": 1.5098,
+      "step": 2840
+    },
+    {
+      "epoch": 0.09088300009566631,
+      "grad_norm": 0.37215662002563477,
+      "learning_rate": 0.001,
+      "loss": 1.4903,
+      "step": 2850
+    },
+    {
+      "epoch": 0.09120188781530024,
+      "grad_norm": 0.35044750571250916,
+      "learning_rate": 0.001,
+      "loss": 1.5006,
+      "step": 2860
+    },
+    {
+      "epoch": 0.09152077553493415,
+      "grad_norm": 0.3568022847175598,
+      "learning_rate": 0.001,
+      "loss": 1.4942,
+      "step": 2870
+    },
+    {
+      "epoch": 0.09183966325456806,
+      "grad_norm": 0.3533008098602295,
+      "learning_rate": 0.001,
+      "loss": 1.5072,
+      "step": 2880
+    },
+    {
+      "epoch": 0.09215855097420199,
+      "grad_norm": 0.3491477370262146,
+      "learning_rate": 0.001,
+      "loss": 1.4951,
+      "step": 2890
+    },
+    {
+      "epoch": 0.0924774386938359,
+      "grad_norm": 0.3570086658000946,
+      "learning_rate": 0.001,
+      "loss": 1.4967,
+      "step": 2900
+    },
+    {
+      "epoch": 0.09279632641346981,
+      "grad_norm": 0.3539246916770935,
+      "learning_rate": 0.001,
+      "loss": 1.503,
+      "step": 2910
+    },
+    {
+      "epoch": 0.09311521413310374,
+      "grad_norm": 0.37031111121177673,
+      "learning_rate": 0.001,
+      "loss": 1.5041,
+      "step": 2920
+    },
+    {
+      "epoch": 0.09343410185273765,
+      "grad_norm": 0.34995830059051514,
+      "learning_rate": 0.001,
+      "loss": 1.5063,
+      "step": 2930
+    },
+    {
+      "epoch": 0.09375298957237156,
+      "grad_norm": 0.36037734150886536,
+      "learning_rate": 0.001,
+      "loss": 1.5144,
+      "step": 2940
+    },
+    {
+      "epoch": 0.09407187729200549,
+      "grad_norm": 0.382372111082077,
+      "learning_rate": 0.001,
+      "loss": 1.4877,
+      "step": 2950
+    },
+    {
+      "epoch": 0.0943907650116394,
+      "grad_norm": 0.3281106650829315,
+      "learning_rate": 0.001,
+      "loss": 1.4862,
+      "step": 2960
+    },
+    {
+      "epoch": 0.09470965273127332,
+      "grad_norm": 0.36586999893188477,
+      "learning_rate": 0.001,
+      "loss": 1.497,
+      "step": 2970
+    },
+    {
+      "epoch": 0.09502854045090724,
+      "grad_norm": 0.3418881893157959,
+      "learning_rate": 0.001,
+      "loss": 1.4877,
+      "step": 2980
+    },
+    {
+      "epoch": 0.09534742817054115,
+      "grad_norm": 0.3603076934814453,
+      "learning_rate": 0.001,
+      "loss": 1.4896,
+      "step": 2990
+    },
+    {
+      "epoch": 0.09566631589017507,
+      "grad_norm": 0.3709716498851776,
+      "learning_rate": 0.001,
+      "loss": 1.4994,
+      "step": 3000
+    },
+    {
+      "epoch": 0.09598520360980899,
+      "grad_norm": 0.34613725543022156,
+      "learning_rate": 0.001,
+      "loss": 1.4909,
+      "step": 3010
+    },
+    {
+      "epoch": 0.0963040913294429,
+      "grad_norm": 0.34238988161087036,
+      "learning_rate": 0.001,
+      "loss": 1.477,
+      "step": 3020
+    },
+    {
+      "epoch": 0.09662297904907682,
+      "grad_norm": 0.357252299785614,
+      "learning_rate": 0.001,
+      "loss": 1.4807,
+      "step": 3030
+    },
+    {
+      "epoch": 0.09694186676871074,
+      "grad_norm": 0.34529972076416016,
+      "learning_rate": 0.001,
+      "loss": 1.4892,
+      "step": 3040
+    },
+    {
+      "epoch": 0.09726075448834465,
+      "grad_norm": 0.3669123947620392,
+      "learning_rate": 0.001,
+      "loss": 1.5015,
+      "step": 3050
+    },
+    {
+      "epoch": 0.09757964220797857,
+      "grad_norm": 0.35653454065322876,
+      "learning_rate": 0.001,
+      "loss": 1.4943,
+      "step": 3060
+    },
+    {
+      "epoch": 0.09789852992761248,
+      "grad_norm": 0.3683283030986786,
+      "learning_rate": 0.001,
+      "loss": 1.4762,
+      "step": 3070
+    },
+    {
+      "epoch": 0.09821741764724641,
+      "grad_norm": 0.3327406346797943,
+      "learning_rate": 0.001,
+      "loss": 1.4843,
+      "step": 3080
+    },
+    {
+      "epoch": 0.09853630536688032,
+      "grad_norm": 0.3510376513004303,
+      "learning_rate": 0.001,
+      "loss": 1.4725,
+      "step": 3090
+    },
+    {
+      "epoch": 0.09885519308651423,
+      "grad_norm": 0.3618435263633728,
+      "learning_rate": 0.001,
+      "loss": 1.4636,
+      "step": 3100
+    },
+    {
+      "epoch": 0.09917408080614816,
+      "grad_norm": 0.3902658224105835,
+      "learning_rate": 0.001,
+      "loss": 1.4647,
+      "step": 3110
+    },
+    {
+      "epoch": 0.09949296852578207,
+      "grad_norm": 0.336622416973114,
+      "learning_rate": 0.001,
+      "loss": 1.4864,
+      "step": 3120
+    },
+    {
+      "epoch": 0.09981185624541598,
+      "grad_norm": 0.3628654181957245,
+      "learning_rate": 0.001,
+      "loss": 1.4914,
+      "step": 3130
+    },
+    {
+      "epoch": 0.10013074396504991,
+      "grad_norm": 0.34092777967453003,
+      "learning_rate": 0.001,
+      "loss": 1.478,
+      "step": 3140
+    },
+    {
+      "epoch": 0.10044963168468382,
+      "grad_norm": 0.3362221419811249,
+      "learning_rate": 0.001,
+      "loss": 1.4862,
+      "step": 3150
+    },
+    {
+      "epoch": 0.10076851940431773,
+      "grad_norm": 0.37139853835105896,
+      "learning_rate": 0.001,
+      "loss": 1.4655,
+      "step": 3160
+    },
+    {
+      "epoch": 0.10108740712395166,
+      "grad_norm": 0.3196049630641937,
+      "learning_rate": 0.001,
+      "loss": 1.468,
+      "step": 3170
+    },
+    {
+      "epoch": 0.10140629484358557,
+      "grad_norm": 0.34154269099235535,
+      "learning_rate": 0.001,
+      "loss": 1.4603,
+      "step": 3180
+    },
+    {
+      "epoch": 0.1017251825632195,
+      "grad_norm": 0.33629539608955383,
+      "learning_rate": 0.001,
+      "loss": 1.4422,
+      "step": 3190
+    },
+    {
+      "epoch": 0.10204407028285341,
+      "grad_norm": 0.32535743713378906,
+      "learning_rate": 0.001,
+      "loss": 1.4495,
+      "step": 3200
+    },
+    {
+      "epoch": 0.10236295800248732,
+      "grad_norm": 0.3403289020061493,
+      "learning_rate": 0.001,
+      "loss": 1.4856,
+      "step": 3210
+    },
+    {
+      "epoch": 0.10268184572212125,
+      "grad_norm": 0.3382999300956726,
+      "learning_rate": 0.001,
+      "loss": 1.4719,
+      "step": 3220
+    },
+    {
+      "epoch": 0.10300073344175516,
+      "grad_norm": 0.3283659815788269,
+      "learning_rate": 0.001,
+      "loss": 1.4374,
+      "step": 3230
+    },
+    {
+      "epoch": 0.10331962116138907,
+      "grad_norm": 0.3706674873828888,
+      "learning_rate": 0.001,
+      "loss": 1.4259,
+      "step": 3240
+    },
+    {
+      "epoch": 0.103638508881023,
+      "grad_norm": 0.3296601474285126,
+      "learning_rate": 0.001,
+      "loss": 1.4563,
+      "step": 3250
+    },
+    {
+      "epoch": 0.1039573966006569,
+      "grad_norm": 0.3546639680862427,
+      "learning_rate": 0.001,
+      "loss": 1.4707,
+      "step": 3260
+    },
+    {
+      "epoch": 0.10427628432029083,
+      "grad_norm": 0.33120912313461304,
+      "learning_rate": 0.001,
+      "loss": 1.4324,
+      "step": 3270
+    },
+    {
+      "epoch": 0.10459517203992474,
+      "grad_norm": 0.36581435799598694,
+      "learning_rate": 0.001,
+      "loss": 1.4687,
+      "step": 3280
+    },
+    {
+      "epoch": 0.10491405975955866,
+      "grad_norm": 0.34121185541152954,
+      "learning_rate": 0.001,
+      "loss": 1.4292,
+      "step": 3290
+    },
+    {
+      "epoch": 0.10523294747919258,
+      "grad_norm": 0.3378187119960785,
+      "learning_rate": 0.001,
+      "loss": 1.4586,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1055518351988265,
+      "grad_norm": 0.34159472584724426,
+      "learning_rate": 0.001,
+      "loss": 1.4506,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1058707229184604,
+      "grad_norm": 0.3327382504940033,
+      "learning_rate": 0.001,
+      "loss": 1.4283,
+      "step": 3320
+    },
+    {
+      "epoch": 0.10618961063809433,
+      "grad_norm": 0.3393729329109192,
+      "learning_rate": 0.001,
+      "loss": 1.4972,
+      "step": 3330
+    },
+    {
+      "epoch": 0.10650849835772824,
+      "grad_norm": 0.3339972198009491,
+      "learning_rate": 0.001,
+      "loss": 1.4593,
+      "step": 3340
+    },
+    {
+      "epoch": 0.10682738607736215,
+      "grad_norm": 0.34833648800849915,
+      "learning_rate": 0.001,
+      "loss": 1.4697,
+      "step": 3350
+    },
+    {
+      "epoch": 0.10714627379699608,
+      "grad_norm": 0.3432925045490265,
+      "learning_rate": 0.001,
+      "loss": 1.451,
+      "step": 3360
+    },
+    {
+      "epoch": 0.10746516151662999,
+      "grad_norm": 0.3279559910297394,
+      "learning_rate": 0.001,
+      "loss": 1.4578,
+      "step": 3370
+    },
+    {
+      "epoch": 0.10778404923626392,
+      "grad_norm": 0.3407198488712311,
+      "learning_rate": 0.001,
+      "loss": 1.4476,
+      "step": 3380
+    },
+    {
+      "epoch": 0.10810293695589783,
+      "grad_norm": 0.35291847586631775,
+      "learning_rate": 0.001,
+      "loss": 1.4141,
+      "step": 3390
+    },
+    {
+      "epoch": 0.10842182467553174,
+      "grad_norm": 0.3389386832714081,
+      "learning_rate": 0.001,
+      "loss": 1.4367,
+      "step": 3400
+    },
+    {
+      "epoch": 0.10874071239516567,
+      "grad_norm": 0.3285823464393616,
+      "learning_rate": 0.001,
+      "loss": 1.4468,
+      "step": 3410
+    },
+    {
+      "epoch": 0.10905960011479958,
+      "grad_norm": 0.3264278471469879,
+      "learning_rate": 0.001,
+      "loss": 1.4572,
+      "step": 3420
+    },
+    {
+      "epoch": 0.10937848783443349,
+      "grad_norm": 0.3490174412727356,
+      "learning_rate": 0.001,
+      "loss": 1.4502,
+      "step": 3430
+    },
+    {
+      "epoch": 0.10969737555406742,
+      "grad_norm": 0.31935915350914,
+      "learning_rate": 0.001,
+      "loss": 1.4519,
+      "step": 3440
+    },
+    {
+      "epoch": 0.11001626327370133,
+      "grad_norm": 0.325456440448761,
+      "learning_rate": 0.001,
+      "loss": 1.4553,
+      "step": 3450
+    },
+    {
+      "epoch": 0.11033515099333524,
+      "grad_norm": 0.33232828974723816,
+      "learning_rate": 0.001,
+      "loss": 1.4221,
+      "step": 3460
+    },
+    {
+      "epoch": 0.11065403871296917,
+      "grad_norm": 0.3299655318260193,
+      "learning_rate": 0.001,
+      "loss": 1.4496,
+      "step": 3470
+    },
+    {
+      "epoch": 0.11097292643260308,
+      "grad_norm": 0.33184587955474854,
+      "learning_rate": 0.001,
+      "loss": 1.4256,
+      "step": 3480
+    },
+    {
+      "epoch": 0.111291814152237,
+      "grad_norm": 0.33328554034233093,
+      "learning_rate": 0.001,
+      "loss": 1.4229,
+      "step": 3490
+    },
+    {
+      "epoch": 0.11161070187187092,
+      "grad_norm": 0.33121776580810547,
+      "learning_rate": 0.001,
+      "loss": 1.4369,
+      "step": 3500
+    },
+    {
+      "epoch": 0.11192958959150483,
+      "grad_norm": 0.33411258459091187,
+      "learning_rate": 0.001,
+      "loss": 1.4346,
+      "step": 3510
+    },
+    {
+      "epoch": 0.11224847731113875,
+      "grad_norm": 0.3087165355682373,
+      "learning_rate": 0.001,
+      "loss": 1.4262,
+      "step": 3520
+    },
+    {
+      "epoch": 0.11256736503077266,
+      "grad_norm": 0.3170263469219208,
+      "learning_rate": 0.001,
+      "loss": 1.4212,
+      "step": 3530
+    },
+    {
+      "epoch": 0.11288625275040658,
+      "grad_norm": 0.33870166540145874,
+      "learning_rate": 0.001,
+      "loss": 1.4373,
+      "step": 3540
+    },
+    {
+      "epoch": 0.1132051404700405,
+      "grad_norm": 0.34458819031715393,
+      "learning_rate": 0.001,
+      "loss": 1.4329,
+      "step": 3550
+    },
+    {
+      "epoch": 0.11352402818967441,
+      "grad_norm": 0.3231976628303528,
+      "learning_rate": 0.001,
+      "loss": 1.4343,
+      "step": 3560
+    },
+    {
+      "epoch": 0.11384291590930833,
+      "grad_norm": 0.33272790908813477,
+      "learning_rate": 0.001,
+      "loss": 1.4098,
+      "step": 3570
+    },
+    {
+      "epoch": 0.11416180362894225,
+      "grad_norm": 0.31947606801986694,
+      "learning_rate": 0.001,
+      "loss": 1.4049,
+      "step": 3580
+    },
+    {
+      "epoch": 0.11448069134857616,
+      "grad_norm": 0.34905490279197693,
+      "learning_rate": 0.001,
+      "loss": 1.444,
+      "step": 3590
+    },
+    {
+      "epoch": 0.11479957906821009,
+      "grad_norm": 0.3256986737251282,
+      "learning_rate": 0.001,
+      "loss": 1.4312,
+      "step": 3600
+    },
+    {
+      "epoch": 0.115118466787844,
+      "grad_norm": 0.3320602476596832,
+      "learning_rate": 0.001,
+      "loss": 1.3994,
+      "step": 3610
+    },
+    {
+      "epoch": 0.11543735450747791,
+      "grad_norm": 0.32200315594673157,
+      "learning_rate": 0.001,
+      "loss": 1.3982,
+      "step": 3620
+    },
+    {
+      "epoch": 0.11575624222711184,
+      "grad_norm": 0.3169730007648468,
+      "learning_rate": 0.001,
+      "loss": 1.4297,
+      "step": 3630
+    },
+    {
+      "epoch": 0.11607512994674575,
+      "grad_norm": 0.33035051822662354,
+      "learning_rate": 0.001,
+      "loss": 1.4403,
+      "step": 3640
+    },
+    {
+      "epoch": 0.11639401766637966,
+      "grad_norm": 0.3387161195278168,
+      "learning_rate": 0.001,
+      "loss": 1.4163,
+      "step": 3650
+    },
+    {
+      "epoch": 0.11671290538601359,
+      "grad_norm": 0.31775781512260437,
+      "learning_rate": 0.001,
+      "loss": 1.4008,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1170317931056475,
+      "grad_norm": 0.3159157931804657,
+      "learning_rate": 0.001,
+      "loss": 1.4304,
+      "step": 3670
+    },
+    {
+      "epoch": 0.11735068082528141,
+      "grad_norm": 0.3200733959674835,
+      "learning_rate": 0.001,
+      "loss": 1.4446,
+      "step": 3680
+    },
+    {
+      "epoch": 0.11766956854491534,
+      "grad_norm": 0.3100077211856842,
+      "learning_rate": 0.001,
+      "loss": 1.4135,
+      "step": 3690
+    },
+    {
+      "epoch": 0.11798845626454925,
+      "grad_norm": 0.3178308606147766,
+      "learning_rate": 0.001,
+      "loss": 1.4192,
+      "step": 3700
+    },
+    {
+      "epoch": 0.11830734398418317,
+      "grad_norm": 0.3008234202861786,
+      "learning_rate": 0.001,
+      "loss": 1.4215,
+      "step": 3710
+    },
+    {
+      "epoch": 0.11862623170381709,
+      "grad_norm": 0.31308454275131226,
+      "learning_rate": 0.001,
+      "loss": 1.4019,
+      "step": 3720
+    },
+    {
+      "epoch": 0.118945119423451,
+      "grad_norm": 0.32719430327415466,
+      "learning_rate": 0.001,
+      "loss": 1.4209,
+      "step": 3730
+    },
+    {
+      "epoch": 0.11926400714308492,
+      "grad_norm": 0.3134723901748657,
+      "learning_rate": 0.001,
+      "loss": 1.4274,
+      "step": 3740
+    },
+    {
+      "epoch": 0.11958289486271884,
+      "grad_norm": 0.31542280316352844,
+      "learning_rate": 0.001,
+      "loss": 1.401,
+      "step": 3750
+    },
+    {
+      "epoch": 0.11990178258235275,
+      "grad_norm": 0.3206769526004791,
+      "learning_rate": 0.001,
+      "loss": 1.4101,
+      "step": 3760
+    },
+    {
+      "epoch": 0.12022067030198667,
+      "grad_norm": 0.31581780314445496,
+      "learning_rate": 0.001,
+      "loss": 1.4272,
+      "step": 3770
+    },
+    {
+      "epoch": 0.12053955802162059,
+      "grad_norm": 0.31411078572273254,
+      "learning_rate": 0.001,
+      "loss": 1.44,
+      "step": 3780
+    },
+    {
+      "epoch": 0.12085844574125451,
+      "grad_norm": 0.30180829763412476,
+      "learning_rate": 0.001,
+      "loss": 1.3919,
+      "step": 3790
+    },
+    {
+      "epoch": 0.12117733346088842,
+      "grad_norm": 0.30781787633895874,
+      "learning_rate": 0.001,
+      "loss": 1.3889,
+      "step": 3800
+    },
+    {
+      "epoch": 0.12149622118052233,
+      "grad_norm": 0.32143038511276245,
+      "learning_rate": 0.001,
+      "loss": 1.3944,
+      "step": 3810
+    },
+    {
+      "epoch": 0.12181510890015626,
+      "grad_norm": 0.34171000123023987,
+      "learning_rate": 0.001,
+      "loss": 1.4061,
+      "step": 3820
+    },
+    {
+      "epoch": 0.12213399661979017,
+      "grad_norm": 0.31811872124671936,
+      "learning_rate": 0.001,
+      "loss": 1.3958,
+      "step": 3830
+    },
+    {
+      "epoch": 0.12245288433942408,
+      "grad_norm": 0.33086341619491577,
+      "learning_rate": 0.001,
+      "loss": 1.4189,
+      "step": 3840
+    },
+    {
+      "epoch": 0.12277177205905801,
+      "grad_norm": 0.30563294887542725,
+      "learning_rate": 0.001,
+      "loss": 1.4161,
+      "step": 3850
+    },
+    {
+      "epoch": 0.12309065977869192,
+      "grad_norm": 0.30680814385414124,
+      "learning_rate": 0.001,
+      "loss": 1.4156,
+      "step": 3860
+    },
+    {
+      "epoch": 0.12340954749832583,
+      "grad_norm": 0.32353097200393677,
+      "learning_rate": 0.001,
+      "loss": 1.4239,
+      "step": 3870
+    },
+    {
+      "epoch": 0.12372843521795976,
+      "grad_norm": 0.3182706832885742,
+      "learning_rate": 0.001,
+      "loss": 1.3936,
+      "step": 3880
+    },
+    {
+      "epoch": 0.12404732293759367,
+      "grad_norm": 0.309923380613327,
+      "learning_rate": 0.001,
+      "loss": 1.4099,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1243662106572276,
+      "grad_norm": 0.31839415431022644,
+      "learning_rate": 0.001,
+      "loss": 1.4066,
+      "step": 3900
+    },
+    {
+      "epoch": 0.12468509837686151,
+      "grad_norm": 0.33320650458335876,
+      "learning_rate": 0.001,
+      "loss": 1.4058,
+      "step": 3910
+    },
+    {
+      "epoch": 0.12500398609649543,
+      "grad_norm": 0.31085923314094543,
+      "learning_rate": 0.001,
+      "loss": 1.4109,
+      "step": 3920
+    },
+    {
+      "epoch": 0.12532287381612933,
+      "grad_norm": 0.308795303106308,
+      "learning_rate": 0.001,
+      "loss": 1.3775,
+      "step": 3930
+    },
+    {
+      "epoch": 0.12564176153576326,
+      "grad_norm": 0.3182823657989502,
+      "learning_rate": 0.001,
+      "loss": 1.3809,
+      "step": 3940
+    },
+    {
+      "epoch": 0.12596064925539718,
+      "grad_norm": 0.2957512140274048,
+      "learning_rate": 0.001,
+      "loss": 1.3972,
+      "step": 3950
+    },
+    {
+      "epoch": 0.12627953697503108,
+      "grad_norm": 0.32786622643470764,
+      "learning_rate": 0.001,
+      "loss": 1.3852,
+      "step": 3960
+    },
+    {
+      "epoch": 0.126598424694665,
+      "grad_norm": 0.3133629560470581,
+      "learning_rate": 0.001,
+      "loss": 1.4012,
+      "step": 3970
+    },
+    {
+      "epoch": 0.12691731241429893,
+      "grad_norm": 0.3149418830871582,
+      "learning_rate": 0.001,
+      "loss": 1.384,
+      "step": 3980
+    },
+    {
+      "epoch": 0.12723620013393283,
+      "grad_norm": 0.31080493330955505,
+      "learning_rate": 0.001,
+      "loss": 1.3909,
+      "step": 3990
+    },
+    {
+      "epoch": 0.12755508785356676,
+      "grad_norm": 0.3017434775829315,
+      "learning_rate": 0.001,
+      "loss": 1.389,
+      "step": 4000
+    },
+    {
+      "epoch": 0.12787397557320068,
+      "grad_norm": 0.3071618378162384,
+      "learning_rate": 0.001,
+      "loss": 1.4003,
+      "step": 4010
+    },
+    {
+      "epoch": 0.12819286329283458,
+      "grad_norm": 0.3157201409339905,
+      "learning_rate": 0.001,
+      "loss": 1.3902,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1285117510124685,
+      "grad_norm": 0.32032763957977295,
+      "learning_rate": 0.001,
+      "loss": 1.3973,
+      "step": 4030
+    },
+    {
+      "epoch": 0.12883063873210243,
+      "grad_norm": 0.3034627437591553,
+      "learning_rate": 0.001,
+      "loss": 1.3938,
+      "step": 4040
+    },
+    {
+      "epoch": 0.12914952645173636,
+      "grad_norm": 0.3229222893714905,
+      "learning_rate": 0.001,
+      "loss": 1.3955,
+      "step": 4050
+    },
+    {
+      "epoch": 0.12946841417137026,
+      "grad_norm": 0.3058593273162842,
+      "learning_rate": 0.001,
+      "loss": 1.4061,
+      "step": 4060
+    },
+    {
+      "epoch": 0.12978730189100418,
+      "grad_norm": 0.30278199911117554,
+      "learning_rate": 0.001,
+      "loss": 1.3985,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1301061896106381,
+      "grad_norm": 0.3108304440975189,
+      "learning_rate": 0.001,
+      "loss": 1.379,
+      "step": 4080
+    },
+    {
+      "epoch": 0.130425077330272,
+      "grad_norm": 0.3051409125328064,
+      "learning_rate": 0.001,
+      "loss": 1.3886,
+      "step": 4090
+    },
+    {
+      "epoch": 0.13074396504990593,
+      "grad_norm": 0.32630011439323425,
+      "learning_rate": 0.001,
+      "loss": 1.3905,
+      "step": 4100
+    },
+    {
+      "epoch": 0.13106285276953986,
+      "grad_norm": 0.31419435143470764,
+      "learning_rate": 0.001,
+      "loss": 1.3871,
+      "step": 4110
+    },
+    {
+      "epoch": 0.13138174048917375,
+      "grad_norm": 0.32161471247673035,
+      "learning_rate": 0.001,
+      "loss": 1.3701,
+      "step": 4120
+    },
+    {
+      "epoch": 0.13170062820880768,
+      "grad_norm": 0.29289132356643677,
+      "learning_rate": 0.001,
+      "loss": 1.3881,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1320195159284416,
+      "grad_norm": 0.3462681174278259,
+      "learning_rate": 0.001,
+      "loss": 1.3798,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1323384036480755,
+      "grad_norm": 0.3089030086994171,
+      "learning_rate": 0.001,
+      "loss": 1.3721,
+      "step": 4150
+    },
+    {
+      "epoch": 0.13265729136770943,
+      "grad_norm": 0.2968080937862396,
+      "learning_rate": 0.001,
+      "loss": 1.3774,
+      "step": 4160
+    },
+    {
+      "epoch": 0.13297617908734335,
+      "grad_norm": 0.3071855306625366,
+      "learning_rate": 0.001,
+      "loss": 1.3786,
+      "step": 4170
+    },
+    {
+      "epoch": 0.13329506680697725,
+      "grad_norm": 0.3024219870567322,
+      "learning_rate": 0.001,
+      "loss": 1.3618,
+      "step": 4180
+    },
+    {
+      "epoch": 0.13361395452661118,
+      "grad_norm": 0.3031976521015167,
+      "learning_rate": 0.001,
+      "loss": 1.3712,
+      "step": 4190
+    },
+    {
+      "epoch": 0.1339328422462451,
+      "grad_norm": 0.30467966198921204,
+      "learning_rate": 0.001,
+      "loss": 1.3776,
+      "step": 4200
+    },
+    {
+      "epoch": 0.134251729965879,
+      "grad_norm": 0.3077302575111389,
+      "learning_rate": 0.001,
+      "loss": 1.3746,
+      "step": 4210
+    },
+    {
+      "epoch": 0.13457061768551293,
+      "grad_norm": 0.28658169507980347,
+      "learning_rate": 0.001,
+      "loss": 1.3679,
+      "step": 4220
+    },
+    {
+      "epoch": 0.13488950540514685,
+      "grad_norm": 0.2950628399848938,
+      "learning_rate": 0.001,
+      "loss": 1.3658,
+      "step": 4230
+    },
+    {
+      "epoch": 0.13520839312478075,
+      "grad_norm": 0.3013644516468048,
+      "learning_rate": 0.001,
+      "loss": 1.4024,
+      "step": 4240
+    },
+    {
+      "epoch": 0.13552728084441468,
+      "grad_norm": 0.30451542139053345,
+      "learning_rate": 0.001,
+      "loss": 1.3746,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1358461685640486,
+      "grad_norm": 0.30308064818382263,
+      "learning_rate": 0.001,
+      "loss": 1.3883,
+      "step": 4260
+    },
+    {
+      "epoch": 0.13616505628368253,
+      "grad_norm": 0.2882744371891022,
+      "learning_rate": 0.001,
+      "loss": 1.3283,
+      "step": 4270
+    },
+    {
+      "epoch": 0.13648394400331643,
+      "grad_norm": 0.3060248792171478,
+      "learning_rate": 0.001,
+      "loss": 1.3688,
+      "step": 4280
+    },
+    {
+      "epoch": 0.13680283172295035,
+      "grad_norm": 0.2914104461669922,
+      "learning_rate": 0.001,
+      "loss": 1.3523,
+      "step": 4290
+    },
+    {
+      "epoch": 0.13712171944258428,
+      "grad_norm": 0.2941143810749054,
+      "learning_rate": 0.001,
+      "loss": 1.3662,
+      "step": 4300
+    },
+    {
+      "epoch": 0.13744060716221818,
+      "grad_norm": 0.31439968943595886,
+      "learning_rate": 0.001,
+      "loss": 1.3765,
+      "step": 4310
+    },
+    {
+      "epoch": 0.1377594948818521,
+      "grad_norm": 0.2949891686439514,
+      "learning_rate": 0.001,
+      "loss": 1.357,
+      "step": 4320
+    },
+    {
+      "epoch": 0.13807838260148603,
+      "grad_norm": 0.3023112416267395,
+      "learning_rate": 0.001,
+      "loss": 1.3667,
+      "step": 4330
+    },
+    {
+      "epoch": 0.13839727032111993,
+      "grad_norm": 0.2970675528049469,
+      "learning_rate": 0.001,
+      "loss": 1.363,
+      "step": 4340
+    },
+    {
+      "epoch": 0.13871615804075385,
+      "grad_norm": 0.28739404678344727,
+      "learning_rate": 0.001,
+      "loss": 1.3696,
+      "step": 4350
+    },
+    {
+      "epoch": 0.13903504576038778,
+      "grad_norm": 0.30378416180610657,
+      "learning_rate": 0.001,
+      "loss": 1.366,
+      "step": 4360
+    },
+    {
+      "epoch": 0.13935393348002167,
+      "grad_norm": 0.2950860559940338,
+      "learning_rate": 0.001,
+      "loss": 1.3759,
+      "step": 4370
+    },
+    {
+      "epoch": 0.1396728211996556,
+      "grad_norm": 0.2949100732803345,
+      "learning_rate": 0.001,
+      "loss": 1.357,
+      "step": 4380
+    },
+    {
+      "epoch": 0.13999170891928953,
+      "grad_norm": 0.3102058172225952,
+      "learning_rate": 0.001,
+      "loss": 1.3654,
+      "step": 4390
+    },
+    {
+      "epoch": 0.14031059663892342,
+      "grad_norm": 0.3096378445625305,
+      "learning_rate": 0.001,
+      "loss": 1.3792,
+      "step": 4400
+    },
+    {
+      "epoch": 0.14062948435855735,
+      "grad_norm": 0.29340341687202454,
+      "learning_rate": 0.001,
+      "loss": 1.3812,
+      "step": 4410
+    },
+    {
+      "epoch": 0.14094837207819128,
+      "grad_norm": 0.28675857186317444,
+      "learning_rate": 0.001,
+      "loss": 1.3428,
+      "step": 4420
+    },
+    {
+      "epoch": 0.14126725979782517,
+      "grad_norm": 0.2995110750198364,
+      "learning_rate": 0.001,
+      "loss": 1.3632,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1415861475174591,
+      "grad_norm": 0.2885999381542206,
+      "learning_rate": 0.001,
+      "loss": 1.347,
+      "step": 4440
+    },
+    {
+      "epoch": 0.14190503523709302,
+      "grad_norm": 0.29834800958633423,
+      "learning_rate": 0.001,
+      "loss": 1.3329,
+      "step": 4450
+    },
+    {
+      "epoch": 0.14222392295672695,
+      "grad_norm": 0.3012324273586273,
+      "learning_rate": 0.001,
+      "loss": 1.3406,
+      "step": 4460
+    },
+    {
+      "epoch": 0.14254281067636085,
+      "grad_norm": 0.3000870943069458,
+      "learning_rate": 0.001,
+      "loss": 1.3511,
+      "step": 4470
+    },
+    {
+      "epoch": 0.14286169839599477,
+      "grad_norm": 0.28593116998672485,
+      "learning_rate": 0.001,
+      "loss": 1.3563,
+      "step": 4480
+    },
+    {
+      "epoch": 0.1431805861156287,
+      "grad_norm": 0.2958747148513794,
+      "learning_rate": 0.001,
+      "loss": 1.3523,
+      "step": 4490
+    },
+    {
+      "epoch": 0.1434994738352626,
+      "grad_norm": 0.2945686876773834,
+      "learning_rate": 0.001,
+      "loss": 1.383,
+      "step": 4500
+    },
+    {
+      "epoch": 0.14381836155489652,
+      "grad_norm": 0.2951934039592743,
+      "learning_rate": 0.001,
+      "loss": 1.3828,
+      "step": 4510
+    },
+    {
+      "epoch": 0.14413724927453045,
+      "grad_norm": 0.2921164035797119,
+      "learning_rate": 0.001,
+      "loss": 1.353,
+      "step": 4520
+    },
+    {
+      "epoch": 0.14445613699416435,
+      "grad_norm": 0.27818742394447327,
+      "learning_rate": 0.001,
+      "loss": 1.348,
+      "step": 4530
+    },
+    {
+      "epoch": 0.14477502471379827,
+      "grad_norm": 0.2907786965370178,
+      "learning_rate": 0.001,
+      "loss": 1.3581,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1450939124334322,
+      "grad_norm": 0.283187597990036,
+      "learning_rate": 0.001,
+      "loss": 1.3547,
+      "step": 4550
+    },
+    {
+      "epoch": 0.1454128001530661,
+      "grad_norm": 0.2873290479183197,
+      "learning_rate": 0.001,
+      "loss": 1.3735,
+      "step": 4560
+    },
+    {
+      "epoch": 0.14573168787270002,
+      "grad_norm": 0.30278632044792175,
+      "learning_rate": 0.001,
+      "loss": 1.3545,
+      "step": 4570
+    },
+    {
+      "epoch": 0.14605057559233395,
+      "grad_norm": 0.28067076206207275,
+      "learning_rate": 0.001,
+      "loss": 1.3435,
+      "step": 4580
+    },
+    {
+      "epoch": 0.14636946331196785,
+      "grad_norm": 0.29316315054893494,
+      "learning_rate": 0.001,
+      "loss": 1.3462,
+      "step": 4590
+    },
+    {
+      "epoch": 0.14668835103160177,
+      "grad_norm": 0.31010857224464417,
+      "learning_rate": 0.001,
+      "loss": 1.3656,
+      "step": 4600
+    },
+    {
+      "epoch": 0.1470072387512357,
+      "grad_norm": 0.2965826690196991,
+      "learning_rate": 0.001,
+      "loss": 1.3534,
+      "step": 4610
+    },
+    {
+      "epoch": 0.1473261264708696,
+      "grad_norm": 0.27814781665802,
+      "learning_rate": 0.001,
+      "loss": 1.3557,
+      "step": 4620
+    },
+    {
+      "epoch": 0.14764501419050352,
+      "grad_norm": 0.28145530819892883,
+      "learning_rate": 0.001,
+      "loss": 1.3534,
+      "step": 4630
+    },
+    {
+      "epoch": 0.14796390191013745,
+      "grad_norm": 0.2858860492706299,
+      "learning_rate": 0.001,
+      "loss": 1.3653,
+      "step": 4640
+    },
+    {
+      "epoch": 0.14828278962977134,
+      "grad_norm": 0.29033321142196655,
+      "learning_rate": 0.001,
+      "loss": 1.3598,
+      "step": 4650
+    },
+    {
+      "epoch": 0.14860167734940527,
+      "grad_norm": 0.33033251762390137,
+      "learning_rate": 0.001,
+      "loss": 1.3537,
+      "step": 4660
+    },
+    {
+      "epoch": 0.1489205650690392,
+      "grad_norm": 0.2896032929420471,
+      "learning_rate": 0.001,
+      "loss": 1.3405,
+      "step": 4670
+    },
+    {
+      "epoch": 0.14923945278867312,
+      "grad_norm": 0.29234012961387634,
+      "learning_rate": 0.001,
+      "loss": 1.3518,
+      "step": 4680
+    },
+    {
+      "epoch": 0.14955834050830702,
+      "grad_norm": 0.2925928831100464,
+      "learning_rate": 0.001,
+      "loss": 1.3544,
+      "step": 4690
+    },
+    {
+      "epoch": 0.14987722822794095,
+      "grad_norm": 0.3034275472164154,
+      "learning_rate": 0.001,
+      "loss": 1.3266,
+      "step": 4700
+    },
+    {
+      "epoch": 0.15019611594757487,
+      "grad_norm": 0.29347124695777893,
+      "learning_rate": 0.001,
+      "loss": 1.3444,
+      "step": 4710
+    },
+    {
+      "epoch": 0.15051500366720877,
+      "grad_norm": 0.30910998582839966,
+      "learning_rate": 0.001,
+      "loss": 1.333,
+      "step": 4720
+    },
+    {
+      "epoch": 0.1508338913868427,
+      "grad_norm": 0.292319655418396,
+      "learning_rate": 0.001,
+      "loss": 1.3427,
+      "step": 4730
+    },
+    {
+      "epoch": 0.15115277910647662,
+      "grad_norm": 0.29981669783592224,
+      "learning_rate": 0.001,
+      "loss": 1.3413,
+      "step": 4740
+    },
+    {
+      "epoch": 0.15147166682611052,
+      "grad_norm": 0.29529082775115967,
+      "learning_rate": 0.001,
+      "loss": 1.3301,
+      "step": 4750
+    },
+    {
+      "epoch": 0.15179055454574444,
+      "grad_norm": 0.26876968145370483,
+      "learning_rate": 0.001,
+      "loss": 1.3274,
+      "step": 4760
+    },
+    {
+      "epoch": 0.15210944226537837,
+      "grad_norm": 0.2886004149913788,
+      "learning_rate": 0.001,
+      "loss": 1.3412,
+      "step": 4770
+    },
+    {
+      "epoch": 0.15242832998501227,
+      "grad_norm": 0.27672141790390015,
+      "learning_rate": 0.001,
+      "loss": 1.3268,
+      "step": 4780
+    },
+    {
+      "epoch": 0.1527472177046462,
+      "grad_norm": 0.3041819930076599,
+      "learning_rate": 0.001,
+      "loss": 1.3546,
+      "step": 4790
+    },
+    {
+      "epoch": 0.15306610542428012,
+      "grad_norm": 0.2882234454154968,
+      "learning_rate": 0.001,
+      "loss": 1.3284,
+      "step": 4800
+    },
+    {
+      "epoch": 0.15338499314391402,
+      "grad_norm": 0.2742898464202881,
+      "learning_rate": 0.001,
+      "loss": 1.3522,
+      "step": 4810
+    },
+    {
+      "epoch": 0.15370388086354794,
+      "grad_norm": 0.2767809331417084,
+      "learning_rate": 0.001,
+      "loss": 1.3441,
+      "step": 4820
+    },
+    {
+      "epoch": 0.15402276858318187,
+      "grad_norm": 0.2767125070095062,
+      "learning_rate": 0.001,
+      "loss": 1.3277,
+      "step": 4830
+    },
+    {
+      "epoch": 0.15434165630281577,
+      "grad_norm": 0.28059667348861694,
+      "learning_rate": 0.001,
+      "loss": 1.3457,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1546605440224497,
+      "grad_norm": 0.28245338797569275,
+      "learning_rate": 0.001,
+      "loss": 1.3371,
+      "step": 4850
+    },
+    {
+      "epoch": 0.15497943174208362,
+      "grad_norm": 0.27916160225868225,
+      "learning_rate": 0.001,
+      "loss": 1.3402,
+      "step": 4860
+    },
+    {
+      "epoch": 0.15529831946171752,
+      "grad_norm": 0.2856876850128174,
+      "learning_rate": 0.001,
+      "loss": 1.3189,
+      "step": 4870
+    },
+    {
+      "epoch": 0.15561720718135144,
+      "grad_norm": 0.2746897339820862,
+      "learning_rate": 0.001,
+      "loss": 1.3146,
+      "step": 4880
+    },
+    {
+      "epoch": 0.15593609490098537,
+      "grad_norm": 0.2816491723060608,
+      "learning_rate": 0.001,
+      "loss": 1.3357,
+      "step": 4890
+    },
+    {
+      "epoch": 0.1562549826206193,
+      "grad_norm": 0.2790864408016205,
+      "learning_rate": 0.001,
+      "loss": 1.3435,
+      "step": 4900
+    },
+    {
+      "epoch": 0.1565738703402532,
+      "grad_norm": 0.2920440137386322,
+      "learning_rate": 0.001,
+      "loss": 1.3517,
+      "step": 4910
+    },
+    {
+      "epoch": 0.15689275805988712,
+      "grad_norm": 0.28280529379844666,
+      "learning_rate": 0.001,
+      "loss": 1.3371,
+      "step": 4920
+    },
+    {
+      "epoch": 0.15721164577952104,
+      "grad_norm": 0.27726370096206665,
+      "learning_rate": 0.001,
+      "loss": 1.3402,
+      "step": 4930
+    },
+    {
+      "epoch": 0.15753053349915494,
+      "grad_norm": 0.27465248107910156,
+      "learning_rate": 0.001,
+      "loss": 1.315,
+      "step": 4940
+    },
+    {
+      "epoch": 0.15784942121878887,
+      "grad_norm": 0.281573086977005,
+      "learning_rate": 0.001,
+      "loss": 1.313,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1581683089384228,
+      "grad_norm": 0.2761354446411133,
+      "learning_rate": 0.001,
+      "loss": 1.3402,
+      "step": 4960
+    },
+    {
+      "epoch": 0.1584871966580567,
+      "grad_norm": 0.28072458505630493,
+      "learning_rate": 0.001,
+      "loss": 1.3357,
+      "step": 4970
+    },
+    {
+      "epoch": 0.15880608437769062,
+      "grad_norm": 0.2800447344779968,
+      "learning_rate": 0.001,
+      "loss": 1.3229,
+      "step": 4980
+    },
+    {
+      "epoch": 0.15912497209732454,
+      "grad_norm": 0.28519025444984436,
+      "learning_rate": 0.001,
+      "loss": 1.3226,
+      "step": 4990
+    },
+    {
+      "epoch": 0.15944385981695844,
+      "grad_norm": 0.27568909525871277,
+      "learning_rate": 0.001,
+      "loss": 1.3009,
+      "step": 5000
+    },
+    {
+      "epoch": 0.15976274753659236,
+      "grad_norm": 0.28567227721214294,
+      "learning_rate": 0.001,
+      "loss": 1.3363,
+      "step": 5010
+    },
+    {
+      "epoch": 0.1600816352562263,
+      "grad_norm": 0.2807476222515106,
+      "learning_rate": 0.001,
+      "loss": 1.3158,
+      "step": 5020
+    },
+    {
+      "epoch": 0.1604005229758602,
+      "grad_norm": 0.2860855758190155,
+      "learning_rate": 0.001,
+      "loss": 1.3546,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1607194106954941,
+      "grad_norm": 0.27725300192832947,
+      "learning_rate": 0.001,
+      "loss": 1.3362,
+      "step": 5040
+    },
+    {
+      "epoch": 0.16103829841512804,
+      "grad_norm": 0.27752354741096497,
+      "learning_rate": 0.001,
+      "loss": 1.3208,
+      "step": 5050
+    },
+    {
+      "epoch": 0.16135718613476194,
+      "grad_norm": 0.27968016266822815,
+      "learning_rate": 0.001,
+      "loss": 1.3253,
+      "step": 5060
+    },
+    {
+      "epoch": 0.16167607385439586,
+      "grad_norm": 0.2815000116825104,
+      "learning_rate": 0.001,
+      "loss": 1.316,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1619949615740298,
+      "grad_norm": 0.28594574332237244,
+      "learning_rate": 0.001,
+      "loss": 1.3247,
+      "step": 5080
+    },
+    {
+      "epoch": 0.16231384929366371,
+      "grad_norm": 0.2727883756160736,
+      "learning_rate": 0.001,
+      "loss": 1.3486,
+      "step": 5090
+    },
+    {
+      "epoch": 0.1626327370132976,
+      "grad_norm": 0.27950167655944824,
+      "learning_rate": 0.001,
+      "loss": 1.3368,
+      "step": 5100
+    },
+    {
+      "epoch": 0.16295162473293154,
+      "grad_norm": 0.28539350628852844,
+      "learning_rate": 0.001,
+      "loss": 1.328,
+      "step": 5110
+    },
+    {
+      "epoch": 0.16327051245256546,
+      "grad_norm": 0.280464231967926,
+      "learning_rate": 0.001,
+      "loss": 1.3091,
+      "step": 5120
+    },
+    {
+      "epoch": 0.16358940017219936,
+      "grad_norm": 0.2751573622226715,
+      "learning_rate": 0.001,
+      "loss": 1.3242,
+      "step": 5130
+    },
+    {
+      "epoch": 0.1639082878918333,
+      "grad_norm": 0.27065640687942505,
+      "learning_rate": 0.001,
+      "loss": 1.3189,
+      "step": 5140
+    },
+    {
+      "epoch": 0.1642271756114672,
+      "grad_norm": 0.27752992510795593,
+      "learning_rate": 0.001,
+      "loss": 1.325,
+      "step": 5150
+    },
+    {
+      "epoch": 0.1645460633311011,
+      "grad_norm": 0.2825421988964081,
+      "learning_rate": 0.001,
+      "loss": 1.3221,
+      "step": 5160
+    },
+    {
+      "epoch": 0.16486495105073504,
+      "grad_norm": 0.2743771970272064,
+      "learning_rate": 0.001,
+      "loss": 1.3272,
+      "step": 5170
+    },
+    {
+      "epoch": 0.16518383877036896,
+      "grad_norm": 0.2885359823703766,
+      "learning_rate": 0.001,
+      "loss": 1.3362,
+      "step": 5180
+    },
+    {
+      "epoch": 0.16550272649000286,
+      "grad_norm": 0.289696604013443,
+      "learning_rate": 0.001,
+      "loss": 1.345,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1658216142096368,
+      "grad_norm": 0.26721301674842834,
+      "learning_rate": 0.001,
+      "loss": 1.3037,
+      "step": 5200
+    },
+    {
+      "epoch": 0.1661405019292707,
+      "grad_norm": 0.27076032757759094,
+      "learning_rate": 0.001,
+      "loss": 1.3044,
+      "step": 5210
+    },
+    {
+      "epoch": 0.1664593896489046,
+      "grad_norm": 0.27800723910331726,
+      "learning_rate": 0.001,
+      "loss": 1.308,
+      "step": 5220
+    },
+    {
+      "epoch": 0.16677827736853854,
+      "grad_norm": 0.28856852650642395,
+      "learning_rate": 0.001,
+      "loss": 1.3066,
+      "step": 5230
+    },
+    {
+      "epoch": 0.16709716508817246,
+      "grad_norm": 0.27235889434814453,
+      "learning_rate": 0.001,
+      "loss": 1.3236,
+      "step": 5240
+    },
+    {
+      "epoch": 0.16741605280780636,
+      "grad_norm": 0.2833673655986786,
+      "learning_rate": 0.001,
+      "loss": 1.3194,
+      "step": 5250
+    },
+    {
+      "epoch": 0.16773494052744028,
+      "grad_norm": 0.28075218200683594,
+      "learning_rate": 0.001,
+      "loss": 1.313,
+      "step": 5260
+    },
+    {
+      "epoch": 0.1680538282470742,
+      "grad_norm": 0.28121885657310486,
+      "learning_rate": 0.001,
+      "loss": 1.3013,
+      "step": 5270
+    },
+    {
+      "epoch": 0.1683727159667081,
+      "grad_norm": 0.28455910086631775,
+      "learning_rate": 0.001,
+      "loss": 1.3183,
+      "step": 5280
+    },
+    {
+      "epoch": 0.16869160368634203,
+      "grad_norm": 0.27447423338890076,
+      "learning_rate": 0.001,
+      "loss": 1.3102,
+      "step": 5290
+    },
+    {
+      "epoch": 0.16901049140597596,
+      "grad_norm": 0.2774691581726074,
+      "learning_rate": 0.001,
+      "loss": 1.3086,
+      "step": 5300
+    },
+    {
+      "epoch": 0.16932937912560989,
+      "grad_norm": 0.27635547518730164,
+      "learning_rate": 0.001,
+      "loss": 1.3152,
+      "step": 5310
+    },
+    {
+      "epoch": 0.16964826684524378,
+      "grad_norm": 0.28113260865211487,
+      "learning_rate": 0.001,
+      "loss": 1.3087,
+      "step": 5320
+    },
+    {
+      "epoch": 0.1699671545648777,
+      "grad_norm": 0.27385708689689636,
+      "learning_rate": 0.001,
+      "loss": 1.3259,
+      "step": 5330
+    },
+    {
+      "epoch": 0.17028604228451164,
+      "grad_norm": 0.2752876579761505,
+      "learning_rate": 0.001,
+      "loss": 1.3323,
+      "step": 5340
+    },
+    {
+      "epoch": 0.17060493000414553,
+      "grad_norm": 0.29049137234687805,
+      "learning_rate": 0.001,
+      "loss": 1.3128,
+      "step": 5350
+    },
+    {
+      "epoch": 0.17092381772377946,
+      "grad_norm": 0.2888701558113098,
+      "learning_rate": 0.001,
+      "loss": 1.3111,
+      "step": 5360
+    },
+    {
+      "epoch": 0.17124270544341338,
+      "grad_norm": 0.27347761392593384,
+      "learning_rate": 0.001,
+      "loss": 1.2978,
+      "step": 5370
+    },
+    {
+      "epoch": 0.17156159316304728,
+      "grad_norm": 0.26778650283813477,
+      "learning_rate": 0.001,
+      "loss": 1.2873,
+      "step": 5380
+    },
+    {
+      "epoch": 0.1718804808826812,
+      "grad_norm": 0.26770761609077454,
+      "learning_rate": 0.001,
+      "loss": 1.3033,
+      "step": 5390
+    },
+    {
+      "epoch": 0.17219936860231513,
+      "grad_norm": 0.27429458498954773,
+      "learning_rate": 0.001,
+      "loss": 1.3285,
+      "step": 5400
+    },
+    {
+      "epoch": 0.17251825632194903,
+      "grad_norm": 0.26610255241394043,
+      "learning_rate": 0.001,
+      "loss": 1.2999,
+      "step": 5410
+    },
+    {
+      "epoch": 0.17283714404158296,
+      "grad_norm": 0.2606316804885864,
+      "learning_rate": 0.001,
+      "loss": 1.3159,
+      "step": 5420
+    },
+    {
+      "epoch": 0.17315603176121688,
+      "grad_norm": 0.2717187702655792,
+      "learning_rate": 0.001,
+      "loss": 1.3083,
+      "step": 5430
+    },
+    {
+      "epoch": 0.17347491948085078,
+      "grad_norm": 0.2680508494377136,
+      "learning_rate": 0.001,
+      "loss": 1.3116,
+      "step": 5440
+    },
+    {
+      "epoch": 0.1737938072004847,
+      "grad_norm": 0.2761911153793335,
+      "learning_rate": 0.001,
+      "loss": 1.3151,
+      "step": 5450
+    },
+    {
+      "epoch": 0.17411269492011863,
+      "grad_norm": 0.28334981203079224,
+      "learning_rate": 0.001,
+      "loss": 1.2956,
+      "step": 5460
+    },
+    {
+      "epoch": 0.17443158263975253,
+      "grad_norm": 0.2760418653488159,
+      "learning_rate": 0.001,
+      "loss": 1.3197,
+      "step": 5470
+    },
+    {
+      "epoch": 0.17475047035938646,
+      "grad_norm": 0.2959541380405426,
+      "learning_rate": 0.001,
+      "loss": 1.3219,
+      "step": 5480
+    },
+    {
+      "epoch": 0.17506935807902038,
+      "grad_norm": 0.27441099286079407,
+      "learning_rate": 0.001,
+      "loss": 1.2796,
+      "step": 5490
+    },
+    {
+      "epoch": 0.1753882457986543,
+      "grad_norm": 0.27085989713668823,
+      "learning_rate": 0.001,
+      "loss": 1.285,
+      "step": 5500
+    },
+    {
+      "epoch": 0.1757071335182882,
+      "grad_norm": 0.27273258566856384,
+      "learning_rate": 0.001,
+      "loss": 1.3222,
+      "step": 5510
+    },
+    {
+      "epoch": 0.17602602123792213,
+      "grad_norm": 0.2756020724773407,
+      "learning_rate": 0.001,
+      "loss": 1.318,
+      "step": 5520
+    },
+    {
+      "epoch": 0.17634490895755606,
+      "grad_norm": 0.276136577129364,
+      "learning_rate": 0.001,
+      "loss": 1.2945,
+      "step": 5530
+    },
+    {
+      "epoch": 0.17666379667718995,
+      "grad_norm": 0.27136778831481934,
+      "learning_rate": 0.001,
+      "loss": 1.3012,
+      "step": 5540
+    },
+    {
+      "epoch": 0.17698268439682388,
+      "grad_norm": 0.2842538058757782,
+      "learning_rate": 0.001,
+      "loss": 1.315,
+      "step": 5550
+    },
+    {
+      "epoch": 0.1773015721164578,
+      "grad_norm": 0.27779433131217957,
+      "learning_rate": 0.001,
+      "loss": 1.3073,
+      "step": 5560
+    },
+    {
+      "epoch": 0.1776204598360917,
+      "grad_norm": 0.27324429154396057,
+      "learning_rate": 0.001,
+      "loss": 1.2957,
+      "step": 5570
+    },
+    {
+      "epoch": 0.17793934755572563,
+      "grad_norm": 0.272482305765152,
+      "learning_rate": 0.001,
+      "loss": 1.2816,
+      "step": 5580
+    },
+    {
+      "epoch": 0.17825823527535956,
+      "grad_norm": 0.2856181561946869,
+      "learning_rate": 0.001,
+      "loss": 1.322,
+      "step": 5590
+    },
+    {
+      "epoch": 0.17857712299499345,
+      "grad_norm": 0.27352333068847656,
+      "learning_rate": 0.001,
+      "loss": 1.3114,
+      "step": 5600
+    },
+    {
+      "epoch": 0.17889601071462738,
+      "grad_norm": 0.27149438858032227,
+      "learning_rate": 0.001,
+      "loss": 1.3057,
+      "step": 5610
+    },
+    {
+      "epoch": 0.1792148984342613,
+      "grad_norm": 0.2803710997104645,
+      "learning_rate": 0.001,
+      "loss": 1.3219,
+      "step": 5620
+    },
+    {
+      "epoch": 0.1795337861538952,
+      "grad_norm": 0.26612791419029236,
+      "learning_rate": 0.001,
+      "loss": 1.3059,
+      "step": 5630
+    },
+    {
+      "epoch": 0.17985267387352913,
+      "grad_norm": 0.2668563425540924,
+      "learning_rate": 0.001,
+      "loss": 1.2957,
+      "step": 5640
+    },
+    {
+      "epoch": 0.18017156159316305,
+      "grad_norm": 0.2617603540420532,
+      "learning_rate": 0.001,
+      "loss": 1.3134,
+      "step": 5650
+    },
+    {
+      "epoch": 0.18049044931279695,
+      "grad_norm": 0.26398953795433044,
+      "learning_rate": 0.001,
+      "loss": 1.3042,
+      "step": 5660
+    },
+    {
+      "epoch": 0.18080933703243088,
+      "grad_norm": 0.2688938081264496,
+      "learning_rate": 0.001,
+      "loss": 1.3085,
+      "step": 5670
+    },
+    {
+      "epoch": 0.1811282247520648,
+      "grad_norm": 0.2666449248790741,
+      "learning_rate": 0.001,
+      "loss": 1.2839,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1814471124716987,
+      "grad_norm": 0.2589948773384094,
+      "learning_rate": 0.001,
+      "loss": 1.2923,
+      "step": 5690
+    },
+    {
+      "epoch": 0.18176600019133263,
+      "grad_norm": 0.2649987041950226,
+      "learning_rate": 0.001,
+      "loss": 1.2997,
+      "step": 5700
+    },
+    {
+      "epoch": 0.18208488791096655,
+      "grad_norm": 0.26647627353668213,
+      "learning_rate": 0.001,
+      "loss": 1.307,
+      "step": 5710
+    },
+    {
+      "epoch": 0.18240377563060048,
+      "grad_norm": 0.2552083432674408,
+      "learning_rate": 0.001,
+      "loss": 1.3178,
+      "step": 5720
+    },
+    {
+      "epoch": 0.18272266335023438,
+      "grad_norm": 0.28030291199684143,
+      "learning_rate": 0.001,
+      "loss": 1.2808,
+      "step": 5730
+    },
+    {
+      "epoch": 0.1830415510698683,
+      "grad_norm": 0.2664260268211365,
+      "learning_rate": 0.001,
+      "loss": 1.2746,
+      "step": 5740
+    },
+    {
+      "epoch": 0.18336043878950223,
+      "grad_norm": 0.258434921503067,
+      "learning_rate": 0.001,
+      "loss": 1.3166,
+      "step": 5750
+    },
+    {
+      "epoch": 0.18367932650913613,
+      "grad_norm": 0.27148333191871643,
+      "learning_rate": 0.001,
+      "loss": 1.3001,
+      "step": 5760
+    },
+    {
+      "epoch": 0.18399821422877005,
+      "grad_norm": 0.2850393056869507,
+      "learning_rate": 0.001,
+      "loss": 1.296,
+      "step": 5770
+    },
+    {
+      "epoch": 0.18431710194840398,
+      "grad_norm": 0.2537064254283905,
+      "learning_rate": 0.001,
+      "loss": 1.2861,
+      "step": 5780
+    },
+    {
+      "epoch": 0.18463598966803788,
+      "grad_norm": 0.26095375418663025,
+      "learning_rate": 0.001,
+      "loss": 1.2824,
+      "step": 5790
+    },
+    {
+      "epoch": 0.1849548773876718,
+      "grad_norm": 0.2742947041988373,
+      "learning_rate": 0.001,
+      "loss": 1.294,
+      "step": 5800
+    },
+    {
+      "epoch": 0.18527376510730573,
+      "grad_norm": 0.2614261209964752,
+      "learning_rate": 0.001,
+      "loss": 1.2955,
+      "step": 5810
+    },
+    {
+      "epoch": 0.18559265282693962,
+      "grad_norm": 0.2708014249801636,
+      "learning_rate": 0.001,
+      "loss": 1.2847,
+      "step": 5820
+    },
+    {
+      "epoch": 0.18591154054657355,
+      "grad_norm": 0.25221359729766846,
+      "learning_rate": 0.001,
+      "loss": 1.2916,
+      "step": 5830
+    },
+    {
+      "epoch": 0.18623042826620748,
+      "grad_norm": 0.2672080397605896,
+      "learning_rate": 0.001,
+      "loss": 1.2905,
+      "step": 5840
+    },
+    {
+      "epoch": 0.18654931598584137,
+      "grad_norm": 0.27737823128700256,
+      "learning_rate": 0.001,
+      "loss": 1.2926,
+      "step": 5850
+    },
+    {
+      "epoch": 0.1868682037054753,
+      "grad_norm": 0.2654053270816803,
+      "learning_rate": 0.001,
+      "loss": 1.2916,
+      "step": 5860
+    },
+    {
+      "epoch": 0.18718709142510923,
+      "grad_norm": 0.25915852189064026,
+      "learning_rate": 0.001,
+      "loss": 1.2842,
+      "step": 5870
+    },
+    {
+      "epoch": 0.18750597914474312,
+      "grad_norm": 0.26840394735336304,
+      "learning_rate": 0.001,
+      "loss": 1.3053,
+      "step": 5880
+    },
+    {
+      "epoch": 0.18782486686437705,
+      "grad_norm": 0.2692258954048157,
+      "learning_rate": 0.001,
+      "loss": 1.2909,
+      "step": 5890
+    },
+    {
+      "epoch": 0.18814375458401097,
+      "grad_norm": 0.2546669840812683,
+      "learning_rate": 0.001,
+      "loss": 1.2853,
+      "step": 5900
+    },
+    {
+      "epoch": 0.18846264230364487,
+      "grad_norm": 0.27450940012931824,
+      "learning_rate": 0.001,
+      "loss": 1.2816,
+      "step": 5910
+    },
+    {
+      "epoch": 0.1887815300232788,
+      "grad_norm": 0.2487523853778839,
+      "learning_rate": 0.001,
+      "loss": 1.2828,
+      "step": 5920
+    },
+    {
+      "epoch": 0.18910041774291272,
+      "grad_norm": 0.2829253375530243,
+      "learning_rate": 0.001,
+      "loss": 1.2841,
+      "step": 5930
+    },
+    {
+      "epoch": 0.18941930546254665,
+      "grad_norm": 0.26952558755874634,
+      "learning_rate": 0.001,
+      "loss": 1.2958,
+      "step": 5940
+    },
+    {
+      "epoch": 0.18973819318218055,
+      "grad_norm": 0.25571727752685547,
+      "learning_rate": 0.001,
+      "loss": 1.2796,
+      "step": 5950
+    },
+    {
+      "epoch": 0.19005708090181447,
+      "grad_norm": 0.26292943954467773,
+      "learning_rate": 0.001,
+      "loss": 1.3109,
+      "step": 5960
+    },
+    {
+      "epoch": 0.1903759686214484,
+      "grad_norm": 0.2683652937412262,
+      "learning_rate": 0.001,
+      "loss": 1.2942,
+      "step": 5970
+    },
+    {
+      "epoch": 0.1906948563410823,
+      "grad_norm": 0.2673659324645996,
+      "learning_rate": 0.001,
+      "loss": 1.2799,
+      "step": 5980
+    },
+    {
+      "epoch": 0.19101374406071622,
+      "grad_norm": 0.2598000764846802,
+      "learning_rate": 0.001,
+      "loss": 1.2904,
+      "step": 5990
+    },
+    {
+      "epoch": 0.19133263178035015,
+      "grad_norm": 0.2635900676250458,
+      "learning_rate": 0.001,
+      "loss": 1.2705,
+      "step": 6000
+    },
+    {
+      "epoch": 0.19165151949998405,
+      "grad_norm": 0.25515443086624146,
+      "learning_rate": 0.001,
+      "loss": 1.2951,
+      "step": 6010
+    },
+    {
+      "epoch": 0.19197040721961797,
+      "grad_norm": 0.2600856423377991,
+      "learning_rate": 0.001,
+      "loss": 1.2733,
+      "step": 6020
+    },
+    {
+      "epoch": 0.1922892949392519,
+      "grad_norm": 0.2564580738544464,
+      "learning_rate": 0.001,
+      "loss": 1.2617,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1926081826588858,
+      "grad_norm": 0.2747453451156616,
+      "learning_rate": 0.001,
+      "loss": 1.2833,
+      "step": 6040
+    },
+    {
+      "epoch": 0.19292707037851972,
+      "grad_norm": 0.26666703820228577,
+      "learning_rate": 0.001,
+      "loss": 1.2765,
+      "step": 6050
+    },
+    {
+      "epoch": 0.19324595809815365,
+      "grad_norm": 0.27212536334991455,
+      "learning_rate": 0.001,
+      "loss": 1.2701,
+      "step": 6060
+    },
+    {
+      "epoch": 0.19356484581778755,
+      "grad_norm": 0.2613239288330078,
+      "learning_rate": 0.001,
+      "loss": 1.278,
+      "step": 6070
+    },
+    {
+      "epoch": 0.19388373353742147,
+      "grad_norm": 0.27437734603881836,
+      "learning_rate": 0.001,
+      "loss": 1.275,
+      "step": 6080
+    },
+    {
+      "epoch": 0.1942026212570554,
+      "grad_norm": 0.26033124327659607,
+      "learning_rate": 0.001,
+      "loss": 1.2795,
+      "step": 6090
+    },
+    {
+      "epoch": 0.1945215089766893,
+      "grad_norm": 0.25559207797050476,
+      "learning_rate": 0.001,
+      "loss": 1.2808,
+      "step": 6100
+    },
+    {
+      "epoch": 0.19484039669632322,
+      "grad_norm": 0.26534515619277954,
+      "learning_rate": 0.001,
+      "loss": 1.2914,
+      "step": 6110
+    },
+    {
+      "epoch": 0.19515928441595715,
+      "grad_norm": 0.27692893147468567,
+      "learning_rate": 0.001,
+      "loss": 1.2815,
+      "step": 6120
+    },
+    {
+      "epoch": 0.19547817213559107,
+      "grad_norm": 0.27069565653800964,
+      "learning_rate": 0.001,
+      "loss": 1.2594,
+      "step": 6130
+    },
+    {
+      "epoch": 0.19579705985522497,
+      "grad_norm": 0.2529548108577728,
+      "learning_rate": 0.001,
+      "loss": 1.2881,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1961159475748589,
+      "grad_norm": 0.25786975026130676,
+      "learning_rate": 0.001,
+      "loss": 1.2543,
+      "step": 6150
+    },
+    {
+      "epoch": 0.19643483529449282,
+      "grad_norm": 0.25865936279296875,
+      "learning_rate": 0.001,
+      "loss": 1.2623,
+      "step": 6160
+    },
+    {
+      "epoch": 0.19675372301412672,
+      "grad_norm": 0.2520538568496704,
+      "learning_rate": 0.001,
+      "loss": 1.2958,
+      "step": 6170
+    },
+    {
+      "epoch": 0.19707261073376064,
+      "grad_norm": 0.26968657970428467,
+      "learning_rate": 0.001,
+      "loss": 1.2757,
+      "step": 6180
+    },
+    {
+      "epoch": 0.19739149845339457,
+      "grad_norm": 0.2608780264854431,
+      "learning_rate": 0.001,
+      "loss": 1.2955,
+      "step": 6190
+    },
+    {
+      "epoch": 0.19771038617302847,
+      "grad_norm": 0.2636774182319641,
+      "learning_rate": 0.001,
+      "loss": 1.281,
+      "step": 6200
+    },
+    {
+      "epoch": 0.1980292738926624,
+      "grad_norm": 0.250747948884964,
+      "learning_rate": 0.001,
+      "loss": 1.286,
+      "step": 6210
+    },
+    {
+      "epoch": 0.19834816161229632,
+      "grad_norm": 0.2567255198955536,
+      "learning_rate": 0.001,
+      "loss": 1.258,
+      "step": 6220
+    },
+    {
+      "epoch": 0.19866704933193022,
+      "grad_norm": 0.2622133195400238,
+      "learning_rate": 0.001,
+      "loss": 1.2623,
+      "step": 6230
+    },
+    {
+      "epoch": 0.19898593705156414,
+      "grad_norm": 0.2595137655735016,
+      "learning_rate": 0.001,
+      "loss": 1.2697,
+      "step": 6240
+    },
+    {
+      "epoch": 0.19930482477119807,
+      "grad_norm": 0.25806212425231934,
+      "learning_rate": 0.001,
+      "loss": 1.2863,
+      "step": 6250
+    },
+    {
+      "epoch": 0.19962371249083197,
+      "grad_norm": 0.2540823817253113,
+      "learning_rate": 0.001,
+      "loss": 1.2602,
+      "step": 6260
+    },
+    {
+      "epoch": 0.1999426002104659,
+      "grad_norm": 0.26109036803245544,
+      "learning_rate": 0.001,
+      "loss": 1.277,
+      "step": 6270
+    },
+    {
+      "epoch": 0.20026148793009982,
+      "grad_norm": 0.25082194805145264,
+      "learning_rate": 0.001,
+      "loss": 1.2713,
+      "step": 6280
+    },
+    {
+      "epoch": 0.20058037564973372,
+      "grad_norm": 0.26120105385780334,
+      "learning_rate": 0.001,
+      "loss": 1.2857,
+      "step": 6290
+    },
+    {
+      "epoch": 0.20089926336936764,
+      "grad_norm": 0.2576233446598053,
+      "learning_rate": 0.001,
+      "loss": 1.2767,
+      "step": 6300
+    },
+    {
+      "epoch": 0.20121815108900157,
+      "grad_norm": 0.2628646194934845,
+      "learning_rate": 0.001,
+      "loss": 1.272,
+      "step": 6310
+    },
+    {
+      "epoch": 0.20153703880863547,
+      "grad_norm": 0.2471473067998886,
+      "learning_rate": 0.001,
+      "loss": 1.2702,
+      "step": 6320
+    },
+    {
+      "epoch": 0.2018559265282694,
+      "grad_norm": 0.2571963369846344,
+      "learning_rate": 0.001,
+      "loss": 1.2662,
+      "step": 6330
+    },
+    {
+      "epoch": 0.20217481424790332,
+      "grad_norm": 0.25734174251556396,
+      "learning_rate": 0.001,
+      "loss": 1.2718,
+      "step": 6340
+    },
+    {
+      "epoch": 0.20249370196753724,
+      "grad_norm": 0.25071901082992554,
+      "learning_rate": 0.001,
+      "loss": 1.2618,
+      "step": 6350
+    },
+    {
+      "epoch": 0.20281258968717114,
+      "grad_norm": 0.25671815872192383,
+      "learning_rate": 0.001,
+      "loss": 1.2696,
+      "step": 6360
+    },
+    {
+      "epoch": 0.20313147740680507,
+      "grad_norm": 0.2598605453968048,
+      "learning_rate": 0.001,
+      "loss": 1.2859,
+      "step": 6370
+    },
+    {
+      "epoch": 0.203450365126439,
+      "grad_norm": 0.26527634263038635,
+      "learning_rate": 0.001,
+      "loss": 1.2794,
+      "step": 6380
+    },
+    {
+      "epoch": 0.2037692528460729,
+      "grad_norm": 0.26094022393226624,
+      "learning_rate": 0.001,
+      "loss": 1.2636,
+      "step": 6390
+    },
+    {
+      "epoch": 0.20408814056570682,
+      "grad_norm": 0.2590426802635193,
+      "learning_rate": 0.001,
+      "loss": 1.2553,
+      "step": 6400
+    },
+    {
+      "epoch": 0.20440702828534074,
+      "grad_norm": 0.25953012704849243,
+      "learning_rate": 0.001,
+      "loss": 1.2787,
+      "step": 6410
+    },
+    {
+      "epoch": 0.20472591600497464,
+      "grad_norm": 0.25189971923828125,
+      "learning_rate": 0.001,
+      "loss": 1.2597,
+      "step": 6420
+    },
+    {
+      "epoch": 0.20504480372460857,
+      "grad_norm": 0.2543141841888428,
+      "learning_rate": 0.001,
+      "loss": 1.2896,
+      "step": 6430
+    },
+    {
+      "epoch": 0.2053636914442425,
+      "grad_norm": 0.24777406454086304,
+      "learning_rate": 0.001,
+      "loss": 1.2656,
+      "step": 6440
+    },
+    {
+      "epoch": 0.2056825791638764,
+      "grad_norm": 0.25355079770088196,
+      "learning_rate": 0.001,
+      "loss": 1.2605,
+      "step": 6450
+    },
+    {
+      "epoch": 0.20600146688351031,
+      "grad_norm": 0.2523370683193207,
+      "learning_rate": 0.001,
+      "loss": 1.2617,
+      "step": 6460
+    },
+    {
+      "epoch": 0.20632035460314424,
+      "grad_norm": 0.2629269063472748,
+      "learning_rate": 0.001,
+      "loss": 1.2741,
+      "step": 6470
+    },
+    {
+      "epoch": 0.20663924232277814,
+      "grad_norm": 0.2504100799560547,
+      "learning_rate": 0.001,
+      "loss": 1.2715,
+      "step": 6480
+    },
+    {
+      "epoch": 0.20695813004241206,
+      "grad_norm": 0.2450273036956787,
+      "learning_rate": 0.001,
+      "loss": 1.2705,
+      "step": 6490
+    },
+    {
+      "epoch": 0.207277017762046,
+      "grad_norm": 0.2537860870361328,
+      "learning_rate": 0.001,
+      "loss": 1.2565,
+      "step": 6500
+    },
+    {
+      "epoch": 0.2075959054816799,
+      "grad_norm": 0.24981524050235748,
+      "learning_rate": 0.001,
+      "loss": 1.2587,
+      "step": 6510
+    },
+    {
+      "epoch": 0.2079147932013138,
+      "grad_norm": 0.24786381423473358,
+      "learning_rate": 0.001,
+      "loss": 1.2628,
+      "step": 6520
+    },
+    {
+      "epoch": 0.20823368092094774,
+      "grad_norm": 0.25516125559806824,
+      "learning_rate": 0.001,
+      "loss": 1.2695,
+      "step": 6530
+    },
+    {
+      "epoch": 0.20855256864058166,
+      "grad_norm": 0.24738313257694244,
+      "learning_rate": 0.001,
+      "loss": 1.2857,
+      "step": 6540
+    },
+    {
+      "epoch": 0.20887145636021556,
+      "grad_norm": 0.25449132919311523,
+      "learning_rate": 0.001,
+      "loss": 1.2663,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2091903440798495,
+      "grad_norm": 0.24622990190982819,
+      "learning_rate": 0.001,
+      "loss": 1.2639,
+      "step": 6560
+    },
+    {
+      "epoch": 0.20950923179948341,
+      "grad_norm": 0.24932384490966797,
+      "learning_rate": 0.001,
+      "loss": 1.2662,
+      "step": 6570
+    },
+    {
+      "epoch": 0.2098281195191173,
+      "grad_norm": 0.24670833349227905,
+      "learning_rate": 0.001,
+      "loss": 1.2637,
+      "step": 6580
+    },
+    {
+      "epoch": 0.21014700723875124,
+      "grad_norm": 0.24999317526817322,
+      "learning_rate": 0.001,
+      "loss": 1.2673,
+      "step": 6590
+    },
+    {
+      "epoch": 0.21046589495838516,
+      "grad_norm": 0.2563895881175995,
+      "learning_rate": 0.001,
+      "loss": 1.249,
+      "step": 6600
+    },
+    {
+      "epoch": 0.21078478267801906,
+      "grad_norm": 0.2394222617149353,
+      "learning_rate": 0.001,
+      "loss": 1.2653,
+      "step": 6610
+    },
+    {
+      "epoch": 0.211103670397653,
+      "grad_norm": 0.24972309172153473,
+      "learning_rate": 0.001,
+      "loss": 1.2585,
+      "step": 6620
+    },
+    {
+      "epoch": 0.2114225581172869,
+      "grad_norm": 0.2598929703235626,
+      "learning_rate": 0.001,
+      "loss": 1.2682,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2117414458369208,
+      "grad_norm": 0.2623753249645233,
+      "learning_rate": 0.001,
+      "loss": 1.2789,
+      "step": 6640
+    },
+    {
+      "epoch": 0.21206033355655474,
+      "grad_norm": 0.2436663657426834,
+      "learning_rate": 0.001,
+      "loss": 1.2495,
+      "step": 6650
+    },
+    {
+      "epoch": 0.21237922127618866,
+      "grad_norm": 0.25046613812446594,
+      "learning_rate": 0.001,
+      "loss": 1.2666,
+      "step": 6660
+    },
+    {
+      "epoch": 0.21269810899582256,
+      "grad_norm": 0.24722261726856232,
+      "learning_rate": 0.001,
+      "loss": 1.2628,
+      "step": 6670
+    },
+    {
+      "epoch": 0.21301699671545649,
+      "grad_norm": 0.26566001772880554,
+      "learning_rate": 0.001,
+      "loss": 1.2589,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2133358844350904,
+      "grad_norm": 0.259066104888916,
+      "learning_rate": 0.001,
+      "loss": 1.2713,
+      "step": 6690
+    },
+    {
+      "epoch": 0.2136547721547243,
+      "grad_norm": 0.25356295704841614,
+      "learning_rate": 0.001,
+      "loss": 1.2642,
+      "step": 6700
+    },
+    {
+      "epoch": 0.21397365987435824,
+      "grad_norm": 0.24706615507602692,
+      "learning_rate": 0.001,
+      "loss": 1.2488,
+      "step": 6710
+    },
+    {
+      "epoch": 0.21429254759399216,
+      "grad_norm": 0.2540133595466614,
+      "learning_rate": 0.001,
+      "loss": 1.2659,
+      "step": 6720
+    },
+    {
+      "epoch": 0.21461143531362606,
+      "grad_norm": 0.2536522448062897,
+      "learning_rate": 0.001,
+      "loss": 1.2542,
+      "step": 6730
+    },
+    {
+      "epoch": 0.21493032303325998,
+      "grad_norm": 0.24283340573310852,
+      "learning_rate": 0.001,
+      "loss": 1.2612,
+      "step": 6740
+    },
+    {
+      "epoch": 0.2152492107528939,
+      "grad_norm": 0.2563481330871582,
+      "learning_rate": 0.001,
+      "loss": 1.2744,
+      "step": 6750
+    },
+    {
+      "epoch": 0.21556809847252784,
+      "grad_norm": 0.24372535943984985,
+      "learning_rate": 0.001,
+      "loss": 1.2694,
+      "step": 6760
+    },
+    {
+      "epoch": 0.21588698619216173,
+      "grad_norm": 0.24880190193653107,
+      "learning_rate": 0.001,
+      "loss": 1.2522,
+      "step": 6770
+    },
+    {
+      "epoch": 0.21620587391179566,
+      "grad_norm": 0.25291043519973755,
+      "learning_rate": 0.001,
+      "loss": 1.2463,
+      "step": 6780
+    },
+    {
+      "epoch": 0.21652476163142959,
+      "grad_norm": 0.24317456781864166,
+      "learning_rate": 0.001,
+      "loss": 1.2431,
+      "step": 6790
+    },
+    {
+      "epoch": 0.21684364935106348,
+      "grad_norm": 0.2563396990299225,
+      "learning_rate": 0.001,
+      "loss": 1.2359,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2171625370706974,
+      "grad_norm": 0.24072889983654022,
+      "learning_rate": 0.001,
+      "loss": 1.2708,
+      "step": 6810
+    },
+    {
+      "epoch": 0.21748142479033133,
+      "grad_norm": 0.25073298811912537,
+      "learning_rate": 0.001,
+      "loss": 1.2582,
+      "step": 6820
+    },
+    {
+      "epoch": 0.21780031250996523,
+      "grad_norm": 0.25314459204673767,
+      "learning_rate": 0.001,
+      "loss": 1.2544,
+      "step": 6830
+    },
+    {
+      "epoch": 0.21811920022959916,
+      "grad_norm": 0.2519735097885132,
+      "learning_rate": 0.001,
+      "loss": 1.2522,
+      "step": 6840
+    },
+    {
+      "epoch": 0.21843808794923308,
+      "grad_norm": 0.24390500783920288,
+      "learning_rate": 0.001,
+      "loss": 1.2366,
+      "step": 6850
+    },
+    {
+      "epoch": 0.21875697566886698,
+      "grad_norm": 0.2493201494216919,
+      "learning_rate": 0.001,
+      "loss": 1.2402,
+      "step": 6860
+    },
+    {
+      "epoch": 0.2190758633885009,
+      "grad_norm": 0.24233953654766083,
+      "learning_rate": 0.001,
+      "loss": 1.2553,
+      "step": 6870
+    },
+    {
+      "epoch": 0.21939475110813483,
+      "grad_norm": 0.2534473240375519,
+      "learning_rate": 0.001,
+      "loss": 1.2575,
+      "step": 6880
+    },
+    {
+      "epoch": 0.21971363882776873,
+      "grad_norm": 0.24528414011001587,
+      "learning_rate": 0.001,
+      "loss": 1.2536,
+      "step": 6890
+    },
+    {
+      "epoch": 0.22003252654740266,
+      "grad_norm": 0.24577198922634125,
+      "learning_rate": 0.001,
+      "loss": 1.262,
+      "step": 6900
+    },
+    {
+      "epoch": 0.22035141426703658,
+      "grad_norm": 0.25635457038879395,
+      "learning_rate": 0.001,
+      "loss": 1.2592,
+      "step": 6910
+    },
+    {
+      "epoch": 0.22067030198667048,
+      "grad_norm": 0.24953170120716095,
+      "learning_rate": 0.001,
+      "loss": 1.2712,
+      "step": 6920
+    },
+    {
+      "epoch": 0.2209891897063044,
+      "grad_norm": 0.24969792366027832,
+      "learning_rate": 0.001,
+      "loss": 1.261,
+      "step": 6930
+    },
+    {
+      "epoch": 0.22130807742593833,
+      "grad_norm": 0.2447703331708908,
+      "learning_rate": 0.001,
+      "loss": 1.2449,
+      "step": 6940
+    },
+    {
+      "epoch": 0.22162696514557226,
+      "grad_norm": 0.25282493233680725,
+      "learning_rate": 0.001,
+      "loss": 1.2434,
+      "step": 6950
+    },
+    {
+      "epoch": 0.22194585286520616,
+      "grad_norm": 0.24152617156505585,
+      "learning_rate": 0.001,
+      "loss": 1.2362,
+      "step": 6960
+    },
+    {
+      "epoch": 0.22226474058484008,
+      "grad_norm": 0.2417408972978592,
+      "learning_rate": 0.001,
+      "loss": 1.2521,
+      "step": 6970
+    },
+    {
+      "epoch": 0.222583628304474,
+      "grad_norm": 0.2589294910430908,
+      "learning_rate": 0.001,
+      "loss": 1.2462,
+      "step": 6980
+    },
+    {
+      "epoch": 0.2229025160241079,
+      "grad_norm": 0.24300368130207062,
+      "learning_rate": 0.001,
+      "loss": 1.2606,
+      "step": 6990
+    },
+    {
+      "epoch": 0.22322140374374183,
+      "grad_norm": 0.2448713481426239,
+      "learning_rate": 0.001,
+      "loss": 1.2741,
+      "step": 7000
+    },
+    {
+      "epoch": 0.22354029146337576,
+      "grad_norm": 0.2419728934764862,
+      "learning_rate": 0.001,
+      "loss": 1.2359,
+      "step": 7010
+    },
+    {
+      "epoch": 0.22385917918300965,
+      "grad_norm": 0.2582404911518097,
+      "learning_rate": 0.001,
+      "loss": 1.2584,
+      "step": 7020
+    },
+    {
+      "epoch": 0.22417806690264358,
+      "grad_norm": 0.2443564236164093,
+      "learning_rate": 0.001,
+      "loss": 1.2251,
+      "step": 7030
+    },
+    {
+      "epoch": 0.2244969546222775,
+      "grad_norm": 0.2606917917728424,
+      "learning_rate": 0.001,
+      "loss": 1.2312,
+      "step": 7040
+    },
+    {
+      "epoch": 0.2248158423419114,
+      "grad_norm": 0.24423828721046448,
+      "learning_rate": 0.001,
+      "loss": 1.2456,
+      "step": 7050
+    },
+    {
+      "epoch": 0.22513473006154533,
+      "grad_norm": 0.23391813039779663,
+      "learning_rate": 0.001,
+      "loss": 1.2435,
+      "step": 7060
+    },
+    {
+      "epoch": 0.22545361778117926,
+      "grad_norm": 0.2510395050048828,
+      "learning_rate": 0.001,
+      "loss": 1.243,
+      "step": 7070
+    },
+    {
+      "epoch": 0.22577250550081315,
+      "grad_norm": 0.24144716560840607,
+      "learning_rate": 0.001,
+      "loss": 1.2645,
+      "step": 7080
+    },
+    {
+      "epoch": 0.22609139322044708,
+      "grad_norm": 0.2472696304321289,
+      "learning_rate": 0.001,
+      "loss": 1.2681,
+      "step": 7090
+    },
+    {
+      "epoch": 0.226410280940081,
+      "grad_norm": 0.2584279179573059,
+      "learning_rate": 0.001,
+      "loss": 1.2481,
+      "step": 7100
+    },
+    {
+      "epoch": 0.2267291686597149,
+      "grad_norm": 0.24365928769111633,
+      "learning_rate": 0.001,
+      "loss": 1.2629,
+      "step": 7110
+    },
+    {
+      "epoch": 0.22704805637934883,
+      "grad_norm": 0.23801958560943604,
+      "learning_rate": 0.001,
+      "loss": 1.24,
+      "step": 7120
+    },
+    {
+      "epoch": 0.22736694409898275,
+      "grad_norm": 0.2477579265832901,
+      "learning_rate": 0.001,
+      "loss": 1.2501,
+      "step": 7130
+    },
+    {
+      "epoch": 0.22768583181861665,
+      "grad_norm": 0.244456946849823,
+      "learning_rate": 0.001,
+      "loss": 1.2524,
+      "step": 7140
+    },
+    {
+      "epoch": 0.22800471953825058,
+      "grad_norm": 0.24829816818237305,
+      "learning_rate": 0.001,
+      "loss": 1.2458,
+      "step": 7150
+    },
+    {
+      "epoch": 0.2283236072578845,
+      "grad_norm": 0.2522318661212921,
+      "learning_rate": 0.001,
+      "loss": 1.2503,
+      "step": 7160
+    },
+    {
+      "epoch": 0.22864249497751843,
+      "grad_norm": 0.24825812876224518,
+      "learning_rate": 0.001,
+      "loss": 1.2666,
+      "step": 7170
+    },
+    {
+      "epoch": 0.22896138269715233,
+      "grad_norm": 0.2433803528547287,
+      "learning_rate": 0.001,
+      "loss": 1.2335,
+      "step": 7180
+    },
+    {
+      "epoch": 0.22928027041678625,
+      "grad_norm": 0.24022041261196136,
+      "learning_rate": 0.001,
+      "loss": 1.2387,
+      "step": 7190
+    },
+    {
+      "epoch": 0.22959915813642018,
+      "grad_norm": 0.2453920543193817,
+      "learning_rate": 0.001,
+      "loss": 1.2411,
+      "step": 7200
+    },
+    {
+      "epoch": 0.22991804585605408,
+      "grad_norm": 0.25992751121520996,
+      "learning_rate": 0.001,
+      "loss": 1.2367,
+      "step": 7210
+    },
+    {
+      "epoch": 0.230236933575688,
+      "grad_norm": 0.24138158559799194,
+      "learning_rate": 0.001,
+      "loss": 1.2157,
+      "step": 7220
+    },
+    {
+      "epoch": 0.23055582129532193,
+      "grad_norm": 0.23895490169525146,
+      "learning_rate": 0.001,
+      "loss": 1.2484,
+      "step": 7230
+    },
+    {
+      "epoch": 0.23087470901495583,
+      "grad_norm": 0.24053625762462616,
+      "learning_rate": 0.001,
+      "loss": 1.2554,
+      "step": 7240
+    },
+    {
+      "epoch": 0.23119359673458975,
+      "grad_norm": 0.24910859763622284,
+      "learning_rate": 0.001,
+      "loss": 1.2431,
+      "step": 7250
+    },
+    {
+      "epoch": 0.23151248445422368,
+      "grad_norm": 0.24892625212669373,
+      "learning_rate": 0.001,
+      "loss": 1.2313,
+      "step": 7260
+    },
+    {
+      "epoch": 0.23183137217385758,
+      "grad_norm": 0.24436773359775543,
+      "learning_rate": 0.001,
+      "loss": 1.2304,
+      "step": 7270
+    },
+    {
+      "epoch": 0.2321502598934915,
+      "grad_norm": 0.2511202096939087,
+      "learning_rate": 0.001,
+      "loss": 1.2574,
+      "step": 7280
+    },
+    {
+      "epoch": 0.23246914761312543,
+      "grad_norm": 0.24174362421035767,
+      "learning_rate": 0.001,
+      "loss": 1.2455,
+      "step": 7290
+    },
+    {
+      "epoch": 0.23278803533275932,
+      "grad_norm": 0.240260049700737,
+      "learning_rate": 0.001,
+      "loss": 1.2362,
+      "step": 7300
+    },
+    {
+      "epoch": 0.23310692305239325,
+      "grad_norm": 0.23553872108459473,
+      "learning_rate": 0.001,
+      "loss": 1.2563,
+      "step": 7310
+    },
+    {
+      "epoch": 0.23342581077202718,
+      "grad_norm": 0.24824745953083038,
+      "learning_rate": 0.001,
+      "loss": 1.2357,
+      "step": 7320
+    },
+    {
+      "epoch": 0.23374469849166107,
+      "grad_norm": 0.24948884546756744,
+      "learning_rate": 0.001,
+      "loss": 1.2236,
+      "step": 7330
+    },
+    {
+      "epoch": 0.234063586211295,
+      "grad_norm": 0.23782293498516083,
+      "learning_rate": 0.001,
+      "loss": 1.2443,
+      "step": 7340
+    },
+    {
+      "epoch": 0.23438247393092893,
+      "grad_norm": 0.23776574432849884,
+      "learning_rate": 0.001,
+      "loss": 1.2445,
+      "step": 7350
+    },
+    {
+      "epoch": 0.23470136165056282,
+      "grad_norm": 0.24230794608592987,
+      "learning_rate": 0.001,
+      "loss": 1.2339,
+      "step": 7360
+    },
+    {
+      "epoch": 0.23502024937019675,
+      "grad_norm": 0.23560328781604767,
+      "learning_rate": 0.001,
+      "loss": 1.2359,
+      "step": 7370
+    },
+    {
+      "epoch": 0.23533913708983067,
+      "grad_norm": 0.23826272785663605,
+      "learning_rate": 0.001,
+      "loss": 1.2416,
+      "step": 7380
+    },
+    {
+      "epoch": 0.2356580248094646,
+      "grad_norm": 0.2352159023284912,
+      "learning_rate": 0.001,
+      "loss": 1.2457,
+      "step": 7390
+    },
+    {
+      "epoch": 0.2359769125290985,
+      "grad_norm": 0.24641361832618713,
+      "learning_rate": 0.001,
+      "loss": 1.2282,
+      "step": 7400
+    },
+    {
+      "epoch": 0.23629580024873242,
+      "grad_norm": 0.23941314220428467,
+      "learning_rate": 0.001,
+      "loss": 1.2317,
+      "step": 7410
+    },
+    {
+      "epoch": 0.23661468796836635,
+      "grad_norm": 0.25254225730895996,
+      "learning_rate": 0.001,
+      "loss": 1.2565,
+      "step": 7420
+    },
+    {
+      "epoch": 0.23693357568800025,
+      "grad_norm": 0.2551426291465759,
+      "learning_rate": 0.001,
+      "loss": 1.2174,
+      "step": 7430
+    },
+    {
+      "epoch": 0.23725246340763417,
+      "grad_norm": 0.2395494133234024,
+      "learning_rate": 0.001,
+      "loss": 1.2315,
+      "step": 7440
+    },
+    {
+      "epoch": 0.2375713511272681,
+      "grad_norm": 0.24731141328811646,
+      "learning_rate": 0.001,
+      "loss": 1.2533,
+      "step": 7450
+    },
+    {
+      "epoch": 0.237890238846902,
+      "grad_norm": 0.24077807366847992,
+      "learning_rate": 0.001,
+      "loss": 1.2508,
+      "step": 7460
+    },
+    {
+      "epoch": 0.23820912656653592,
+      "grad_norm": 0.23398423194885254,
+      "learning_rate": 0.001,
+      "loss": 1.2408,
+      "step": 7470
+    },
+    {
+      "epoch": 0.23852801428616985,
+      "grad_norm": 0.25103890895843506,
+      "learning_rate": 0.001,
+      "loss": 1.2552,
+      "step": 7480
+    },
+    {
+      "epoch": 0.23884690200580375,
+      "grad_norm": 0.24333910644054413,
+      "learning_rate": 0.001,
+      "loss": 1.2284,
+      "step": 7490
+    },
+    {
+      "epoch": 0.23916578972543767,
+      "grad_norm": 0.24267978966236115,
+      "learning_rate": 0.001,
+      "loss": 1.2541,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2394846774450716,
+      "grad_norm": 0.24442200362682343,
+      "learning_rate": 0.001,
+      "loss": 1.2373,
+      "step": 7510
+    },
+    {
+      "epoch": 0.2398035651647055,
+      "grad_norm": 0.23441925644874573,
+      "learning_rate": 0.001,
+      "loss": 1.2411,
+      "step": 7520
+    },
+    {
+      "epoch": 0.24012245288433942,
+      "grad_norm": 0.2286960482597351,
+      "learning_rate": 0.001,
+      "loss": 1.2428,
+      "step": 7530
+    },
+    {
+      "epoch": 0.24044134060397335,
+      "grad_norm": 0.23532751202583313,
+      "learning_rate": 0.001,
+      "loss": 1.2352,
+      "step": 7540
+    },
+    {
+      "epoch": 0.24076022832360724,
+      "grad_norm": 0.24091851711273193,
+      "learning_rate": 0.001,
+      "loss": 1.2205,
+      "step": 7550
+    },
+    {
+      "epoch": 0.24107911604324117,
+      "grad_norm": 0.23746101558208466,
+      "learning_rate": 0.001,
+      "loss": 1.2307,
+      "step": 7560
+    },
+    {
+      "epoch": 0.2413980037628751,
+      "grad_norm": 0.25066760182380676,
+      "learning_rate": 0.001,
+      "loss": 1.2517,
+      "step": 7570
+    },
+    {
+      "epoch": 0.24171689148250902,
+      "grad_norm": 0.2366878241300583,
+      "learning_rate": 0.001,
+      "loss": 1.2313,
+      "step": 7580
+    },
+    {
+      "epoch": 0.24203577920214292,
+      "grad_norm": 0.2423858940601349,
+      "learning_rate": 0.001,
+      "loss": 1.2344,
+      "step": 7590
+    },
+    {
+      "epoch": 0.24235466692177685,
+      "grad_norm": 0.2289690375328064,
+      "learning_rate": 0.001,
+      "loss": 1.2471,
+      "step": 7600
+    },
+    {
+      "epoch": 0.24267355464141077,
+      "grad_norm": 0.23971182107925415,
+      "learning_rate": 0.001,
+      "loss": 1.2137,
+      "step": 7610
+    },
+    {
+      "epoch": 0.24299244236104467,
+      "grad_norm": 0.252573162317276,
+      "learning_rate": 0.001,
+      "loss": 1.2214,
+      "step": 7620
+    },
+    {
+      "epoch": 0.2433113300806786,
+      "grad_norm": 0.23575535416603088,
+      "learning_rate": 0.001,
+      "loss": 1.2391,
+      "step": 7630
+    },
+    {
+      "epoch": 0.24363021780031252,
+      "grad_norm": 0.23862920701503754,
+      "learning_rate": 0.001,
+      "loss": 1.2582,
+      "step": 7640
+    },
+    {
+      "epoch": 0.24394910551994642,
+      "grad_norm": 0.23157523572444916,
+      "learning_rate": 0.001,
+      "loss": 1.2299,
+      "step": 7650
+    },
+    {
+      "epoch": 0.24426799323958034,
+      "grad_norm": 0.24095961451530457,
+      "learning_rate": 0.001,
+      "loss": 1.2256,
+      "step": 7660
+    },
+    {
+      "epoch": 0.24458688095921427,
+      "grad_norm": 0.23619233071804047,
+      "learning_rate": 0.001,
+      "loss": 1.2291,
+      "step": 7670
+    },
+    {
+      "epoch": 0.24490576867884817,
+      "grad_norm": 0.23273909091949463,
+      "learning_rate": 0.001,
+      "loss": 1.2175,
+      "step": 7680
+    },
+    {
+      "epoch": 0.2452246563984821,
+      "grad_norm": 0.24084524810314178,
+      "learning_rate": 0.001,
+      "loss": 1.2414,
+      "step": 7690
+    },
+    {
+      "epoch": 0.24554354411811602,
+      "grad_norm": 0.25185689330101013,
+      "learning_rate": 0.001,
+      "loss": 1.2254,
+      "step": 7700
+    },
+    {
+      "epoch": 0.24586243183774992,
+      "grad_norm": 0.22970926761627197,
+      "learning_rate": 0.001,
+      "loss": 1.2073,
+      "step": 7710
+    },
+    {
+      "epoch": 0.24618131955738384,
+      "grad_norm": 0.23903733491897583,
+      "learning_rate": 0.001,
+      "loss": 1.2254,
+      "step": 7720
+    },
+    {
+      "epoch": 0.24650020727701777,
+      "grad_norm": 0.23516680300235748,
+      "learning_rate": 0.001,
+      "loss": 1.2347,
+      "step": 7730
+    },
+    {
+      "epoch": 0.24681909499665167,
+      "grad_norm": 0.24322205781936646,
+      "learning_rate": 0.001,
+      "loss": 1.2315,
+      "step": 7740
+    },
+    {
+      "epoch": 0.2471379827162856,
+      "grad_norm": 0.2339169830083847,
+      "learning_rate": 0.001,
+      "loss": 1.2511,
+      "step": 7750
+    },
+    {
+      "epoch": 0.24745687043591952,
+      "grad_norm": 0.23135200142860413,
+      "learning_rate": 0.001,
+      "loss": 1.2234,
+      "step": 7760
+    },
+    {
+      "epoch": 0.24777575815555342,
+      "grad_norm": 0.23176640272140503,
+      "learning_rate": 0.001,
+      "loss": 1.2359,
+      "step": 7770
+    },
+    {
+      "epoch": 0.24809464587518734,
+      "grad_norm": 0.24246467649936676,
+      "learning_rate": 0.001,
+      "loss": 1.211,
+      "step": 7780
+    },
+    {
+      "epoch": 0.24841353359482127,
+      "grad_norm": 0.23513196408748627,
+      "learning_rate": 0.001,
+      "loss": 1.2461,
+      "step": 7790
+    },
+    {
+      "epoch": 0.2487324213144552,
+      "grad_norm": 0.2445318102836609,
+      "learning_rate": 0.001,
+      "loss": 1.2237,
+      "step": 7800
+    },
+    {
+      "epoch": 0.2490513090340891,
+      "grad_norm": 0.2518177628517151,
+      "learning_rate": 0.001,
+      "loss": 1.2365,
+      "step": 7810
+    },
+    {
+      "epoch": 0.24937019675372302,
+      "grad_norm": 0.2446645200252533,
+      "learning_rate": 0.001,
+      "loss": 1.2237,
+      "step": 7820
+    },
+    {
+      "epoch": 0.24968908447335694,
+      "grad_norm": 0.229459747672081,
+      "learning_rate": 0.001,
+      "loss": 1.2074,
+      "step": 7830
+    },
+    {
+      "epoch": 0.25000797219299087,
+      "grad_norm": 0.2381288856267929,
+      "learning_rate": 0.001,
+      "loss": 1.214,
+      "step": 7840
+    },
+    {
+      "epoch": 0.25032685991262477,
+      "grad_norm": 0.2357947826385498,
+      "learning_rate": 0.001,
+      "loss": 1.2024,
+      "step": 7850
+    },
+    {
+      "epoch": 0.25064574763225866,
+      "grad_norm": 0.22776131331920624,
+      "learning_rate": 0.001,
+      "loss": 1.2011,
+      "step": 7860
+    },
+    {
+      "epoch": 0.2509646353518926,
+      "grad_norm": 0.23940008878707886,
+      "learning_rate": 0.001,
+      "loss": 1.2098,
+      "step": 7870
+    },
+    {
+      "epoch": 0.2512835230715265,
+      "grad_norm": 0.23641014099121094,
+      "learning_rate": 0.001,
+      "loss": 1.2112,
+      "step": 7880
+    },
+    {
+      "epoch": 0.2516024107911604,
+      "grad_norm": 0.2425474226474762,
+      "learning_rate": 0.001,
+      "loss": 1.2518,
+      "step": 7890
+    },
+    {
+      "epoch": 0.25192129851079437,
+      "grad_norm": 0.23014706373214722,
+      "learning_rate": 0.001,
+      "loss": 1.2273,
+      "step": 7900
+    },
+    {
+      "epoch": 0.25224018623042826,
+      "grad_norm": 0.24386554956436157,
+      "learning_rate": 0.001,
+      "loss": 1.2195,
+      "step": 7910
+    },
+    {
+      "epoch": 0.25255907395006216,
+      "grad_norm": 0.2404145896434784,
+      "learning_rate": 0.001,
+      "loss": 1.2307,
+      "step": 7920
+    },
+    {
+      "epoch": 0.2528779616696961,
+      "grad_norm": 0.25573667883872986,
+      "learning_rate": 0.001,
+      "loss": 1.2186,
+      "step": 7930
+    },
+    {
+      "epoch": 0.25319684938933,
+      "grad_norm": 0.24742534756660461,
+      "learning_rate": 0.001,
+      "loss": 1.2188,
+      "step": 7940
+    },
+    {
+      "epoch": 0.2535157371089639,
+      "grad_norm": 0.22783447802066803,
+      "learning_rate": 0.001,
+      "loss": 1.2203,
+      "step": 7950
+    },
+    {
+      "epoch": 0.25383462482859787,
+      "grad_norm": 0.24916143715381622,
+      "learning_rate": 0.001,
+      "loss": 1.2348,
+      "step": 7960
+    },
+    {
+      "epoch": 0.25415351254823176,
+      "grad_norm": 0.23791836202144623,
+      "learning_rate": 0.001,
+      "loss": 1.2223,
+      "step": 7970
+    },
+    {
+      "epoch": 0.25447240026786566,
+      "grad_norm": 0.2505311667919159,
+      "learning_rate": 0.001,
+      "loss": 1.2459,
+      "step": 7980
+    },
+    {
+      "epoch": 0.2547912879874996,
+      "grad_norm": 0.23071864247322083,
+      "learning_rate": 0.001,
+      "loss": 1.2412,
+      "step": 7990
+    },
+    {
+      "epoch": 0.2551101757071335,
+      "grad_norm": 0.23231177031993866,
+      "learning_rate": 0.001,
+      "loss": 1.2135,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2554290634267674,
+      "grad_norm": 0.2333109825849533,
+      "learning_rate": 0.001,
+      "loss": 1.208,
+      "step": 8010
+    },
+    {
+      "epoch": 0.25574795114640136,
+      "grad_norm": 0.22724169492721558,
+      "learning_rate": 0.001,
+      "loss": 1.224,
+      "step": 8020
+    },
+    {
+      "epoch": 0.25606683886603526,
+      "grad_norm": 0.21973343193531036,
+      "learning_rate": 0.001,
+      "loss": 1.2057,
+      "step": 8030
+    },
+    {
+      "epoch": 0.25638572658566916,
+      "grad_norm": 0.23393845558166504,
+      "learning_rate": 0.001,
+      "loss": 1.2247,
+      "step": 8040
+    },
+    {
+      "epoch": 0.2567046143053031,
+      "grad_norm": 0.23509255051612854,
+      "learning_rate": 0.001,
+      "loss": 1.2381,
+      "step": 8050
+    },
+    {
+      "epoch": 0.257023502024937,
+      "grad_norm": 0.2345188558101654,
+      "learning_rate": 0.001,
+      "loss": 1.2221,
+      "step": 8060
+    },
+    {
+      "epoch": 0.2573423897445709,
+      "grad_norm": 0.2355467826128006,
+      "learning_rate": 0.001,
+      "loss": 1.2338,
+      "step": 8070
+    },
+    {
+      "epoch": 0.25766127746420486,
+      "grad_norm": 0.2355579137802124,
+      "learning_rate": 0.001,
+      "loss": 1.2197,
+      "step": 8080
+    },
+    {
+      "epoch": 0.25798016518383876,
+      "grad_norm": 0.23406566679477692,
+      "learning_rate": 0.001,
+      "loss": 1.2519,
+      "step": 8090
+    },
+    {
+      "epoch": 0.2582990529034727,
+      "grad_norm": 0.22977030277252197,
+      "learning_rate": 0.001,
+      "loss": 1.2381,
+      "step": 8100
+    },
+    {
+      "epoch": 0.2586179406231066,
+      "grad_norm": 0.23623408377170563,
+      "learning_rate": 0.001,
+      "loss": 1.2299,
+      "step": 8110
+    },
+    {
+      "epoch": 0.2589368283427405,
+      "grad_norm": 0.22804781794548035,
+      "learning_rate": 0.001,
+      "loss": 1.2062,
+      "step": 8120
+    },
+    {
+      "epoch": 0.25925571606237446,
+      "grad_norm": 0.23730701208114624,
+      "learning_rate": 0.001,
+      "loss": 1.2091,
+      "step": 8130
+    },
+    {
+      "epoch": 0.25957460378200836,
+      "grad_norm": 0.23314902186393738,
+      "learning_rate": 0.001,
+      "loss": 1.1943,
+      "step": 8140
+    },
+    {
+      "epoch": 0.25989349150164226,
+      "grad_norm": 0.23350052535533905,
+      "learning_rate": 0.001,
+      "loss": 1.2201,
+      "step": 8150
+    },
+    {
+      "epoch": 0.2602123792212762,
+      "grad_norm": 0.2341201901435852,
+      "learning_rate": 0.001,
+      "loss": 1.2133,
+      "step": 8160
+    },
+    {
+      "epoch": 0.2605312669409101,
+      "grad_norm": 0.2353760153055191,
+      "learning_rate": 0.001,
+      "loss": 1.2034,
+      "step": 8170
+    },
+    {
+      "epoch": 0.260850154660544,
+      "grad_norm": 0.23472733795642853,
+      "learning_rate": 0.001,
+      "loss": 1.2234,
+      "step": 8180
+    },
+    {
+      "epoch": 0.26116904238017796,
+      "grad_norm": 0.23768886923789978,
+      "learning_rate": 0.001,
+      "loss": 1.2087,
+      "step": 8190
+    },
+    {
+      "epoch": 0.26148793009981186,
+      "grad_norm": 0.2352430671453476,
+      "learning_rate": 0.001,
+      "loss": 1.2135,
+      "step": 8200
+    },
+    {
+      "epoch": 0.26180681781944576,
+      "grad_norm": 0.22320342063903809,
+      "learning_rate": 0.001,
+      "loss": 1.2301,
+      "step": 8210
+    },
+    {
+      "epoch": 0.2621257055390797,
+      "grad_norm": 0.2318148910999298,
+      "learning_rate": 0.001,
+      "loss": 1.2281,
+      "step": 8220
+    },
+    {
+      "epoch": 0.2624445932587136,
+      "grad_norm": 0.22970159351825714,
+      "learning_rate": 0.001,
+      "loss": 1.2129,
+      "step": 8230
+    },
+    {
+      "epoch": 0.2627634809783475,
+      "grad_norm": 0.2396457940340042,
+      "learning_rate": 0.001,
+      "loss": 1.2104,
+      "step": 8240
+    },
+    {
+      "epoch": 0.26308236869798146,
+      "grad_norm": 0.22637683153152466,
+      "learning_rate": 0.001,
+      "loss": 1.2003,
+      "step": 8250
+    },
+    {
+      "epoch": 0.26340125641761536,
+      "grad_norm": 0.23237474262714386,
+      "learning_rate": 0.001,
+      "loss": 1.2245,
+      "step": 8260
+    },
+    {
+      "epoch": 0.26372014413724926,
+      "grad_norm": 0.22723113000392914,
+      "learning_rate": 0.001,
+      "loss": 1.2201,
+      "step": 8270
+    },
+    {
+      "epoch": 0.2640390318568832,
+      "grad_norm": 0.23084807395935059,
+      "learning_rate": 0.001,
+      "loss": 1.2184,
+      "step": 8280
+    },
+    {
+      "epoch": 0.2643579195765171,
+      "grad_norm": 0.2344745695590973,
+      "learning_rate": 0.001,
+      "loss": 1.22,
+      "step": 8290
+    },
+    {
+      "epoch": 0.264676807296151,
+      "grad_norm": 0.23803043365478516,
+      "learning_rate": 0.001,
+      "loss": 1.2093,
+      "step": 8300
+    },
+    {
+      "epoch": 0.26499569501578496,
+      "grad_norm": 0.2341003566980362,
+      "learning_rate": 0.001,
+      "loss": 1.2055,
+      "step": 8310
+    },
+    {
+      "epoch": 0.26531458273541886,
+      "grad_norm": 0.21944069862365723,
+      "learning_rate": 0.001,
+      "loss": 1.202,
+      "step": 8320
+    },
+    {
+      "epoch": 0.26563347045505276,
+      "grad_norm": 0.2393493801355362,
+      "learning_rate": 0.001,
+      "loss": 1.1969,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2659523581746867,
+      "grad_norm": 0.23989616334438324,
+      "learning_rate": 0.001,
+      "loss": 1.2248,
+      "step": 8340
+    },
+    {
+      "epoch": 0.2662712458943206,
+      "grad_norm": 0.24278509616851807,
+      "learning_rate": 0.001,
+      "loss": 1.2123,
+      "step": 8350
+    },
+    {
+      "epoch": 0.2665901336139545,
+      "grad_norm": 0.23201477527618408,
+      "learning_rate": 0.001,
+      "loss": 1.2448,
+      "step": 8360
+    },
+    {
+      "epoch": 0.26690902133358846,
+      "grad_norm": 0.23349370062351227,
+      "learning_rate": 0.001,
+      "loss": 1.2317,
+      "step": 8370
+    },
+    {
+      "epoch": 0.26722790905322236,
+      "grad_norm": 0.2284669727087021,
+      "learning_rate": 0.001,
+      "loss": 1.2232,
+      "step": 8380
+    },
+    {
+      "epoch": 0.26754679677285625,
+      "grad_norm": 0.23704218864440918,
+      "learning_rate": 0.001,
+      "loss": 1.2192,
+      "step": 8390
+    },
+    {
+      "epoch": 0.2678656844924902,
+      "grad_norm": 0.2318849116563797,
+      "learning_rate": 0.001,
+      "loss": 1.2119,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2681845722121241,
+      "grad_norm": 0.2331661731004715,
+      "learning_rate": 0.001,
+      "loss": 1.2088,
+      "step": 8410
+    },
+    {
+      "epoch": 0.268503459931758,
+      "grad_norm": 0.22507762908935547,
+      "learning_rate": 0.001,
+      "loss": 1.2048,
+      "step": 8420
+    },
+    {
+      "epoch": 0.26882234765139196,
+      "grad_norm": 0.2239762246608734,
+      "learning_rate": 0.001,
+      "loss": 1.2033,
+      "step": 8430
+    },
+    {
+      "epoch": 0.26914123537102586,
+      "grad_norm": 0.22471489012241364,
+      "learning_rate": 0.001,
+      "loss": 1.2004,
+      "step": 8440
+    },
+    {
+      "epoch": 0.26946012309065975,
+      "grad_norm": 0.22484731674194336,
+      "learning_rate": 0.001,
+      "loss": 1.2154,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2697790108102937,
+      "grad_norm": 0.2240588217973709,
+      "learning_rate": 0.001,
+      "loss": 1.2067,
+      "step": 8460
+    },
+    {
+      "epoch": 0.2700978985299276,
+      "grad_norm": 0.2298148274421692,
+      "learning_rate": 0.001,
+      "loss": 1.2215,
+      "step": 8470
+    },
+    {
+      "epoch": 0.2704167862495615,
+      "grad_norm": 0.22547484934329987,
+      "learning_rate": 0.001,
+      "loss": 1.2017,
+      "step": 8480
+    },
+    {
+      "epoch": 0.27073567396919546,
+      "grad_norm": 0.23824118077754974,
+      "learning_rate": 0.001,
+      "loss": 1.2013,
+      "step": 8490
+    },
+    {
+      "epoch": 0.27105456168882935,
+      "grad_norm": 0.22975823283195496,
+      "learning_rate": 0.001,
+      "loss": 1.2293,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2713734494084633,
+      "grad_norm": 0.22520454227924347,
+      "learning_rate": 0.001,
+      "loss": 1.2064,
+      "step": 8510
+    },
+    {
+      "epoch": 0.2716923371280972,
+      "grad_norm": 0.22511954605579376,
+      "learning_rate": 0.001,
+      "loss": 1.1975,
+      "step": 8520
+    },
+    {
+      "epoch": 0.2720112248477311,
+      "grad_norm": 0.22746147215366364,
+      "learning_rate": 0.001,
+      "loss": 1.1967,
+      "step": 8530
+    },
+    {
+      "epoch": 0.27233011256736506,
+      "grad_norm": 0.2236718386411667,
+      "learning_rate": 0.001,
+      "loss": 1.2091,
+      "step": 8540
+    },
+    {
+      "epoch": 0.27264900028699895,
+      "grad_norm": 0.23237386345863342,
+      "learning_rate": 0.001,
+      "loss": 1.2299,
+      "step": 8550
+    },
+    {
+      "epoch": 0.27296788800663285,
+      "grad_norm": 0.22995509207248688,
+      "learning_rate": 0.001,
+      "loss": 1.2193,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2732867757262668,
+      "grad_norm": 0.231089785695076,
+      "learning_rate": 0.001,
+      "loss": 1.2297,
+      "step": 8570
+    },
+    {
+      "epoch": 0.2736056634459007,
+      "grad_norm": 0.2250330150127411,
+      "learning_rate": 0.001,
+      "loss": 1.1958,
+      "step": 8580
+    },
+    {
+      "epoch": 0.2739245511655346,
+      "grad_norm": 0.2263891100883484,
+      "learning_rate": 0.001,
+      "loss": 1.2029,
+      "step": 8590
+    },
+    {
+      "epoch": 0.27424343888516856,
+      "grad_norm": 0.2355814427137375,
+      "learning_rate": 0.001,
+      "loss": 1.2225,
+      "step": 8600
+    },
+    {
+      "epoch": 0.27456232660480245,
+      "grad_norm": 0.23153933882713318,
+      "learning_rate": 0.001,
+      "loss": 1.2279,
+      "step": 8610
+    },
+    {
+      "epoch": 0.27488121432443635,
+      "grad_norm": 0.23251959681510925,
+      "learning_rate": 0.001,
+      "loss": 1.1949,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2752001020440703,
+      "grad_norm": 0.22222910821437836,
+      "learning_rate": 0.001,
+      "loss": 1.1986,
+      "step": 8630
+    },
+    {
+      "epoch": 0.2755189897637042,
+      "grad_norm": 0.22941821813583374,
+      "learning_rate": 0.001,
+      "loss": 1.1969,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2758378774833381,
+      "grad_norm": 0.2258928269147873,
+      "learning_rate": 0.001,
+      "loss": 1.2217,
+      "step": 8650
+    },
+    {
+      "epoch": 0.27615676520297205,
+      "grad_norm": 0.22354252636432648,
+      "learning_rate": 0.001,
+      "loss": 1.2045,
+      "step": 8660
+    },
+    {
+      "epoch": 0.27647565292260595,
+      "grad_norm": 0.22870154678821564,
+      "learning_rate": 0.001,
+      "loss": 1.2092,
+      "step": 8670
+    },
+    {
+      "epoch": 0.27679454064223985,
+      "grad_norm": 0.2278323620557785,
+      "learning_rate": 0.001,
+      "loss": 1.1935,
+      "step": 8680
+    },
+    {
+      "epoch": 0.2771134283618738,
+      "grad_norm": 0.2331470400094986,
+      "learning_rate": 0.001,
+      "loss": 1.1874,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2774323160815077,
+      "grad_norm": 0.23780789971351624,
+      "learning_rate": 0.001,
+      "loss": 1.2081,
+      "step": 8700
+    },
+    {
+      "epoch": 0.2777512038011416,
+      "grad_norm": 0.22317449748516083,
+      "learning_rate": 0.001,
+      "loss": 1.2126,
+      "step": 8710
+    },
+    {
+      "epoch": 0.27807009152077555,
+      "grad_norm": 0.22962188720703125,
+      "learning_rate": 0.001,
+      "loss": 1.209,
+      "step": 8720
+    },
+    {
+      "epoch": 0.27838897924040945,
+      "grad_norm": 0.2307327538728714,
+      "learning_rate": 0.001,
+      "loss": 1.1841,
+      "step": 8730
+    },
+    {
+      "epoch": 0.27870786696004335,
+      "grad_norm": 0.23068435490131378,
+      "learning_rate": 0.001,
+      "loss": 1.2017,
+      "step": 8740
+    },
+    {
+      "epoch": 0.2790267546796773,
+      "grad_norm": 0.24138770997524261,
+      "learning_rate": 0.001,
+      "loss": 1.2083,
+      "step": 8750
+    },
+    {
+      "epoch": 0.2793456423993112,
+      "grad_norm": 0.23563891649246216,
+      "learning_rate": 0.001,
+      "loss": 1.2027,
+      "step": 8760
+    },
+    {
+      "epoch": 0.2796645301189451,
+      "grad_norm": 0.22962354123592377,
+      "learning_rate": 0.001,
+      "loss": 1.2111,
+      "step": 8770
+    },
+    {
+      "epoch": 0.27998341783857905,
+      "grad_norm": 0.2287108451128006,
+      "learning_rate": 0.001,
+      "loss": 1.2177,
+      "step": 8780
+    },
+    {
+      "epoch": 0.28030230555821295,
+      "grad_norm": 0.22435401380062103,
+      "learning_rate": 0.001,
+      "loss": 1.2016,
+      "step": 8790
+    },
+    {
+      "epoch": 0.28062119327784685,
+      "grad_norm": 0.23022402822971344,
+      "learning_rate": 0.001,
+      "loss": 1.1926,
+      "step": 8800
+    },
+    {
+      "epoch": 0.2809400809974808,
+      "grad_norm": 0.2214146852493286,
+      "learning_rate": 0.001,
+      "loss": 1.2209,
+      "step": 8810
+    },
+    {
+      "epoch": 0.2812589687171147,
+      "grad_norm": 0.22423280775547028,
+      "learning_rate": 0.001,
+      "loss": 1.205,
+      "step": 8820
+    },
+    {
+      "epoch": 0.2815778564367486,
+      "grad_norm": 0.2278013676404953,
+      "learning_rate": 0.001,
+      "loss": 1.2274,
+      "step": 8830
+    },
+    {
+      "epoch": 0.28189674415638255,
+      "grad_norm": 0.22766144573688507,
+      "learning_rate": 0.001,
+      "loss": 1.2044,
+      "step": 8840
+    },
+    {
+      "epoch": 0.28221563187601645,
+      "grad_norm": 0.2270435094833374,
+      "learning_rate": 0.001,
+      "loss": 1.2072,
+      "step": 8850
+    },
+    {
+      "epoch": 0.28253451959565035,
+      "grad_norm": 0.219182550907135,
+      "learning_rate": 0.001,
+      "loss": 1.2071,
+      "step": 8860
+    },
+    {
+      "epoch": 0.2828534073152843,
+      "grad_norm": 0.22934895753860474,
+      "learning_rate": 0.001,
+      "loss": 1.192,
+      "step": 8870
+    },
+    {
+      "epoch": 0.2831722950349182,
+      "grad_norm": 0.22396302223205566,
+      "learning_rate": 0.001,
+      "loss": 1.2112,
+      "step": 8880
+    },
+    {
+      "epoch": 0.2834911827545521,
+      "grad_norm": 0.22344917058944702,
+      "learning_rate": 0.001,
+      "loss": 1.1915,
+      "step": 8890
+    },
+    {
+      "epoch": 0.28381007047418605,
+      "grad_norm": 0.23081816732883453,
+      "learning_rate": 0.001,
+      "loss": 1.2115,
+      "step": 8900
+    },
+    {
+      "epoch": 0.28412895819381995,
+      "grad_norm": 0.22885209321975708,
+      "learning_rate": 0.001,
+      "loss": 1.1975,
+      "step": 8910
+    },
+    {
+      "epoch": 0.2844478459134539,
+      "grad_norm": 0.2284027636051178,
+      "learning_rate": 0.001,
+      "loss": 1.1961,
+      "step": 8920
+    },
+    {
+      "epoch": 0.2847667336330878,
+      "grad_norm": 0.21693913638591766,
+      "learning_rate": 0.001,
+      "loss": 1.2041,
+      "step": 8930
+    },
+    {
+      "epoch": 0.2850856213527217,
+      "grad_norm": 0.22245806455612183,
+      "learning_rate": 0.001,
+      "loss": 1.2045,
+      "step": 8940
+    },
+    {
+      "epoch": 0.28540450907235565,
+      "grad_norm": 0.22184446454048157,
+      "learning_rate": 0.001,
+      "loss": 1.2137,
+      "step": 8950
+    },
+    {
+      "epoch": 0.28572339679198955,
+      "grad_norm": 0.23200933635234833,
+      "learning_rate": 0.001,
+      "loss": 1.1944,
+      "step": 8960
+    },
+    {
+      "epoch": 0.28604228451162345,
+      "grad_norm": 0.22749820351600647,
+      "learning_rate": 0.001,
+      "loss": 1.1989,
+      "step": 8970
+    },
+    {
+      "epoch": 0.2863611722312574,
+      "grad_norm": 0.2261044979095459,
+      "learning_rate": 0.001,
+      "loss": 1.2006,
+      "step": 8980
+    },
+    {
+      "epoch": 0.2866800599508913,
+      "grad_norm": 0.23911482095718384,
+      "learning_rate": 0.001,
+      "loss": 1.1997,
+      "step": 8990
+    },
+    {
+      "epoch": 0.2869989476705252,
+      "grad_norm": 0.2479439377784729,
+      "learning_rate": 0.001,
+      "loss": 1.2261,
+      "step": 9000
+    },
+    {
+      "epoch": 0.28731783539015915,
+      "grad_norm": 0.22959910333156586,
+      "learning_rate": 0.001,
+      "loss": 1.182,
+      "step": 9010
+    },
+    {
+      "epoch": 0.28763672310979305,
+      "grad_norm": 0.2242193967103958,
+      "learning_rate": 0.001,
+      "loss": 1.2011,
+      "step": 9020
+    },
+    {
+      "epoch": 0.28795561082942694,
+      "grad_norm": 0.22244490683078766,
+      "learning_rate": 0.001,
+      "loss": 1.1991,
+      "step": 9030
+    },
+    {
+      "epoch": 0.2882744985490609,
+      "grad_norm": 0.22404681146144867,
+      "learning_rate": 0.001,
+      "loss": 1.2082,
+      "step": 9040
+    },
+    {
+      "epoch": 0.2885933862686948,
+      "grad_norm": 0.2188567817211151,
+      "learning_rate": 0.001,
+      "loss": 1.1923,
+      "step": 9050
+    },
+    {
+      "epoch": 0.2889122739883287,
+      "grad_norm": 0.22456391155719757,
+      "learning_rate": 0.001,
+      "loss": 1.1907,
+      "step": 9060
+    },
+    {
+      "epoch": 0.28923116170796265,
+      "grad_norm": 0.22750291228294373,
+      "learning_rate": 0.001,
+      "loss": 1.196,
+      "step": 9070
+    },
+    {
+      "epoch": 0.28955004942759655,
+      "grad_norm": 0.22974838316440582,
+      "learning_rate": 0.001,
+      "loss": 1.2025,
+      "step": 9080
+    },
+    {
+      "epoch": 0.28986893714723044,
+      "grad_norm": 0.23487216234207153,
+      "learning_rate": 0.001,
+      "loss": 1.2249,
+      "step": 9090
+    },
+    {
+      "epoch": 0.2901878248668644,
+      "grad_norm": 0.2250378280878067,
+      "learning_rate": 0.001,
+      "loss": 1.1928,
+      "step": 9100
+    },
+    {
+      "epoch": 0.2905067125864983,
+      "grad_norm": 0.22203022241592407,
+      "learning_rate": 0.001,
+      "loss": 1.1971,
+      "step": 9110
+    },
+    {
+      "epoch": 0.2908256003061322,
+      "grad_norm": 0.23717299103736877,
+      "learning_rate": 0.001,
+      "loss": 1.2061,
+      "step": 9120
+    },
+    {
+      "epoch": 0.29114448802576615,
+      "grad_norm": 0.22907336056232452,
+      "learning_rate": 0.001,
+      "loss": 1.2061,
+      "step": 9130
+    },
+    {
+      "epoch": 0.29146337574540004,
+      "grad_norm": 0.22980521619319916,
+      "learning_rate": 0.001,
+      "loss": 1.1919,
+      "step": 9140
+    },
+    {
+      "epoch": 0.29178226346503394,
+      "grad_norm": 0.2266918420791626,
+      "learning_rate": 0.001,
+      "loss": 1.1871,
+      "step": 9150
+    },
+    {
+      "epoch": 0.2921011511846679,
+      "grad_norm": 0.2265349179506302,
+      "learning_rate": 0.001,
+      "loss": 1.2098,
+      "step": 9160
+    },
+    {
+      "epoch": 0.2924200389043018,
+      "grad_norm": 0.23188531398773193,
+      "learning_rate": 0.001,
+      "loss": 1.2114,
+      "step": 9170
+    },
+    {
+      "epoch": 0.2927389266239357,
+      "grad_norm": 0.21577180922031403,
+      "learning_rate": 0.001,
+      "loss": 1.2019,
+      "step": 9180
+    },
+    {
+      "epoch": 0.29305781434356964,
+      "grad_norm": 0.2256147712469101,
+      "learning_rate": 0.001,
+      "loss": 1.1884,
+      "step": 9190
+    },
+    {
+      "epoch": 0.29337670206320354,
+      "grad_norm": 0.22333773970603943,
+      "learning_rate": 0.001,
+      "loss": 1.1911,
+      "step": 9200
+    },
+    {
+      "epoch": 0.29369558978283744,
+      "grad_norm": 0.2238871157169342,
+      "learning_rate": 0.001,
+      "loss": 1.1881,
+      "step": 9210
+    },
+    {
+      "epoch": 0.2940144775024714,
+      "grad_norm": 0.22535620629787445,
+      "learning_rate": 0.001,
+      "loss": 1.1906,
+      "step": 9220
+    },
+    {
+      "epoch": 0.2943333652221053,
+      "grad_norm": 0.22432921826839447,
+      "learning_rate": 0.001,
+      "loss": 1.2011,
+      "step": 9230
+    },
+    {
+      "epoch": 0.2946522529417392,
+      "grad_norm": 0.22004643082618713,
+      "learning_rate": 0.001,
+      "loss": 1.1829,
+      "step": 9240
+    },
+    {
+      "epoch": 0.29497114066137314,
+      "grad_norm": 0.223861962556839,
+      "learning_rate": 0.001,
+      "loss": 1.1977,
+      "step": 9250
+    },
+    {
+      "epoch": 0.29529002838100704,
+      "grad_norm": 0.22514422237873077,
+      "learning_rate": 0.001,
+      "loss": 1.1897,
+      "step": 9260
+    },
+    {
+      "epoch": 0.29560891610064094,
+      "grad_norm": 0.214486762881279,
+      "learning_rate": 0.001,
+      "loss": 1.1842,
+      "step": 9270
+    },
+    {
+      "epoch": 0.2959278038202749,
+      "grad_norm": 0.22228612005710602,
+      "learning_rate": 0.001,
+      "loss": 1.193,
+      "step": 9280
+    },
+    {
+      "epoch": 0.2962466915399088,
+      "grad_norm": 0.2186896950006485,
+      "learning_rate": 0.001,
+      "loss": 1.1751,
+      "step": 9290
+    },
+    {
+      "epoch": 0.2965655792595427,
+      "grad_norm": 0.22547854483127594,
+      "learning_rate": 0.001,
+      "loss": 1.1931,
+      "step": 9300
+    },
+    {
+      "epoch": 0.29688446697917664,
+      "grad_norm": 0.22223107516765594,
+      "learning_rate": 0.001,
+      "loss": 1.2041,
+      "step": 9310
+    },
+    {
+      "epoch": 0.29720335469881054,
+      "grad_norm": 0.21878202259540558,
+      "learning_rate": 0.001,
+      "loss": 1.1737,
+      "step": 9320
+    },
+    {
+      "epoch": 0.29752224241844444,
+      "grad_norm": 0.22275276482105255,
+      "learning_rate": 0.001,
+      "loss": 1.2027,
+      "step": 9330
+    },
+    {
+      "epoch": 0.2978411301380784,
+      "grad_norm": 0.21805565059185028,
+      "learning_rate": 0.001,
+      "loss": 1.1769,
+      "step": 9340
+    },
+    {
+      "epoch": 0.2981600178577123,
+      "grad_norm": 0.21685688197612762,
+      "learning_rate": 0.001,
+      "loss": 1.1857,
+      "step": 9350
+    },
+    {
+      "epoch": 0.29847890557734624,
+      "grad_norm": 0.22610515356063843,
+      "learning_rate": 0.001,
+      "loss": 1.1919,
+      "step": 9360
+    },
+    {
+      "epoch": 0.29879779329698014,
+      "grad_norm": 0.23180913925170898,
+      "learning_rate": 0.001,
+      "loss": 1.1976,
+      "step": 9370
+    },
+    {
+      "epoch": 0.29911668101661404,
+      "grad_norm": 0.2190287858247757,
+      "learning_rate": 0.001,
+      "loss": 1.1931,
+      "step": 9380
+    },
+    {
+      "epoch": 0.299435568736248,
+      "grad_norm": 0.22509071230888367,
+      "learning_rate": 0.001,
+      "loss": 1.1801,
+      "step": 9390
+    },
+    {
+      "epoch": 0.2997544564558819,
+      "grad_norm": 0.23787586390972137,
+      "learning_rate": 0.001,
+      "loss": 1.1912,
+      "step": 9400
+    },
+    {
+      "epoch": 0.3000733441755158,
+      "grad_norm": 0.22887369990348816,
+      "learning_rate": 0.001,
+      "loss": 1.1961,
+      "step": 9410
+    },
+    {
+      "epoch": 0.30039223189514974,
+      "grad_norm": 0.2291015237569809,
+      "learning_rate": 0.001,
+      "loss": 1.1952,
+      "step": 9420
+    },
+    {
+      "epoch": 0.30071111961478364,
+      "grad_norm": 0.22769349813461304,
+      "learning_rate": 0.001,
+      "loss": 1.1844,
+      "step": 9430
+    },
+    {
+      "epoch": 0.30103000733441754,
+      "grad_norm": 0.22478654980659485,
+      "learning_rate": 0.001,
+      "loss": 1.1955,
+      "step": 9440
+    },
+    {
+      "epoch": 0.3013488950540515,
+      "grad_norm": 0.2176402360200882,
+      "learning_rate": 0.001,
+      "loss": 1.1987,
+      "step": 9450
+    },
+    {
+      "epoch": 0.3016677827736854,
+      "grad_norm": 0.22715549170970917,
+      "learning_rate": 0.001,
+      "loss": 1.1986,
+      "step": 9460
+    },
+    {
+      "epoch": 0.3019866704933193,
+      "grad_norm": 0.22686997056007385,
+      "learning_rate": 0.001,
+      "loss": 1.2123,
+      "step": 9470
+    },
+    {
+      "epoch": 0.30230555821295324,
+      "grad_norm": 0.22678443789482117,
+      "learning_rate": 0.001,
+      "loss": 1.192,
+      "step": 9480
+    },
+    {
+      "epoch": 0.30262444593258714,
+      "grad_norm": 0.22301578521728516,
+      "learning_rate": 0.001,
+      "loss": 1.192,
+      "step": 9490
+    },
+    {
+      "epoch": 0.30294333365222104,
+      "grad_norm": 0.23071426153182983,
+      "learning_rate": 0.001,
+      "loss": 1.194,
+      "step": 9500
+    },
+    {
+      "epoch": 0.303262221371855,
+      "grad_norm": 0.21674400568008423,
+      "learning_rate": 0.001,
+      "loss": 1.1759,
+      "step": 9510
+    },
+    {
+      "epoch": 0.3035811090914889,
+      "grad_norm": 0.23387682437896729,
+      "learning_rate": 0.001,
+      "loss": 1.1884,
+      "step": 9520
+    },
+    {
+      "epoch": 0.3038999968111228,
+      "grad_norm": 0.21519413590431213,
+      "learning_rate": 0.001,
+      "loss": 1.173,
+      "step": 9530
+    },
+    {
+      "epoch": 0.30421888453075674,
+      "grad_norm": 0.2206396609544754,
+      "learning_rate": 0.001,
+      "loss": 1.1847,
+      "step": 9540
+    },
+    {
+      "epoch": 0.30453777225039064,
+      "grad_norm": 0.2154570370912552,
+      "learning_rate": 0.001,
+      "loss": 1.1651,
+      "step": 9550
+    },
+    {
+      "epoch": 0.30485665997002453,
+      "grad_norm": 0.21689386665821075,
+      "learning_rate": 0.001,
+      "loss": 1.185,
+      "step": 9560
+    },
+    {
+      "epoch": 0.3051755476896585,
+      "grad_norm": 0.21970142424106598,
+      "learning_rate": 0.001,
+      "loss": 1.1835,
+      "step": 9570
+    },
+    {
+      "epoch": 0.3054944354092924,
+      "grad_norm": 0.2152683436870575,
+      "learning_rate": 0.001,
+      "loss": 1.201,
+      "step": 9580
+    },
+    {
+      "epoch": 0.3058133231289263,
+      "grad_norm": 0.21406981348991394,
+      "learning_rate": 0.001,
+      "loss": 1.1853,
+      "step": 9590
+    },
+    {
+      "epoch": 0.30613221084856024,
+      "grad_norm": 0.22537820041179657,
+      "learning_rate": 0.001,
+      "loss": 1.1782,
+      "step": 9600
+    },
+    {
+      "epoch": 0.30645109856819414,
+      "grad_norm": 0.2190530151128769,
+      "learning_rate": 0.001,
+      "loss": 1.1859,
+      "step": 9610
+    },
+    {
+      "epoch": 0.30676998628782803,
+      "grad_norm": 0.22278915345668793,
+      "learning_rate": 0.001,
+      "loss": 1.1933,
+      "step": 9620
+    },
+    {
+      "epoch": 0.307088874007462,
+      "grad_norm": 0.22467225790023804,
+      "learning_rate": 0.001,
+      "loss": 1.1582,
+      "step": 9630
+    },
+    {
+      "epoch": 0.3074077617270959,
+      "grad_norm": 0.21523691713809967,
+      "learning_rate": 0.001,
+      "loss": 1.1914,
+      "step": 9640
+    },
+    {
+      "epoch": 0.3077266494467298,
+      "grad_norm": 0.21753321588039398,
+      "learning_rate": 0.001,
+      "loss": 1.1819,
+      "step": 9650
+    },
+    {
+      "epoch": 0.30804553716636374,
+      "grad_norm": 0.21796949207782745,
+      "learning_rate": 0.001,
+      "loss": 1.1754,
+      "step": 9660
+    },
+    {
+      "epoch": 0.30836442488599763,
+      "grad_norm": 0.23503394424915314,
+      "learning_rate": 0.001,
+      "loss": 1.1749,
+      "step": 9670
+    },
+    {
+      "epoch": 0.30868331260563153,
+      "grad_norm": 0.2093040943145752,
+      "learning_rate": 0.001,
+      "loss": 1.1836,
+      "step": 9680
+    },
+    {
+      "epoch": 0.3090022003252655,
+      "grad_norm": 0.22570084035396576,
+      "learning_rate": 0.001,
+      "loss": 1.1994,
+      "step": 9690
+    },
+    {
+      "epoch": 0.3093210880448994,
+      "grad_norm": 0.2191409170627594,
+      "learning_rate": 0.001,
+      "loss": 1.1873,
+      "step": 9700
+    },
+    {
+      "epoch": 0.3096399757645333,
+      "grad_norm": 0.21672411262989044,
+      "learning_rate": 0.001,
+      "loss": 1.1996,
+      "step": 9710
+    },
+    {
+      "epoch": 0.30995886348416724,
+      "grad_norm": 0.22737610340118408,
+      "learning_rate": 0.001,
+      "loss": 1.1935,
+      "step": 9720
+    },
+    {
+      "epoch": 0.31027775120380113,
+      "grad_norm": 0.22928710281848907,
+      "learning_rate": 0.001,
+      "loss": 1.1962,
+      "step": 9730
+    },
+    {
+      "epoch": 0.31059663892343503,
+      "grad_norm": 0.22526556253433228,
+      "learning_rate": 0.001,
+      "loss": 1.1886,
+      "step": 9740
+    },
+    {
+      "epoch": 0.310915526643069,
+      "grad_norm": 0.2220076620578766,
+      "learning_rate": 0.001,
+      "loss": 1.1757,
+      "step": 9750
+    },
+    {
+      "epoch": 0.3112344143627029,
+      "grad_norm": 0.210331529378891,
+      "learning_rate": 0.001,
+      "loss": 1.1849,
+      "step": 9760
+    },
+    {
+      "epoch": 0.31155330208233684,
+      "grad_norm": 0.22174914181232452,
+      "learning_rate": 0.001,
+      "loss": 1.2025,
+      "step": 9770
+    },
+    {
+      "epoch": 0.31187218980197073,
+      "grad_norm": 0.21560117602348328,
+      "learning_rate": 0.001,
+      "loss": 1.1687,
+      "step": 9780
+    },
+    {
+      "epoch": 0.31219107752160463,
+      "grad_norm": 0.21781526505947113,
+      "learning_rate": 0.001,
+      "loss": 1.1778,
+      "step": 9790
+    },
+    {
+      "epoch": 0.3125099652412386,
+      "grad_norm": 0.21552933752536774,
+      "learning_rate": 0.001,
+      "loss": 1.1908,
+      "step": 9800
+    },
+    {
+      "epoch": 0.3128288529608725,
+      "grad_norm": 0.2246668040752411,
+      "learning_rate": 0.001,
+      "loss": 1.1889,
+      "step": 9810
+    },
+    {
+      "epoch": 0.3131477406805064,
+      "grad_norm": 0.2357940971851349,
+      "learning_rate": 0.001,
+      "loss": 1.1943,
+      "step": 9820
+    },
+    {
+      "epoch": 0.31346662840014033,
+      "grad_norm": 0.22036802768707275,
+      "learning_rate": 0.001,
+      "loss": 1.1787,
+      "step": 9830
+    },
+    {
+      "epoch": 0.31378551611977423,
+      "grad_norm": 0.21087811887264252,
+      "learning_rate": 0.001,
+      "loss": 1.1806,
+      "step": 9840
+    },
+    {
+      "epoch": 0.31410440383940813,
+      "grad_norm": 0.22128775715827942,
+      "learning_rate": 0.001,
+      "loss": 1.1631,
+      "step": 9850
+    },
+    {
+      "epoch": 0.3144232915590421,
+      "grad_norm": 0.21533682942390442,
+      "learning_rate": 0.001,
+      "loss": 1.1952,
+      "step": 9860
+    },
+    {
+      "epoch": 0.314742179278676,
+      "grad_norm": 0.21292875707149506,
+      "learning_rate": 0.001,
+      "loss": 1.1896,
+      "step": 9870
+    },
+    {
+      "epoch": 0.3150610669983099,
+      "grad_norm": 0.2134082019329071,
+      "learning_rate": 0.001,
+      "loss": 1.1911,
+      "step": 9880
+    },
+    {
+      "epoch": 0.31537995471794383,
+      "grad_norm": 0.2236315906047821,
+      "learning_rate": 0.001,
+      "loss": 1.1733,
+      "step": 9890
+    },
+    {
+      "epoch": 0.31569884243757773,
+      "grad_norm": 0.2129340022802353,
+      "learning_rate": 0.001,
+      "loss": 1.178,
+      "step": 9900
+    },
+    {
+      "epoch": 0.31601773015721163,
+      "grad_norm": 0.21715772151947021,
+      "learning_rate": 0.001,
+      "loss": 1.1675,
+      "step": 9910
+    },
+    {
+      "epoch": 0.3163366178768456,
+      "grad_norm": 0.22146131098270416,
+      "learning_rate": 0.001,
+      "loss": 1.1753,
+      "step": 9920
+    },
+    {
+      "epoch": 0.3166555055964795,
+      "grad_norm": 0.21491622924804688,
+      "learning_rate": 0.001,
+      "loss": 1.1903,
+      "step": 9930
+    },
+    {
+      "epoch": 0.3169743933161134,
+      "grad_norm": 0.22050689160823822,
+      "learning_rate": 0.001,
+      "loss": 1.1906,
+      "step": 9940
+    },
+    {
+      "epoch": 0.31729328103574733,
+      "grad_norm": 0.2244400829076767,
+      "learning_rate": 0.001,
+      "loss": 1.1912,
+      "step": 9950
+    },
+    {
+      "epoch": 0.31761216875538123,
+      "grad_norm": 0.21547940373420715,
+      "learning_rate": 0.001,
+      "loss": 1.1796,
+      "step": 9960
+    },
+    {
+      "epoch": 0.31793105647501513,
+      "grad_norm": 0.21386213600635529,
+      "learning_rate": 0.001,
+      "loss": 1.2017,
+      "step": 9970
+    },
+    {
+      "epoch": 0.3182499441946491,
+      "grad_norm": 0.21595734357833862,
+      "learning_rate": 0.001,
+      "loss": 1.1788,
+      "step": 9980
+    },
+    {
+      "epoch": 0.318568831914283,
+      "grad_norm": 0.22140143811702728,
+      "learning_rate": 0.001,
+      "loss": 1.1862,
+      "step": 9990
+    },
+    {
+      "epoch": 0.3188877196339169,
+      "grad_norm": 0.2175026386976242,
+      "learning_rate": 0.001,
+      "loss": 1.1879,
+      "step": 10000
+    },
+    {
+      "epoch": 0.31920660735355083,
+      "grad_norm": 0.22484996914863586,
+      "learning_rate": 0.001,
+      "loss": 1.1876,
+      "step": 10010
+    },
+    {
+      "epoch": 0.31952549507318473,
+      "grad_norm": 0.21960212290287018,
+      "learning_rate": 0.001,
+      "loss": 1.174,
+      "step": 10020
+    },
+    {
+      "epoch": 0.3198443827928186,
+      "grad_norm": 0.21397638320922852,
+      "learning_rate": 0.001,
+      "loss": 1.181,
+      "step": 10030
+    },
+    {
+      "epoch": 0.3201632705124526,
+      "grad_norm": 0.22058583796024323,
+      "learning_rate": 0.001,
+      "loss": 1.1784,
+      "step": 10040
+    },
+    {
+      "epoch": 0.3204821582320865,
+      "grad_norm": 0.21691805124282837,
+      "learning_rate": 0.001,
+      "loss": 1.1997,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3208010459517204,
+      "grad_norm": 0.21837332844734192,
+      "learning_rate": 0.001,
+      "loss": 1.1817,
+      "step": 10060
+    },
+    {
+      "epoch": 0.32111993367135433,
+      "grad_norm": 0.21407966315746307,
+      "learning_rate": 0.001,
+      "loss": 1.1682,
+      "step": 10070
+    },
+    {
+      "epoch": 0.3214388213909882,
+      "grad_norm": 0.22078247368335724,
+      "learning_rate": 0.001,
+      "loss": 1.1843,
+      "step": 10080
+    },
+    {
+      "epoch": 0.3217577091106221,
+      "grad_norm": 0.22645726799964905,
+      "learning_rate": 0.001,
+      "loss": 1.1804,
+      "step": 10090
+    },
+    {
+      "epoch": 0.3220765968302561,
+      "grad_norm": 0.23695030808448792,
+      "learning_rate": 0.001,
+      "loss": 1.1888,
+      "step": 10100
+    },
+    {
+      "epoch": 0.32239548454989,
+      "grad_norm": 0.21678394079208374,
+      "learning_rate": 0.001,
+      "loss": 1.1851,
+      "step": 10110
+    },
+    {
+      "epoch": 0.3227143722695239,
+      "grad_norm": 0.2129441499710083,
+      "learning_rate": 0.001,
+      "loss": 1.1582,
+      "step": 10120
+    },
+    {
+      "epoch": 0.32303325998915783,
+      "grad_norm": 0.23205362260341644,
+      "learning_rate": 0.001,
+      "loss": 1.1673,
+      "step": 10130
+    },
+    {
+      "epoch": 0.3233521477087917,
+      "grad_norm": 0.21547579765319824,
+      "learning_rate": 0.001,
+      "loss": 1.1743,
+      "step": 10140
+    },
+    {
+      "epoch": 0.3236710354284256,
+      "grad_norm": 0.2152055948972702,
+      "learning_rate": 0.001,
+      "loss": 1.1855,
+      "step": 10150
+    },
+    {
+      "epoch": 0.3239899231480596,
+      "grad_norm": 0.21947774291038513,
+      "learning_rate": 0.001,
+      "loss": 1.1733,
+      "step": 10160
+    },
+    {
+      "epoch": 0.3243088108676935,
+      "grad_norm": 0.22177624702453613,
+      "learning_rate": 0.001,
+      "loss": 1.172,
+      "step": 10170
+    },
+    {
+      "epoch": 0.32462769858732743,
+      "grad_norm": 0.21848328411579132,
+      "learning_rate": 0.001,
+      "loss": 1.1694,
+      "step": 10180
+    },
+    {
+      "epoch": 0.3249465863069613,
+      "grad_norm": 0.2304374724626541,
+      "learning_rate": 0.001,
+      "loss": 1.1636,
+      "step": 10190
+    },
+    {
+      "epoch": 0.3252654740265952,
+      "grad_norm": 0.21279925107955933,
+      "learning_rate": 0.001,
+      "loss": 1.1744,
+      "step": 10200
+    },
+    {
+      "epoch": 0.3255843617462292,
+      "grad_norm": 0.22444838285446167,
+      "learning_rate": 0.001,
+      "loss": 1.1749,
+      "step": 10210
+    },
+    {
+      "epoch": 0.3259032494658631,
+      "grad_norm": 0.22544382512569427,
+      "learning_rate": 0.001,
+      "loss": 1.1733,
+      "step": 10220
+    },
+    {
+      "epoch": 0.326222137185497,
+      "grad_norm": 0.20965062081813812,
+      "learning_rate": 0.001,
+      "loss": 1.2065,
+      "step": 10230
+    },
+    {
+      "epoch": 0.3265410249051309,
+      "grad_norm": 0.2140328288078308,
+      "learning_rate": 0.001,
+      "loss": 1.1949,
+      "step": 10240
+    },
+    {
+      "epoch": 0.3268599126247648,
+      "grad_norm": 0.2196279913187027,
+      "learning_rate": 0.001,
+      "loss": 1.154,
+      "step": 10250
+    },
+    {
+      "epoch": 0.3271788003443987,
+      "grad_norm": 0.22404052317142487,
+      "learning_rate": 0.001,
+      "loss": 1.1803,
+      "step": 10260
+    },
+    {
+      "epoch": 0.3274976880640327,
+      "grad_norm": 0.21107755601406097,
+      "learning_rate": 0.001,
+      "loss": 1.1745,
+      "step": 10270
+    },
+    {
+      "epoch": 0.3278165757836666,
+      "grad_norm": 0.22213582694530487,
+      "learning_rate": 0.001,
+      "loss": 1.1809,
+      "step": 10280
+    },
+    {
+      "epoch": 0.3281354635033005,
+      "grad_norm": 0.2129000425338745,
+      "learning_rate": 0.001,
+      "loss": 1.1832,
+      "step": 10290
+    },
+    {
+      "epoch": 0.3284543512229344,
+      "grad_norm": 0.21625253558158875,
+      "learning_rate": 0.001,
+      "loss": 1.1723,
+      "step": 10300
+    },
+    {
+      "epoch": 0.3287732389425683,
+      "grad_norm": 0.2259497046470642,
+      "learning_rate": 0.001,
+      "loss": 1.199,
+      "step": 10310
+    },
+    {
+      "epoch": 0.3290921266622022,
+      "grad_norm": 0.22692418098449707,
+      "learning_rate": 0.001,
+      "loss": 1.1668,
+      "step": 10320
+    },
+    {
+      "epoch": 0.3294110143818362,
+      "grad_norm": 0.2245110273361206,
+      "learning_rate": 0.001,
+      "loss": 1.199,
+      "step": 10330
+    },
+    {
+      "epoch": 0.3297299021014701,
+      "grad_norm": 0.22499682009220123,
+      "learning_rate": 0.001,
+      "loss": 1.1874,
+      "step": 10340
+    },
+    {
+      "epoch": 0.33004878982110397,
+      "grad_norm": 0.2121761441230774,
+      "learning_rate": 0.001,
+      "loss": 1.1635,
+      "step": 10350
+    },
+    {
+      "epoch": 0.3303676775407379,
+      "grad_norm": 0.22067177295684814,
+      "learning_rate": 0.001,
+      "loss": 1.1735,
+      "step": 10360
+    },
+    {
+      "epoch": 0.3306865652603718,
+      "grad_norm": 0.21458233892917633,
+      "learning_rate": 0.001,
+      "loss": 1.1928,
+      "step": 10370
+    },
+    {
+      "epoch": 0.3310054529800057,
+      "grad_norm": 0.21691793203353882,
+      "learning_rate": 0.001,
+      "loss": 1.1694,
+      "step": 10380
+    },
+    {
+      "epoch": 0.3313243406996397,
+      "grad_norm": 0.21943393349647522,
+      "learning_rate": 0.001,
+      "loss": 1.1591,
+      "step": 10390
+    },
+    {
+      "epoch": 0.3316432284192736,
+      "grad_norm": 0.20909449458122253,
+      "learning_rate": 0.001,
+      "loss": 1.1801,
+      "step": 10400
+    },
+    {
+      "epoch": 0.33196211613890747,
+      "grad_norm": 0.20836131274700165,
+      "learning_rate": 0.001,
+      "loss": 1.1636,
+      "step": 10410
+    },
+    {
+      "epoch": 0.3322810038585414,
+      "grad_norm": 0.21603058278560638,
+      "learning_rate": 0.001,
+      "loss": 1.1708,
+      "step": 10420
+    },
+    {
+      "epoch": 0.3325998915781753,
+      "grad_norm": 0.20854921638965607,
+      "learning_rate": 0.001,
+      "loss": 1.1604,
+      "step": 10430
+    },
+    {
+      "epoch": 0.3329187792978092,
+      "grad_norm": 0.21955880522727966,
+      "learning_rate": 0.001,
+      "loss": 1.1963,
+      "step": 10440
+    },
+    {
+      "epoch": 0.3332376670174432,
+      "grad_norm": 0.21727994084358215,
+      "learning_rate": 0.001,
+      "loss": 1.1648,
+      "step": 10450
+    },
+    {
+      "epoch": 0.33355655473707707,
+      "grad_norm": 0.21586079895496368,
+      "learning_rate": 0.001,
+      "loss": 1.1633,
+      "step": 10460
+    },
+    {
+      "epoch": 0.33387544245671097,
+      "grad_norm": 0.22296805679798126,
+      "learning_rate": 0.001,
+      "loss": 1.1799,
+      "step": 10470
+    },
+    {
+      "epoch": 0.3341943301763449,
+      "grad_norm": 0.21028365194797516,
+      "learning_rate": 0.001,
+      "loss": 1.1718,
+      "step": 10480
+    },
+    {
+      "epoch": 0.3345132178959788,
+      "grad_norm": 0.21386286616325378,
+      "learning_rate": 0.001,
+      "loss": 1.1644,
+      "step": 10490
+    },
+    {
+      "epoch": 0.3348321056156127,
+      "grad_norm": 0.22596792876720428,
+      "learning_rate": 0.001,
+      "loss": 1.1676,
+      "step": 10500
+    },
+    {
+      "epoch": 0.33515099333524667,
+      "grad_norm": 0.21902908384799957,
+      "learning_rate": 0.001,
+      "loss": 1.178,
+      "step": 10510
+    },
+    {
+      "epoch": 0.33546988105488057,
+      "grad_norm": 0.22614213824272156,
+      "learning_rate": 0.001,
+      "loss": 1.1866,
+      "step": 10520
+    },
+    {
+      "epoch": 0.33578876877451447,
+      "grad_norm": 0.21819354593753815,
+      "learning_rate": 0.001,
+      "loss": 1.1793,
+      "step": 10530
+    },
+    {
+      "epoch": 0.3361076564941484,
+      "grad_norm": 0.20999178290367126,
+      "learning_rate": 0.001,
+      "loss": 1.1555,
+      "step": 10540
+    },
+    {
+      "epoch": 0.3364265442137823,
+      "grad_norm": 0.2194094955921173,
+      "learning_rate": 0.001,
+      "loss": 1.1641,
+      "step": 10550
+    },
+    {
+      "epoch": 0.3367454319334162,
+      "grad_norm": 0.21941301226615906,
+      "learning_rate": 0.001,
+      "loss": 1.1692,
+      "step": 10560
+    },
+    {
+      "epoch": 0.33706431965305017,
+      "grad_norm": 0.2173202931880951,
+      "learning_rate": 0.001,
+      "loss": 1.1612,
+      "step": 10570
+    },
+    {
+      "epoch": 0.33738320737268407,
+      "grad_norm": 0.21929022669792175,
+      "learning_rate": 0.001,
+      "loss": 1.168,
+      "step": 10580
+    },
+    {
+      "epoch": 0.337702095092318,
+      "grad_norm": 0.22014977037906647,
+      "learning_rate": 0.001,
+      "loss": 1.1395,
+      "step": 10590
+    },
+    {
+      "epoch": 0.3380209828119519,
+      "grad_norm": 0.22155609726905823,
+      "learning_rate": 0.001,
+      "loss": 1.1624,
+      "step": 10600
+    },
+    {
+      "epoch": 0.3383398705315858,
+      "grad_norm": 0.23078712821006775,
+      "learning_rate": 0.001,
+      "loss": 1.1666,
+      "step": 10610
+    },
+    {
+      "epoch": 0.33865875825121977,
+      "grad_norm": 0.2146841287612915,
+      "learning_rate": 0.001,
+      "loss": 1.1943,
+      "step": 10620
+    },
+    {
+      "epoch": 0.33897764597085367,
+      "grad_norm": 0.21783071756362915,
+      "learning_rate": 0.001,
+      "loss": 1.1676,
+      "step": 10630
+    },
+    {
+      "epoch": 0.33929653369048757,
+      "grad_norm": 0.21608036756515503,
+      "learning_rate": 0.001,
+      "loss": 1.1781,
+      "step": 10640
+    },
+    {
+      "epoch": 0.3396154214101215,
+      "grad_norm": 0.22186727821826935,
+      "learning_rate": 0.001,
+      "loss": 1.1702,
+      "step": 10650
+    },
+    {
+      "epoch": 0.3399343091297554,
+      "grad_norm": 0.2069188356399536,
+      "learning_rate": 0.001,
+      "loss": 1.1704,
+      "step": 10660
+    },
+    {
+      "epoch": 0.3402531968493893,
+      "grad_norm": 0.20236782729625702,
+      "learning_rate": 0.001,
+      "loss": 1.1769,
+      "step": 10670
+    },
+    {
+      "epoch": 0.34057208456902327,
+      "grad_norm": 0.21130560338497162,
+      "learning_rate": 0.001,
+      "loss": 1.1815,
+      "step": 10680
+    },
+    {
+      "epoch": 0.34089097228865717,
+      "grad_norm": 0.2142263948917389,
+      "learning_rate": 0.001,
+      "loss": 1.1666,
+      "step": 10690
+    },
+    {
+      "epoch": 0.34120986000829107,
+      "grad_norm": 0.21785111725330353,
+      "learning_rate": 0.001,
+      "loss": 1.1745,
+      "step": 10700
+    },
+    {
+      "epoch": 0.341528747727925,
+      "grad_norm": 0.20937016606330872,
+      "learning_rate": 0.001,
+      "loss": 1.1686,
+      "step": 10710
+    },
+    {
+      "epoch": 0.3418476354475589,
+      "grad_norm": 0.21203775703907013,
+      "learning_rate": 0.001,
+      "loss": 1.1754,
+      "step": 10720
+    },
+    {
+      "epoch": 0.3421665231671928,
+      "grad_norm": 0.21170243620872498,
+      "learning_rate": 0.001,
+      "loss": 1.1505,
+      "step": 10730
+    },
+    {
+      "epoch": 0.34248541088682677,
+      "grad_norm": 0.21524657309055328,
+      "learning_rate": 0.001,
+      "loss": 1.1787,
+      "step": 10740
+    },
+    {
+      "epoch": 0.34280429860646067,
+      "grad_norm": 0.2171153426170349,
+      "learning_rate": 0.001,
+      "loss": 1.1622,
+      "step": 10750
+    },
+    {
+      "epoch": 0.34312318632609456,
+      "grad_norm": 0.21716740727424622,
+      "learning_rate": 0.001,
+      "loss": 1.1576,
+      "step": 10760
+    },
+    {
+      "epoch": 0.3434420740457285,
+      "grad_norm": 0.21167676150798798,
+      "learning_rate": 0.001,
+      "loss": 1.1459,
+      "step": 10770
+    },
+    {
+      "epoch": 0.3437609617653624,
+      "grad_norm": 0.2166099101305008,
+      "learning_rate": 0.001,
+      "loss": 1.1741,
+      "step": 10780
+    },
+    {
+      "epoch": 0.3440798494849963,
+      "grad_norm": 0.22174325585365295,
+      "learning_rate": 0.001,
+      "loss": 1.1635,
+      "step": 10790
+    },
+    {
+      "epoch": 0.34439873720463027,
+      "grad_norm": 0.217443585395813,
+      "learning_rate": 0.001,
+      "loss": 1.1738,
+      "step": 10800
+    },
+    {
+      "epoch": 0.34471762492426417,
+      "grad_norm": 0.2111525535583496,
+      "learning_rate": 0.001,
+      "loss": 1.1894,
+      "step": 10810
+    },
+    {
+      "epoch": 0.34503651264389806,
+      "grad_norm": 0.22243352234363556,
+      "learning_rate": 0.001,
+      "loss": 1.178,
+      "step": 10820
+    },
+    {
+      "epoch": 0.345355400363532,
+      "grad_norm": 0.21252469718456268,
+      "learning_rate": 0.001,
+      "loss": 1.1657,
+      "step": 10830
+    },
+    {
+      "epoch": 0.3456742880831659,
+      "grad_norm": 0.20956046879291534,
+      "learning_rate": 0.001,
+      "loss": 1.1531,
+      "step": 10840
+    },
+    {
+      "epoch": 0.3459931758027998,
+      "grad_norm": 0.20752333104610443,
+      "learning_rate": 0.001,
+      "loss": 1.1582,
+      "step": 10850
+    },
+    {
+      "epoch": 0.34631206352243377,
+      "grad_norm": 0.2089800238609314,
+      "learning_rate": 0.001,
+      "loss": 1.1547,
+      "step": 10860
+    },
+    {
+      "epoch": 0.34663095124206766,
+      "grad_norm": 0.21103443205356598,
+      "learning_rate": 0.001,
+      "loss": 1.1723,
+      "step": 10870
+    },
+    {
+      "epoch": 0.34694983896170156,
+      "grad_norm": 0.21871396899223328,
+      "learning_rate": 0.001,
+      "loss": 1.1838,
+      "step": 10880
+    },
+    {
+      "epoch": 0.3472687266813355,
+      "grad_norm": 0.21302813291549683,
+      "learning_rate": 0.001,
+      "loss": 1.1713,
+      "step": 10890
+    },
+    {
+      "epoch": 0.3475876144009694,
+      "grad_norm": 0.22152137756347656,
+      "learning_rate": 0.001,
+      "loss": 1.1773,
+      "step": 10900
+    },
+    {
+      "epoch": 0.3479065021206033,
+      "grad_norm": 0.2116517275571823,
+      "learning_rate": 0.001,
+      "loss": 1.1505,
+      "step": 10910
+    },
+    {
+      "epoch": 0.34822538984023726,
+      "grad_norm": 0.21990253031253815,
+      "learning_rate": 0.001,
+      "loss": 1.1738,
+      "step": 10920
+    },
+    {
+      "epoch": 0.34854427755987116,
+      "grad_norm": 0.22551077604293823,
+      "learning_rate": 0.001,
+      "loss": 1.1788,
+      "step": 10930
+    },
+    {
+      "epoch": 0.34886316527950506,
+      "grad_norm": 0.21925295889377594,
+      "learning_rate": 0.001,
+      "loss": 1.1585,
+      "step": 10940
+    },
+    {
+      "epoch": 0.349182052999139,
+      "grad_norm": 0.2133491337299347,
+      "learning_rate": 0.001,
+      "loss": 1.1832,
+      "step": 10950
+    },
+    {
+      "epoch": 0.3495009407187729,
+      "grad_norm": 0.20826612412929535,
+      "learning_rate": 0.001,
+      "loss": 1.1685,
+      "step": 10960
+    },
+    {
+      "epoch": 0.3498198284384068,
+      "grad_norm": 0.2133883684873581,
+      "learning_rate": 0.001,
+      "loss": 1.1692,
+      "step": 10970
+    },
+    {
+      "epoch": 0.35013871615804076,
+      "grad_norm": 0.21480365097522736,
+      "learning_rate": 0.001,
+      "loss": 1.1698,
+      "step": 10980
+    },
+    {
+      "epoch": 0.35045760387767466,
+      "grad_norm": 0.21561360359191895,
+      "learning_rate": 0.001,
+      "loss": 1.1494,
+      "step": 10990
+    },
+    {
+      "epoch": 0.3507764915973086,
+      "grad_norm": 0.21999944746494293,
+      "learning_rate": 0.001,
+      "loss": 1.161,
+      "step": 11000
+    },
+    {
+      "epoch": 0.3510953793169425,
+      "grad_norm": 0.21331818401813507,
+      "learning_rate": 0.001,
+      "loss": 1.1743,
+      "step": 11010
+    },
+    {
+      "epoch": 0.3514142670365764,
+      "grad_norm": 0.2172483205795288,
+      "learning_rate": 0.001,
+      "loss": 1.1575,
+      "step": 11020
+    },
+    {
+      "epoch": 0.35173315475621036,
+      "grad_norm": 0.20713630318641663,
+      "learning_rate": 0.001,
+      "loss": 1.139,
+      "step": 11030
+    },
+    {
+      "epoch": 0.35205204247584426,
+      "grad_norm": 0.21412107348442078,
+      "learning_rate": 0.001,
+      "loss": 1.1533,
+      "step": 11040
+    },
+    {
+      "epoch": 0.35237093019547816,
+      "grad_norm": 0.21454772353172302,
+      "learning_rate": 0.001,
+      "loss": 1.1401,
+      "step": 11050
+    },
+    {
+      "epoch": 0.3526898179151121,
+      "grad_norm": 0.20491410791873932,
+      "learning_rate": 0.001,
+      "loss": 1.1654,
+      "step": 11060
+    },
+    {
+      "epoch": 0.353008705634746,
+      "grad_norm": 0.20974987745285034,
+      "learning_rate": 0.001,
+      "loss": 1.156,
+      "step": 11070
+    },
+    {
+      "epoch": 0.3533275933543799,
+      "grad_norm": 0.2226075381040573,
+      "learning_rate": 0.001,
+      "loss": 1.1584,
+      "step": 11080
+    },
+    {
+      "epoch": 0.35364648107401386,
+      "grad_norm": 0.21573099493980408,
+      "learning_rate": 0.001,
+      "loss": 1.1468,
+      "step": 11090
+    },
+    {
+      "epoch": 0.35396536879364776,
+      "grad_norm": 0.2131635844707489,
+      "learning_rate": 0.001,
+      "loss": 1.1439,
+      "step": 11100
+    },
+    {
+      "epoch": 0.35428425651328166,
+      "grad_norm": 0.2120184600353241,
+      "learning_rate": 0.001,
+      "loss": 1.1671,
+      "step": 11110
+    },
+    {
+      "epoch": 0.3546031442329156,
+      "grad_norm": 0.21845276653766632,
+      "learning_rate": 0.001,
+      "loss": 1.1712,
+      "step": 11120
+    },
+    {
+      "epoch": 0.3549220319525495,
+      "grad_norm": 0.2118763029575348,
+      "learning_rate": 0.001,
+      "loss": 1.1679,
+      "step": 11130
+    },
+    {
+      "epoch": 0.3552409196721834,
+      "grad_norm": 0.21182695031166077,
+      "learning_rate": 0.001,
+      "loss": 1.1633,
+      "step": 11140
+    },
+    {
+      "epoch": 0.35555980739181736,
+      "grad_norm": 0.2156038135290146,
+      "learning_rate": 0.001,
+      "loss": 1.1644,
+      "step": 11150
+    },
+    {
+      "epoch": 0.35587869511145126,
+      "grad_norm": 0.20957325398921967,
+      "learning_rate": 0.001,
+      "loss": 1.1658,
+      "step": 11160
+    },
+    {
+      "epoch": 0.35619758283108516,
+      "grad_norm": 0.20978455245494843,
+      "learning_rate": 0.001,
+      "loss": 1.1679,
+      "step": 11170
+    },
+    {
+      "epoch": 0.3565164705507191,
+      "grad_norm": 0.21704687178134918,
+      "learning_rate": 0.001,
+      "loss": 1.1588,
+      "step": 11180
+    },
+    {
+      "epoch": 0.356835358270353,
+      "grad_norm": 0.2168053388595581,
+      "learning_rate": 0.001,
+      "loss": 1.1589,
+      "step": 11190
+    },
+    {
+      "epoch": 0.3571542459899869,
+      "grad_norm": 0.2075323611497879,
+      "learning_rate": 0.001,
+      "loss": 1.1368,
+      "step": 11200
+    },
+    {
+      "epoch": 0.35747313370962086,
+      "grad_norm": 0.21764330565929413,
+      "learning_rate": 0.001,
+      "loss": 1.1713,
+      "step": 11210
+    },
+    {
+      "epoch": 0.35779202142925476,
+      "grad_norm": 0.2073378711938858,
+      "learning_rate": 0.001,
+      "loss": 1.1598,
+      "step": 11220
+    },
+    {
+      "epoch": 0.35811090914888866,
+      "grad_norm": 0.21896976232528687,
+      "learning_rate": 0.001,
+      "loss": 1.1781,
+      "step": 11230
+    },
+    {
+      "epoch": 0.3584297968685226,
+      "grad_norm": 0.21934987604618073,
+      "learning_rate": 0.001,
+      "loss": 1.1788,
+      "step": 11240
+    },
+    {
+      "epoch": 0.3587486845881565,
+      "grad_norm": 0.2167501598596573,
+      "learning_rate": 0.001,
+      "loss": 1.155,
+      "step": 11250
+    },
+    {
+      "epoch": 0.3590675723077904,
+      "grad_norm": 0.20548115670681,
+      "learning_rate": 0.001,
+      "loss": 1.1457,
+      "step": 11260
+    },
+    {
+      "epoch": 0.35938646002742436,
+      "grad_norm": 0.2144881784915924,
+      "learning_rate": 0.001,
+      "loss": 1.1708,
+      "step": 11270
+    },
+    {
+      "epoch": 0.35970534774705826,
+      "grad_norm": 0.20482467114925385,
+      "learning_rate": 0.001,
+      "loss": 1.141,
+      "step": 11280
+    },
+    {
+      "epoch": 0.36002423546669216,
+      "grad_norm": 0.21546229720115662,
+      "learning_rate": 0.001,
+      "loss": 1.1507,
+      "step": 11290
+    },
+    {
+      "epoch": 0.3603431231863261,
+      "grad_norm": 0.2183115929365158,
+      "learning_rate": 0.001,
+      "loss": 1.1648,
+      "step": 11300
+    },
+    {
+      "epoch": 0.36066201090596,
+      "grad_norm": 0.20953315496444702,
+      "learning_rate": 0.001,
+      "loss": 1.1527,
+      "step": 11310
+    },
+    {
+      "epoch": 0.3609808986255939,
+      "grad_norm": 0.21324065327644348,
+      "learning_rate": 0.001,
+      "loss": 1.1472,
+      "step": 11320
+    },
+    {
+      "epoch": 0.36129978634522786,
+      "grad_norm": 0.20822182297706604,
+      "learning_rate": 0.001,
+      "loss": 1.145,
+      "step": 11330
+    },
+    {
+      "epoch": 0.36161867406486176,
+      "grad_norm": 0.21162369847297668,
+      "learning_rate": 0.001,
+      "loss": 1.1699,
+      "step": 11340
+    },
+    {
+      "epoch": 0.36193756178449565,
+      "grad_norm": 0.21036869287490845,
+      "learning_rate": 0.001,
+      "loss": 1.1592,
+      "step": 11350
+    },
+    {
+      "epoch": 0.3622564495041296,
+      "grad_norm": 0.21742284297943115,
+      "learning_rate": 0.001,
+      "loss": 1.1698,
+      "step": 11360
+    },
+    {
+      "epoch": 0.3625753372237635,
+      "grad_norm": 0.20884467661380768,
+      "learning_rate": 0.001,
+      "loss": 1.155,
+      "step": 11370
+    },
+    {
+      "epoch": 0.3628942249433974,
+      "grad_norm": 0.21736779808998108,
+      "learning_rate": 0.001,
+      "loss": 1.1549,
+      "step": 11380
+    },
+    {
+      "epoch": 0.36321311266303136,
+      "grad_norm": 0.2081272453069687,
+      "learning_rate": 0.001,
+      "loss": 1.1599,
+      "step": 11390
+    },
+    {
+      "epoch": 0.36353200038266525,
+      "grad_norm": 0.21277934312820435,
+      "learning_rate": 0.001,
+      "loss": 1.1681,
+      "step": 11400
+    },
+    {
+      "epoch": 0.3638508881022992,
+      "grad_norm": 0.21342188119888306,
+      "learning_rate": 0.001,
+      "loss": 1.1478,
+      "step": 11410
+    },
+    {
+      "epoch": 0.3641697758219331,
+      "grad_norm": 0.21795684099197388,
+      "learning_rate": 0.001,
+      "loss": 1.1564,
+      "step": 11420
+    },
+    {
+      "epoch": 0.364488663541567,
+      "grad_norm": 0.2076573669910431,
+      "learning_rate": 0.001,
+      "loss": 1.1331,
+      "step": 11430
+    },
+    {
+      "epoch": 0.36480755126120096,
+      "grad_norm": 0.22240781784057617,
+      "learning_rate": 0.001,
+      "loss": 1.1609,
+      "step": 11440
+    },
+    {
+      "epoch": 0.36512643898083486,
+      "grad_norm": 0.208001971244812,
+      "learning_rate": 0.001,
+      "loss": 1.1581,
+      "step": 11450
+    },
+    {
+      "epoch": 0.36544532670046875,
+      "grad_norm": 0.20453792810440063,
+      "learning_rate": 0.001,
+      "loss": 1.1437,
+      "step": 11460
+    },
+    {
+      "epoch": 0.3657642144201027,
+      "grad_norm": 0.2149641364812851,
+      "learning_rate": 0.001,
+      "loss": 1.1582,
+      "step": 11470
+    },
+    {
+      "epoch": 0.3660831021397366,
+      "grad_norm": 0.21504372358322144,
+      "learning_rate": 0.001,
+      "loss": 1.1506,
+      "step": 11480
+    },
+    {
+      "epoch": 0.3664019898593705,
+      "grad_norm": 0.21626321971416473,
+      "learning_rate": 0.001,
+      "loss": 1.157,
+      "step": 11490
+    },
+    {
+      "epoch": 0.36672087757900446,
+      "grad_norm": 0.22062943875789642,
+      "learning_rate": 0.001,
+      "loss": 1.1718,
+      "step": 11500
+    },
+    {
+      "epoch": 0.36703976529863835,
+      "grad_norm": 0.20778238773345947,
+      "learning_rate": 0.001,
+      "loss": 1.1441,
+      "step": 11510
+    },
+    {
+      "epoch": 0.36735865301827225,
+      "grad_norm": 0.21755580604076385,
+      "learning_rate": 0.001,
+      "loss": 1.1711,
+      "step": 11520
+    },
+    {
+      "epoch": 0.3676775407379062,
+      "grad_norm": 0.21479518711566925,
+      "learning_rate": 0.001,
+      "loss": 1.1441,
+      "step": 11530
+    },
+    {
+      "epoch": 0.3679964284575401,
+      "grad_norm": 0.2130654752254486,
+      "learning_rate": 0.001,
+      "loss": 1.1366,
+      "step": 11540
+    },
+    {
+      "epoch": 0.368315316177174,
+      "grad_norm": 0.21198046207427979,
+      "learning_rate": 0.001,
+      "loss": 1.1578,
+      "step": 11550
+    },
+    {
+      "epoch": 0.36863420389680795,
+      "grad_norm": 0.2044605165719986,
+      "learning_rate": 0.001,
+      "loss": 1.1484,
+      "step": 11560
+    },
+    {
+      "epoch": 0.36895309161644185,
+      "grad_norm": 0.21563421189785004,
+      "learning_rate": 0.001,
+      "loss": 1.1645,
+      "step": 11570
+    },
+    {
+      "epoch": 0.36927197933607575,
+      "grad_norm": 0.20926934480667114,
+      "learning_rate": 0.001,
+      "loss": 1.131,
+      "step": 11580
+    },
+    {
+      "epoch": 0.3695908670557097,
+      "grad_norm": 0.21198894083499908,
+      "learning_rate": 0.001,
+      "loss": 1.1424,
+      "step": 11590
+    },
+    {
+      "epoch": 0.3699097547753436,
+      "grad_norm": 0.20869193971157074,
+      "learning_rate": 0.001,
+      "loss": 1.1747,
+      "step": 11600
+    },
+    {
+      "epoch": 0.3702286424949775,
+      "grad_norm": 0.20894703269004822,
+      "learning_rate": 0.001,
+      "loss": 1.1578,
+      "step": 11610
+    },
+    {
+      "epoch": 0.37054753021461145,
+      "grad_norm": 0.2092287838459015,
+      "learning_rate": 0.001,
+      "loss": 1.1677,
+      "step": 11620
+    },
+    {
+      "epoch": 0.37086641793424535,
+      "grad_norm": 0.21486634016036987,
+      "learning_rate": 0.001,
+      "loss": 1.1626,
+      "step": 11630
+    },
+    {
+      "epoch": 0.37118530565387925,
+      "grad_norm": 0.2121204286813736,
+      "learning_rate": 0.001,
+      "loss": 1.1816,
+      "step": 11640
+    },
+    {
+      "epoch": 0.3715041933735132,
+      "grad_norm": 0.20252400636672974,
+      "learning_rate": 0.001,
+      "loss": 1.1521,
+      "step": 11650
+    },
+    {
+      "epoch": 0.3718230810931471,
+      "grad_norm": 0.22328728437423706,
+      "learning_rate": 0.001,
+      "loss": 1.1693,
+      "step": 11660
+    },
+    {
+      "epoch": 0.372141968812781,
+      "grad_norm": 0.2148735076189041,
+      "learning_rate": 0.001,
+      "loss": 1.1723,
+      "step": 11670
+    },
+    {
+      "epoch": 0.37246085653241495,
+      "grad_norm": 0.20686081051826477,
+      "learning_rate": 0.001,
+      "loss": 1.1521,
+      "step": 11680
+    },
+    {
+      "epoch": 0.37277974425204885,
+      "grad_norm": 0.2087356299161911,
+      "learning_rate": 0.001,
+      "loss": 1.1516,
+      "step": 11690
+    },
+    {
+      "epoch": 0.37309863197168275,
+      "grad_norm": 0.20420542359352112,
+      "learning_rate": 0.001,
+      "loss": 1.1394,
+      "step": 11700
+    },
+    {
+      "epoch": 0.3734175196913167,
+      "grad_norm": 0.20632268488407135,
+      "learning_rate": 0.001,
+      "loss": 1.152,
+      "step": 11710
+    },
+    {
+      "epoch": 0.3737364074109506,
+      "grad_norm": 0.20878000557422638,
+      "learning_rate": 0.001,
+      "loss": 1.1492,
+      "step": 11720
+    },
+    {
+      "epoch": 0.3740552951305845,
+      "grad_norm": 0.21679173409938812,
+      "learning_rate": 0.001,
+      "loss": 1.1608,
+      "step": 11730
+    },
+    {
+      "epoch": 0.37437418285021845,
+      "grad_norm": 0.2118385136127472,
+      "learning_rate": 0.001,
+      "loss": 1.1604,
+      "step": 11740
+    },
+    {
+      "epoch": 0.37469307056985235,
+      "grad_norm": 0.2076609879732132,
+      "learning_rate": 0.001,
+      "loss": 1.1706,
+      "step": 11750
+    },
+    {
+      "epoch": 0.37501195828948625,
+      "grad_norm": 0.2135295271873474,
+      "learning_rate": 0.001,
+      "loss": 1.1611,
+      "step": 11760
+    },
+    {
+      "epoch": 0.3753308460091202,
+      "grad_norm": 0.21652574837207794,
+      "learning_rate": 0.001,
+      "loss": 1.1626,
+      "step": 11770
+    },
+    {
+      "epoch": 0.3756497337287541,
+      "grad_norm": 0.20713207125663757,
+      "learning_rate": 0.001,
+      "loss": 1.1533,
+      "step": 11780
+    },
+    {
+      "epoch": 0.375968621448388,
+      "grad_norm": 0.21347831189632416,
+      "learning_rate": 0.001,
+      "loss": 1.1775,
+      "step": 11790
+    },
+    {
+      "epoch": 0.37628750916802195,
+      "grad_norm": 0.2028295397758484,
+      "learning_rate": 0.001,
+      "loss": 1.1549,
+      "step": 11800
+    },
+    {
+      "epoch": 0.37660639688765585,
+      "grad_norm": 0.20557062327861786,
+      "learning_rate": 0.001,
+      "loss": 1.159,
+      "step": 11810
+    },
+    {
+      "epoch": 0.37692528460728975,
+      "grad_norm": 0.20521752536296844,
+      "learning_rate": 0.001,
+      "loss": 1.1624,
+      "step": 11820
+    },
+    {
+      "epoch": 0.3772441723269237,
+      "grad_norm": 0.20790110528469086,
+      "learning_rate": 0.001,
+      "loss": 1.1359,
+      "step": 11830
+    },
+    {
+      "epoch": 0.3775630600465576,
+      "grad_norm": 0.20268124341964722,
+      "learning_rate": 0.001,
+      "loss": 1.1597,
+      "step": 11840
+    },
+    {
+      "epoch": 0.37788194776619155,
+      "grad_norm": 0.20956410467624664,
+      "learning_rate": 0.001,
+      "loss": 1.1579,
+      "step": 11850
+    },
+    {
+      "epoch": 0.37820083548582545,
+      "grad_norm": 0.20633342862129211,
+      "learning_rate": 0.001,
+      "loss": 1.165,
+      "step": 11860
+    },
+    {
+      "epoch": 0.37851972320545935,
+      "grad_norm": 0.20639200508594513,
+      "learning_rate": 0.001,
+      "loss": 1.1598,
+      "step": 11870
+    },
+    {
+      "epoch": 0.3788386109250933,
+      "grad_norm": 0.2101225107908249,
+      "learning_rate": 0.001,
+      "loss": 1.1587,
+      "step": 11880
+    },
+    {
+      "epoch": 0.3791574986447272,
+      "grad_norm": 0.2012529969215393,
+      "learning_rate": 0.001,
+      "loss": 1.1504,
+      "step": 11890
+    },
+    {
+      "epoch": 0.3794763863643611,
+      "grad_norm": 0.2067694514989853,
+      "learning_rate": 0.001,
+      "loss": 1.1442,
+      "step": 11900
+    },
+    {
+      "epoch": 0.37979527408399505,
+      "grad_norm": 0.20460672676563263,
+      "learning_rate": 0.001,
+      "loss": 1.1604,
+      "step": 11910
+    },
+    {
+      "epoch": 0.38011416180362895,
+      "grad_norm": 0.20649559795856476,
+      "learning_rate": 0.001,
+      "loss": 1.164,
+      "step": 11920
+    },
+    {
+      "epoch": 0.38043304952326285,
+      "grad_norm": 0.20930494368076324,
+      "learning_rate": 0.001,
+      "loss": 1.1323,
+      "step": 11930
+    },
+    {
+      "epoch": 0.3807519372428968,
+      "grad_norm": 0.20886340737342834,
+      "learning_rate": 0.001,
+      "loss": 1.1271,
+      "step": 11940
+    },
+    {
+      "epoch": 0.3810708249625307,
+      "grad_norm": 0.21348540484905243,
+      "learning_rate": 0.001,
+      "loss": 1.1522,
+      "step": 11950
+    },
+    {
+      "epoch": 0.3813897126821646,
+      "grad_norm": 0.21220609545707703,
+      "learning_rate": 0.001,
+      "loss": 1.1615,
+      "step": 11960
+    },
+    {
+      "epoch": 0.38170860040179855,
+      "grad_norm": 0.2051163762807846,
+      "learning_rate": 0.001,
+      "loss": 1.1568,
+      "step": 11970
+    },
+    {
+      "epoch": 0.38202748812143245,
+      "grad_norm": 0.20502525568008423,
+      "learning_rate": 0.001,
+      "loss": 1.1374,
+      "step": 11980
+    },
+    {
+      "epoch": 0.38234637584106634,
+      "grad_norm": 0.21495133638381958,
+      "learning_rate": 0.001,
+      "loss": 1.1492,
+      "step": 11990
+    },
+    {
+      "epoch": 0.3826652635607003,
+      "grad_norm": 0.2174481600522995,
+      "learning_rate": 0.001,
+      "loss": 1.1474,
+      "step": 12000
+    },
+    {
+      "epoch": 0.3829841512803342,
+      "grad_norm": 0.21735884249210358,
+      "learning_rate": 0.001,
+      "loss": 1.1598,
+      "step": 12010
+    },
+    {
+      "epoch": 0.3833030389999681,
+      "grad_norm": 0.21421964466571808,
+      "learning_rate": 0.001,
+      "loss": 1.1757,
+      "step": 12020
+    },
+    {
+      "epoch": 0.38362192671960205,
+      "grad_norm": 0.2076413929462433,
+      "learning_rate": 0.001,
+      "loss": 1.1523,
+      "step": 12030
+    },
+    {
+      "epoch": 0.38394081443923594,
+      "grad_norm": 0.2062806487083435,
+      "learning_rate": 0.001,
+      "loss": 1.1421,
+      "step": 12040
+    },
+    {
+      "epoch": 0.38425970215886984,
+      "grad_norm": 0.2101665884256363,
+      "learning_rate": 0.001,
+      "loss": 1.1509,
+      "step": 12050
+    },
+    {
+      "epoch": 0.3845785898785038,
+      "grad_norm": 0.2052222043275833,
+      "learning_rate": 0.001,
+      "loss": 1.1387,
+      "step": 12060
+    },
+    {
+      "epoch": 0.3848974775981377,
+      "grad_norm": 0.20605748891830444,
+      "learning_rate": 0.001,
+      "loss": 1.1564,
+      "step": 12070
+    },
+    {
+      "epoch": 0.3852163653177716,
+      "grad_norm": 0.2109629064798355,
+      "learning_rate": 0.001,
+      "loss": 1.1539,
+      "step": 12080
+    },
+    {
+      "epoch": 0.38553525303740555,
+      "grad_norm": 0.21306735277175903,
+      "learning_rate": 0.001,
+      "loss": 1.1447,
+      "step": 12090
+    },
+    {
+      "epoch": 0.38585414075703944,
+      "grad_norm": 0.2046939879655838,
+      "learning_rate": 0.001,
+      "loss": 1.1549,
+      "step": 12100
+    },
+    {
+      "epoch": 0.38617302847667334,
+      "grad_norm": 0.20633463561534882,
+      "learning_rate": 0.001,
+      "loss": 1.1547,
+      "step": 12110
+    },
+    {
+      "epoch": 0.3864919161963073,
+      "grad_norm": 0.20908667147159576,
+      "learning_rate": 0.001,
+      "loss": 1.1626,
+      "step": 12120
+    },
+    {
+      "epoch": 0.3868108039159412,
+      "grad_norm": 0.2130550891160965,
+      "learning_rate": 0.001,
+      "loss": 1.1633,
+      "step": 12130
+    },
+    {
+      "epoch": 0.3871296916355751,
+      "grad_norm": 0.2054852545261383,
+      "learning_rate": 0.001,
+      "loss": 1.1264,
+      "step": 12140
+    },
+    {
+      "epoch": 0.38744857935520904,
+      "grad_norm": 0.2040906846523285,
+      "learning_rate": 0.001,
+      "loss": 1.1707,
+      "step": 12150
+    },
+    {
+      "epoch": 0.38776746707484294,
+      "grad_norm": 0.20435848832130432,
+      "learning_rate": 0.001,
+      "loss": 1.165,
+      "step": 12160
+    },
+    {
+      "epoch": 0.38808635479447684,
+      "grad_norm": 0.2084130495786667,
+      "learning_rate": 0.001,
+      "loss": 1.1264,
+      "step": 12170
+    },
+    {
+      "epoch": 0.3884052425141108,
+      "grad_norm": 0.2129623293876648,
+      "learning_rate": 0.001,
+      "loss": 1.1697,
+      "step": 12180
+    },
+    {
+      "epoch": 0.3887241302337447,
+      "grad_norm": 0.21030132472515106,
+      "learning_rate": 0.001,
+      "loss": 1.1461,
+      "step": 12190
+    },
+    {
+      "epoch": 0.3890430179533786,
+      "grad_norm": 0.20669963955879211,
+      "learning_rate": 0.001,
+      "loss": 1.1445,
+      "step": 12200
+    },
+    {
+      "epoch": 0.38936190567301254,
+      "grad_norm": 0.2081628292798996,
+      "learning_rate": 0.001,
+      "loss": 1.1528,
+      "step": 12210
+    },
+    {
+      "epoch": 0.38968079339264644,
+      "grad_norm": 0.20809587836265564,
+      "learning_rate": 0.001,
+      "loss": 1.1561,
+      "step": 12220
+    },
+    {
+      "epoch": 0.38999968111228034,
+      "grad_norm": 0.20185744762420654,
+      "learning_rate": 0.001,
+      "loss": 1.1413,
+      "step": 12230
+    },
+    {
+      "epoch": 0.3903185688319143,
+      "grad_norm": 0.20260155200958252,
+      "learning_rate": 0.001,
+      "loss": 1.1399,
+      "step": 12240
+    },
+    {
+      "epoch": 0.3906374565515482,
+      "grad_norm": 0.21670649945735931,
+      "learning_rate": 0.001,
+      "loss": 1.1275,
+      "step": 12250
+    },
+    {
+      "epoch": 0.39095634427118214,
+      "grad_norm": 0.21641764044761658,
+      "learning_rate": 0.001,
+      "loss": 1.1474,
+      "step": 12260
+    },
+    {
+      "epoch": 0.39127523199081604,
+      "grad_norm": 0.20669713616371155,
+      "learning_rate": 0.001,
+      "loss": 1.1597,
+      "step": 12270
+    },
+    {
+      "epoch": 0.39159411971044994,
+      "grad_norm": 0.21705591678619385,
+      "learning_rate": 0.001,
+      "loss": 1.1473,
+      "step": 12280
+    },
+    {
+      "epoch": 0.3919130074300839,
+      "grad_norm": 0.2091154009103775,
+      "learning_rate": 0.001,
+      "loss": 1.1463,
+      "step": 12290
+    },
+    {
+      "epoch": 0.3922318951497178,
+      "grad_norm": 0.20971107482910156,
+      "learning_rate": 0.001,
+      "loss": 1.1617,
+      "step": 12300
+    },
+    {
+      "epoch": 0.3925507828693517,
+      "grad_norm": 0.2061934471130371,
+      "learning_rate": 0.001,
+      "loss": 1.1493,
+      "step": 12310
+    },
+    {
+      "epoch": 0.39286967058898564,
+      "grad_norm": 0.20891344547271729,
+      "learning_rate": 0.001,
+      "loss": 1.1559,
+      "step": 12320
+    },
+    {
+      "epoch": 0.39318855830861954,
+      "grad_norm": 0.21326862275600433,
+      "learning_rate": 0.001,
+      "loss": 1.1546,
+      "step": 12330
+    },
+    {
+      "epoch": 0.39350744602825344,
+      "grad_norm": 0.2107415497303009,
+      "learning_rate": 0.001,
+      "loss": 1.1388,
+      "step": 12340
+    },
+    {
+      "epoch": 0.3938263337478874,
+      "grad_norm": 0.20264574885368347,
+      "learning_rate": 0.001,
+      "loss": 1.1338,
+      "step": 12350
+    },
+    {
+      "epoch": 0.3941452214675213,
+      "grad_norm": 0.20951254665851593,
+      "learning_rate": 0.001,
+      "loss": 1.1702,
+      "step": 12360
+    },
+    {
+      "epoch": 0.3944641091871552,
+      "grad_norm": 0.2009592056274414,
+      "learning_rate": 0.001,
+      "loss": 1.1421,
+      "step": 12370
+    },
+    {
+      "epoch": 0.39478299690678914,
+      "grad_norm": 0.2120770961046219,
+      "learning_rate": 0.001,
+      "loss": 1.1384,
+      "step": 12380
+    },
+    {
+      "epoch": 0.39510188462642304,
+      "grad_norm": 0.22238101065158844,
+      "learning_rate": 0.001,
+      "loss": 1.1391,
+      "step": 12390
+    },
+    {
+      "epoch": 0.39542077234605694,
+      "grad_norm": 0.20835790038108826,
+      "learning_rate": 0.001,
+      "loss": 1.154,
+      "step": 12400
+    },
+    {
+      "epoch": 0.3957396600656909,
+      "grad_norm": 0.20659910142421722,
+      "learning_rate": 0.001,
+      "loss": 1.1403,
+      "step": 12410
+    },
+    {
+      "epoch": 0.3960585477853248,
+      "grad_norm": 0.21030953526496887,
+      "learning_rate": 0.001,
+      "loss": 1.1501,
+      "step": 12420
+    },
+    {
+      "epoch": 0.3963774355049587,
+      "grad_norm": 0.20341965556144714,
+      "learning_rate": 0.001,
+      "loss": 1.1516,
+      "step": 12430
+    },
+    {
+      "epoch": 0.39669632322459264,
+      "grad_norm": 0.20903363823890686,
+      "learning_rate": 0.001,
+      "loss": 1.129,
+      "step": 12440
+    },
+    {
+      "epoch": 0.39701521094422654,
+      "grad_norm": 0.210405170917511,
+      "learning_rate": 0.001,
+      "loss": 1.1531,
+      "step": 12450
+    },
+    {
+      "epoch": 0.39733409866386044,
+      "grad_norm": 0.20492255687713623,
+      "learning_rate": 0.001,
+      "loss": 1.1362,
+      "step": 12460
+    },
+    {
+      "epoch": 0.3976529863834944,
+      "grad_norm": 0.21225130558013916,
+      "learning_rate": 0.001,
+      "loss": 1.1609,
+      "step": 12470
+    },
+    {
+      "epoch": 0.3979718741031283,
+      "grad_norm": 0.2099357396364212,
+      "learning_rate": 0.001,
+      "loss": 1.158,
+      "step": 12480
+    },
+    {
+      "epoch": 0.3982907618227622,
+      "grad_norm": 0.2044241577386856,
+      "learning_rate": 0.001,
+      "loss": 1.1476,
+      "step": 12490
+    },
+    {
+      "epoch": 0.39860964954239614,
+      "grad_norm": 0.20494481921195984,
+      "learning_rate": 0.001,
+      "loss": 1.1621,
+      "step": 12500
+    },
+    {
+      "epoch": 0.39892853726203004,
+      "grad_norm": 0.2142401486635208,
+      "learning_rate": 0.001,
+      "loss": 1.1352,
+      "step": 12510
+    },
+    {
+      "epoch": 0.39924742498166393,
+      "grad_norm": 0.2000083476305008,
+      "learning_rate": 0.001,
+      "loss": 1.1388,
+      "step": 12520
+    },
+    {
+      "epoch": 0.3995663127012979,
+      "grad_norm": 0.2018578201532364,
+      "learning_rate": 0.001,
+      "loss": 1.1563,
+      "step": 12530
+    },
+    {
+      "epoch": 0.3998852004209318,
+      "grad_norm": 0.21176375448703766,
+      "learning_rate": 0.001,
+      "loss": 1.1445,
+      "step": 12540
+    },
+    {
+      "epoch": 0.4002040881405657,
+      "grad_norm": 0.21311414241790771,
+      "learning_rate": 0.001,
+      "loss": 1.1257,
+      "step": 12550
+    },
+    {
+      "epoch": 0.40052297586019964,
+      "grad_norm": 0.19823896884918213,
+      "learning_rate": 0.001,
+      "loss": 1.1407,
+      "step": 12560
+    },
+    {
+      "epoch": 0.40084186357983353,
+      "grad_norm": 0.2098669707775116,
+      "learning_rate": 0.001,
+      "loss": 1.1684,
+      "step": 12570
+    },
+    {
+      "epoch": 0.40116075129946743,
+      "grad_norm": 0.20935393869876862,
+      "learning_rate": 0.001,
+      "loss": 1.1175,
+      "step": 12580
+    },
+    {
+      "epoch": 0.4014796390191014,
+      "grad_norm": 0.20373448729515076,
+      "learning_rate": 0.001,
+      "loss": 1.1666,
+      "step": 12590
+    },
+    {
+      "epoch": 0.4017985267387353,
+      "grad_norm": 0.2081720381975174,
+      "learning_rate": 0.001,
+      "loss": 1.1402,
+      "step": 12600
+    },
+    {
+      "epoch": 0.4021174144583692,
+      "grad_norm": 0.20864781737327576,
+      "learning_rate": 0.001,
+      "loss": 1.1447,
+      "step": 12610
+    },
+    {
+      "epoch": 0.40243630217800314,
+      "grad_norm": 0.20449693500995636,
+      "learning_rate": 0.001,
+      "loss": 1.1525,
+      "step": 12620
+    },
+    {
+      "epoch": 0.40275518989763703,
+      "grad_norm": 0.20613113045692444,
+      "learning_rate": 0.001,
+      "loss": 1.1263,
+      "step": 12630
+    },
+    {
+      "epoch": 0.40307407761727093,
+      "grad_norm": 0.20619328320026398,
+      "learning_rate": 0.001,
+      "loss": 1.127,
+      "step": 12640
+    },
+    {
+      "epoch": 0.4033929653369049,
+      "grad_norm": 0.20716503262519836,
+      "learning_rate": 0.001,
+      "loss": 1.1516,
+      "step": 12650
+    },
+    {
+      "epoch": 0.4037118530565388,
+      "grad_norm": 0.20824389159679413,
+      "learning_rate": 0.001,
+      "loss": 1.141,
+      "step": 12660
+    },
+    {
+      "epoch": 0.40403074077617274,
+      "grad_norm": 0.2075486034154892,
+      "learning_rate": 0.001,
+      "loss": 1.138,
+      "step": 12670
+    },
+    {
+      "epoch": 0.40434962849580663,
+      "grad_norm": 0.21800272166728973,
+      "learning_rate": 0.001,
+      "loss": 1.1466,
+      "step": 12680
+    },
+    {
+      "epoch": 0.40466851621544053,
+      "grad_norm": 0.2054111361503601,
+      "learning_rate": 0.001,
+      "loss": 1.1323,
+      "step": 12690
+    },
+    {
+      "epoch": 0.4049874039350745,
+      "grad_norm": 0.20511585474014282,
+      "learning_rate": 0.001,
+      "loss": 1.1468,
+      "step": 12700
+    },
+    {
+      "epoch": 0.4053062916547084,
+      "grad_norm": 0.21086783707141876,
+      "learning_rate": 0.001,
+      "loss": 1.1305,
+      "step": 12710
+    },
+    {
+      "epoch": 0.4056251793743423,
+      "grad_norm": 0.2127915620803833,
+      "learning_rate": 0.001,
+      "loss": 1.1521,
+      "step": 12720
+    },
+    {
+      "epoch": 0.40594406709397624,
+      "grad_norm": 0.2181047797203064,
+      "learning_rate": 0.001,
+      "loss": 1.1347,
+      "step": 12730
+    },
+    {
+      "epoch": 0.40626295481361013,
+      "grad_norm": 0.21023984253406525,
+      "learning_rate": 0.001,
+      "loss": 1.1454,
+      "step": 12740
+    },
+    {
+      "epoch": 0.40658184253324403,
+      "grad_norm": 0.1992928385734558,
+      "learning_rate": 0.001,
+      "loss": 1.1357,
+      "step": 12750
+    },
+    {
+      "epoch": 0.406900730252878,
+      "grad_norm": 0.20112110674381256,
+      "learning_rate": 0.001,
+      "loss": 1.141,
+      "step": 12760
+    },
+    {
+      "epoch": 0.4072196179725119,
+      "grad_norm": 0.2041330635547638,
+      "learning_rate": 0.001,
+      "loss": 1.1541,
+      "step": 12770
+    },
+    {
+      "epoch": 0.4075385056921458,
+      "grad_norm": 0.20971736311912537,
+      "learning_rate": 0.001,
+      "loss": 1.1343,
+      "step": 12780
+    },
+    {
+      "epoch": 0.40785739341177973,
+      "grad_norm": 0.20947255194187164,
+      "learning_rate": 0.001,
+      "loss": 1.1403,
+      "step": 12790
+    },
+    {
+      "epoch": 0.40817628113141363,
+      "grad_norm": 0.2103559523820877,
+      "learning_rate": 0.001,
+      "loss": 1.1574,
+      "step": 12800
+    },
+    {
+      "epoch": 0.40849516885104753,
+      "grad_norm": 0.20097647607326508,
+      "learning_rate": 0.001,
+      "loss": 1.1358,
+      "step": 12810
+    },
+    {
+      "epoch": 0.4088140565706815,
+      "grad_norm": 0.20684273540973663,
+      "learning_rate": 0.001,
+      "loss": 1.1376,
+      "step": 12820
+    },
+    {
+      "epoch": 0.4091329442903154,
+      "grad_norm": 0.2013324797153473,
+      "learning_rate": 0.001,
+      "loss": 1.1315,
+      "step": 12830
+    },
+    {
+      "epoch": 0.4094518320099493,
+      "grad_norm": 0.20758014917373657,
+      "learning_rate": 0.001,
+      "loss": 1.1097,
+      "step": 12840
+    },
+    {
+      "epoch": 0.40977071972958323,
+      "grad_norm": 0.21744020283222198,
+      "learning_rate": 0.001,
+      "loss": 1.1434,
+      "step": 12850
+    },
+    {
+      "epoch": 0.41008960744921713,
+      "grad_norm": 0.19597770273685455,
+      "learning_rate": 0.001,
+      "loss": 1.1233,
+      "step": 12860
+    },
+    {
+      "epoch": 0.41040849516885103,
+      "grad_norm": 0.19693255424499512,
+      "learning_rate": 0.001,
+      "loss": 1.1297,
+      "step": 12870
+    },
+    {
+      "epoch": 0.410727382888485,
+      "grad_norm": 0.2022457867860794,
+      "learning_rate": 0.001,
+      "loss": 1.144,
+      "step": 12880
+    },
+    {
+      "epoch": 0.4110462706081189,
+      "grad_norm": 0.20683251321315765,
+      "learning_rate": 0.001,
+      "loss": 1.149,
+      "step": 12890
+    },
+    {
+      "epoch": 0.4113651583277528,
+      "grad_norm": 0.202144593000412,
+      "learning_rate": 0.001,
+      "loss": 1.1299,
+      "step": 12900
+    },
+    {
+      "epoch": 0.41168404604738673,
+      "grad_norm": 0.1991073489189148,
+      "learning_rate": 0.001,
+      "loss": 1.1262,
+      "step": 12910
+    },
+    {
+      "epoch": 0.41200293376702063,
+      "grad_norm": 0.20432505011558533,
+      "learning_rate": 0.001,
+      "loss": 1.137,
+      "step": 12920
+    },
+    {
+      "epoch": 0.4123218214866545,
+      "grad_norm": 0.2065206915140152,
+      "learning_rate": 0.001,
+      "loss": 1.1301,
+      "step": 12930
+    },
+    {
+      "epoch": 0.4126407092062885,
+      "grad_norm": 0.21112370491027832,
+      "learning_rate": 0.001,
+      "loss": 1.1481,
+      "step": 12940
+    },
+    {
+      "epoch": 0.4129595969259224,
+      "grad_norm": 0.215112566947937,
+      "learning_rate": 0.001,
+      "loss": 1.1468,
+      "step": 12950
+    },
+    {
+      "epoch": 0.4132784846455563,
+      "grad_norm": 0.20078018307685852,
+      "learning_rate": 0.001,
+      "loss": 1.1475,
+      "step": 12960
+    },
+    {
+      "epoch": 0.41359737236519023,
+      "grad_norm": 0.21014413237571716,
+      "learning_rate": 0.001,
+      "loss": 1.1507,
+      "step": 12970
+    },
+    {
+      "epoch": 0.41391626008482413,
+      "grad_norm": 0.2088528871536255,
+      "learning_rate": 0.001,
+      "loss": 1.1526,
+      "step": 12980
+    },
+    {
+      "epoch": 0.414235147804458,
+      "grad_norm": 0.2205217182636261,
+      "learning_rate": 0.001,
+      "loss": 1.1377,
+      "step": 12990
+    },
+    {
+      "epoch": 0.414554035524092,
+      "grad_norm": 0.20459334552288055,
+      "learning_rate": 0.001,
+      "loss": 1.1384,
+      "step": 13000
+    },
+    {
+      "epoch": 0.4148729232437259,
+      "grad_norm": 0.21253494918346405,
+      "learning_rate": 0.001,
+      "loss": 1.1215,
+      "step": 13010
+    },
+    {
+      "epoch": 0.4151918109633598,
+      "grad_norm": 0.20328834652900696,
+      "learning_rate": 0.001,
+      "loss": 1.1553,
+      "step": 13020
+    },
+    {
+      "epoch": 0.41551069868299373,
+      "grad_norm": 0.2183370292186737,
+      "learning_rate": 0.001,
+      "loss": 1.1477,
+      "step": 13030
+    },
+    {
+      "epoch": 0.4158295864026276,
+      "grad_norm": 0.2050125151872635,
+      "learning_rate": 0.001,
+      "loss": 1.1318,
+      "step": 13040
+    },
+    {
+      "epoch": 0.4161484741222615,
+      "grad_norm": 0.20545805990695953,
+      "learning_rate": 0.001,
+      "loss": 1.1359,
+      "step": 13050
+    },
+    {
+      "epoch": 0.4164673618418955,
+      "grad_norm": 0.2132645845413208,
+      "learning_rate": 0.001,
+      "loss": 1.156,
+      "step": 13060
+    },
+    {
+      "epoch": 0.4167862495615294,
+      "grad_norm": 0.2059139609336853,
+      "learning_rate": 0.001,
+      "loss": 1.1395,
+      "step": 13070
+    },
+    {
+      "epoch": 0.41710513728116333,
+      "grad_norm": 0.2027749866247177,
+      "learning_rate": 0.001,
+      "loss": 1.1437,
+      "step": 13080
+    },
+    {
+      "epoch": 0.4174240250007972,
+      "grad_norm": 0.20797421038150787,
+      "learning_rate": 0.001,
+      "loss": 1.1403,
+      "step": 13090
+    },
+    {
+      "epoch": 0.4177429127204311,
+      "grad_norm": 0.20062734186649323,
+      "learning_rate": 0.001,
+      "loss": 1.1325,
+      "step": 13100
+    },
+    {
+      "epoch": 0.4180618004400651,
+      "grad_norm": 0.21262063086032867,
+      "learning_rate": 0.001,
+      "loss": 1.1396,
+      "step": 13110
+    },
+    {
+      "epoch": 0.418380688159699,
+      "grad_norm": 0.2075001448392868,
+      "learning_rate": 0.001,
+      "loss": 1.149,
+      "step": 13120
+    },
+    {
+      "epoch": 0.4186995758793329,
+      "grad_norm": 0.20594367384910583,
+      "learning_rate": 0.001,
+      "loss": 1.1424,
+      "step": 13130
+    },
+    {
+      "epoch": 0.41901846359896683,
+      "grad_norm": 0.20854982733726501,
+      "learning_rate": 0.001,
+      "loss": 1.1376,
+      "step": 13140
+    },
+    {
+      "epoch": 0.4193373513186007,
+      "grad_norm": 0.20273441076278687,
+      "learning_rate": 0.001,
+      "loss": 1.148,
+      "step": 13150
+    },
+    {
+      "epoch": 0.4196562390382346,
+      "grad_norm": 0.20512281358242035,
+      "learning_rate": 0.001,
+      "loss": 1.1313,
+      "step": 13160
+    },
+    {
+      "epoch": 0.4199751267578686,
+      "grad_norm": 0.20326954126358032,
+      "learning_rate": 0.001,
+      "loss": 1.1415,
+      "step": 13170
+    },
+    {
+      "epoch": 0.4202940144775025,
+      "grad_norm": 0.21317848563194275,
+      "learning_rate": 0.001,
+      "loss": 1.1229,
+      "step": 13180
+    },
+    {
+      "epoch": 0.4206129021971364,
+      "grad_norm": 0.2062990963459015,
+      "learning_rate": 0.001,
+      "loss": 1.1389,
+      "step": 13190
+    },
+    {
+      "epoch": 0.4209317899167703,
+      "grad_norm": 0.2079552263021469,
+      "learning_rate": 0.001,
+      "loss": 1.1538,
+      "step": 13200
+    },
+    {
+      "epoch": 0.4212506776364042,
+      "grad_norm": 0.20866861939430237,
+      "learning_rate": 0.001,
+      "loss": 1.1403,
+      "step": 13210
+    },
+    {
+      "epoch": 0.4215695653560381,
+      "grad_norm": 0.2059970647096634,
+      "learning_rate": 0.001,
+      "loss": 1.1388,
+      "step": 13220
+    },
+    {
+      "epoch": 0.4218884530756721,
+      "grad_norm": 0.20208317041397095,
+      "learning_rate": 0.001,
+      "loss": 1.1436,
+      "step": 13230
+    },
+    {
+      "epoch": 0.422207340795306,
+      "grad_norm": 0.20684318244457245,
+      "learning_rate": 0.001,
+      "loss": 1.1254,
+      "step": 13240
+    },
+    {
+      "epoch": 0.42252622851493987,
+      "grad_norm": 0.2068798840045929,
+      "learning_rate": 0.001,
+      "loss": 1.1176,
+      "step": 13250
+    },
+    {
+      "epoch": 0.4228451162345738,
+      "grad_norm": 0.21011073887348175,
+      "learning_rate": 0.001,
+      "loss": 1.1267,
+      "step": 13260
+    },
+    {
+      "epoch": 0.4231640039542077,
+      "grad_norm": 0.20945926010608673,
+      "learning_rate": 0.001,
+      "loss": 1.1438,
+      "step": 13270
+    },
+    {
+      "epoch": 0.4234828916738416,
+      "grad_norm": 0.20559492707252502,
+      "learning_rate": 0.001,
+      "loss": 1.1207,
+      "step": 13280
+    },
+    {
+      "epoch": 0.4238017793934756,
+      "grad_norm": 0.19994908571243286,
+      "learning_rate": 0.001,
+      "loss": 1.1322,
+      "step": 13290
+    },
+    {
+      "epoch": 0.4241206671131095,
+      "grad_norm": 0.2100694179534912,
+      "learning_rate": 0.001,
+      "loss": 1.1563,
+      "step": 13300
+    },
+    {
+      "epoch": 0.42443955483274337,
+      "grad_norm": 0.2056904286146164,
+      "learning_rate": 0.001,
+      "loss": 1.1523,
+      "step": 13310
+    },
+    {
+      "epoch": 0.4247584425523773,
+      "grad_norm": 0.20013177394866943,
+      "learning_rate": 0.001,
+      "loss": 1.1321,
+      "step": 13320
+    },
+    {
+      "epoch": 0.4250773302720112,
+      "grad_norm": 0.20355407893657684,
+      "learning_rate": 0.001,
+      "loss": 1.1393,
+      "step": 13330
+    },
+    {
+      "epoch": 0.4253962179916451,
+      "grad_norm": 0.2022884041070938,
+      "learning_rate": 0.001,
+      "loss": 1.1307,
+      "step": 13340
+    },
+    {
+      "epoch": 0.4257151057112791,
+      "grad_norm": 0.2134474664926529,
+      "learning_rate": 0.001,
+      "loss": 1.1358,
+      "step": 13350
+    },
+    {
+      "epoch": 0.42603399343091297,
+      "grad_norm": 0.21307939291000366,
+      "learning_rate": 0.001,
+      "loss": 1.1268,
+      "step": 13360
+    },
+    {
+      "epoch": 0.42635288115054687,
+      "grad_norm": 0.21968849003314972,
+      "learning_rate": 0.001,
+      "loss": 1.1401,
+      "step": 13370
+    },
+    {
+      "epoch": 0.4266717688701808,
+      "grad_norm": 0.2057107537984848,
+      "learning_rate": 0.001,
+      "loss": 1.1221,
+      "step": 13380
+    },
+    {
+      "epoch": 0.4269906565898147,
+      "grad_norm": 0.20316410064697266,
+      "learning_rate": 0.001,
+      "loss": 1.1284,
+      "step": 13390
+    },
+    {
+      "epoch": 0.4273095443094486,
+      "grad_norm": 0.201827734708786,
+      "learning_rate": 0.001,
+      "loss": 1.1229,
+      "step": 13400
+    },
+    {
+      "epoch": 0.4276284320290826,
+      "grad_norm": 0.1960659623146057,
+      "learning_rate": 0.001,
+      "loss": 1.1161,
+      "step": 13410
+    },
+    {
+      "epoch": 0.42794731974871647,
+      "grad_norm": 0.209204763174057,
+      "learning_rate": 0.001,
+      "loss": 1.1459,
+      "step": 13420
+    },
+    {
+      "epoch": 0.42826620746835037,
+      "grad_norm": 0.2040780484676361,
+      "learning_rate": 0.001,
+      "loss": 1.1405,
+      "step": 13430
+    },
+    {
+      "epoch": 0.4285850951879843,
+      "grad_norm": 0.2102985382080078,
+      "learning_rate": 0.001,
+      "loss": 1.1436,
+      "step": 13440
+    },
+    {
+      "epoch": 0.4289039829076182,
+      "grad_norm": 0.2021404355764389,
+      "learning_rate": 0.001,
+      "loss": 1.1329,
+      "step": 13450
+    },
+    {
+      "epoch": 0.4292228706272521,
+      "grad_norm": 0.1990777999162674,
+      "learning_rate": 0.001,
+      "loss": 1.134,
+      "step": 13460
+    },
+    {
+      "epoch": 0.42954175834688607,
+      "grad_norm": 0.20001259446144104,
+      "learning_rate": 0.001,
+      "loss": 1.1359,
+      "step": 13470
+    },
+    {
+      "epoch": 0.42986064606651997,
+      "grad_norm": 0.2053236961364746,
+      "learning_rate": 0.001,
+      "loss": 1.1493,
+      "step": 13480
+    },
+    {
+      "epoch": 0.4301795337861539,
+      "grad_norm": 0.19334106147289276,
+      "learning_rate": 0.001,
+      "loss": 1.129,
+      "step": 13490
+    },
+    {
+      "epoch": 0.4304984215057878,
+      "grad_norm": 0.20170816779136658,
+      "learning_rate": 0.001,
+      "loss": 1.1302,
+      "step": 13500
+    },
+    {
+      "epoch": 0.4308173092254217,
+      "grad_norm": 0.2028067409992218,
+      "learning_rate": 0.001,
+      "loss": 1.1287,
+      "step": 13510
+    },
+    {
+      "epoch": 0.43113619694505567,
+      "grad_norm": 0.20177865028381348,
+      "learning_rate": 0.001,
+      "loss": 1.1547,
+      "step": 13520
+    },
+    {
+      "epoch": 0.43145508466468957,
+      "grad_norm": 0.20576360821723938,
+      "learning_rate": 0.001,
+      "loss": 1.1246,
+      "step": 13530
+    },
+    {
+      "epoch": 0.43177397238432347,
+      "grad_norm": 0.2029898315668106,
+      "learning_rate": 0.001,
+      "loss": 1.1383,
+      "step": 13540
+    },
+    {
+      "epoch": 0.4320928601039574,
+      "grad_norm": 0.2129756659269333,
+      "learning_rate": 0.001,
+      "loss": 1.1465,
+      "step": 13550
+    },
+    {
+      "epoch": 0.4324117478235913,
+      "grad_norm": 0.20829495787620544,
+      "learning_rate": 0.001,
+      "loss": 1.1352,
+      "step": 13560
+    },
+    {
+      "epoch": 0.4327306355432252,
+      "grad_norm": 0.20578406751155853,
+      "learning_rate": 0.001,
+      "loss": 1.1344,
+      "step": 13570
+    },
+    {
+      "epoch": 0.43304952326285917,
+      "grad_norm": 0.20792439579963684,
+      "learning_rate": 0.001,
+      "loss": 1.144,
+      "step": 13580
+    },
+    {
+      "epoch": 0.43336841098249307,
+      "grad_norm": 0.207744300365448,
+      "learning_rate": 0.001,
+      "loss": 1.1303,
+      "step": 13590
+    },
+    {
+      "epoch": 0.43368729870212697,
+      "grad_norm": 0.2021283656358719,
+      "learning_rate": 0.001,
+      "loss": 1.1297,
+      "step": 13600
+    },
+    {
+      "epoch": 0.4340061864217609,
+      "grad_norm": 0.2113650143146515,
+      "learning_rate": 0.001,
+      "loss": 1.1322,
+      "step": 13610
+    },
+    {
+      "epoch": 0.4343250741413948,
+      "grad_norm": 0.20138460397720337,
+      "learning_rate": 0.001,
+      "loss": 1.1487,
+      "step": 13620
+    },
+    {
+      "epoch": 0.4346439618610287,
+      "grad_norm": 0.19764484465122223,
+      "learning_rate": 0.001,
+      "loss": 1.1179,
+      "step": 13630
+    },
+    {
+      "epoch": 0.43496284958066267,
+      "grad_norm": 0.20085632801055908,
+      "learning_rate": 0.001,
+      "loss": 1.1391,
+      "step": 13640
+    },
+    {
+      "epoch": 0.43528173730029657,
+      "grad_norm": 0.1991196572780609,
+      "learning_rate": 0.001,
+      "loss": 1.1327,
+      "step": 13650
+    },
+    {
+      "epoch": 0.43560062501993047,
+      "grad_norm": 0.1970689445734024,
+      "learning_rate": 0.001,
+      "loss": 1.1204,
+      "step": 13660
+    },
+    {
+      "epoch": 0.4359195127395644,
+      "grad_norm": 0.2032547891139984,
+      "learning_rate": 0.001,
+      "loss": 1.1384,
+      "step": 13670
+    },
+    {
+      "epoch": 0.4362384004591983,
+      "grad_norm": 0.20708051323890686,
+      "learning_rate": 0.001,
+      "loss": 1.1258,
+      "step": 13680
+    },
+    {
+      "epoch": 0.4365572881788322,
+      "grad_norm": 0.21240440011024475,
+      "learning_rate": 0.001,
+      "loss": 1.129,
+      "step": 13690
+    },
+    {
+      "epoch": 0.43687617589846617,
+      "grad_norm": 0.21483121812343597,
+      "learning_rate": 0.001,
+      "loss": 1.1412,
+      "step": 13700
+    },
+    {
+      "epoch": 0.43719506361810007,
+      "grad_norm": 0.20378629863262177,
+      "learning_rate": 0.001,
+      "loss": 1.1283,
+      "step": 13710
+    },
+    {
+      "epoch": 0.43751395133773396,
+      "grad_norm": 0.19939197599887848,
+      "learning_rate": 0.001,
+      "loss": 1.1122,
+      "step": 13720
+    },
+    {
+      "epoch": 0.4378328390573679,
+      "grad_norm": 0.19610774517059326,
+      "learning_rate": 0.001,
+      "loss": 1.1482,
+      "step": 13730
+    },
+    {
+      "epoch": 0.4381517267770018,
+      "grad_norm": 0.20175138115882874,
+      "learning_rate": 0.001,
+      "loss": 1.1367,
+      "step": 13740
+    },
+    {
+      "epoch": 0.4384706144966357,
+      "grad_norm": 0.20364269614219666,
+      "learning_rate": 0.001,
+      "loss": 1.1276,
+      "step": 13750
+    },
+    {
+      "epoch": 0.43878950221626967,
+      "grad_norm": 0.20994988083839417,
+      "learning_rate": 0.001,
+      "loss": 1.126,
+      "step": 13760
+    },
+    {
+      "epoch": 0.43910838993590356,
+      "grad_norm": 0.2069726586341858,
+      "learning_rate": 0.001,
+      "loss": 1.1354,
+      "step": 13770
+    },
+    {
+      "epoch": 0.43942727765553746,
+      "grad_norm": 0.2140015810728073,
+      "learning_rate": 0.001,
+      "loss": 1.1374,
+      "step": 13780
+    },
+    {
+      "epoch": 0.4397461653751714,
+      "grad_norm": 0.20899558067321777,
+      "learning_rate": 0.001,
+      "loss": 1.1374,
+      "step": 13790
+    },
+    {
+      "epoch": 0.4400650530948053,
+      "grad_norm": 0.20549681782722473,
+      "learning_rate": 0.001,
+      "loss": 1.1326,
+      "step": 13800
+    },
+    {
+      "epoch": 0.4403839408144392,
+      "grad_norm": 0.20833928883075714,
+      "learning_rate": 0.001,
+      "loss": 1.1228,
+      "step": 13810
+    },
+    {
+      "epoch": 0.44070282853407317,
+      "grad_norm": 0.20258395373821259,
+      "learning_rate": 0.001,
+      "loss": 1.1241,
+      "step": 13820
+    },
+    {
+      "epoch": 0.44102171625370706,
+      "grad_norm": 0.20737484097480774,
+      "learning_rate": 0.001,
+      "loss": 1.1264,
+      "step": 13830
+    },
+    {
+      "epoch": 0.44134060397334096,
+      "grad_norm": 0.21442458033561707,
+      "learning_rate": 0.001,
+      "loss": 1.1135,
+      "step": 13840
+    },
+    {
+      "epoch": 0.4416594916929749,
+      "grad_norm": 0.20314937829971313,
+      "learning_rate": 0.001,
+      "loss": 1.1266,
+      "step": 13850
+    },
+    {
+      "epoch": 0.4419783794126088,
+      "grad_norm": 0.2008056789636612,
+      "learning_rate": 0.001,
+      "loss": 1.1277,
+      "step": 13860
+    },
+    {
+      "epoch": 0.4422972671322427,
+      "grad_norm": 0.20103736221790314,
+      "learning_rate": 0.001,
+      "loss": 1.118,
+      "step": 13870
+    },
+    {
+      "epoch": 0.44261615485187666,
+      "grad_norm": 0.20233802497386932,
+      "learning_rate": 0.001,
+      "loss": 1.109,
+      "step": 13880
+    },
+    {
+      "epoch": 0.44293504257151056,
+      "grad_norm": 0.19475015997886658,
+      "learning_rate": 0.001,
+      "loss": 1.1169,
+      "step": 13890
+    },
+    {
+      "epoch": 0.4432539302911445,
+      "grad_norm": 0.20203594863414764,
+      "learning_rate": 0.001,
+      "loss": 1.1366,
+      "step": 13900
+    },
+    {
+      "epoch": 0.4435728180107784,
+      "grad_norm": 0.19808048009872437,
+      "learning_rate": 0.001,
+      "loss": 1.1152,
+      "step": 13910
+    },
+    {
+      "epoch": 0.4438917057304123,
+      "grad_norm": 0.19753964245319366,
+      "learning_rate": 0.001,
+      "loss": 1.1312,
+      "step": 13920
+    },
+    {
+      "epoch": 0.44421059345004626,
+      "grad_norm": 0.20471017062664032,
+      "learning_rate": 0.001,
+      "loss": 1.1402,
+      "step": 13930
+    },
+    {
+      "epoch": 0.44452948116968016,
+      "grad_norm": 0.2034641057252884,
+      "learning_rate": 0.001,
+      "loss": 1.1154,
+      "step": 13940
+    },
+    {
+      "epoch": 0.44484836888931406,
+      "grad_norm": 0.20656068623065948,
+      "learning_rate": 0.001,
+      "loss": 1.1271,
+      "step": 13950
+    },
+    {
+      "epoch": 0.445167256608948,
+      "grad_norm": 0.19708023965358734,
+      "learning_rate": 0.001,
+      "loss": 1.1462,
+      "step": 13960
+    },
+    {
+      "epoch": 0.4454861443285819,
+      "grad_norm": 0.20702509582042694,
+      "learning_rate": 0.001,
+      "loss": 1.1216,
+      "step": 13970
+    },
+    {
+      "epoch": 0.4458050320482158,
+      "grad_norm": 0.20129531621932983,
+      "learning_rate": 0.001,
+      "loss": 1.127,
+      "step": 13980
+    },
+    {
+      "epoch": 0.44612391976784976,
+      "grad_norm": 0.19751425087451935,
+      "learning_rate": 0.001,
+      "loss": 1.1247,
+      "step": 13990
+    },
+    {
+      "epoch": 0.44644280748748366,
+      "grad_norm": 0.20703715085983276,
+      "learning_rate": 0.001,
+      "loss": 1.121,
+      "step": 14000
+    },
+    {
+      "epoch": 0.44676169520711756,
+      "grad_norm": 0.20253314077854156,
+      "learning_rate": 0.001,
+      "loss": 1.1366,
+      "step": 14010
+    },
+    {
+      "epoch": 0.4470805829267515,
+      "grad_norm": 0.21088387072086334,
+      "learning_rate": 0.001,
+      "loss": 1.1354,
+      "step": 14020
+    },
+    {
+      "epoch": 0.4473994706463854,
+      "grad_norm": 0.212107852101326,
+      "learning_rate": 0.001,
+      "loss": 1.1329,
+      "step": 14030
+    },
+    {
+      "epoch": 0.4477183583660193,
+      "grad_norm": 0.19477473199367523,
+      "learning_rate": 0.001,
+      "loss": 1.1186,
+      "step": 14040
+    },
+    {
+      "epoch": 0.44803724608565326,
+      "grad_norm": 0.20723167061805725,
+      "learning_rate": 0.001,
+      "loss": 1.1134,
+      "step": 14050
+    },
+    {
+      "epoch": 0.44835613380528716,
+      "grad_norm": 0.20093318819999695,
+      "learning_rate": 0.001,
+      "loss": 1.1169,
+      "step": 14060
+    },
+    {
+      "epoch": 0.44867502152492106,
+      "grad_norm": 0.2053205668926239,
+      "learning_rate": 0.001,
+      "loss": 1.1109,
+      "step": 14070
+    },
+    {
+      "epoch": 0.448993909244555,
+      "grad_norm": 0.20331956446170807,
+      "learning_rate": 0.001,
+      "loss": 1.125,
+      "step": 14080
+    },
+    {
+      "epoch": 0.4493127969641889,
+      "grad_norm": 0.2000836879014969,
+      "learning_rate": 0.001,
+      "loss": 1.1215,
+      "step": 14090
+    },
+    {
+      "epoch": 0.4496316846838228,
+      "grad_norm": 0.1976880431175232,
+      "learning_rate": 0.001,
+      "loss": 1.1243,
+      "step": 14100
+    },
+    {
+      "epoch": 0.44995057240345676,
+      "grad_norm": 0.20531907677650452,
+      "learning_rate": 0.001,
+      "loss": 1.1278,
+      "step": 14110
+    },
+    {
+      "epoch": 0.45026946012309066,
+      "grad_norm": 0.2220134139060974,
+      "learning_rate": 0.001,
+      "loss": 1.1162,
+      "step": 14120
+    },
+    {
+      "epoch": 0.45058834784272456,
+      "grad_norm": 0.19986401498317719,
+      "learning_rate": 0.001,
+      "loss": 1.1199,
+      "step": 14130
+    },
+    {
+      "epoch": 0.4509072355623585,
+      "grad_norm": 0.20359410345554352,
+      "learning_rate": 0.001,
+      "loss": 1.1143,
+      "step": 14140
+    },
+    {
+      "epoch": 0.4512261232819924,
+      "grad_norm": 0.19630028307437897,
+      "learning_rate": 0.001,
+      "loss": 1.147,
+      "step": 14150
+    },
+    {
+      "epoch": 0.4515450110016263,
+      "grad_norm": 0.20296724140644073,
+      "learning_rate": 0.001,
+      "loss": 1.1201,
+      "step": 14160
+    },
+    {
+      "epoch": 0.45186389872126026,
+      "grad_norm": 0.19895164668560028,
+      "learning_rate": 0.001,
+      "loss": 1.1155,
+      "step": 14170
+    },
+    {
+      "epoch": 0.45218278644089416,
+      "grad_norm": 0.19755735993385315,
+      "learning_rate": 0.001,
+      "loss": 1.1226,
+      "step": 14180
+    },
+    {
+      "epoch": 0.45250167416052806,
+      "grad_norm": 0.19897565245628357,
+      "learning_rate": 0.001,
+      "loss": 1.106,
+      "step": 14190
+    },
+    {
+      "epoch": 0.452820561880162,
+      "grad_norm": 0.20602868497371674,
+      "learning_rate": 0.001,
+      "loss": 1.1359,
+      "step": 14200
+    },
+    {
+      "epoch": 0.4531394495997959,
+      "grad_norm": 0.19570337235927582,
+      "learning_rate": 0.001,
+      "loss": 1.1332,
+      "step": 14210
+    },
+    {
+      "epoch": 0.4534583373194298,
+      "grad_norm": 0.198469340801239,
+      "learning_rate": 0.001,
+      "loss": 1.1235,
+      "step": 14220
+    },
+    {
+      "epoch": 0.45377722503906376,
+      "grad_norm": 0.21264946460723877,
+      "learning_rate": 0.001,
+      "loss": 1.1112,
+      "step": 14230
+    },
+    {
+      "epoch": 0.45409611275869766,
+      "grad_norm": 0.2016083151102066,
+      "learning_rate": 0.001,
+      "loss": 1.1111,
+      "step": 14240
+    },
+    {
+      "epoch": 0.45441500047833155,
+      "grad_norm": 0.20173093676567078,
+      "learning_rate": 0.001,
+      "loss": 1.1295,
+      "step": 14250
+    },
+    {
+      "epoch": 0.4547338881979655,
+      "grad_norm": 0.20293079316616058,
+      "learning_rate": 0.001,
+      "loss": 1.1402,
+      "step": 14260
+    },
+    {
+      "epoch": 0.4550527759175994,
+      "grad_norm": 0.19944404065608978,
+      "learning_rate": 0.001,
+      "loss": 1.1199,
+      "step": 14270
+    },
+    {
+      "epoch": 0.4553716636372333,
+      "grad_norm": 0.20182962715625763,
+      "learning_rate": 0.001,
+      "loss": 1.1321,
+      "step": 14280
+    },
+    {
+      "epoch": 0.45569055135686726,
+      "grad_norm": 0.21239547431468964,
+      "learning_rate": 0.001,
+      "loss": 1.1333,
+      "step": 14290
+    },
+    {
+      "epoch": 0.45600943907650116,
+      "grad_norm": 0.1999959498643875,
+      "learning_rate": 0.001,
+      "loss": 1.1371,
+      "step": 14300
+    },
+    {
+      "epoch": 0.45632832679613505,
+      "grad_norm": 0.20423386991024017,
+      "learning_rate": 0.001,
+      "loss": 1.1355,
+      "step": 14310
+    },
+    {
+      "epoch": 0.456647214515769,
+      "grad_norm": 0.205377459526062,
+      "learning_rate": 0.001,
+      "loss": 1.1277,
+      "step": 14320
+    },
+    {
+      "epoch": 0.4569661022354029,
+      "grad_norm": 0.20549103617668152,
+      "learning_rate": 0.001,
+      "loss": 1.1211,
+      "step": 14330
+    },
+    {
+      "epoch": 0.45728498995503686,
+      "grad_norm": 0.20159433782100677,
+      "learning_rate": 0.001,
+      "loss": 1.1114,
+      "step": 14340
+    },
+    {
+      "epoch": 0.45760387767467076,
+      "grad_norm": 0.19585607945919037,
+      "learning_rate": 0.001,
+      "loss": 1.1233,
+      "step": 14350
+    },
+    {
+      "epoch": 0.45792276539430465,
+      "grad_norm": 0.19485332071781158,
+      "learning_rate": 0.001,
+      "loss": 1.1195,
+      "step": 14360
+    },
+    {
+      "epoch": 0.4582416531139386,
+      "grad_norm": 0.20004454255104065,
+      "learning_rate": 0.001,
+      "loss": 1.148,
+      "step": 14370
+    },
+    {
+      "epoch": 0.4585605408335725,
+      "grad_norm": 0.19846779108047485,
+      "learning_rate": 0.001,
+      "loss": 1.1371,
+      "step": 14380
+    },
+    {
+      "epoch": 0.4588794285532064,
+      "grad_norm": 0.20324277877807617,
+      "learning_rate": 0.001,
+      "loss": 1.1138,
+      "step": 14390
+    },
+    {
+      "epoch": 0.45919831627284036,
+      "grad_norm": 0.20810678601264954,
+      "learning_rate": 0.001,
+      "loss": 1.1397,
+      "step": 14400
+    },
+    {
+      "epoch": 0.45951720399247425,
+      "grad_norm": 0.20195460319519043,
+      "learning_rate": 0.001,
+      "loss": 1.0977,
+      "step": 14410
+    },
+    {
+      "epoch": 0.45983609171210815,
+      "grad_norm": 0.208050936460495,
+      "learning_rate": 0.001,
+      "loss": 1.1208,
+      "step": 14420
+    },
+    {
+      "epoch": 0.4601549794317421,
+      "grad_norm": 0.21316717565059662,
+      "learning_rate": 0.001,
+      "loss": 1.1178,
+      "step": 14430
+    },
+    {
+      "epoch": 0.460473867151376,
+      "grad_norm": 0.20156152546405792,
+      "learning_rate": 0.001,
+      "loss": 1.1368,
+      "step": 14440
+    },
+    {
+      "epoch": 0.4607927548710099,
+      "grad_norm": 0.19752919673919678,
+      "learning_rate": 0.001,
+      "loss": 1.1164,
+      "step": 14450
+    },
+    {
+      "epoch": 0.46111164259064386,
+      "grad_norm": 0.2068374752998352,
+      "learning_rate": 0.001,
+      "loss": 1.1285,
+      "step": 14460
+    },
+    {
+      "epoch": 0.46143053031027775,
+      "grad_norm": 0.20846422016620636,
+      "learning_rate": 0.001,
+      "loss": 1.1185,
+      "step": 14470
+    },
+    {
+      "epoch": 0.46174941802991165,
+      "grad_norm": 0.2047877013683319,
+      "learning_rate": 0.001,
+      "loss": 1.1047,
+      "step": 14480
+    },
+    {
+      "epoch": 0.4620683057495456,
+      "grad_norm": 0.2022201269865036,
+      "learning_rate": 0.001,
+      "loss": 1.1189,
+      "step": 14490
+    },
+    {
+      "epoch": 0.4623871934691795,
+      "grad_norm": 0.2064102441072464,
+      "learning_rate": 0.001,
+      "loss": 1.119,
+      "step": 14500
+    },
+    {
+      "epoch": 0.4627060811888134,
+      "grad_norm": 0.20427323877811432,
+      "learning_rate": 0.001,
+      "loss": 1.1152,
+      "step": 14510
+    },
+    {
+      "epoch": 0.46302496890844735,
+      "grad_norm": 0.20223671197891235,
+      "learning_rate": 0.001,
+      "loss": 1.1162,
+      "step": 14520
+    },
+    {
+      "epoch": 0.46334385662808125,
+      "grad_norm": 0.21089716255664825,
+      "learning_rate": 0.001,
+      "loss": 1.1452,
+      "step": 14530
+    },
+    {
+      "epoch": 0.46366274434771515,
+      "grad_norm": 0.19083629548549652,
+      "learning_rate": 0.001,
+      "loss": 1.1092,
+      "step": 14540
+    },
+    {
+      "epoch": 0.4639816320673491,
+      "grad_norm": 0.20347940921783447,
+      "learning_rate": 0.001,
+      "loss": 1.1336,
+      "step": 14550
+    },
+    {
+      "epoch": 0.464300519786983,
+      "grad_norm": 0.19843998551368713,
+      "learning_rate": 0.001,
+      "loss": 1.1206,
+      "step": 14560
+    },
+    {
+      "epoch": 0.4646194075066169,
+      "grad_norm": 0.2033349722623825,
+      "learning_rate": 0.001,
+      "loss": 1.1205,
+      "step": 14570
+    },
+    {
+      "epoch": 0.46493829522625085,
+      "grad_norm": 0.20412671566009521,
+      "learning_rate": 0.001,
+      "loss": 1.1181,
+      "step": 14580
+    },
+    {
+      "epoch": 0.46525718294588475,
+      "grad_norm": 0.19806016981601715,
+      "learning_rate": 0.001,
+      "loss": 1.1177,
+      "step": 14590
+    },
+    {
+      "epoch": 0.46557607066551865,
+      "grad_norm": 0.2021852433681488,
+      "learning_rate": 0.001,
+      "loss": 1.1228,
+      "step": 14600
+    },
+    {
+      "epoch": 0.4658949583851526,
+      "grad_norm": 0.20214885473251343,
+      "learning_rate": 0.001,
+      "loss": 1.1265,
+      "step": 14610
+    },
+    {
+      "epoch": 0.4662138461047865,
+      "grad_norm": 0.1979108452796936,
+      "learning_rate": 0.001,
+      "loss": 1.1219,
+      "step": 14620
+    },
+    {
+      "epoch": 0.4665327338244204,
+      "grad_norm": 0.2091963291168213,
+      "learning_rate": 0.001,
+      "loss": 1.1226,
+      "step": 14630
+    },
+    {
+      "epoch": 0.46685162154405435,
+      "grad_norm": 0.19643768668174744,
+      "learning_rate": 0.001,
+      "loss": 1.0983,
+      "step": 14640
+    },
+    {
+      "epoch": 0.46717050926368825,
+      "grad_norm": 0.20225919783115387,
+      "learning_rate": 0.001,
+      "loss": 1.1236,
+      "step": 14650
+    },
+    {
+      "epoch": 0.46748939698332215,
+      "grad_norm": 0.20334993302822113,
+      "learning_rate": 0.001,
+      "loss": 1.1202,
+      "step": 14660
+    },
+    {
+      "epoch": 0.4678082847029561,
+      "grad_norm": 0.20838162302970886,
+      "learning_rate": 0.001,
+      "loss": 1.1189,
+      "step": 14670
+    },
+    {
+      "epoch": 0.46812717242259,
+      "grad_norm": 0.19877314567565918,
+      "learning_rate": 0.001,
+      "loss": 1.1088,
+      "step": 14680
+    },
+    {
+      "epoch": 0.4684460601422239,
+      "grad_norm": 0.19828420877456665,
+      "learning_rate": 0.001,
+      "loss": 1.1117,
+      "step": 14690
+    },
+    {
+      "epoch": 0.46876494786185785,
+      "grad_norm": 0.20643897354602814,
+      "learning_rate": 0.001,
+      "loss": 1.1153,
+      "step": 14700
+    },
+    {
+      "epoch": 0.46908383558149175,
+      "grad_norm": 0.20390458405017853,
+      "learning_rate": 0.001,
+      "loss": 1.1338,
+      "step": 14710
+    },
+    {
+      "epoch": 0.46940272330112565,
+      "grad_norm": 0.20136936008930206,
+      "learning_rate": 0.001,
+      "loss": 1.1234,
+      "step": 14720
+    },
+    {
+      "epoch": 0.4697216110207596,
+      "grad_norm": 0.2019462287425995,
+      "learning_rate": 0.001,
+      "loss": 1.1193,
+      "step": 14730
+    },
+    {
+      "epoch": 0.4700404987403935,
+      "grad_norm": 0.20261630415916443,
+      "learning_rate": 0.001,
+      "loss": 1.1325,
+      "step": 14740
+    },
+    {
+      "epoch": 0.47035938646002745,
+      "grad_norm": 0.2060079276561737,
+      "learning_rate": 0.001,
+      "loss": 1.1314,
+      "step": 14750
+    },
+    {
+      "epoch": 0.47067827417966135,
+      "grad_norm": 0.19139251112937927,
+      "learning_rate": 0.001,
+      "loss": 1.1073,
+      "step": 14760
+    },
+    {
+      "epoch": 0.47099716189929525,
+      "grad_norm": 0.1992751955986023,
+      "learning_rate": 0.001,
+      "loss": 1.0953,
+      "step": 14770
+    },
+    {
+      "epoch": 0.4713160496189292,
+      "grad_norm": 0.19770146906375885,
+      "learning_rate": 0.001,
+      "loss": 1.1258,
+      "step": 14780
+    },
+    {
+      "epoch": 0.4716349373385631,
+      "grad_norm": 0.2079428881406784,
+      "learning_rate": 0.001,
+      "loss": 1.1327,
+      "step": 14790
+    },
+    {
+      "epoch": 0.471953825058197,
+      "grad_norm": 0.20283734798431396,
+      "learning_rate": 0.001,
+      "loss": 1.1325,
+      "step": 14800
+    },
+    {
+      "epoch": 0.47227271277783095,
+      "grad_norm": 0.19868163764476776,
+      "learning_rate": 0.001,
+      "loss": 1.1156,
+      "step": 14810
+    },
+    {
+      "epoch": 0.47259160049746485,
+      "grad_norm": 0.20224575698375702,
+      "learning_rate": 0.001,
+      "loss": 1.1284,
+      "step": 14820
+    },
+    {
+      "epoch": 0.47291048821709875,
+      "grad_norm": 0.19769024848937988,
+      "learning_rate": 0.001,
+      "loss": 1.1313,
+      "step": 14830
+    },
+    {
+      "epoch": 0.4732293759367327,
+      "grad_norm": 0.19698044657707214,
+      "learning_rate": 0.001,
+      "loss": 1.1305,
+      "step": 14840
+    },
+    {
+      "epoch": 0.4735482636563666,
+      "grad_norm": 0.19292454421520233,
+      "learning_rate": 0.001,
+      "loss": 1.1227,
+      "step": 14850
+    },
+    {
+      "epoch": 0.4738671513760005,
+      "grad_norm": 0.20179186761379242,
+      "learning_rate": 0.001,
+      "loss": 1.11,
+      "step": 14860
+    },
+    {
+      "epoch": 0.47418603909563445,
+      "grad_norm": 0.19742195308208466,
+      "learning_rate": 0.001,
+      "loss": 1.1086,
+      "step": 14870
+    },
+    {
+      "epoch": 0.47450492681526835,
+      "grad_norm": 0.2044827789068222,
+      "learning_rate": 0.001,
+      "loss": 1.1171,
+      "step": 14880
+    },
+    {
+      "epoch": 0.47482381453490224,
+      "grad_norm": 0.20579135417938232,
+      "learning_rate": 0.001,
+      "loss": 1.1231,
+      "step": 14890
+    },
+    {
+      "epoch": 0.4751427022545362,
+      "grad_norm": 0.20538045465946198,
+      "learning_rate": 0.001,
+      "loss": 1.1226,
+      "step": 14900
+    },
+    {
+      "epoch": 0.4754615899741701,
+      "grad_norm": 0.2013697773218155,
+      "learning_rate": 0.001,
+      "loss": 1.1342,
+      "step": 14910
+    },
+    {
+      "epoch": 0.475780477693804,
+      "grad_norm": 0.20036698877811432,
+      "learning_rate": 0.001,
+      "loss": 1.1222,
+      "step": 14920
+    },
+    {
+      "epoch": 0.47609936541343795,
+      "grad_norm": 0.20143401622772217,
+      "learning_rate": 0.001,
+      "loss": 1.1294,
+      "step": 14930
+    },
+    {
+      "epoch": 0.47641825313307185,
+      "grad_norm": 0.1989293098449707,
+      "learning_rate": 0.001,
+      "loss": 1.1357,
+      "step": 14940
+    },
+    {
+      "epoch": 0.47673714085270574,
+      "grad_norm": 0.19735567271709442,
+      "learning_rate": 0.001,
+      "loss": 1.1111,
+      "step": 14950
+    },
+    {
+      "epoch": 0.4770560285723397,
+      "grad_norm": 0.19401957094669342,
+      "learning_rate": 0.001,
+      "loss": 1.1239,
+      "step": 14960
+    },
+    {
+      "epoch": 0.4773749162919736,
+      "grad_norm": 0.2115725427865982,
+      "learning_rate": 0.001,
+      "loss": 1.1239,
+      "step": 14970
+    },
+    {
+      "epoch": 0.4776938040116075,
+      "grad_norm": 0.20321685075759888,
+      "learning_rate": 0.001,
+      "loss": 1.112,
+      "step": 14980
+    },
+    {
+      "epoch": 0.47801269173124145,
+      "grad_norm": 0.19996105134487152,
+      "learning_rate": 0.001,
+      "loss": 1.1056,
+      "step": 14990
+    },
+    {
+      "epoch": 0.47833157945087534,
+      "grad_norm": 0.20889067649841309,
+      "learning_rate": 0.001,
+      "loss": 1.1347,
+      "step": 15000
+    },
+    {
+      "epoch": 0.47865046717050924,
+      "grad_norm": 0.20725928246974945,
+      "learning_rate": 0.001,
+      "loss": 1.1317,
+      "step": 15010
+    },
+    {
+      "epoch": 0.4789693548901432,
+      "grad_norm": 0.21042446792125702,
+      "learning_rate": 0.001,
+      "loss": 1.1133,
+      "step": 15020
+    },
+    {
+      "epoch": 0.4792882426097771,
+      "grad_norm": 0.2007140964269638,
+      "learning_rate": 0.001,
+      "loss": 1.1308,
+      "step": 15030
+    },
+    {
+      "epoch": 0.479607130329411,
+      "grad_norm": 0.19597560167312622,
+      "learning_rate": 0.001,
+      "loss": 1.1055,
+      "step": 15040
+    },
+    {
+      "epoch": 0.47992601804904494,
+      "grad_norm": 0.20281028747558594,
+      "learning_rate": 0.001,
+      "loss": 1.1226,
+      "step": 15050
+    },
+    {
+      "epoch": 0.48024490576867884,
+      "grad_norm": 0.1976708173751831,
+      "learning_rate": 0.001,
+      "loss": 1.1159,
+      "step": 15060
+    },
+    {
+      "epoch": 0.48056379348831274,
+      "grad_norm": 0.19908636808395386,
+      "learning_rate": 0.001,
+      "loss": 1.1279,
+      "step": 15070
+    },
+    {
+      "epoch": 0.4808826812079467,
+      "grad_norm": 0.20221206545829773,
+      "learning_rate": 0.001,
+      "loss": 1.123,
+      "step": 15080
+    },
+    {
+      "epoch": 0.4812015689275806,
+      "grad_norm": 0.2094741016626358,
+      "learning_rate": 0.001,
+      "loss": 1.115,
+      "step": 15090
+    },
+    {
+      "epoch": 0.4815204566472145,
+      "grad_norm": 0.20078401267528534,
+      "learning_rate": 0.001,
+      "loss": 1.1298,
+      "step": 15100
+    },
+    {
+      "epoch": 0.48183934436684844,
+      "grad_norm": 0.20339038968086243,
+      "learning_rate": 0.001,
+      "loss": 1.1008,
+      "step": 15110
+    },
+    {
+      "epoch": 0.48215823208648234,
+      "grad_norm": 0.20455828309059143,
+      "learning_rate": 0.001,
+      "loss": 1.0992,
+      "step": 15120
+    },
+    {
+      "epoch": 0.48247711980611624,
+      "grad_norm": 0.2017097920179367,
+      "learning_rate": 0.001,
+      "loss": 1.1076,
+      "step": 15130
+    },
+    {
+      "epoch": 0.4827960075257502,
+      "grad_norm": 0.19957958161830902,
+      "learning_rate": 0.001,
+      "loss": 1.1125,
+      "step": 15140
+    },
+    {
+      "epoch": 0.4831148952453841,
+      "grad_norm": 0.19430184364318848,
+      "learning_rate": 0.001,
+      "loss": 1.1012,
+      "step": 15150
+    },
+    {
+      "epoch": 0.48343378296501804,
+      "grad_norm": 0.196171373128891,
+      "learning_rate": 0.001,
+      "loss": 1.1116,
+      "step": 15160
+    },
+    {
+      "epoch": 0.48375267068465194,
+      "grad_norm": 0.2169899195432663,
+      "learning_rate": 0.001,
+      "loss": 1.1154,
+      "step": 15170
+    },
+    {
+      "epoch": 0.48407155840428584,
+      "grad_norm": 0.20589181780815125,
+      "learning_rate": 0.001,
+      "loss": 1.1101,
+      "step": 15180
+    },
+    {
+      "epoch": 0.4843904461239198,
+      "grad_norm": 0.20493051409721375,
+      "learning_rate": 0.001,
+      "loss": 1.103,
+      "step": 15190
+    },
+    {
+      "epoch": 0.4847093338435537,
+      "grad_norm": 0.2098941057920456,
+      "learning_rate": 0.001,
+      "loss": 1.1414,
+      "step": 15200
+    },
+    {
+      "epoch": 0.4850282215631876,
+      "grad_norm": 0.1904533952474594,
+      "learning_rate": 0.001,
+      "loss": 1.1027,
+      "step": 15210
+    },
+    {
+      "epoch": 0.48534710928282154,
+      "grad_norm": 0.1936672180891037,
+      "learning_rate": 0.001,
+      "loss": 1.1141,
+      "step": 15220
+    },
+    {
+      "epoch": 0.48566599700245544,
+      "grad_norm": 0.20026364922523499,
+      "learning_rate": 0.001,
+      "loss": 1.1243,
+      "step": 15230
+    },
+    {
+      "epoch": 0.48598488472208934,
+      "grad_norm": 0.20649927854537964,
+      "learning_rate": 0.001,
+      "loss": 1.1259,
+      "step": 15240
+    },
+    {
+      "epoch": 0.4863037724417233,
+      "grad_norm": 0.19340914487838745,
+      "learning_rate": 0.001,
+      "loss": 1.1079,
+      "step": 15250
+    },
+    {
+      "epoch": 0.4866226601613572,
+      "grad_norm": 0.19627441465854645,
+      "learning_rate": 0.001,
+      "loss": 1.1149,
+      "step": 15260
+    },
+    {
+      "epoch": 0.4869415478809911,
+      "grad_norm": 0.19537758827209473,
+      "learning_rate": 0.001,
+      "loss": 1.1113,
+      "step": 15270
+    },
+    {
+      "epoch": 0.48726043560062504,
+      "grad_norm": 0.19534988701343536,
+      "learning_rate": 0.001,
+      "loss": 1.0957,
+      "step": 15280
+    },
+    {
+      "epoch": 0.48757932332025894,
+      "grad_norm": 0.19658777117729187,
+      "learning_rate": 0.001,
+      "loss": 1.1178,
+      "step": 15290
+    },
+    {
+      "epoch": 0.48789821103989284,
+      "grad_norm": 0.19526197016239166,
+      "learning_rate": 0.001,
+      "loss": 1.1025,
+      "step": 15300
+    },
+    {
+      "epoch": 0.4882170987595268,
+      "grad_norm": 0.2003452330827713,
+      "learning_rate": 0.001,
+      "loss": 1.0961,
+      "step": 15310
+    },
+    {
+      "epoch": 0.4885359864791607,
+      "grad_norm": 0.19825130701065063,
+      "learning_rate": 0.001,
+      "loss": 1.1145,
+      "step": 15320
+    },
+    {
+      "epoch": 0.4888548741987946,
+      "grad_norm": 0.19999739527702332,
+      "learning_rate": 0.001,
+      "loss": 1.1157,
+      "step": 15330
+    },
+    {
+      "epoch": 0.48917376191842854,
+      "grad_norm": 0.2152135968208313,
+      "learning_rate": 0.001,
+      "loss": 1.1051,
+      "step": 15340
+    },
+    {
+      "epoch": 0.48949264963806244,
+      "grad_norm": 0.21462447941303253,
+      "learning_rate": 0.001,
+      "loss": 1.1062,
+      "step": 15350
+    },
+    {
+      "epoch": 0.48981153735769634,
+      "grad_norm": 0.1989239752292633,
+      "learning_rate": 0.001,
+      "loss": 1.0992,
+      "step": 15360
+    },
+    {
+      "epoch": 0.4901304250773303,
+      "grad_norm": 0.19589900970458984,
+      "learning_rate": 0.001,
+      "loss": 1.1241,
+      "step": 15370
+    },
+    {
+      "epoch": 0.4904493127969642,
+      "grad_norm": 0.2006465196609497,
+      "learning_rate": 0.001,
+      "loss": 1.1142,
+      "step": 15380
+    },
+    {
+      "epoch": 0.4907682005165981,
+      "grad_norm": 0.19966904819011688,
+      "learning_rate": 0.001,
+      "loss": 1.1233,
+      "step": 15390
+    },
+    {
+      "epoch": 0.49108708823623204,
+      "grad_norm": 0.20915593206882477,
+      "learning_rate": 0.001,
+      "loss": 1.1042,
+      "step": 15400
+    },
+    {
+      "epoch": 0.49140597595586594,
+      "grad_norm": 0.19926536083221436,
+      "learning_rate": 0.001,
+      "loss": 1.1226,
+      "step": 15410
+    },
+    {
+      "epoch": 0.49172486367549983,
+      "grad_norm": 0.20816414058208466,
+      "learning_rate": 0.001,
+      "loss": 1.121,
+      "step": 15420
+    },
+    {
+      "epoch": 0.4920437513951338,
+      "grad_norm": 0.19813525676727295,
+      "learning_rate": 0.001,
+      "loss": 1.1168,
+      "step": 15430
+    },
+    {
+      "epoch": 0.4923626391147677,
+      "grad_norm": 0.19787755608558655,
+      "learning_rate": 0.001,
+      "loss": 1.1333,
+      "step": 15440
+    },
+    {
+      "epoch": 0.4926815268344016,
+      "grad_norm": 0.20607894659042358,
+      "learning_rate": 0.001,
+      "loss": 1.1423,
+      "step": 15450
+    },
+    {
+      "epoch": 0.49300041455403554,
+      "grad_norm": 0.19217760860919952,
+      "learning_rate": 0.001,
+      "loss": 1.1317,
+      "step": 15460
+    },
+    {
+      "epoch": 0.49331930227366944,
+      "grad_norm": 0.215703085064888,
+      "learning_rate": 0.001,
+      "loss": 1.1366,
+      "step": 15470
+    },
+    {
+      "epoch": 0.49363818999330333,
+      "grad_norm": 0.20476995408535004,
+      "learning_rate": 0.001,
+      "loss": 1.1027,
+      "step": 15480
+    },
+    {
+      "epoch": 0.4939570777129373,
+      "grad_norm": 0.2130293995141983,
+      "learning_rate": 0.001,
+      "loss": 1.1285,
+      "step": 15490
+    },
+    {
+      "epoch": 0.4942759654325712,
+      "grad_norm": 0.1988312005996704,
+      "learning_rate": 0.001,
+      "loss": 1.108,
+      "step": 15500
+    },
+    {
+      "epoch": 0.4945948531522051,
+      "grad_norm": 0.21203584969043732,
+      "learning_rate": 0.001,
+      "loss": 1.1159,
+      "step": 15510
+    },
+    {
+      "epoch": 0.49491374087183904,
+      "grad_norm": 0.194093719124794,
+      "learning_rate": 0.001,
+      "loss": 1.127,
+      "step": 15520
+    },
+    {
+      "epoch": 0.49523262859147293,
+      "grad_norm": 0.19752824306488037,
+      "learning_rate": 0.001,
+      "loss": 1.1186,
+      "step": 15530
+    },
+    {
+      "epoch": 0.49555151631110683,
+      "grad_norm": 0.19278797507286072,
+      "learning_rate": 0.001,
+      "loss": 1.1151,
+      "step": 15540
+    },
+    {
+      "epoch": 0.4958704040307408,
+      "grad_norm": 0.20309525728225708,
+      "learning_rate": 0.001,
+      "loss": 1.1221,
+      "step": 15550
+    },
+    {
+      "epoch": 0.4961892917503747,
+      "grad_norm": 0.19723108410835266,
+      "learning_rate": 0.001,
+      "loss": 1.1124,
+      "step": 15560
+    },
+    {
+      "epoch": 0.49650817947000864,
+      "grad_norm": 0.19733686745166779,
+      "learning_rate": 0.001,
+      "loss": 1.1239,
+      "step": 15570
+    },
+    {
+      "epoch": 0.49682706718964254,
+      "grad_norm": 0.19719557464122772,
+      "learning_rate": 0.001,
+      "loss": 1.1165,
+      "step": 15580
+    },
+    {
+      "epoch": 0.49714595490927643,
+      "grad_norm": 0.19925282895565033,
+      "learning_rate": 0.001,
+      "loss": 1.1261,
+      "step": 15590
+    },
+    {
+      "epoch": 0.4974648426289104,
+      "grad_norm": 0.1964748650789261,
+      "learning_rate": 0.001,
+      "loss": 1.1078,
+      "step": 15600
+    },
+    {
+      "epoch": 0.4977837303485443,
+      "grad_norm": 0.20201486349105835,
+      "learning_rate": 0.001,
+      "loss": 1.108,
+      "step": 15610
+    },
+    {
+      "epoch": 0.4981026180681782,
+      "grad_norm": 0.19340449571609497,
+      "learning_rate": 0.001,
+      "loss": 1.1143,
+      "step": 15620
+    },
+    {
+      "epoch": 0.49842150578781214,
+      "grad_norm": 0.1944858431816101,
+      "learning_rate": 0.001,
+      "loss": 1.1001,
+      "step": 15630
+    },
+    {
+      "epoch": 0.49874039350744603,
+      "grad_norm": 0.20309315621852875,
+      "learning_rate": 0.001,
+      "loss": 1.1148,
+      "step": 15640
+    },
+    {
+      "epoch": 0.49905928122707993,
+      "grad_norm": 0.20255059003829956,
+      "learning_rate": 0.001,
+      "loss": 1.1152,
+      "step": 15650
+    },
+    {
+      "epoch": 0.4993781689467139,
+      "grad_norm": 0.2079618126153946,
+      "learning_rate": 0.001,
+      "loss": 1.1346,
+      "step": 15660
+    },
+    {
+      "epoch": 0.4996970566663478,
+      "grad_norm": 0.20329873263835907,
+      "learning_rate": 0.001,
+      "loss": 1.1143,
+      "step": 15670
+    },
+    {
+      "epoch": 0.5000159443859817,
+      "grad_norm": 0.2044350504875183,
+      "learning_rate": 0.001,
+      "loss": 1.1169,
+      "step": 15680
+    },
+    {
+      "epoch": 0.5003348321056156,
+      "grad_norm": 0.19735576212406158,
+      "learning_rate": 0.001,
+      "loss": 1.1092,
+      "step": 15690
+    },
+    {
+      "epoch": 0.5006537198252495,
+      "grad_norm": 0.19718340039253235,
+      "learning_rate": 0.001,
+      "loss": 1.1114,
+      "step": 15700
+    },
+    {
+      "epoch": 0.5009726075448835,
+      "grad_norm": 0.20076552033424377,
+      "learning_rate": 0.001,
+      "loss": 1.1162,
+      "step": 15710
+    },
+    {
+      "epoch": 0.5012914952645173,
+      "grad_norm": 0.19438549876213074,
+      "learning_rate": 0.001,
+      "loss": 1.1284,
+      "step": 15720
+    },
+    {
+      "epoch": 0.5016103829841513,
+      "grad_norm": 0.2000940889120102,
+      "learning_rate": 0.001,
+      "loss": 1.105,
+      "step": 15730
+    },
+    {
+      "epoch": 0.5019292707037852,
+      "grad_norm": 0.19995105266571045,
+      "learning_rate": 0.001,
+      "loss": 1.1332,
+      "step": 15740
+    },
+    {
+      "epoch": 0.5022481584234191,
+      "grad_norm": 0.1938658207654953,
+      "learning_rate": 0.001,
+      "loss": 1.1162,
+      "step": 15750
+    },
+    {
+      "epoch": 0.502567046143053,
+      "grad_norm": 0.20208021998405457,
+      "learning_rate": 0.001,
+      "loss": 1.0963,
+      "step": 15760
+    },
+    {
+      "epoch": 0.502885933862687,
+      "grad_norm": 0.19796696305274963,
+      "learning_rate": 0.001,
+      "loss": 1.1174,
+      "step": 15770
+    },
+    {
+      "epoch": 0.5032048215823208,
+      "grad_norm": 0.20823107659816742,
+      "learning_rate": 0.001,
+      "loss": 1.115,
+      "step": 15780
+    },
+    {
+      "epoch": 0.5035237093019548,
+      "grad_norm": 0.19293028116226196,
+      "learning_rate": 0.001,
+      "loss": 1.107,
+      "step": 15790
+    },
+    {
+      "epoch": 0.5038425970215887,
+      "grad_norm": 0.19398380815982819,
+      "learning_rate": 0.001,
+      "loss": 1.1052,
+      "step": 15800
+    },
+    {
+      "epoch": 0.5041614847412226,
+      "grad_norm": 0.19862490892410278,
+      "learning_rate": 0.001,
+      "loss": 1.1215,
+      "step": 15810
+    },
+    {
+      "epoch": 0.5044803724608565,
+      "grad_norm": 0.1949176788330078,
+      "learning_rate": 0.001,
+      "loss": 1.1116,
+      "step": 15820
+    },
+    {
+      "epoch": 0.5047992601804905,
+      "grad_norm": 0.20239092409610748,
+      "learning_rate": 0.001,
+      "loss": 1.1019,
+      "step": 15830
+    },
+    {
+      "epoch": 0.5051181479001243,
+      "grad_norm": 0.19988751411437988,
+      "learning_rate": 0.001,
+      "loss": 1.1158,
+      "step": 15840
+    },
+    {
+      "epoch": 0.5054370356197583,
+      "grad_norm": 0.2054993063211441,
+      "learning_rate": 0.001,
+      "loss": 1.1217,
+      "step": 15850
+    },
+    {
+      "epoch": 0.5057559233393922,
+      "grad_norm": 0.20914222300052643,
+      "learning_rate": 0.001,
+      "loss": 1.1258,
+      "step": 15860
+    },
+    {
+      "epoch": 0.5060748110590261,
+      "grad_norm": 0.19964182376861572,
+      "learning_rate": 0.001,
+      "loss": 1.106,
+      "step": 15870
+    },
+    {
+      "epoch": 0.50639369877866,
+      "grad_norm": 0.20104973018169403,
+      "learning_rate": 0.001,
+      "loss": 1.1128,
+      "step": 15880
+    },
+    {
+      "epoch": 0.506712586498294,
+      "grad_norm": 0.20262691378593445,
+      "learning_rate": 0.001,
+      "loss": 1.0998,
+      "step": 15890
+    },
+    {
+      "epoch": 0.5070314742179278,
+      "grad_norm": 0.20666274428367615,
+      "learning_rate": 0.001,
+      "loss": 1.1046,
+      "step": 15900
+    },
+    {
+      "epoch": 0.5073503619375618,
+      "grad_norm": 0.19419854879379272,
+      "learning_rate": 0.001,
+      "loss": 1.1065,
+      "step": 15910
+    },
+    {
+      "epoch": 0.5076692496571957,
+      "grad_norm": 0.20134715735912323,
+      "learning_rate": 0.001,
+      "loss": 1.1036,
+      "step": 15920
+    },
+    {
+      "epoch": 0.5079881373768296,
+      "grad_norm": 0.20258976519107819,
+      "learning_rate": 0.001,
+      "loss": 1.1118,
+      "step": 15930
+    },
+    {
+      "epoch": 0.5083070250964635,
+      "grad_norm": 0.19312818348407745,
+      "learning_rate": 0.001,
+      "loss": 1.1039,
+      "step": 15940
+    },
+    {
+      "epoch": 0.5086259128160975,
+      "grad_norm": 0.20441237092018127,
+      "learning_rate": 0.001,
+      "loss": 1.106,
+      "step": 15950
+    },
+    {
+      "epoch": 0.5089448005357313,
+      "grad_norm": 0.19846174120903015,
+      "learning_rate": 0.001,
+      "loss": 1.1146,
+      "step": 15960
+    },
+    {
+      "epoch": 0.5092636882553653,
+      "grad_norm": 0.2000012993812561,
+      "learning_rate": 0.001,
+      "loss": 1.1157,
+      "step": 15970
+    },
+    {
+      "epoch": 0.5095825759749992,
+      "grad_norm": 0.19293838739395142,
+      "learning_rate": 0.001,
+      "loss": 1.1043,
+      "step": 15980
+    },
+    {
+      "epoch": 0.5099014636946331,
+      "grad_norm": 0.1996237188577652,
+      "learning_rate": 0.001,
+      "loss": 1.0916,
+      "step": 15990
+    },
+    {
+      "epoch": 0.510220351414267,
+      "grad_norm": 0.20133572816848755,
+      "learning_rate": 0.001,
+      "loss": 1.101,
+      "step": 16000
+    },
+    {
+      "epoch": 0.510539239133901,
+      "grad_norm": 0.20171783864498138,
+      "learning_rate": 0.001,
+      "loss": 1.1077,
+      "step": 16010
+    },
+    {
+      "epoch": 0.5108581268535348,
+      "grad_norm": 0.20236876606941223,
+      "learning_rate": 0.001,
+      "loss": 1.1173,
+      "step": 16020
+    },
+    {
+      "epoch": 0.5111770145731688,
+      "grad_norm": 0.19765743613243103,
+      "learning_rate": 0.001,
+      "loss": 1.0923,
+      "step": 16030
+    },
+    {
+      "epoch": 0.5114959022928027,
+      "grad_norm": 0.2046414166688919,
+      "learning_rate": 0.001,
+      "loss": 1.1292,
+      "step": 16040
+    },
+    {
+      "epoch": 0.5118147900124366,
+      "grad_norm": 0.20035295188426971,
+      "learning_rate": 0.001,
+      "loss": 1.1027,
+      "step": 16050
+    },
+    {
+      "epoch": 0.5121336777320705,
+      "grad_norm": 0.20384572446346283,
+      "learning_rate": 0.001,
+      "loss": 1.1259,
+      "step": 16060
+    },
+    {
+      "epoch": 0.5124525654517045,
+      "grad_norm": 0.19551537930965424,
+      "learning_rate": 0.001,
+      "loss": 1.081,
+      "step": 16070
+    },
+    {
+      "epoch": 0.5127714531713383,
+      "grad_norm": 0.19394861161708832,
+      "learning_rate": 0.001,
+      "loss": 1.1006,
+      "step": 16080
+    },
+    {
+      "epoch": 0.5130903408909723,
+      "grad_norm": 0.19792471826076508,
+      "learning_rate": 0.001,
+      "loss": 1.1112,
+      "step": 16090
+    },
+    {
+      "epoch": 0.5134092286106062,
+      "grad_norm": 0.19944746792316437,
+      "learning_rate": 0.001,
+      "loss": 1.1147,
+      "step": 16100
+    },
+    {
+      "epoch": 0.5137281163302401,
+      "grad_norm": 0.20174691081047058,
+      "learning_rate": 0.001,
+      "loss": 1.1072,
+      "step": 16110
+    },
+    {
+      "epoch": 0.514047004049874,
+      "grad_norm": 0.1961384117603302,
+      "learning_rate": 0.001,
+      "loss": 1.1048,
+      "step": 16120
+    },
+    {
+      "epoch": 0.514365891769508,
+      "grad_norm": 0.19337189197540283,
+      "learning_rate": 0.001,
+      "loss": 1.1041,
+      "step": 16130
+    },
+    {
+      "epoch": 0.5146847794891418,
+      "grad_norm": 0.1975313425064087,
+      "learning_rate": 0.001,
+      "loss": 1.1047,
+      "step": 16140
+    },
+    {
+      "epoch": 0.5150036672087758,
+      "grad_norm": 0.20119927823543549,
+      "learning_rate": 0.001,
+      "loss": 1.1132,
+      "step": 16150
+    },
+    {
+      "epoch": 0.5153225549284097,
+      "grad_norm": 0.19607770442962646,
+      "learning_rate": 0.001,
+      "loss": 1.1088,
+      "step": 16160
+    },
+    {
+      "epoch": 0.5156414426480436,
+      "grad_norm": 0.2076045274734497,
+      "learning_rate": 0.001,
+      "loss": 1.0965,
+      "step": 16170
+    },
+    {
+      "epoch": 0.5159603303676775,
+      "grad_norm": 0.20277442038059235,
+      "learning_rate": 0.001,
+      "loss": 1.0985,
+      "step": 16180
+    },
+    {
+      "epoch": 0.5162792180873115,
+      "grad_norm": 0.19827525317668915,
+      "learning_rate": 0.001,
+      "loss": 1.1024,
+      "step": 16190
+    },
+    {
+      "epoch": 0.5165981058069454,
+      "grad_norm": 0.19448797404766083,
+      "learning_rate": 0.001,
+      "loss": 1.1168,
+      "step": 16200
+    },
+    {
+      "epoch": 0.5169169935265793,
+      "grad_norm": 0.20615480840206146,
+      "learning_rate": 0.001,
+      "loss": 1.1078,
+      "step": 16210
+    },
+    {
+      "epoch": 0.5172358812462132,
+      "grad_norm": 0.19415822625160217,
+      "learning_rate": 0.001,
+      "loss": 1.1123,
+      "step": 16220
+    },
+    {
+      "epoch": 0.5175547689658472,
+      "grad_norm": 0.19522392749786377,
+      "learning_rate": 0.001,
+      "loss": 1.1041,
+      "step": 16230
+    },
+    {
+      "epoch": 0.517873656685481,
+      "grad_norm": 0.199362650513649,
+      "learning_rate": 0.001,
+      "loss": 1.1114,
+      "step": 16240
+    },
+    {
+      "epoch": 0.518192544405115,
+      "grad_norm": 0.19673673808574677,
+      "learning_rate": 0.001,
+      "loss": 1.1173,
+      "step": 16250
+    },
+    {
+      "epoch": 0.5185114321247489,
+      "grad_norm": 0.19982314109802246,
+      "learning_rate": 0.001,
+      "loss": 1.1089,
+      "step": 16260
+    },
+    {
+      "epoch": 0.5188303198443828,
+      "grad_norm": 0.19413277506828308,
+      "learning_rate": 0.001,
+      "loss": 1.0947,
+      "step": 16270
+    },
+    {
+      "epoch": 0.5191492075640167,
+      "grad_norm": 0.19558317959308624,
+      "learning_rate": 0.001,
+      "loss": 1.0951,
+      "step": 16280
+    },
+    {
+      "epoch": 0.5194680952836507,
+      "grad_norm": 0.20333553850650787,
+      "learning_rate": 0.001,
+      "loss": 1.1017,
+      "step": 16290
+    },
+    {
+      "epoch": 0.5197869830032845,
+      "grad_norm": 0.1913626492023468,
+      "learning_rate": 0.001,
+      "loss": 1.1263,
+      "step": 16300
+    },
+    {
+      "epoch": 0.5201058707229185,
+      "grad_norm": 0.19983579218387604,
+      "learning_rate": 0.001,
+      "loss": 1.1234,
+      "step": 16310
+    },
+    {
+      "epoch": 0.5204247584425524,
+      "grad_norm": 0.1915716677904129,
+      "learning_rate": 0.001,
+      "loss": 1.102,
+      "step": 16320
+    },
+    {
+      "epoch": 0.5207436461621863,
+      "grad_norm": 0.2049056887626648,
+      "learning_rate": 0.001,
+      "loss": 1.107,
+      "step": 16330
+    },
+    {
+      "epoch": 0.5210625338818202,
+      "grad_norm": 0.20463578402996063,
+      "learning_rate": 0.001,
+      "loss": 1.1209,
+      "step": 16340
+    },
+    {
+      "epoch": 0.5213814216014542,
+      "grad_norm": 0.20603425800800323,
+      "learning_rate": 0.001,
+      "loss": 1.0963,
+      "step": 16350
+    },
+    {
+      "epoch": 0.521700309321088,
+      "grad_norm": 0.19194631278514862,
+      "learning_rate": 0.001,
+      "loss": 1.0881,
+      "step": 16360
+    },
+    {
+      "epoch": 0.522019197040722,
+      "grad_norm": 0.20453418791294098,
+      "learning_rate": 0.001,
+      "loss": 1.098,
+      "step": 16370
+    },
+    {
+      "epoch": 0.5223380847603559,
+      "grad_norm": 0.1975255161523819,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 16380
+    },
+    {
+      "epoch": 0.5226569724799898,
+      "grad_norm": 0.20485936105251312,
+      "learning_rate": 0.001,
+      "loss": 1.1387,
+      "step": 16390
+    },
+    {
+      "epoch": 0.5229758601996237,
+      "grad_norm": 0.20453734695911407,
+      "learning_rate": 0.001,
+      "loss": 1.0976,
+      "step": 16400
+    },
+    {
+      "epoch": 0.5232947479192577,
+      "grad_norm": 0.20409847795963287,
+      "learning_rate": 0.001,
+      "loss": 1.0998,
+      "step": 16410
+    },
+    {
+      "epoch": 0.5236136356388915,
+      "grad_norm": 0.19635087251663208,
+      "learning_rate": 0.001,
+      "loss": 1.0953,
+      "step": 16420
+    },
+    {
+      "epoch": 0.5239325233585255,
+      "grad_norm": 0.19614706933498383,
+      "learning_rate": 0.001,
+      "loss": 1.1247,
+      "step": 16430
+    },
+    {
+      "epoch": 0.5242514110781594,
+      "grad_norm": 0.19600021839141846,
+      "learning_rate": 0.001,
+      "loss": 1.115,
+      "step": 16440
+    },
+    {
+      "epoch": 0.5245702987977933,
+      "grad_norm": 0.19973431527614594,
+      "learning_rate": 0.001,
+      "loss": 1.1194,
+      "step": 16450
+    },
+    {
+      "epoch": 0.5248891865174272,
+      "grad_norm": 0.20397818088531494,
+      "learning_rate": 0.001,
+      "loss": 1.1036,
+      "step": 16460
+    },
+    {
+      "epoch": 0.5252080742370612,
+      "grad_norm": 0.19828130304813385,
+      "learning_rate": 0.001,
+      "loss": 1.1153,
+      "step": 16470
+    },
+    {
+      "epoch": 0.525526961956695,
+      "grad_norm": 0.19535043835639954,
+      "learning_rate": 0.001,
+      "loss": 1.1259,
+      "step": 16480
+    },
+    {
+      "epoch": 0.525845849676329,
+      "grad_norm": 0.20396125316619873,
+      "learning_rate": 0.001,
+      "loss": 1.1176,
+      "step": 16490
+    },
+    {
+      "epoch": 0.5261647373959629,
+      "grad_norm": 0.1998380720615387,
+      "learning_rate": 0.001,
+      "loss": 1.0899,
+      "step": 16500
+    },
+    {
+      "epoch": 0.5264836251155968,
+      "grad_norm": 0.1916278749704361,
+      "learning_rate": 0.001,
+      "loss": 1.0945,
+      "step": 16510
+    },
+    {
+      "epoch": 0.5268025128352307,
+      "grad_norm": 0.2070852816104889,
+      "learning_rate": 0.001,
+      "loss": 1.1125,
+      "step": 16520
+    },
+    {
+      "epoch": 0.5271214005548647,
+      "grad_norm": 0.19235506653785706,
+      "learning_rate": 0.001,
+      "loss": 1.0999,
+      "step": 16530
+    },
+    {
+      "epoch": 0.5274402882744985,
+      "grad_norm": 0.2007400244474411,
+      "learning_rate": 0.001,
+      "loss": 1.1357,
+      "step": 16540
+    },
+    {
+      "epoch": 0.5277591759941325,
+      "grad_norm": 0.19115379452705383,
+      "learning_rate": 0.001,
+      "loss": 1.0922,
+      "step": 16550
+    },
+    {
+      "epoch": 0.5280780637137664,
+      "grad_norm": 0.19835355877876282,
+      "learning_rate": 0.001,
+      "loss": 1.1089,
+      "step": 16560
+    },
+    {
+      "epoch": 0.5283969514334003,
+      "grad_norm": 0.19772939383983612,
+      "learning_rate": 0.001,
+      "loss": 1.0973,
+      "step": 16570
+    },
+    {
+      "epoch": 0.5287158391530342,
+      "grad_norm": 0.20019714534282684,
+      "learning_rate": 0.001,
+      "loss": 1.0867,
+      "step": 16580
+    },
+    {
+      "epoch": 0.5290347268726682,
+      "grad_norm": 0.2011161893606186,
+      "learning_rate": 0.001,
+      "loss": 1.096,
+      "step": 16590
+    },
+    {
+      "epoch": 0.529353614592302,
+      "grad_norm": 0.1947372555732727,
+      "learning_rate": 0.001,
+      "loss": 1.1014,
+      "step": 16600
+    },
+    {
+      "epoch": 0.529672502311936,
+      "grad_norm": 0.19452321529388428,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 16610
+    },
+    {
+      "epoch": 0.5299913900315699,
+      "grad_norm": 0.19821277260780334,
+      "learning_rate": 0.001,
+      "loss": 1.1306,
+      "step": 16620
+    },
+    {
+      "epoch": 0.5303102777512038,
+      "grad_norm": 0.1981065422296524,
+      "learning_rate": 0.001,
+      "loss": 1.1072,
+      "step": 16630
+    },
+    {
+      "epoch": 0.5306291654708377,
+      "grad_norm": 0.20271334052085876,
+      "learning_rate": 0.001,
+      "loss": 1.1066,
+      "step": 16640
+    },
+    {
+      "epoch": 0.5309480531904717,
+      "grad_norm": 0.20248804986476898,
+      "learning_rate": 0.001,
+      "loss": 1.1099,
+      "step": 16650
+    },
+    {
+      "epoch": 0.5312669409101055,
+      "grad_norm": 0.1929582953453064,
+      "learning_rate": 0.001,
+      "loss": 1.113,
+      "step": 16660
+    },
+    {
+      "epoch": 0.5315858286297395,
+      "grad_norm": 0.19067426025867462,
+      "learning_rate": 0.001,
+      "loss": 1.0873,
+      "step": 16670
+    },
+    {
+      "epoch": 0.5319047163493734,
+      "grad_norm": 0.2004399299621582,
+      "learning_rate": 0.001,
+      "loss": 1.1371,
+      "step": 16680
+    },
+    {
+      "epoch": 0.5322236040690073,
+      "grad_norm": 0.19235877692699432,
+      "learning_rate": 0.001,
+      "loss": 1.0983,
+      "step": 16690
+    },
+    {
+      "epoch": 0.5325424917886412,
+      "grad_norm": 0.19555066525936127,
+      "learning_rate": 0.001,
+      "loss": 1.0975,
+      "step": 16700
+    },
+    {
+      "epoch": 0.5328613795082752,
+      "grad_norm": 0.19806230068206787,
+      "learning_rate": 0.001,
+      "loss": 1.093,
+      "step": 16710
+    },
+    {
+      "epoch": 0.533180267227909,
+      "grad_norm": 0.20040126144886017,
+      "learning_rate": 0.001,
+      "loss": 1.1206,
+      "step": 16720
+    },
+    {
+      "epoch": 0.533499154947543,
+      "grad_norm": 0.21600055694580078,
+      "learning_rate": 0.001,
+      "loss": 1.11,
+      "step": 16730
+    },
+    {
+      "epoch": 0.5338180426671769,
+      "grad_norm": 0.19331254065036774,
+      "learning_rate": 0.001,
+      "loss": 1.1009,
+      "step": 16740
+    },
+    {
+      "epoch": 0.5341369303868108,
+      "grad_norm": 0.1930786818265915,
+      "learning_rate": 0.001,
+      "loss": 1.0901,
+      "step": 16750
+    },
+    {
+      "epoch": 0.5344558181064447,
+      "grad_norm": 0.19137626886367798,
+      "learning_rate": 0.001,
+      "loss": 1.1151,
+      "step": 16760
+    },
+    {
+      "epoch": 0.5347747058260787,
+      "grad_norm": 0.19468235969543457,
+      "learning_rate": 0.001,
+      "loss": 1.1088,
+      "step": 16770
+    },
+    {
+      "epoch": 0.5350935935457125,
+      "grad_norm": 0.20413386821746826,
+      "learning_rate": 0.001,
+      "loss": 1.1125,
+      "step": 16780
+    },
+    {
+      "epoch": 0.5354124812653465,
+      "grad_norm": 0.20103423297405243,
+      "learning_rate": 0.001,
+      "loss": 1.1109,
+      "step": 16790
+    },
+    {
+      "epoch": 0.5357313689849804,
+      "grad_norm": 0.19415734708309174,
+      "learning_rate": 0.001,
+      "loss": 1.1001,
+      "step": 16800
+    },
+    {
+      "epoch": 0.5360502567046143,
+      "grad_norm": 0.20362050831317902,
+      "learning_rate": 0.001,
+      "loss": 1.1277,
+      "step": 16810
+    },
+    {
+      "epoch": 0.5363691444242482,
+      "grad_norm": 0.1981305480003357,
+      "learning_rate": 0.001,
+      "loss": 1.106,
+      "step": 16820
+    },
+    {
+      "epoch": 0.5366880321438822,
+      "grad_norm": 0.19732020795345306,
+      "learning_rate": 0.001,
+      "loss": 1.1032,
+      "step": 16830
+    },
+    {
+      "epoch": 0.537006919863516,
+      "grad_norm": 0.2011566162109375,
+      "learning_rate": 0.001,
+      "loss": 1.1282,
+      "step": 16840
+    },
+    {
+      "epoch": 0.53732580758315,
+      "grad_norm": 0.19841685891151428,
+      "learning_rate": 0.001,
+      "loss": 1.1094,
+      "step": 16850
+    },
+    {
+      "epoch": 0.5376446953027839,
+      "grad_norm": 0.1937408298254013,
+      "learning_rate": 0.001,
+      "loss": 1.1098,
+      "step": 16860
+    },
+    {
+      "epoch": 0.5379635830224178,
+      "grad_norm": 0.20316290855407715,
+      "learning_rate": 0.001,
+      "loss": 1.1002,
+      "step": 16870
+    },
+    {
+      "epoch": 0.5382824707420517,
+      "grad_norm": 0.1968275010585785,
+      "learning_rate": 0.001,
+      "loss": 1.0835,
+      "step": 16880
+    },
+    {
+      "epoch": 0.5386013584616857,
+      "grad_norm": 0.1885264664888382,
+      "learning_rate": 0.001,
+      "loss": 1.114,
+      "step": 16890
+    },
+    {
+      "epoch": 0.5389202461813195,
+      "grad_norm": 0.19835618138313293,
+      "learning_rate": 0.001,
+      "loss": 1.1116,
+      "step": 16900
+    },
+    {
+      "epoch": 0.5392391339009535,
+      "grad_norm": 0.20179498195648193,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 16910
+    },
+    {
+      "epoch": 0.5395580216205874,
+      "grad_norm": 0.20029352605342865,
+      "learning_rate": 0.001,
+      "loss": 1.1197,
+      "step": 16920
+    },
+    {
+      "epoch": 0.5398769093402213,
+      "grad_norm": 0.19566185772418976,
+      "learning_rate": 0.001,
+      "loss": 1.105,
+      "step": 16930
+    },
+    {
+      "epoch": 0.5401957970598552,
+      "grad_norm": 0.19709070026874542,
+      "learning_rate": 0.001,
+      "loss": 1.1057,
+      "step": 16940
+    },
+    {
+      "epoch": 0.5405146847794892,
+      "grad_norm": 0.20236507058143616,
+      "learning_rate": 0.001,
+      "loss": 1.0964,
+      "step": 16950
+    },
+    {
+      "epoch": 0.540833572499123,
+      "grad_norm": 0.2061866670846939,
+      "learning_rate": 0.001,
+      "loss": 1.104,
+      "step": 16960
+    },
+    {
+      "epoch": 0.541152460218757,
+      "grad_norm": 0.1935662180185318,
+      "learning_rate": 0.001,
+      "loss": 1.1032,
+      "step": 16970
+    },
+    {
+      "epoch": 0.5414713479383909,
+      "grad_norm": 0.19856195151805878,
+      "learning_rate": 0.001,
+      "loss": 1.0977,
+      "step": 16980
+    },
+    {
+      "epoch": 0.5417902356580248,
+      "grad_norm": 0.19449791312217712,
+      "learning_rate": 0.001,
+      "loss": 1.1232,
+      "step": 16990
+    },
+    {
+      "epoch": 0.5421091233776587,
+      "grad_norm": 0.18957769870758057,
+      "learning_rate": 0.001,
+      "loss": 1.0844,
+      "step": 17000
+    },
+    {
+      "epoch": 0.5424280110972927,
+      "grad_norm": 0.19556500017642975,
+      "learning_rate": 0.001,
+      "loss": 1.0877,
+      "step": 17010
+    },
+    {
+      "epoch": 0.5427468988169266,
+      "grad_norm": 0.19305263459682465,
+      "learning_rate": 0.001,
+      "loss": 1.0766,
+      "step": 17020
+    },
+    {
+      "epoch": 0.5430657865365605,
+      "grad_norm": 0.20026710629463196,
+      "learning_rate": 0.001,
+      "loss": 1.1082,
+      "step": 17030
+    },
+    {
+      "epoch": 0.5433846742561944,
+      "grad_norm": 0.19406114518642426,
+      "learning_rate": 0.001,
+      "loss": 1.0823,
+      "step": 17040
+    },
+    {
+      "epoch": 0.5437035619758284,
+      "grad_norm": 0.1891794353723526,
+      "learning_rate": 0.001,
+      "loss": 1.0884,
+      "step": 17050
+    },
+    {
+      "epoch": 0.5440224496954622,
+      "grad_norm": 0.20411112904548645,
+      "learning_rate": 0.001,
+      "loss": 1.1138,
+      "step": 17060
+    },
+    {
+      "epoch": 0.5443413374150962,
+      "grad_norm": 0.19278182089328766,
+      "learning_rate": 0.001,
+      "loss": 1.1148,
+      "step": 17070
+    },
+    {
+      "epoch": 0.5446602251347301,
+      "grad_norm": 0.1992575079202652,
+      "learning_rate": 0.001,
+      "loss": 1.1188,
+      "step": 17080
+    },
+    {
+      "epoch": 0.544979112854364,
+      "grad_norm": 0.21864908933639526,
+      "learning_rate": 0.001,
+      "loss": 1.1079,
+      "step": 17090
+    },
+    {
+      "epoch": 0.5452980005739979,
+      "grad_norm": 0.1993020623922348,
+      "learning_rate": 0.001,
+      "loss": 1.0993,
+      "step": 17100
+    },
+    {
+      "epoch": 0.5456168882936319,
+      "grad_norm": 0.19873854517936707,
+      "learning_rate": 0.001,
+      "loss": 1.1108,
+      "step": 17110
+    },
+    {
+      "epoch": 0.5459357760132657,
+      "grad_norm": 0.200069859623909,
+      "learning_rate": 0.001,
+      "loss": 1.1162,
+      "step": 17120
+    },
+    {
+      "epoch": 0.5462546637328997,
+      "grad_norm": 0.19639627635478973,
+      "learning_rate": 0.001,
+      "loss": 1.1054,
+      "step": 17130
+    },
+    {
+      "epoch": 0.5465735514525336,
+      "grad_norm": 0.19923728704452515,
+      "learning_rate": 0.001,
+      "loss": 1.0938,
+      "step": 17140
+    },
+    {
+      "epoch": 0.5468924391721675,
+      "grad_norm": 0.20334571599960327,
+      "learning_rate": 0.001,
+      "loss": 1.1029,
+      "step": 17150
+    },
+    {
+      "epoch": 0.5472113268918014,
+      "grad_norm": 0.19749508798122406,
+      "learning_rate": 0.001,
+      "loss": 1.0991,
+      "step": 17160
+    },
+    {
+      "epoch": 0.5475302146114354,
+      "grad_norm": 0.18736779689788818,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 17170
+    },
+    {
+      "epoch": 0.5478491023310692,
+      "grad_norm": 0.20096038281917572,
+      "learning_rate": 0.001,
+      "loss": 1.1176,
+      "step": 17180
+    },
+    {
+      "epoch": 0.5481679900507032,
+      "grad_norm": 0.20153848826885223,
+      "learning_rate": 0.001,
+      "loss": 1.0939,
+      "step": 17190
+    },
+    {
+      "epoch": 0.5484868777703371,
+      "grad_norm": 0.19858387112617493,
+      "learning_rate": 0.001,
+      "loss": 1.11,
+      "step": 17200
+    },
+    {
+      "epoch": 0.548805765489971,
+      "grad_norm": 0.196171373128891,
+      "learning_rate": 0.001,
+      "loss": 1.1042,
+      "step": 17210
+    },
+    {
+      "epoch": 0.5491246532096049,
+      "grad_norm": 0.19899526238441467,
+      "learning_rate": 0.001,
+      "loss": 1.093,
+      "step": 17220
+    },
+    {
+      "epoch": 0.5494435409292389,
+      "grad_norm": 0.19485172629356384,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 17230
+    },
+    {
+      "epoch": 0.5497624286488727,
+      "grad_norm": 0.1970384567975998,
+      "learning_rate": 0.001,
+      "loss": 1.1095,
+      "step": 17240
+    },
+    {
+      "epoch": 0.5500813163685067,
+      "grad_norm": 0.19277705252170563,
+      "learning_rate": 0.001,
+      "loss": 1.1061,
+      "step": 17250
+    },
+    {
+      "epoch": 0.5504002040881406,
+      "grad_norm": 0.204323410987854,
+      "learning_rate": 0.001,
+      "loss": 1.1005,
+      "step": 17260
+    },
+    {
+      "epoch": 0.5507190918077745,
+      "grad_norm": 0.19751951098442078,
+      "learning_rate": 0.001,
+      "loss": 1.1238,
+      "step": 17270
+    },
+    {
+      "epoch": 0.5510379795274084,
+      "grad_norm": 0.20313957333564758,
+      "learning_rate": 0.001,
+      "loss": 1.0872,
+      "step": 17280
+    },
+    {
+      "epoch": 0.5513568672470424,
+      "grad_norm": 0.2005039006471634,
+      "learning_rate": 0.001,
+      "loss": 1.1231,
+      "step": 17290
+    },
+    {
+      "epoch": 0.5516757549666762,
+      "grad_norm": 0.19764991104602814,
+      "learning_rate": 0.001,
+      "loss": 1.0972,
+      "step": 17300
+    },
+    {
+      "epoch": 0.5519946426863102,
+      "grad_norm": 0.19491420686244965,
+      "learning_rate": 0.001,
+      "loss": 1.1061,
+      "step": 17310
+    },
+    {
+      "epoch": 0.5523135304059441,
+      "grad_norm": 0.19485335052013397,
+      "learning_rate": 0.001,
+      "loss": 1.1027,
+      "step": 17320
+    },
+    {
+      "epoch": 0.552632418125578,
+      "grad_norm": 0.1993619054555893,
+      "learning_rate": 0.001,
+      "loss": 1.0933,
+      "step": 17330
+    },
+    {
+      "epoch": 0.5529513058452119,
+      "grad_norm": 0.19590185582637787,
+      "learning_rate": 0.001,
+      "loss": 1.1032,
+      "step": 17340
+    },
+    {
+      "epoch": 0.5532701935648459,
+      "grad_norm": 0.19090668857097626,
+      "learning_rate": 0.001,
+      "loss": 1.1033,
+      "step": 17350
+    },
+    {
+      "epoch": 0.5535890812844797,
+      "grad_norm": 0.1968427151441574,
+      "learning_rate": 0.001,
+      "loss": 1.0788,
+      "step": 17360
+    },
+    {
+      "epoch": 0.5539079690041137,
+      "grad_norm": 0.19366152584552765,
+      "learning_rate": 0.001,
+      "loss": 1.0949,
+      "step": 17370
+    },
+    {
+      "epoch": 0.5542268567237476,
+      "grad_norm": 0.20130929350852966,
+      "learning_rate": 0.001,
+      "loss": 1.1111,
+      "step": 17380
+    },
+    {
+      "epoch": 0.5545457444433814,
+      "grad_norm": 0.2025395929813385,
+      "learning_rate": 0.001,
+      "loss": 1.101,
+      "step": 17390
+    },
+    {
+      "epoch": 0.5548646321630154,
+      "grad_norm": 0.19849498569965363,
+      "learning_rate": 0.001,
+      "loss": 1.127,
+      "step": 17400
+    },
+    {
+      "epoch": 0.5551835198826494,
+      "grad_norm": 0.20587202906608582,
+      "learning_rate": 0.001,
+      "loss": 1.1102,
+      "step": 17410
+    },
+    {
+      "epoch": 0.5555024076022832,
+      "grad_norm": 0.19566045701503754,
+      "learning_rate": 0.001,
+      "loss": 1.0888,
+      "step": 17420
+    },
+    {
+      "epoch": 0.5558212953219172,
+      "grad_norm": 0.2037980705499649,
+      "learning_rate": 0.001,
+      "loss": 1.1007,
+      "step": 17430
+    },
+    {
+      "epoch": 0.5561401830415511,
+      "grad_norm": 0.20354115962982178,
+      "learning_rate": 0.001,
+      "loss": 1.1055,
+      "step": 17440
+    },
+    {
+      "epoch": 0.556459070761185,
+      "grad_norm": 0.19119271636009216,
+      "learning_rate": 0.001,
+      "loss": 1.1044,
+      "step": 17450
+    },
+    {
+      "epoch": 0.5567779584808189,
+      "grad_norm": 0.19770164787769318,
+      "learning_rate": 0.001,
+      "loss": 1.106,
+      "step": 17460
+    },
+    {
+      "epoch": 0.5570968462004529,
+      "grad_norm": 0.19170503318309784,
+      "learning_rate": 0.001,
+      "loss": 1.109,
+      "step": 17470
+    },
+    {
+      "epoch": 0.5574157339200867,
+      "grad_norm": 0.19421005249023438,
+      "learning_rate": 0.001,
+      "loss": 1.083,
+      "step": 17480
+    },
+    {
+      "epoch": 0.5577346216397207,
+      "grad_norm": 0.1924920678138733,
+      "learning_rate": 0.001,
+      "loss": 1.109,
+      "step": 17490
+    },
+    {
+      "epoch": 0.5580535093593546,
+      "grad_norm": 0.1946050226688385,
+      "learning_rate": 0.001,
+      "loss": 1.1043,
+      "step": 17500
+    },
+    {
+      "epoch": 0.5583723970789884,
+      "grad_norm": 0.20246347784996033,
+      "learning_rate": 0.001,
+      "loss": 1.1085,
+      "step": 17510
+    },
+    {
+      "epoch": 0.5586912847986224,
+      "grad_norm": 0.20876635611057281,
+      "learning_rate": 0.001,
+      "loss": 1.1011,
+      "step": 17520
+    },
+    {
+      "epoch": 0.5590101725182564,
+      "grad_norm": 0.2076573222875595,
+      "learning_rate": 0.001,
+      "loss": 1.0885,
+      "step": 17530
+    },
+    {
+      "epoch": 0.5593290602378902,
+      "grad_norm": 0.1925816386938095,
+      "learning_rate": 0.001,
+      "loss": 1.0931,
+      "step": 17540
+    },
+    {
+      "epoch": 0.5596479479575242,
+      "grad_norm": 0.1946088969707489,
+      "learning_rate": 0.001,
+      "loss": 1.1242,
+      "step": 17550
+    },
+    {
+      "epoch": 0.5599668356771581,
+      "grad_norm": 0.2025405466556549,
+      "learning_rate": 0.001,
+      "loss": 1.0851,
+      "step": 17560
+    },
+    {
+      "epoch": 0.560285723396792,
+      "grad_norm": 0.19911693036556244,
+      "learning_rate": 0.001,
+      "loss": 1.1009,
+      "step": 17570
+    },
+    {
+      "epoch": 0.5606046111164259,
+      "grad_norm": 0.2055385559797287,
+      "learning_rate": 0.001,
+      "loss": 1.1024,
+      "step": 17580
+    },
+    {
+      "epoch": 0.5609234988360599,
+      "grad_norm": 0.20296630263328552,
+      "learning_rate": 0.001,
+      "loss": 1.0984,
+      "step": 17590
+    },
+    {
+      "epoch": 0.5612423865556937,
+      "grad_norm": 0.19320446252822876,
+      "learning_rate": 0.001,
+      "loss": 1.0961,
+      "step": 17600
+    },
+    {
+      "epoch": 0.5615612742753276,
+      "grad_norm": 0.1940065175294876,
+      "learning_rate": 0.001,
+      "loss": 1.1012,
+      "step": 17610
+    },
+    {
+      "epoch": 0.5618801619949616,
+      "grad_norm": 0.1887366771697998,
+      "learning_rate": 0.001,
+      "loss": 1.0996,
+      "step": 17620
+    },
+    {
+      "epoch": 0.5621990497145954,
+      "grad_norm": 0.19770963490009308,
+      "learning_rate": 0.001,
+      "loss": 1.1049,
+      "step": 17630
+    },
+    {
+      "epoch": 0.5625179374342294,
+      "grad_norm": 0.20164230465888977,
+      "learning_rate": 0.001,
+      "loss": 1.0906,
+      "step": 17640
+    },
+    {
+      "epoch": 0.5628368251538634,
+      "grad_norm": 0.19402539730072021,
+      "learning_rate": 0.001,
+      "loss": 1.0791,
+      "step": 17650
+    },
+    {
+      "epoch": 0.5631557128734972,
+      "grad_norm": 0.20351935923099518,
+      "learning_rate": 0.001,
+      "loss": 1.1006,
+      "step": 17660
+    },
+    {
+      "epoch": 0.5634746005931311,
+      "grad_norm": 0.2015489935874939,
+      "learning_rate": 0.001,
+      "loss": 1.098,
+      "step": 17670
+    },
+    {
+      "epoch": 0.5637934883127651,
+      "grad_norm": 0.19619329273700714,
+      "learning_rate": 0.001,
+      "loss": 1.0958,
+      "step": 17680
+    },
+    {
+      "epoch": 0.5641123760323989,
+      "grad_norm": 0.20024628937244415,
+      "learning_rate": 0.001,
+      "loss": 1.1097,
+      "step": 17690
+    },
+    {
+      "epoch": 0.5644312637520329,
+      "grad_norm": 0.19571398198604584,
+      "learning_rate": 0.001,
+      "loss": 1.0991,
+      "step": 17700
+    },
+    {
+      "epoch": 0.5647501514716669,
+      "grad_norm": 0.19286087155342102,
+      "learning_rate": 0.001,
+      "loss": 1.1053,
+      "step": 17710
+    },
+    {
+      "epoch": 0.5650690391913007,
+      "grad_norm": 0.19929246604442596,
+      "learning_rate": 0.001,
+      "loss": 1.0978,
+      "step": 17720
+    },
+    {
+      "epoch": 0.5653879269109346,
+      "grad_norm": 0.19413244724273682,
+      "learning_rate": 0.001,
+      "loss": 1.0784,
+      "step": 17730
+    },
+    {
+      "epoch": 0.5657068146305686,
+      "grad_norm": 0.1949404627084732,
+      "learning_rate": 0.001,
+      "loss": 1.0973,
+      "step": 17740
+    },
+    {
+      "epoch": 0.5660257023502024,
+      "grad_norm": 0.1955089569091797,
+      "learning_rate": 0.001,
+      "loss": 1.0846,
+      "step": 17750
+    },
+    {
+      "epoch": 0.5663445900698364,
+      "grad_norm": 0.2049809694290161,
+      "learning_rate": 0.001,
+      "loss": 1.0986,
+      "step": 17760
+    },
+    {
+      "epoch": 0.5666634777894703,
+      "grad_norm": 0.19769684970378876,
+      "learning_rate": 0.001,
+      "loss": 1.0903,
+      "step": 17770
+    },
+    {
+      "epoch": 0.5669823655091042,
+      "grad_norm": 0.2010537087917328,
+      "learning_rate": 0.001,
+      "loss": 1.0989,
+      "step": 17780
+    },
+    {
+      "epoch": 0.5673012532287381,
+      "grad_norm": 0.19881124794483185,
+      "learning_rate": 0.001,
+      "loss": 1.1099,
+      "step": 17790
+    },
+    {
+      "epoch": 0.5676201409483721,
+      "grad_norm": 0.18793614208698273,
+      "learning_rate": 0.001,
+      "loss": 1.0782,
+      "step": 17800
+    },
+    {
+      "epoch": 0.5679390286680059,
+      "grad_norm": 0.20160049200057983,
+      "learning_rate": 0.001,
+      "loss": 1.0984,
+      "step": 17810
+    },
+    {
+      "epoch": 0.5682579163876399,
+      "grad_norm": 0.20136494934558868,
+      "learning_rate": 0.001,
+      "loss": 1.1001,
+      "step": 17820
+    },
+    {
+      "epoch": 0.5685768041072738,
+      "grad_norm": 0.18992014229297638,
+      "learning_rate": 0.001,
+      "loss": 1.0898,
+      "step": 17830
+    },
+    {
+      "epoch": 0.5688956918269078,
+      "grad_norm": 0.20103414356708527,
+      "learning_rate": 0.001,
+      "loss": 1.0877,
+      "step": 17840
+    },
+    {
+      "epoch": 0.5692145795465416,
+      "grad_norm": 0.19369281828403473,
+      "learning_rate": 0.001,
+      "loss": 1.0928,
+      "step": 17850
+    },
+    {
+      "epoch": 0.5695334672661756,
+      "grad_norm": 0.19698695838451385,
+      "learning_rate": 0.001,
+      "loss": 1.0951,
+      "step": 17860
+    },
+    {
+      "epoch": 0.5698523549858096,
+      "grad_norm": 0.1910889446735382,
+      "learning_rate": 0.001,
+      "loss": 1.0902,
+      "step": 17870
+    },
+    {
+      "epoch": 0.5701712427054434,
+      "grad_norm": 0.2007639855146408,
+      "learning_rate": 0.001,
+      "loss": 1.101,
+      "step": 17880
+    },
+    {
+      "epoch": 0.5704901304250773,
+      "grad_norm": 0.19711415469646454,
+      "learning_rate": 0.001,
+      "loss": 1.0923,
+      "step": 17890
+    },
+    {
+      "epoch": 0.5708090181447113,
+      "grad_norm": 0.20394960045814514,
+      "learning_rate": 0.001,
+      "loss": 1.0962,
+      "step": 17900
+    },
+    {
+      "epoch": 0.5711279058643451,
+      "grad_norm": 0.19199271500110626,
+      "learning_rate": 0.001,
+      "loss": 1.0959,
+      "step": 17910
+    },
+    {
+      "epoch": 0.5714467935839791,
+      "grad_norm": 0.19167867302894592,
+      "learning_rate": 0.001,
+      "loss": 1.0927,
+      "step": 17920
+    },
+    {
+      "epoch": 0.571765681303613,
+      "grad_norm": 0.19008177518844604,
+      "learning_rate": 0.001,
+      "loss": 1.0966,
+      "step": 17930
+    },
+    {
+      "epoch": 0.5720845690232469,
+      "grad_norm": 0.19732196629047394,
+      "learning_rate": 0.001,
+      "loss": 1.1091,
+      "step": 17940
+    },
+    {
+      "epoch": 0.5724034567428808,
+      "grad_norm": 0.20157304406166077,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 17950
+    },
+    {
+      "epoch": 0.5727223444625148,
+      "grad_norm": 0.199776753783226,
+      "learning_rate": 0.001,
+      "loss": 1.0737,
+      "step": 17960
+    },
+    {
+      "epoch": 0.5730412321821486,
+      "grad_norm": 0.18558533489704132,
+      "learning_rate": 0.001,
+      "loss": 1.0909,
+      "step": 17970
+    },
+    {
+      "epoch": 0.5733601199017826,
+      "grad_norm": 0.20540328323841095,
+      "learning_rate": 0.001,
+      "loss": 1.1023,
+      "step": 17980
+    },
+    {
+      "epoch": 0.5736790076214165,
+      "grad_norm": 0.1917695701122284,
+      "learning_rate": 0.001,
+      "loss": 1.11,
+      "step": 17990
+    },
+    {
+      "epoch": 0.5739978953410504,
+      "grad_norm": 0.20183104276657104,
+      "learning_rate": 0.001,
+      "loss": 1.0902,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5743167830606843,
+      "grad_norm": 0.19891691207885742,
+      "learning_rate": 0.001,
+      "loss": 1.0862,
+      "step": 18010
+    },
+    {
+      "epoch": 0.5746356707803183,
+      "grad_norm": 0.20543940365314484,
+      "learning_rate": 0.001,
+      "loss": 1.0992,
+      "step": 18020
+    },
+    {
+      "epoch": 0.5749545584999521,
+      "grad_norm": 0.19975398480892181,
+      "learning_rate": 0.001,
+      "loss": 1.1015,
+      "step": 18030
+    },
+    {
+      "epoch": 0.5752734462195861,
+      "grad_norm": 0.18917524814605713,
+      "learning_rate": 0.001,
+      "loss": 1.0802,
+      "step": 18040
+    },
+    {
+      "epoch": 0.57559233393922,
+      "grad_norm": 0.1999102383852005,
+      "learning_rate": 0.001,
+      "loss": 1.0929,
+      "step": 18050
+    },
+    {
+      "epoch": 0.5759112216588539,
+      "grad_norm": 0.20146946609020233,
+      "learning_rate": 0.001,
+      "loss": 1.0856,
+      "step": 18060
+    },
+    {
+      "epoch": 0.5762301093784878,
+      "grad_norm": 0.18728485703468323,
+      "learning_rate": 0.001,
+      "loss": 1.0954,
+      "step": 18070
+    },
+    {
+      "epoch": 0.5765489970981218,
+      "grad_norm": 0.1965542733669281,
+      "learning_rate": 0.001,
+      "loss": 1.0916,
+      "step": 18080
+    },
+    {
+      "epoch": 0.5768678848177556,
+      "grad_norm": 0.18837697803974152,
+      "learning_rate": 0.001,
+      "loss": 1.1121,
+      "step": 18090
+    },
+    {
+      "epoch": 0.5771867725373896,
+      "grad_norm": 0.20012810826301575,
+      "learning_rate": 0.001,
+      "loss": 1.1063,
+      "step": 18100
+    },
+    {
+      "epoch": 0.5775056602570235,
+      "grad_norm": 0.19055570662021637,
+      "learning_rate": 0.001,
+      "loss": 1.0923,
+      "step": 18110
+    },
+    {
+      "epoch": 0.5778245479766574,
+      "grad_norm": 0.20348024368286133,
+      "learning_rate": 0.001,
+      "loss": 1.1026,
+      "step": 18120
+    },
+    {
+      "epoch": 0.5781434356962913,
+      "grad_norm": 0.1999468058347702,
+      "learning_rate": 0.001,
+      "loss": 1.0778,
+      "step": 18130
+    },
+    {
+      "epoch": 0.5784623234159253,
+      "grad_norm": 0.19548684358596802,
+      "learning_rate": 0.001,
+      "loss": 1.0973,
+      "step": 18140
+    },
+    {
+      "epoch": 0.5787812111355591,
+      "grad_norm": 0.19422020018100739,
+      "learning_rate": 0.001,
+      "loss": 1.0645,
+      "step": 18150
+    },
+    {
+      "epoch": 0.5791000988551931,
+      "grad_norm": 0.19890695810317993,
+      "learning_rate": 0.001,
+      "loss": 1.0852,
+      "step": 18160
+    },
+    {
+      "epoch": 0.579418986574827,
+      "grad_norm": 0.19380579888820648,
+      "learning_rate": 0.001,
+      "loss": 1.0773,
+      "step": 18170
+    },
+    {
+      "epoch": 0.5797378742944609,
+      "grad_norm": 0.18737071752548218,
+      "learning_rate": 0.001,
+      "loss": 1.1016,
+      "step": 18180
+    },
+    {
+      "epoch": 0.5800567620140948,
+      "grad_norm": 0.18995201587677002,
+      "learning_rate": 0.001,
+      "loss": 1.0984,
+      "step": 18190
+    },
+    {
+      "epoch": 0.5803756497337288,
+      "grad_norm": 0.19956058263778687,
+      "learning_rate": 0.001,
+      "loss": 1.1032,
+      "step": 18200
+    },
+    {
+      "epoch": 0.5806945374533626,
+      "grad_norm": 0.1915549784898758,
+      "learning_rate": 0.001,
+      "loss": 1.0938,
+      "step": 18210
+    },
+    {
+      "epoch": 0.5810134251729966,
+      "grad_norm": 0.19854103028774261,
+      "learning_rate": 0.001,
+      "loss": 1.0981,
+      "step": 18220
+    },
+    {
+      "epoch": 0.5813323128926305,
+      "grad_norm": 0.19441866874694824,
+      "learning_rate": 0.001,
+      "loss": 1.1146,
+      "step": 18230
+    },
+    {
+      "epoch": 0.5816512006122644,
+      "grad_norm": 0.1921454817056656,
+      "learning_rate": 0.001,
+      "loss": 1.0873,
+      "step": 18240
+    },
+    {
+      "epoch": 0.5819700883318983,
+      "grad_norm": 0.19195596873760223,
+      "learning_rate": 0.001,
+      "loss": 1.0994,
+      "step": 18250
+    },
+    {
+      "epoch": 0.5822889760515323,
+      "grad_norm": 0.1962839663028717,
+      "learning_rate": 0.001,
+      "loss": 1.0879,
+      "step": 18260
+    },
+    {
+      "epoch": 0.5826078637711661,
+      "grad_norm": 0.2027946412563324,
+      "learning_rate": 0.001,
+      "loss": 1.0964,
+      "step": 18270
+    },
+    {
+      "epoch": 0.5829267514908001,
+      "grad_norm": 0.19965171813964844,
+      "learning_rate": 0.001,
+      "loss": 1.0807,
+      "step": 18280
+    },
+    {
+      "epoch": 0.583245639210434,
+      "grad_norm": 0.19153305888175964,
+      "learning_rate": 0.001,
+      "loss": 1.0965,
+      "step": 18290
+    },
+    {
+      "epoch": 0.5835645269300679,
+      "grad_norm": 0.20132271945476532,
+      "learning_rate": 0.001,
+      "loss": 1.0881,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5838834146497018,
+      "grad_norm": 0.19851796329021454,
+      "learning_rate": 0.001,
+      "loss": 1.0796,
+      "step": 18310
+    },
+    {
+      "epoch": 0.5842023023693358,
+      "grad_norm": 0.1932992786169052,
+      "learning_rate": 0.001,
+      "loss": 1.0941,
+      "step": 18320
+    },
+    {
+      "epoch": 0.5845211900889696,
+      "grad_norm": 0.20372772216796875,
+      "learning_rate": 0.001,
+      "loss": 1.097,
+      "step": 18330
+    },
+    {
+      "epoch": 0.5848400778086036,
+      "grad_norm": 0.196427583694458,
+      "learning_rate": 0.001,
+      "loss": 1.0864,
+      "step": 18340
+    },
+    {
+      "epoch": 0.5851589655282375,
+      "grad_norm": 0.19252528250217438,
+      "learning_rate": 0.001,
+      "loss": 1.098,
+      "step": 18350
+    },
+    {
+      "epoch": 0.5854778532478714,
+      "grad_norm": 0.19300012290477753,
+      "learning_rate": 0.001,
+      "loss": 1.1109,
+      "step": 18360
+    },
+    {
+      "epoch": 0.5857967409675053,
+      "grad_norm": 0.19377656280994415,
+      "learning_rate": 0.001,
+      "loss": 1.111,
+      "step": 18370
+    },
+    {
+      "epoch": 0.5861156286871393,
+      "grad_norm": 0.2068767100572586,
+      "learning_rate": 0.001,
+      "loss": 1.1132,
+      "step": 18380
+    },
+    {
+      "epoch": 0.5864345164067731,
+      "grad_norm": 0.20298132300376892,
+      "learning_rate": 0.001,
+      "loss": 1.0819,
+      "step": 18390
+    },
+    {
+      "epoch": 0.5867534041264071,
+      "grad_norm": 0.19519521296024323,
+      "learning_rate": 0.001,
+      "loss": 1.0843,
+      "step": 18400
+    },
+    {
+      "epoch": 0.587072291846041,
+      "grad_norm": 0.19534125924110413,
+      "learning_rate": 0.001,
+      "loss": 1.0902,
+      "step": 18410
+    },
+    {
+      "epoch": 0.5873911795656749,
+      "grad_norm": 0.20693759620189667,
+      "learning_rate": 0.001,
+      "loss": 1.1059,
+      "step": 18420
+    },
+    {
+      "epoch": 0.5877100672853088,
+      "grad_norm": 0.20042476058006287,
+      "learning_rate": 0.001,
+      "loss": 1.1061,
+      "step": 18430
+    },
+    {
+      "epoch": 0.5880289550049428,
+      "grad_norm": 0.20221875607967377,
+      "learning_rate": 0.001,
+      "loss": 1.0991,
+      "step": 18440
+    },
+    {
+      "epoch": 0.5883478427245766,
+      "grad_norm": 0.19484549760818481,
+      "learning_rate": 0.001,
+      "loss": 1.0965,
+      "step": 18450
+    },
+    {
+      "epoch": 0.5886667304442106,
+      "grad_norm": 0.19391950964927673,
+      "learning_rate": 0.001,
+      "loss": 1.0754,
+      "step": 18460
+    },
+    {
+      "epoch": 0.5889856181638445,
+      "grad_norm": 0.19615234434604645,
+      "learning_rate": 0.001,
+      "loss": 1.085,
+      "step": 18470
+    },
+    {
+      "epoch": 0.5893045058834784,
+      "grad_norm": 0.18772046267986298,
+      "learning_rate": 0.001,
+      "loss": 1.1002,
+      "step": 18480
+    },
+    {
+      "epoch": 0.5896233936031123,
+      "grad_norm": 0.19404157996177673,
+      "learning_rate": 0.001,
+      "loss": 1.0865,
+      "step": 18490
+    },
+    {
+      "epoch": 0.5899422813227463,
+      "grad_norm": 0.19353818893432617,
+      "learning_rate": 0.001,
+      "loss": 1.0946,
+      "step": 18500
+    },
+    {
+      "epoch": 0.5902611690423801,
+      "grad_norm": 0.20134851336479187,
+      "learning_rate": 0.001,
+      "loss": 1.0916,
+      "step": 18510
+    },
+    {
+      "epoch": 0.5905800567620141,
+      "grad_norm": 0.20002439618110657,
+      "learning_rate": 0.001,
+      "loss": 1.1061,
+      "step": 18520
+    },
+    {
+      "epoch": 0.590898944481648,
+      "grad_norm": 0.20346671342849731,
+      "learning_rate": 0.001,
+      "loss": 1.0839,
+      "step": 18530
+    },
+    {
+      "epoch": 0.5912178322012819,
+      "grad_norm": 0.19714142382144928,
+      "learning_rate": 0.001,
+      "loss": 1.0961,
+      "step": 18540
+    },
+    {
+      "epoch": 0.5915367199209158,
+      "grad_norm": 0.19142982363700867,
+      "learning_rate": 0.001,
+      "loss": 1.0959,
+      "step": 18550
+    },
+    {
+      "epoch": 0.5918556076405498,
+      "grad_norm": 0.2000853419303894,
+      "learning_rate": 0.001,
+      "loss": 1.1113,
+      "step": 18560
+    },
+    {
+      "epoch": 0.5921744953601836,
+      "grad_norm": 0.19660703837871552,
+      "learning_rate": 0.001,
+      "loss": 1.0753,
+      "step": 18570
+    },
+    {
+      "epoch": 0.5924933830798176,
+      "grad_norm": 0.20565517246723175,
+      "learning_rate": 0.001,
+      "loss": 1.0834,
+      "step": 18580
+    },
+    {
+      "epoch": 0.5928122707994515,
+      "grad_norm": 0.1943969577550888,
+      "learning_rate": 0.001,
+      "loss": 1.0956,
+      "step": 18590
+    },
+    {
+      "epoch": 0.5931311585190854,
+      "grad_norm": 0.19117647409439087,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 18600
+    },
+    {
+      "epoch": 0.5934500462387193,
+      "grad_norm": 0.19650551676750183,
+      "learning_rate": 0.001,
+      "loss": 1.1022,
+      "step": 18610
+    },
+    {
+      "epoch": 0.5937689339583533,
+      "grad_norm": 0.1948203593492508,
+      "learning_rate": 0.001,
+      "loss": 1.0972,
+      "step": 18620
+    },
+    {
+      "epoch": 0.5940878216779871,
+      "grad_norm": 0.20211590826511383,
+      "learning_rate": 0.001,
+      "loss": 1.0944,
+      "step": 18630
+    },
+    {
+      "epoch": 0.5944067093976211,
+      "grad_norm": 0.19122843444347382,
+      "learning_rate": 0.001,
+      "loss": 1.0911,
+      "step": 18640
+    },
+    {
+      "epoch": 0.594725597117255,
+      "grad_norm": 0.18877461552619934,
+      "learning_rate": 0.001,
+      "loss": 1.0961,
+      "step": 18650
+    },
+    {
+      "epoch": 0.5950444848368889,
+      "grad_norm": 0.19857579469680786,
+      "learning_rate": 0.001,
+      "loss": 1.0676,
+      "step": 18660
+    },
+    {
+      "epoch": 0.5953633725565228,
+      "grad_norm": 0.19899848103523254,
+      "learning_rate": 0.001,
+      "loss": 1.0845,
+      "step": 18670
+    },
+    {
+      "epoch": 0.5956822602761568,
+      "grad_norm": 0.19641920924186707,
+      "learning_rate": 0.001,
+      "loss": 1.0935,
+      "step": 18680
+    },
+    {
+      "epoch": 0.5960011479957907,
+      "grad_norm": 0.1865895688533783,
+      "learning_rate": 0.001,
+      "loss": 1.0864,
+      "step": 18690
+    },
+    {
+      "epoch": 0.5963200357154246,
+      "grad_norm": 0.19067290425300598,
+      "learning_rate": 0.001,
+      "loss": 1.0751,
+      "step": 18700
+    },
+    {
+      "epoch": 0.5966389234350585,
+      "grad_norm": 0.19792988896369934,
+      "learning_rate": 0.001,
+      "loss": 1.1009,
+      "step": 18710
+    },
+    {
+      "epoch": 0.5969578111546925,
+      "grad_norm": 0.1916930228471756,
+      "learning_rate": 0.001,
+      "loss": 1.0919,
+      "step": 18720
+    },
+    {
+      "epoch": 0.5972766988743263,
+      "grad_norm": 0.19243399798870087,
+      "learning_rate": 0.001,
+      "loss": 1.0917,
+      "step": 18730
+    },
+    {
+      "epoch": 0.5975955865939603,
+      "grad_norm": 0.1880219578742981,
+      "learning_rate": 0.001,
+      "loss": 1.1085,
+      "step": 18740
+    },
+    {
+      "epoch": 0.5979144743135942,
+      "grad_norm": 0.19094349443912506,
+      "learning_rate": 0.001,
+      "loss": 1.0787,
+      "step": 18750
+    },
+    {
+      "epoch": 0.5982333620332281,
+      "grad_norm": 0.19258365035057068,
+      "learning_rate": 0.001,
+      "loss": 1.0958,
+      "step": 18760
+    },
+    {
+      "epoch": 0.598552249752862,
+      "grad_norm": 0.19618827104568481,
+      "learning_rate": 0.001,
+      "loss": 1.0911,
+      "step": 18770
+    },
+    {
+      "epoch": 0.598871137472496,
+      "grad_norm": 0.1982206553220749,
+      "learning_rate": 0.001,
+      "loss": 1.0841,
+      "step": 18780
+    },
+    {
+      "epoch": 0.5991900251921298,
+      "grad_norm": 0.20072801411151886,
+      "learning_rate": 0.001,
+      "loss": 1.1213,
+      "step": 18790
+    },
+    {
+      "epoch": 0.5995089129117638,
+      "grad_norm": 0.1929350197315216,
+      "learning_rate": 0.001,
+      "loss": 1.0577,
+      "step": 18800
+    },
+    {
+      "epoch": 0.5998278006313977,
+      "grad_norm": 0.19256435334682465,
+      "learning_rate": 0.001,
+      "loss": 1.0812,
+      "step": 18810
+    },
+    {
+      "epoch": 0.6001466883510316,
+      "grad_norm": 0.19631852209568024,
+      "learning_rate": 0.001,
+      "loss": 1.0999,
+      "step": 18820
+    },
+    {
+      "epoch": 0.6004655760706655,
+      "grad_norm": 0.19517488777637482,
+      "learning_rate": 0.001,
+      "loss": 1.0916,
+      "step": 18830
+    },
+    {
+      "epoch": 0.6007844637902995,
+      "grad_norm": 0.19708159565925598,
+      "learning_rate": 0.001,
+      "loss": 1.0882,
+      "step": 18840
+    },
+    {
+      "epoch": 0.6011033515099333,
+      "grad_norm": 0.1962396651506424,
+      "learning_rate": 0.001,
+      "loss": 1.1258,
+      "step": 18850
+    },
+    {
+      "epoch": 0.6014222392295673,
+      "grad_norm": 0.19821253418922424,
+      "learning_rate": 0.001,
+      "loss": 1.0862,
+      "step": 18860
+    },
+    {
+      "epoch": 0.6017411269492012,
+      "grad_norm": 0.1993333101272583,
+      "learning_rate": 0.001,
+      "loss": 1.0901,
+      "step": 18870
+    },
+    {
+      "epoch": 0.6020600146688351,
+      "grad_norm": 0.19593703746795654,
+      "learning_rate": 0.001,
+      "loss": 1.0747,
+      "step": 18880
+    },
+    {
+      "epoch": 0.602378902388469,
+      "grad_norm": 0.18685583770275116,
+      "learning_rate": 0.001,
+      "loss": 1.0908,
+      "step": 18890
+    },
+    {
+      "epoch": 0.602697790108103,
+      "grad_norm": 0.19480600953102112,
+      "learning_rate": 0.001,
+      "loss": 1.0918,
+      "step": 18900
+    },
+    {
+      "epoch": 0.6030166778277368,
+      "grad_norm": 0.19692644476890564,
+      "learning_rate": 0.001,
+      "loss": 1.0936,
+      "step": 18910
+    },
+    {
+      "epoch": 0.6033355655473708,
+      "grad_norm": 0.19970379769802094,
+      "learning_rate": 0.001,
+      "loss": 1.0996,
+      "step": 18920
+    },
+    {
+      "epoch": 0.6036544532670047,
+      "grad_norm": 0.1902437061071396,
+      "learning_rate": 0.001,
+      "loss": 1.0695,
+      "step": 18930
+    },
+    {
+      "epoch": 0.6039733409866386,
+      "grad_norm": 0.20170606672763824,
+      "learning_rate": 0.001,
+      "loss": 1.0699,
+      "step": 18940
+    },
+    {
+      "epoch": 0.6042922287062725,
+      "grad_norm": 0.19433560967445374,
+      "learning_rate": 0.001,
+      "loss": 1.0994,
+      "step": 18950
+    },
+    {
+      "epoch": 0.6046111164259065,
+      "grad_norm": 0.19772832095623016,
+      "learning_rate": 0.001,
+      "loss": 1.0976,
+      "step": 18960
+    },
+    {
+      "epoch": 0.6049300041455403,
+      "grad_norm": 0.2006259709596634,
+      "learning_rate": 0.001,
+      "loss": 1.0949,
+      "step": 18970
+    },
+    {
+      "epoch": 0.6052488918651743,
+      "grad_norm": 0.2032744139432907,
+      "learning_rate": 0.001,
+      "loss": 1.0961,
+      "step": 18980
+    },
+    {
+      "epoch": 0.6055677795848082,
+      "grad_norm": 0.195106640458107,
+      "learning_rate": 0.001,
+      "loss": 1.0955,
+      "step": 18990
+    },
+    {
+      "epoch": 0.6058866673044421,
+      "grad_norm": 0.19616737961769104,
+      "learning_rate": 0.001,
+      "loss": 1.0871,
+      "step": 19000
+    },
+    {
+      "epoch": 0.606205555024076,
+      "grad_norm": 0.2025897055864334,
+      "learning_rate": 0.001,
+      "loss": 1.0822,
+      "step": 19010
+    },
+    {
+      "epoch": 0.60652444274371,
+      "grad_norm": 0.19735278189182281,
+      "learning_rate": 0.001,
+      "loss": 1.1055,
+      "step": 19020
+    },
+    {
+      "epoch": 0.6068433304633438,
+      "grad_norm": 0.18668997287750244,
+      "learning_rate": 0.001,
+      "loss": 1.088,
+      "step": 19030
+    },
+    {
+      "epoch": 0.6071622181829778,
+      "grad_norm": 0.201369971036911,
+      "learning_rate": 0.001,
+      "loss": 1.0767,
+      "step": 19040
+    },
+    {
+      "epoch": 0.6074811059026117,
+      "grad_norm": 0.19418613612651825,
+      "learning_rate": 0.001,
+      "loss": 1.0814,
+      "step": 19050
+    },
+    {
+      "epoch": 0.6077999936222456,
+      "grad_norm": 0.18868911266326904,
+      "learning_rate": 0.001,
+      "loss": 1.0913,
+      "step": 19060
+    },
+    {
+      "epoch": 0.6081188813418795,
+      "grad_norm": 0.19358381628990173,
+      "learning_rate": 0.001,
+      "loss": 1.0936,
+      "step": 19070
+    },
+    {
+      "epoch": 0.6084377690615135,
+      "grad_norm": 0.19428423047065735,
+      "learning_rate": 0.001,
+      "loss": 1.0861,
+      "step": 19080
+    },
+    {
+      "epoch": 0.6087566567811473,
+      "grad_norm": 0.20530322194099426,
+      "learning_rate": 0.001,
+      "loss": 1.068,
+      "step": 19090
+    },
+    {
+      "epoch": 0.6090755445007813,
+      "grad_norm": 0.19999028742313385,
+      "learning_rate": 0.001,
+      "loss": 1.0968,
+      "step": 19100
+    },
+    {
+      "epoch": 0.6093944322204152,
+      "grad_norm": 0.21033933758735657,
+      "learning_rate": 0.001,
+      "loss": 1.0697,
+      "step": 19110
+    },
+    {
+      "epoch": 0.6097133199400491,
+      "grad_norm": 0.20325277745723724,
+      "learning_rate": 0.001,
+      "loss": 1.0923,
+      "step": 19120
+    },
+    {
+      "epoch": 0.610032207659683,
+      "grad_norm": 0.19787181913852692,
+      "learning_rate": 0.001,
+      "loss": 1.1045,
+      "step": 19130
+    },
+    {
+      "epoch": 0.610351095379317,
+      "grad_norm": 0.19693613052368164,
+      "learning_rate": 0.001,
+      "loss": 1.1015,
+      "step": 19140
+    },
+    {
+      "epoch": 0.6106699830989508,
+      "grad_norm": 0.19551625847816467,
+      "learning_rate": 0.001,
+      "loss": 1.1002,
+      "step": 19150
+    },
+    {
+      "epoch": 0.6109888708185848,
+      "grad_norm": 0.1972067505121231,
+      "learning_rate": 0.001,
+      "loss": 1.0734,
+      "step": 19160
+    },
+    {
+      "epoch": 0.6113077585382187,
+      "grad_norm": 0.19269640743732452,
+      "learning_rate": 0.001,
+      "loss": 1.1039,
+      "step": 19170
+    },
+    {
+      "epoch": 0.6116266462578526,
+      "grad_norm": 0.19563859701156616,
+      "learning_rate": 0.001,
+      "loss": 1.0986,
+      "step": 19180
+    },
+    {
+      "epoch": 0.6119455339774865,
+      "grad_norm": 0.19782280921936035,
+      "learning_rate": 0.001,
+      "loss": 1.0991,
+      "step": 19190
+    },
+    {
+      "epoch": 0.6122644216971205,
+      "grad_norm": 0.20272885262966156,
+      "learning_rate": 0.001,
+      "loss": 1.0818,
+      "step": 19200
+    },
+    {
+      "epoch": 0.6125833094167543,
+      "grad_norm": 0.18932965397834778,
+      "learning_rate": 0.001,
+      "loss": 1.0675,
+      "step": 19210
+    },
+    {
+      "epoch": 0.6129021971363883,
+      "grad_norm": 0.19627796113491058,
+      "learning_rate": 0.001,
+      "loss": 1.0917,
+      "step": 19220
+    },
+    {
+      "epoch": 0.6132210848560222,
+      "grad_norm": 0.18708980083465576,
+      "learning_rate": 0.001,
+      "loss": 1.0772,
+      "step": 19230
+    },
+    {
+      "epoch": 0.6135399725756561,
+      "grad_norm": 0.19243952631950378,
+      "learning_rate": 0.001,
+      "loss": 1.0861,
+      "step": 19240
+    },
+    {
+      "epoch": 0.61385886029529,
+      "grad_norm": 0.19430191814899445,
+      "learning_rate": 0.001,
+      "loss": 1.0734,
+      "step": 19250
+    },
+    {
+      "epoch": 0.614177748014924,
+      "grad_norm": 0.1974942982196808,
+      "learning_rate": 0.001,
+      "loss": 1.0971,
+      "step": 19260
+    },
+    {
+      "epoch": 0.6144966357345578,
+      "grad_norm": 0.19094713032245636,
+      "learning_rate": 0.001,
+      "loss": 1.0998,
+      "step": 19270
+    },
+    {
+      "epoch": 0.6148155234541918,
+      "grad_norm": 0.19476744532585144,
+      "learning_rate": 0.001,
+      "loss": 1.0896,
+      "step": 19280
+    },
+    {
+      "epoch": 0.6151344111738257,
+      "grad_norm": 0.1873323768377304,
+      "learning_rate": 0.001,
+      "loss": 1.0716,
+      "step": 19290
+    },
+    {
+      "epoch": 0.6154532988934596,
+      "grad_norm": 0.19868840277194977,
+      "learning_rate": 0.001,
+      "loss": 1.1003,
+      "step": 19300
+    },
+    {
+      "epoch": 0.6157721866130935,
+      "grad_norm": 0.19452963769435883,
+      "learning_rate": 0.001,
+      "loss": 1.097,
+      "step": 19310
+    },
+    {
+      "epoch": 0.6160910743327275,
+      "grad_norm": 0.200517475605011,
+      "learning_rate": 0.001,
+      "loss": 1.0814,
+      "step": 19320
+    },
+    {
+      "epoch": 0.6164099620523613,
+      "grad_norm": 0.1923755258321762,
+      "learning_rate": 0.001,
+      "loss": 1.1073,
+      "step": 19330
+    },
+    {
+      "epoch": 0.6167288497719953,
+      "grad_norm": 0.19483143091201782,
+      "learning_rate": 0.001,
+      "loss": 1.0796,
+      "step": 19340
+    },
+    {
+      "epoch": 0.6170477374916292,
+      "grad_norm": 0.19290411472320557,
+      "learning_rate": 0.001,
+      "loss": 1.0853,
+      "step": 19350
+    },
+    {
+      "epoch": 0.6173666252112631,
+      "grad_norm": 0.19407965242862701,
+      "learning_rate": 0.001,
+      "loss": 1.0913,
+      "step": 19360
+    },
+    {
+      "epoch": 0.617685512930897,
+      "grad_norm": 0.19258002936840057,
+      "learning_rate": 0.001,
+      "loss": 1.0922,
+      "step": 19370
+    },
+    {
+      "epoch": 0.618004400650531,
+      "grad_norm": 0.19973376393318176,
+      "learning_rate": 0.001,
+      "loss": 1.0984,
+      "step": 19380
+    },
+    {
+      "epoch": 0.6183232883701648,
+      "grad_norm": 0.2037927806377411,
+      "learning_rate": 0.001,
+      "loss": 1.1043,
+      "step": 19390
+    },
+    {
+      "epoch": 0.6186421760897988,
+      "grad_norm": 0.1936698704957962,
+      "learning_rate": 0.001,
+      "loss": 1.0796,
+      "step": 19400
+    },
+    {
+      "epoch": 0.6189610638094327,
+      "grad_norm": 0.19478873908519745,
+      "learning_rate": 0.001,
+      "loss": 1.0808,
+      "step": 19410
+    },
+    {
+      "epoch": 0.6192799515290666,
+      "grad_norm": 0.1989588886499405,
+      "learning_rate": 0.001,
+      "loss": 1.1089,
+      "step": 19420
+    },
+    {
+      "epoch": 0.6195988392487005,
+      "grad_norm": 0.19915714859962463,
+      "learning_rate": 0.001,
+      "loss": 1.0871,
+      "step": 19430
+    },
+    {
+      "epoch": 0.6199177269683345,
+      "grad_norm": 0.1963144987821579,
+      "learning_rate": 0.001,
+      "loss": 1.0947,
+      "step": 19440
+    },
+    {
+      "epoch": 0.6202366146879683,
+      "grad_norm": 0.1846805065870285,
+      "learning_rate": 0.001,
+      "loss": 1.0678,
+      "step": 19450
+    },
+    {
+      "epoch": 0.6205555024076023,
+      "grad_norm": 0.1934468150138855,
+      "learning_rate": 0.001,
+      "loss": 1.1098,
+      "step": 19460
+    },
+    {
+      "epoch": 0.6208743901272362,
+      "grad_norm": 0.18973872065544128,
+      "learning_rate": 0.001,
+      "loss": 1.0751,
+      "step": 19470
+    },
+    {
+      "epoch": 0.6211932778468701,
+      "grad_norm": 0.19252535700798035,
+      "learning_rate": 0.001,
+      "loss": 1.0867,
+      "step": 19480
+    },
+    {
+      "epoch": 0.621512165566504,
+      "grad_norm": 0.19858285784721375,
+      "learning_rate": 0.001,
+      "loss": 1.09,
+      "step": 19490
+    },
+    {
+      "epoch": 0.621831053286138,
+      "grad_norm": 0.19829675555229187,
+      "learning_rate": 0.001,
+      "loss": 1.0955,
+      "step": 19500
+    },
+    {
+      "epoch": 0.6221499410057719,
+      "grad_norm": 0.1918306052684784,
+      "learning_rate": 0.001,
+      "loss": 1.0949,
+      "step": 19510
+    },
+    {
+      "epoch": 0.6224688287254058,
+      "grad_norm": 0.1984504908323288,
+      "learning_rate": 0.001,
+      "loss": 1.0865,
+      "step": 19520
+    },
+    {
+      "epoch": 0.6227877164450397,
+      "grad_norm": 0.1985856592655182,
+      "learning_rate": 0.001,
+      "loss": 1.1028,
+      "step": 19530
+    },
+    {
+      "epoch": 0.6231066041646737,
+      "grad_norm": 0.2000814974308014,
+      "learning_rate": 0.001,
+      "loss": 1.0807,
+      "step": 19540
+    },
+    {
+      "epoch": 0.6234254918843075,
+      "grad_norm": 0.19329458475112915,
+      "learning_rate": 0.001,
+      "loss": 1.085,
+      "step": 19550
+    },
+    {
+      "epoch": 0.6237443796039415,
+      "grad_norm": 0.20633338391780853,
+      "learning_rate": 0.001,
+      "loss": 1.1083,
+      "step": 19560
+    },
+    {
+      "epoch": 0.6240632673235754,
+      "grad_norm": 0.1958053559064865,
+      "learning_rate": 0.001,
+      "loss": 1.0988,
+      "step": 19570
+    },
+    {
+      "epoch": 0.6243821550432093,
+      "grad_norm": 0.20357364416122437,
+      "learning_rate": 0.001,
+      "loss": 1.0944,
+      "step": 19580
+    },
+    {
+      "epoch": 0.6247010427628432,
+      "grad_norm": 0.19455161690711975,
+      "learning_rate": 0.001,
+      "loss": 1.0881,
+      "step": 19590
+    },
+    {
+      "epoch": 0.6250199304824772,
+      "grad_norm": 0.19670099020004272,
+      "learning_rate": 0.001,
+      "loss": 1.0812,
+      "step": 19600
+    },
+    {
+      "epoch": 0.625338818202111,
+      "grad_norm": 0.19765827059745789,
+      "learning_rate": 0.001,
+      "loss": 1.0938,
+      "step": 19610
+    },
+    {
+      "epoch": 0.625657705921745,
+      "grad_norm": 0.19898134469985962,
+      "learning_rate": 0.001,
+      "loss": 1.1085,
+      "step": 19620
+    },
+    {
+      "epoch": 0.6259765936413789,
+      "grad_norm": 0.19776412844657898,
+      "learning_rate": 0.001,
+      "loss": 1.0797,
+      "step": 19630
+    },
+    {
+      "epoch": 0.6262954813610128,
+      "grad_norm": 0.20222464203834534,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 19640
+    },
+    {
+      "epoch": 0.6266143690806467,
+      "grad_norm": 0.20248456299304962,
+      "learning_rate": 0.001,
+      "loss": 1.0911,
+      "step": 19650
+    },
+    {
+      "epoch": 0.6269332568002807,
+      "grad_norm": 0.1994301676750183,
+      "learning_rate": 0.001,
+      "loss": 1.0852,
+      "step": 19660
+    },
+    {
+      "epoch": 0.6272521445199145,
+      "grad_norm": 0.19714325666427612,
+      "learning_rate": 0.001,
+      "loss": 1.0835,
+      "step": 19670
+    },
+    {
+      "epoch": 0.6275710322395485,
+      "grad_norm": 0.19362570345401764,
+      "learning_rate": 0.001,
+      "loss": 1.0993,
+      "step": 19680
+    },
+    {
+      "epoch": 0.6278899199591824,
+      "grad_norm": 0.19453924894332886,
+      "learning_rate": 0.001,
+      "loss": 1.0883,
+      "step": 19690
+    },
+    {
+      "epoch": 0.6282088076788163,
+      "grad_norm": 0.19855844974517822,
+      "learning_rate": 0.001,
+      "loss": 1.086,
+      "step": 19700
+    },
+    {
+      "epoch": 0.6285276953984502,
+      "grad_norm": 0.19160586595535278,
+      "learning_rate": 0.001,
+      "loss": 1.0713,
+      "step": 19710
+    },
+    {
+      "epoch": 0.6288465831180842,
+      "grad_norm": 0.1902851015329361,
+      "learning_rate": 0.001,
+      "loss": 1.0884,
+      "step": 19720
+    },
+    {
+      "epoch": 0.629165470837718,
+      "grad_norm": 0.1926606446504593,
+      "learning_rate": 0.001,
+      "loss": 1.0797,
+      "step": 19730
+    },
+    {
+      "epoch": 0.629484358557352,
+      "grad_norm": 0.19630712270736694,
+      "learning_rate": 0.001,
+      "loss": 1.0841,
+      "step": 19740
+    },
+    {
+      "epoch": 0.6298032462769859,
+      "grad_norm": 0.19418714940547943,
+      "learning_rate": 0.001,
+      "loss": 1.0791,
+      "step": 19750
+    },
+    {
+      "epoch": 0.6301221339966198,
+      "grad_norm": 0.19034384191036224,
+      "learning_rate": 0.001,
+      "loss": 1.076,
+      "step": 19760
+    },
+    {
+      "epoch": 0.6304410217162537,
+      "grad_norm": 0.19827674329280853,
+      "learning_rate": 0.001,
+      "loss": 1.0747,
+      "step": 19770
+    },
+    {
+      "epoch": 0.6307599094358877,
+      "grad_norm": 0.1912790685892105,
+      "learning_rate": 0.001,
+      "loss": 1.0814,
+      "step": 19780
+    },
+    {
+      "epoch": 0.6310787971555215,
+      "grad_norm": 0.19014963507652283,
+      "learning_rate": 0.001,
+      "loss": 1.0911,
+      "step": 19790
+    },
+    {
+      "epoch": 0.6313976848751555,
+      "grad_norm": 0.19616587460041046,
+      "learning_rate": 0.001,
+      "loss": 1.0806,
+      "step": 19800
+    },
+    {
+      "epoch": 0.6317165725947894,
+      "grad_norm": 0.18765047192573547,
+      "learning_rate": 0.001,
+      "loss": 1.0855,
+      "step": 19810
+    },
+    {
+      "epoch": 0.6320354603144233,
+      "grad_norm": 0.20021288096904755,
+      "learning_rate": 0.001,
+      "loss": 1.0823,
+      "step": 19820
+    },
+    {
+      "epoch": 0.6323543480340572,
+      "grad_norm": 0.20338086783885956,
+      "learning_rate": 0.001,
+      "loss": 1.087,
+      "step": 19830
+    },
+    {
+      "epoch": 0.6326732357536912,
+      "grad_norm": 0.20663778483867645,
+      "learning_rate": 0.001,
+      "loss": 1.0984,
+      "step": 19840
+    },
+    {
+      "epoch": 0.632992123473325,
+      "grad_norm": 0.18688595294952393,
+      "learning_rate": 0.001,
+      "loss": 1.0799,
+      "step": 19850
+    },
+    {
+      "epoch": 0.633311011192959,
+      "grad_norm": 0.20734763145446777,
+      "learning_rate": 0.001,
+      "loss": 1.0859,
+      "step": 19860
+    },
+    {
+      "epoch": 0.6336298989125929,
+      "grad_norm": 0.19312497973442078,
+      "learning_rate": 0.001,
+      "loss": 1.0801,
+      "step": 19870
+    },
+    {
+      "epoch": 0.6339487866322268,
+      "grad_norm": 0.19226190447807312,
+      "learning_rate": 0.001,
+      "loss": 1.0895,
+      "step": 19880
+    },
+    {
+      "epoch": 0.6342676743518607,
+      "grad_norm": 0.19165466725826263,
+      "learning_rate": 0.001,
+      "loss": 1.0896,
+      "step": 19890
+    },
+    {
+      "epoch": 0.6345865620714947,
+      "grad_norm": 0.19660542905330658,
+      "learning_rate": 0.001,
+      "loss": 1.0864,
+      "step": 19900
+    },
+    {
+      "epoch": 0.6349054497911285,
+      "grad_norm": 0.2044084072113037,
+      "learning_rate": 0.001,
+      "loss": 1.097,
+      "step": 19910
+    },
+    {
+      "epoch": 0.6352243375107625,
+      "grad_norm": 0.20047420263290405,
+      "learning_rate": 0.001,
+      "loss": 1.0935,
+      "step": 19920
+    },
+    {
+      "epoch": 0.6355432252303964,
+      "grad_norm": 0.20268094539642334,
+      "learning_rate": 0.001,
+      "loss": 1.0774,
+      "step": 19930
+    },
+    {
+      "epoch": 0.6358621129500303,
+      "grad_norm": 0.2090209722518921,
+      "learning_rate": 0.001,
+      "loss": 1.0758,
+      "step": 19940
+    },
+    {
+      "epoch": 0.6361810006696642,
+      "grad_norm": 0.19323612749576569,
+      "learning_rate": 0.001,
+      "loss": 1.0709,
+      "step": 19950
+    },
+    {
+      "epoch": 0.6364998883892982,
+      "grad_norm": 0.1982102394104004,
+      "learning_rate": 0.001,
+      "loss": 1.0875,
+      "step": 19960
+    },
+    {
+      "epoch": 0.636818776108932,
+      "grad_norm": 0.20114809274673462,
+      "learning_rate": 0.001,
+      "loss": 1.0995,
+      "step": 19970
+    },
+    {
+      "epoch": 0.637137663828566,
+      "grad_norm": 0.19009973108768463,
+      "learning_rate": 0.001,
+      "loss": 1.1036,
+      "step": 19980
+    },
+    {
+      "epoch": 0.6374565515481999,
+      "grad_norm": 0.19751933217048645,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 19990
+    },
+    {
+      "epoch": 0.6377754392678338,
+      "grad_norm": 0.19855111837387085,
+      "learning_rate": 0.001,
+      "loss": 1.083,
+      "step": 20000
+    },
+    {
+      "epoch": 0.6380943269874677,
+      "grad_norm": 0.18911032378673553,
+      "learning_rate": 0.001,
+      "loss": 1.0974,
+      "step": 20010
+    },
+    {
+      "epoch": 0.6384132147071017,
+      "grad_norm": 0.19520005583763123,
+      "learning_rate": 0.001,
+      "loss": 1.0957,
+      "step": 20020
+    },
+    {
+      "epoch": 0.6387321024267355,
+      "grad_norm": 0.19466760754585266,
+      "learning_rate": 0.001,
+      "loss": 1.1106,
+      "step": 20030
+    },
+    {
+      "epoch": 0.6390509901463695,
+      "grad_norm": 0.18498612940311432,
+      "learning_rate": 0.001,
+      "loss": 1.0813,
+      "step": 20040
+    },
+    {
+      "epoch": 0.6393698778660034,
+      "grad_norm": 0.19077274203300476,
+      "learning_rate": 0.001,
+      "loss": 1.0773,
+      "step": 20050
+    },
+    {
+      "epoch": 0.6396887655856373,
+      "grad_norm": 0.19003081321716309,
+      "learning_rate": 0.001,
+      "loss": 1.0869,
+      "step": 20060
+    },
+    {
+      "epoch": 0.6400076533052712,
+      "grad_norm": 0.19166330993175507,
+      "learning_rate": 0.001,
+      "loss": 1.0862,
+      "step": 20070
+    },
+    {
+      "epoch": 0.6403265410249052,
+      "grad_norm": 0.1930341273546219,
+      "learning_rate": 0.001,
+      "loss": 1.0803,
+      "step": 20080
+    },
+    {
+      "epoch": 0.640645428744539,
+      "grad_norm": 0.18485362827777863,
+      "learning_rate": 0.001,
+      "loss": 1.0921,
+      "step": 20090
+    },
+    {
+      "epoch": 0.640964316464173,
+      "grad_norm": 0.19426341354846954,
+      "learning_rate": 0.001,
+      "loss": 1.0984,
+      "step": 20100
+    },
+    {
+      "epoch": 0.6412832041838069,
+      "grad_norm": 0.19640350341796875,
+      "learning_rate": 0.001,
+      "loss": 1.0716,
+      "step": 20110
+    },
+    {
+      "epoch": 0.6416020919034408,
+      "grad_norm": 0.1945829838514328,
+      "learning_rate": 0.001,
+      "loss": 1.0932,
+      "step": 20120
+    },
+    {
+      "epoch": 0.6419209796230747,
+      "grad_norm": 0.1937423199415207,
+      "learning_rate": 0.001,
+      "loss": 1.0901,
+      "step": 20130
+    },
+    {
+      "epoch": 0.6422398673427087,
+      "grad_norm": 0.19474102556705475,
+      "learning_rate": 0.001,
+      "loss": 1.086,
+      "step": 20140
+    },
+    {
+      "epoch": 0.6425587550623425,
+      "grad_norm": 0.19094392657279968,
+      "learning_rate": 0.001,
+      "loss": 1.0914,
+      "step": 20150
+    },
+    {
+      "epoch": 0.6428776427819765,
+      "grad_norm": 0.18804484605789185,
+      "learning_rate": 0.001,
+      "loss": 1.0939,
+      "step": 20160
+    },
+    {
+      "epoch": 0.6431965305016104,
+      "grad_norm": 0.19421075284481049,
+      "learning_rate": 0.001,
+      "loss": 1.096,
+      "step": 20170
+    },
+    {
+      "epoch": 0.6435154182212443,
+      "grad_norm": 0.19062688946723938,
+      "learning_rate": 0.001,
+      "loss": 1.0921,
+      "step": 20180
+    },
+    {
+      "epoch": 0.6438343059408782,
+      "grad_norm": 0.19020332396030426,
+      "learning_rate": 0.001,
+      "loss": 1.0606,
+      "step": 20190
+    },
+    {
+      "epoch": 0.6441531936605122,
+      "grad_norm": 0.1877630650997162,
+      "learning_rate": 0.001,
+      "loss": 1.0834,
+      "step": 20200
+    },
+    {
+      "epoch": 0.644472081380146,
+      "grad_norm": 0.197085440158844,
+      "learning_rate": 0.001,
+      "loss": 1.1016,
+      "step": 20210
+    },
+    {
+      "epoch": 0.64479096909978,
+      "grad_norm": 0.19835904240608215,
+      "learning_rate": 0.001,
+      "loss": 1.0775,
+      "step": 20220
+    },
+    {
+      "epoch": 0.6451098568194139,
+      "grad_norm": 0.19453178346157074,
+      "learning_rate": 0.001,
+      "loss": 1.0871,
+      "step": 20230
+    },
+    {
+      "epoch": 0.6454287445390477,
+      "grad_norm": 0.19668817520141602,
+      "learning_rate": 0.001,
+      "loss": 1.072,
+      "step": 20240
+    },
+    {
+      "epoch": 0.6457476322586817,
+      "grad_norm": 0.1878117173910141,
+      "learning_rate": 0.001,
+      "loss": 1.0683,
+      "step": 20250
+    },
+    {
+      "epoch": 0.6460665199783157,
+      "grad_norm": 0.20410464704036713,
+      "learning_rate": 0.001,
+      "loss": 1.0824,
+      "step": 20260
+    },
+    {
+      "epoch": 0.6463854076979495,
+      "grad_norm": 0.20036086440086365,
+      "learning_rate": 0.001,
+      "loss": 1.0895,
+      "step": 20270
+    },
+    {
+      "epoch": 0.6467042954175835,
+      "grad_norm": 0.19183671474456787,
+      "learning_rate": 0.001,
+      "loss": 1.1032,
+      "step": 20280
+    },
+    {
+      "epoch": 0.6470231831372174,
+      "grad_norm": 0.19621118903160095,
+      "learning_rate": 0.001,
+      "loss": 1.0708,
+      "step": 20290
+    },
+    {
+      "epoch": 0.6473420708568512,
+      "grad_norm": 0.1952357292175293,
+      "learning_rate": 0.001,
+      "loss": 1.0875,
+      "step": 20300
+    },
+    {
+      "epoch": 0.6476609585764852,
+      "grad_norm": 0.20269818603992462,
+      "learning_rate": 0.001,
+      "loss": 1.0774,
+      "step": 20310
+    },
+    {
+      "epoch": 0.6479798462961192,
+      "grad_norm": 0.19550341367721558,
+      "learning_rate": 0.001,
+      "loss": 1.0756,
+      "step": 20320
+    },
+    {
+      "epoch": 0.6482987340157531,
+      "grad_norm": 0.1954687088727951,
+      "learning_rate": 0.001,
+      "loss": 1.0573,
+      "step": 20330
+    },
+    {
+      "epoch": 0.648617621735387,
+      "grad_norm": 0.20026355981826782,
+      "learning_rate": 0.001,
+      "loss": 1.0903,
+      "step": 20340
+    },
+    {
+      "epoch": 0.6489365094550209,
+      "grad_norm": 0.20052330195903778,
+      "learning_rate": 0.001,
+      "loss": 1.0987,
+      "step": 20350
+    },
+    {
+      "epoch": 0.6492553971746549,
+      "grad_norm": 0.18964596092700958,
+      "learning_rate": 0.001,
+      "loss": 1.0903,
+      "step": 20360
+    },
+    {
+      "epoch": 0.6495742848942887,
+      "grad_norm": 0.1951100081205368,
+      "learning_rate": 0.001,
+      "loss": 1.0782,
+      "step": 20370
+    },
+    {
+      "epoch": 0.6498931726139227,
+      "grad_norm": 0.18879763782024384,
+      "learning_rate": 0.001,
+      "loss": 1.0957,
+      "step": 20380
+    },
+    {
+      "epoch": 0.6502120603335566,
+      "grad_norm": 0.20073334872722626,
+      "learning_rate": 0.0009972136778050526,
+      "loss": 1.059,
+      "step": 20390
+    },
+    {
+      "epoch": 0.6505309480531904,
+      "grad_norm": 0.20074057579040527,
+      "learning_rate": 0.0009930383279382303,
+      "loss": 1.0883,
+      "step": 20400
+    },
+    {
+      "epoch": 0.6508498357728244,
+      "grad_norm": 0.19324301183223724,
+      "learning_rate": 0.000988880460329121,
+      "loss": 1.0649,
+      "step": 20410
+    },
+    {
+      "epoch": 0.6511687234924584,
+      "grad_norm": 0.1927865743637085,
+      "learning_rate": 0.000984740001779228,
+      "loss": 1.0878,
+      "step": 20420
+    },
+    {
+      "epoch": 0.6514876112120922,
+      "grad_norm": 0.19140435755252838,
+      "learning_rate": 0.000980616879396537,
+      "loss": 1.0897,
+      "step": 20430
+    },
+    {
+      "epoch": 0.6518064989317262,
+      "grad_norm": 0.19729351997375488,
+      "learning_rate": 0.0009765110205942349,
+      "loss": 1.1063,
+      "step": 20440
+    },
+    {
+      "epoch": 0.6521253866513601,
+      "grad_norm": 0.19580216705799103,
+      "learning_rate": 0.0009724223530894298,
+      "loss": 1.0841,
+      "step": 20450
+    },
+    {
+      "epoch": 0.652444274370994,
+      "grad_norm": 0.1976882368326187,
+      "learning_rate": 0.00096835080490188,
+      "loss": 1.0777,
+      "step": 20460
+    },
+    {
+      "epoch": 0.6527631620906279,
+      "grad_norm": 0.19940541684627533,
+      "learning_rate": 0.0009642963043527262,
+      "loss": 1.0777,
+      "step": 20470
+    },
+    {
+      "epoch": 0.6530820498102619,
+      "grad_norm": 0.1925906538963318,
+      "learning_rate": 0.0009602587800632295,
+      "loss": 1.05,
+      "step": 20480
+    },
+    {
+      "epoch": 0.6534009375298957,
+      "grad_norm": 0.19239018857479095,
+      "learning_rate": 0.0009562381609535146,
+      "loss": 1.0762,
+      "step": 20490
+    },
+    {
+      "epoch": 0.6537198252495297,
+      "grad_norm": 0.19486668705940247,
+      "learning_rate": 0.0009522343762413196,
+      "loss": 1.0739,
+      "step": 20500
+    },
+    {
+      "epoch": 0.6540387129691636,
+      "grad_norm": 0.19426004588603973,
+      "learning_rate": 0.0009482473554407485,
+      "loss": 1.0741,
+      "step": 20510
+    },
+    {
+      "epoch": 0.6543576006887974,
+      "grad_norm": 0.19421601295471191,
+      "learning_rate": 0.000944277028361031,
+      "loss": 1.0772,
+      "step": 20520
+    },
+    {
+      "epoch": 0.6546764884084314,
+      "grad_norm": 0.19363418221473694,
+      "learning_rate": 0.0009403233251052866,
+      "loss": 1.0623,
+      "step": 20530
+    },
+    {
+      "epoch": 0.6549953761280654,
+      "grad_norm": 0.19872882962226868,
+      "learning_rate": 0.0009363861760692945,
+      "loss": 1.0814,
+      "step": 20540
+    },
+    {
+      "epoch": 0.6553142638476992,
+      "grad_norm": 0.1920090913772583,
+      "learning_rate": 0.0009324655119402678,
+      "loss": 1.0756,
+      "step": 20550
+    },
+    {
+      "epoch": 0.6556331515673332,
+      "grad_norm": 0.19251906871795654,
+      "learning_rate": 0.0009285612636956329,
+      "loss": 1.0625,
+      "step": 20560
+    },
+    {
+      "epoch": 0.6559520392869671,
+      "grad_norm": 0.19410337507724762,
+      "learning_rate": 0.0009246733626018155,
+      "loss": 1.0835,
+      "step": 20570
+    },
+    {
+      "epoch": 0.656270927006601,
+      "grad_norm": 0.19288881123065948,
+      "learning_rate": 0.0009208017402130296,
+      "loss": 1.0844,
+      "step": 20580
+    },
+    {
+      "epoch": 0.6565898147262349,
+      "grad_norm": 0.1908722221851349,
+      "learning_rate": 0.0009169463283700727,
+      "loss": 1.0683,
+      "step": 20590
+    },
+    {
+      "epoch": 0.6569087024458689,
+      "grad_norm": 0.1891368180513382,
+      "learning_rate": 0.0009131070591991262,
+      "loss": 1.0737,
+      "step": 20600
+    },
+    {
+      "epoch": 0.6572275901655027,
+      "grad_norm": 0.18758034706115723,
+      "learning_rate": 0.00090928386511056,
+      "loss": 1.0777,
+      "step": 20610
+    },
+    {
+      "epoch": 0.6575464778851366,
+      "grad_norm": 0.19056786596775055,
+      "learning_rate": 0.0009054766787977433,
+      "loss": 1.0652,
+      "step": 20620
+    },
+    {
+      "epoch": 0.6578653656047706,
+      "grad_norm": 0.19077904522418976,
+      "learning_rate": 0.0009016854332358588,
+      "loss": 1.0512,
+      "step": 20630
+    },
+    {
+      "epoch": 0.6581842533244044,
+      "grad_norm": 0.1994921714067459,
+      "learning_rate": 0.0008979100616807236,
+      "loss": 1.1071,
+      "step": 20640
+    },
+    {
+      "epoch": 0.6585031410440384,
+      "grad_norm": 0.20684292912483215,
+      "learning_rate": 0.0008941504976676136,
+      "loss": 1.0724,
+      "step": 20650
+    },
+    {
+      "epoch": 0.6588220287636724,
+      "grad_norm": 0.18994581699371338,
+      "learning_rate": 0.0008904066750100934,
+      "loss": 1.0524,
+      "step": 20660
+    },
+    {
+      "epoch": 0.6591409164833062,
+      "grad_norm": 0.19218602776527405,
+      "learning_rate": 0.0008866785277988516,
+      "loss": 1.0698,
+      "step": 20670
+    },
+    {
+      "epoch": 0.6594598042029401,
+      "grad_norm": 0.1805793046951294,
+      "learning_rate": 0.00088296599040054,
+      "loss": 1.0691,
+      "step": 20680
+    },
+    {
+      "epoch": 0.6597786919225741,
+      "grad_norm": 0.1912751942873001,
+      "learning_rate": 0.0008792689974566183,
+      "loss": 1.0622,
+      "step": 20690
+    },
+    {
+      "epoch": 0.6600975796422079,
+      "grad_norm": 0.18160144984722137,
+      "learning_rate": 0.0008755874838822034,
+      "loss": 1.0687,
+      "step": 20700
+    },
+    {
+      "epoch": 0.6604164673618419,
+      "grad_norm": 0.19986706972122192,
+      "learning_rate": 0.0008719213848649239,
+      "loss": 1.0675,
+      "step": 20710
+    },
+    {
+      "epoch": 0.6607353550814759,
+      "grad_norm": 0.19376768171787262,
+      "learning_rate": 0.0008682706358637785,
+      "loss": 1.0334,
+      "step": 20720
+    },
+    {
+      "epoch": 0.6610542428011097,
+      "grad_norm": 0.18999677896499634,
+      "learning_rate": 0.0008646351726080005,
+      "loss": 1.0613,
+      "step": 20730
+    },
+    {
+      "epoch": 0.6613731305207436,
+      "grad_norm": 0.19654668867588043,
+      "learning_rate": 0.0008610149310959256,
+      "loss": 1.0636,
+      "step": 20740
+    },
+    {
+      "epoch": 0.6616920182403776,
+      "grad_norm": 0.19226598739624023,
+      "learning_rate": 0.0008574098475938659,
+      "loss": 1.0709,
+      "step": 20750
+    },
+    {
+      "epoch": 0.6620109059600114,
+      "grad_norm": 0.18288561701774597,
+      "learning_rate": 0.000853819858634987,
+      "loss": 1.0513,
+      "step": 20760
+    },
+    {
+      "epoch": 0.6623297936796454,
+      "grad_norm": 0.19555260241031647,
+      "learning_rate": 0.0008502449010181915,
+      "loss": 1.0728,
+      "step": 20770
+    },
+    {
+      "epoch": 0.6626486813992793,
+      "grad_norm": 0.1996360570192337,
+      "learning_rate": 0.0008466849118070059,
+      "loss": 1.0594,
+      "step": 20780
+    },
+    {
+      "epoch": 0.6629675691189132,
+      "grad_norm": 0.19368167221546173,
+      "learning_rate": 0.0008431398283284729,
+      "loss": 1.0615,
+      "step": 20790
+    },
+    {
+      "epoch": 0.6632864568385471,
+      "grad_norm": 0.19357769191265106,
+      "learning_rate": 0.0008396095881720477,
+      "loss": 1.0748,
+      "step": 20800
+    },
+    {
+      "epoch": 0.6636053445581811,
+      "grad_norm": 0.19414184987545013,
+      "learning_rate": 0.0008360941291884995,
+      "loss": 1.0437,
+      "step": 20810
+    },
+    {
+      "epoch": 0.6639242322778149,
+      "grad_norm": 0.1933811902999878,
+      "learning_rate": 0.0008325933894888175,
+      "loss": 1.0687,
+      "step": 20820
+    },
+    {
+      "epoch": 0.6642431199974489,
+      "grad_norm": 0.19190998375415802,
+      "learning_rate": 0.0008291073074431209,
+      "loss": 1.0489,
+      "step": 20830
+    },
+    {
+      "epoch": 0.6645620077170828,
+      "grad_norm": 0.19139911234378815,
+      "learning_rate": 0.0008256358216795744,
+      "loss": 1.0499,
+      "step": 20840
+    },
+    {
+      "epoch": 0.6648808954367167,
+      "grad_norm": 0.18963788449764252,
+      "learning_rate": 0.0008221788710833077,
+      "loss": 1.0551,
+      "step": 20850
+    },
+    {
+      "epoch": 0.6651997831563506,
+      "grad_norm": 0.18441535532474518,
+      "learning_rate": 0.0008187363947953391,
+      "loss": 1.0469,
+      "step": 20860
+    },
+    {
+      "epoch": 0.6655186708759846,
+      "grad_norm": 0.19030874967575073,
+      "learning_rate": 0.0008153083322115049,
+      "loss": 1.0637,
+      "step": 20870
+    },
+    {
+      "epoch": 0.6658375585956184,
+      "grad_norm": 0.19269512593746185,
+      "learning_rate": 0.0008118946229813915,
+      "loss": 1.0512,
+      "step": 20880
+    },
+    {
+      "epoch": 0.6661564463152524,
+      "grad_norm": 0.1916525512933731,
+      "learning_rate": 0.0008084952070072738,
+      "loss": 1.0477,
+      "step": 20890
+    },
+    {
+      "epoch": 0.6664753340348863,
+      "grad_norm": 0.1908433735370636,
+      "learning_rate": 0.0008051100244430569,
+      "loss": 1.0587,
+      "step": 20900
+    },
+    {
+      "epoch": 0.6667942217545202,
+      "grad_norm": 0.19119319319725037,
+      "learning_rate": 0.0008017390156932223,
+      "loss": 1.0501,
+      "step": 20910
+    },
+    {
+      "epoch": 0.6671131094741541,
+      "grad_norm": 0.18825064599514008,
+      "learning_rate": 0.0007983821214117789,
+      "loss": 1.0329,
+      "step": 20920
+    },
+    {
+      "epoch": 0.6674319971937881,
+      "grad_norm": 0.1840713620185852,
+      "learning_rate": 0.0007950392825012183,
+      "loss": 1.052,
+      "step": 20930
+    },
+    {
+      "epoch": 0.6677508849134219,
+      "grad_norm": 0.19763097167015076,
+      "learning_rate": 0.0007917104401114743,
+      "loss": 1.037,
+      "step": 20940
+    },
+    {
+      "epoch": 0.6680697726330559,
+      "grad_norm": 0.19263523817062378,
+      "learning_rate": 0.000788395535638887,
+      "loss": 1.0573,
+      "step": 20950
+    },
+    {
+      "epoch": 0.6683886603526898,
+      "grad_norm": 0.18894965946674347,
+      "learning_rate": 0.0007850945107251707,
+      "loss": 1.0543,
+      "step": 20960
+    },
+    {
+      "epoch": 0.6687075480723237,
+      "grad_norm": 0.18489575386047363,
+      "learning_rate": 0.000781807307256387,
+      "loss": 1.0357,
+      "step": 20970
+    },
+    {
+      "epoch": 0.6690264357919576,
+      "grad_norm": 0.1813621073961258,
+      "learning_rate": 0.0007785338673619216,
+      "loss": 1.041,
+      "step": 20980
+    },
+    {
+      "epoch": 0.6693453235115916,
+      "grad_norm": 0.194569393992424,
+      "learning_rate": 0.0007752741334134652,
+      "loss": 1.0438,
+      "step": 20990
+    },
+    {
+      "epoch": 0.6696642112312254,
+      "grad_norm": 0.18502141535282135,
+      "learning_rate": 0.0007720280480239992,
+      "loss": 1.0595,
+      "step": 21000
+    },
+    {
+      "epoch": 0.6699830989508594,
+      "grad_norm": 0.1821593940258026,
+      "learning_rate": 0.0007687955540467853,
+      "loss": 1.0434,
+      "step": 21010
+    },
+    {
+      "epoch": 0.6703019866704933,
+      "grad_norm": 0.19203370809555054,
+      "learning_rate": 0.0007655765945743598,
+      "loss": 1.0438,
+      "step": 21020
+    },
+    {
+      "epoch": 0.6706208743901272,
+      "grad_norm": 0.20178502798080444,
+      "learning_rate": 0.0007623711129375314,
+      "loss": 1.047,
+      "step": 21030
+    },
+    {
+      "epoch": 0.6709397621097611,
+      "grad_norm": 0.1864270716905594,
+      "learning_rate": 0.0007591790527043832,
+      "loss": 1.0433,
+      "step": 21040
+    },
+    {
+      "epoch": 0.6712586498293951,
+      "grad_norm": 0.18299788236618042,
+      "learning_rate": 0.0007560003576792802,
+      "loss": 1.0466,
+      "step": 21050
+    },
+    {
+      "epoch": 0.6715775375490289,
+      "grad_norm": 0.19573181867599487,
+      "learning_rate": 0.0007528349719018794,
+      "loss": 1.0456,
+      "step": 21060
+    },
+    {
+      "epoch": 0.6718964252686629,
+      "grad_norm": 0.19178634881973267,
+      "learning_rate": 0.0007496828396461442,
+      "loss": 1.0435,
+      "step": 21070
+    },
+    {
+      "epoch": 0.6722153129882968,
+      "grad_norm": 0.1954621821641922,
+      "learning_rate": 0.0007465439054193641,
+      "loss": 1.0665,
+      "step": 21080
+    },
+    {
+      "epoch": 0.6725342007079307,
+      "grad_norm": 0.18990260362625122,
+      "learning_rate": 0.0007434181139611777,
+      "loss": 1.0392,
+      "step": 21090
+    },
+    {
+      "epoch": 0.6728530884275646,
+      "grad_norm": 0.18366816639900208,
+      "learning_rate": 0.0007403054102425991,
+      "loss": 1.0124,
+      "step": 21100
+    },
+    {
+      "epoch": 0.6731719761471986,
+      "grad_norm": 0.1891552358865738,
+      "learning_rate": 0.0007372057394650503,
+      "loss": 1.0624,
+      "step": 21110
+    },
+    {
+      "epoch": 0.6734908638668324,
+      "grad_norm": 0.19928067922592163,
+      "learning_rate": 0.0007341190470593954,
+      "loss": 1.041,
+      "step": 21120
+    },
+    {
+      "epoch": 0.6738097515864664,
+      "grad_norm": 0.18672668933868408,
+      "learning_rate": 0.0007310452786849806,
+      "loss": 1.0245,
+      "step": 21130
+    },
+    {
+      "epoch": 0.6741286393061003,
+      "grad_norm": 0.19206221401691437,
+      "learning_rate": 0.0007279843802286769,
+      "loss": 1.0369,
+      "step": 21140
+    },
+    {
+      "epoch": 0.6744475270257342,
+      "grad_norm": 0.19227498769760132,
+      "learning_rate": 0.0007249362978039282,
+      "loss": 1.0333,
+      "step": 21150
+    },
+    {
+      "epoch": 0.6747664147453681,
+      "grad_norm": 0.18796871602535248,
+      "learning_rate": 0.0007219009777498024,
+      "loss": 1.0591,
+      "step": 21160
+    },
+    {
+      "epoch": 0.6750853024650021,
+      "grad_norm": 0.1877882033586502,
+      "learning_rate": 0.0007188783666300463,
+      "loss": 1.0188,
+      "step": 21170
+    },
+    {
+      "epoch": 0.675404190184636,
+      "grad_norm": 0.192293182015419,
+      "learning_rate": 0.000715868411232145,
+      "loss": 1.0582,
+      "step": 21180
+    },
+    {
+      "epoch": 0.6757230779042699,
+      "grad_norm": 0.19534669816493988,
+      "learning_rate": 0.0007128710585663859,
+      "loss": 1.0109,
+      "step": 21190
+    },
+    {
+      "epoch": 0.6760419656239038,
+      "grad_norm": 0.17991122603416443,
+      "learning_rate": 0.0007098862558649246,
+      "loss": 1.0201,
+      "step": 21200
+    },
+    {
+      "epoch": 0.6763608533435378,
+      "grad_norm": 0.18249261379241943,
+      "learning_rate": 0.000706913950580857,
+      "loss": 1.0056,
+      "step": 21210
+    },
+    {
+      "epoch": 0.6766797410631716,
+      "grad_norm": 0.18837027251720428,
+      "learning_rate": 0.000703954090387293,
+      "loss": 1.0296,
+      "step": 21220
+    },
+    {
+      "epoch": 0.6769986287828056,
+      "grad_norm": 0.18746830523014069,
+      "learning_rate": 0.0007010066231764369,
+      "loss": 1.0128,
+      "step": 21230
+    },
+    {
+      "epoch": 0.6773175165024395,
+      "grad_norm": 0.1960870623588562,
+      "learning_rate": 0.0006980714970586688,
+      "loss": 1.0259,
+      "step": 21240
+    },
+    {
+      "epoch": 0.6776364042220734,
+      "grad_norm": 0.18670812249183655,
+      "learning_rate": 0.0006951486603616313,
+      "loss": 1.0404,
+      "step": 21250
+    },
+    {
+      "epoch": 0.6779552919417073,
+      "grad_norm": 0.1884147822856903,
+      "learning_rate": 0.0006922380616293202,
+      "loss": 1.0233,
+      "step": 21260
+    },
+    {
+      "epoch": 0.6782741796613413,
+      "grad_norm": 0.20132644474506378,
+      "learning_rate": 0.0006893396496211784,
+      "loss": 1.0356,
+      "step": 21270
+    },
+    {
+      "epoch": 0.6785930673809751,
+      "grad_norm": 0.1896122395992279,
+      "learning_rate": 0.0006864533733111942,
+      "loss": 1.0438,
+      "step": 21280
+    },
+    {
+      "epoch": 0.6789119551006091,
+      "grad_norm": 0.18535298109054565,
+      "learning_rate": 0.0006835791818870018,
+      "loss": 1.0126,
+      "step": 21290
+    },
+    {
+      "epoch": 0.679230842820243,
+      "grad_norm": 0.1904032975435257,
+      "learning_rate": 0.0006807170247489883,
+      "loss": 1.0315,
+      "step": 21300
+    },
+    {
+      "epoch": 0.6795497305398769,
+      "grad_norm": 0.18548151850700378,
+      "learning_rate": 0.0006778668515094021,
+      "loss": 1.0316,
+      "step": 21310
+    },
+    {
+      "epoch": 0.6798686182595108,
+      "grad_norm": 0.19118435680866241,
+      "learning_rate": 0.0006750286119914656,
+      "loss": 1.0208,
+      "step": 21320
+    },
+    {
+      "epoch": 0.6801875059791448,
+      "grad_norm": 0.18484553694725037,
+      "learning_rate": 0.0006722022562284926,
+      "loss": 1.0432,
+      "step": 21330
+    },
+    {
+      "epoch": 0.6805063936987786,
+      "grad_norm": 0.18801674246788025,
+      "learning_rate": 0.000669387734463008,
+      "loss": 1.0095,
+      "step": 21340
+    },
+    {
+      "epoch": 0.6808252814184126,
+      "grad_norm": 0.1849338710308075,
+      "learning_rate": 0.0006665849971458721,
+      "loss": 1.0386,
+      "step": 21350
+    },
+    {
+      "epoch": 0.6811441691380465,
+      "grad_norm": 0.18272876739501953,
+      "learning_rate": 0.0006637939949354081,
+      "loss": 1.0148,
+      "step": 21360
+    },
+    {
+      "epoch": 0.6814630568576804,
+      "grad_norm": 0.19171403348445892,
+      "learning_rate": 0.000661014678696534,
+      "loss": 1.0252,
+      "step": 21370
+    },
+    {
+      "epoch": 0.6817819445773143,
+      "grad_norm": 0.18120139837265015,
+      "learning_rate": 0.0006582469994998967,
+      "loss": 1.0499,
+      "step": 21380
+    },
+    {
+      "epoch": 0.6821008322969483,
+      "grad_norm": 0.19342155754566193,
+      "learning_rate": 0.0006554909086210115,
+      "loss": 1.0257,
+      "step": 21390
+    },
+    {
+      "epoch": 0.6824197200165821,
+      "grad_norm": 0.1867322474718094,
+      "learning_rate": 0.0006527463575394037,
+      "loss": 1.0277,
+      "step": 21400
+    },
+    {
+      "epoch": 0.6827386077362161,
+      "grad_norm": 0.18487964570522308,
+      "learning_rate": 0.0006500132979377546,
+      "loss": 1.0246,
+      "step": 21410
+    },
+    {
+      "epoch": 0.68305749545585,
+      "grad_norm": 0.19047372043132782,
+      "learning_rate": 0.0006472916817010511,
+      "loss": 1.0152,
+      "step": 21420
+    },
+    {
+      "epoch": 0.6833763831754839,
+      "grad_norm": 0.1882789582014084,
+      "learning_rate": 0.0006445814609157381,
+      "loss": 1.016,
+      "step": 21430
+    },
+    {
+      "epoch": 0.6836952708951178,
+      "grad_norm": 0.193642720580101,
+      "learning_rate": 0.0006418825878688756,
+      "loss": 1.0261,
+      "step": 21440
+    },
+    {
+      "epoch": 0.6840141586147518,
+      "grad_norm": 0.19018161296844482,
+      "learning_rate": 0.0006391950150472985,
+      "loss": 1.0106,
+      "step": 21450
+    },
+    {
+      "epoch": 0.6843330463343856,
+      "grad_norm": 0.18343989551067352,
+      "learning_rate": 0.0006365186951367798,
+      "loss": 1.0189,
+      "step": 21460
+    },
+    {
+      "epoch": 0.6846519340540196,
+      "grad_norm": 0.19221456348896027,
+      "learning_rate": 0.0006338535810211983,
+      "loss": 1.0154,
+      "step": 21470
+    },
+    {
+      "epoch": 0.6849708217736535,
+      "grad_norm": 0.1819676160812378,
+      "learning_rate": 0.0006311996257817083,
+      "loss": 1.0163,
+      "step": 21480
+    },
+    {
+      "epoch": 0.6852897094932874,
+      "grad_norm": 0.19188915193080902,
+      "learning_rate": 0.0006285567826959148,
+      "loss": 1.0247,
+      "step": 21490
+    },
+    {
+      "epoch": 0.6856085972129213,
+      "grad_norm": 0.1923832893371582,
+      "learning_rate": 0.0006259250052370493,
+      "loss": 1.0384,
+      "step": 21500
+    },
+    {
+      "epoch": 0.6859274849325553,
+      "grad_norm": 0.18359391391277313,
+      "learning_rate": 0.0006233042470731524,
+      "loss": 1.0117,
+      "step": 21510
+    },
+    {
+      "epoch": 0.6862463726521891,
+      "grad_norm": 0.1798015683889389,
+      "learning_rate": 0.0006206944620662569,
+      "loss": 1.0222,
+      "step": 21520
+    },
+    {
+      "epoch": 0.6865652603718231,
+      "grad_norm": 0.1816495805978775,
+      "learning_rate": 0.0006180956042715764,
+      "loss": 1.01,
+      "step": 21530
+    },
+    {
+      "epoch": 0.686884148091457,
+      "grad_norm": 0.19056637585163116,
+      "learning_rate": 0.0006155076279366958,
+      "loss": 1.0215,
+      "step": 21540
+    },
+    {
+      "epoch": 0.6872030358110909,
+      "grad_norm": 0.18547722697257996,
+      "learning_rate": 0.0006129304875007661,
+      "loss": 1.0136,
+      "step": 21550
+    },
+    {
+      "epoch": 0.6875219235307248,
+      "grad_norm": 0.1922054886817932,
+      "learning_rate": 0.0006103641375937023,
+      "loss": 1.0195,
+      "step": 21560
+    },
+    {
+      "epoch": 0.6878408112503588,
+      "grad_norm": 0.1904924213886261,
+      "learning_rate": 0.0006078085330353851,
+      "loss": 1.0025,
+      "step": 21570
+    },
+    {
+      "epoch": 0.6881596989699926,
+      "grad_norm": 0.18738815188407898,
+      "learning_rate": 0.0006052636288348644,
+      "loss": 0.9913,
+      "step": 21580
+    },
+    {
+      "epoch": 0.6884785866896266,
+      "grad_norm": 0.1942305862903595,
+      "learning_rate": 0.0006027293801895685,
+      "loss": 1.0139,
+      "step": 21590
+    },
+    {
+      "epoch": 0.6887974744092605,
+      "grad_norm": 0.18734295666217804,
+      "learning_rate": 0.0006002057424845144,
+      "loss": 1.0081,
+      "step": 21600
+    },
+    {
+      "epoch": 0.6891163621288944,
+      "grad_norm": 0.18953493237495422,
+      "learning_rate": 0.0005976926712915233,
+      "loss": 1.012,
+      "step": 21610
+    },
+    {
+      "epoch": 0.6894352498485283,
+      "grad_norm": 0.1921025663614273,
+      "learning_rate": 0.0005951901223684372,
+      "loss": 1.0153,
+      "step": 21620
+    },
+    {
+      "epoch": 0.6897541375681623,
+      "grad_norm": 0.18728868663311005,
+      "learning_rate": 0.0005926980516583412,
+      "loss": 0.9991,
+      "step": 21630
+    },
+    {
+      "epoch": 0.6900730252877961,
+      "grad_norm": 0.19061149656772614,
+      "learning_rate": 0.0005902164152887875,
+      "loss": 1.0094,
+      "step": 21640
+    },
+    {
+      "epoch": 0.6903919130074301,
+      "grad_norm": 0.18096871674060822,
+      "learning_rate": 0.0005877451695710226,
+      "loss": 1.0133,
+      "step": 21650
+    },
+    {
+      "epoch": 0.690710800727064,
+      "grad_norm": 0.1817445605993271,
+      "learning_rate": 0.0005852842709992187,
+      "loss": 1.0144,
+      "step": 21660
+    },
+    {
+      "epoch": 0.6910296884466979,
+      "grad_norm": 0.19506876170635223,
+      "learning_rate": 0.0005828336762497074,
+      "loss": 1.0025,
+      "step": 21670
+    },
+    {
+      "epoch": 0.6913485761663318,
+      "grad_norm": 0.1860935389995575,
+      "learning_rate": 0.0005803933421802178,
+      "loss": 1.0164,
+      "step": 21680
+    },
+    {
+      "epoch": 0.6916674638859658,
+      "grad_norm": 0.18761244416236877,
+      "learning_rate": 0.0005779632258291156,
+      "loss": 1.0121,
+      "step": 21690
+    },
+    {
+      "epoch": 0.6919863516055996,
+      "grad_norm": 0.18786673247814178,
+      "learning_rate": 0.0005755432844146483,
+      "loss": 1.0137,
+      "step": 21700
+    },
+    {
+      "epoch": 0.6923052393252336,
+      "grad_norm": 0.1867864727973938,
+      "learning_rate": 0.0005731334753341907,
+      "loss": 1.0047,
+      "step": 21710
+    },
+    {
+      "epoch": 0.6926241270448675,
+      "grad_norm": 0.19096152484416962,
+      "learning_rate": 0.0005707337561634957,
+      "loss": 1.0024,
+      "step": 21720
+    },
+    {
+      "epoch": 0.6929430147645014,
+      "grad_norm": 0.1956232190132141,
+      "learning_rate": 0.0005683440846559473,
+      "loss": 1.016,
+      "step": 21730
+    },
+    {
+      "epoch": 0.6932619024841353,
+      "grad_norm": 0.1859396994113922,
+      "learning_rate": 0.0005659644187418168,
+      "loss": 0.9975,
+      "step": 21740
+    },
+    {
+      "epoch": 0.6935807902037693,
+      "grad_norm": 0.18429817259311676,
+      "learning_rate": 0.0005635947165275219,
+      "loss": 1.0059,
+      "step": 21750
+    },
+    {
+      "epoch": 0.6938996779234031,
+      "grad_norm": 0.18661698698997498,
+      "learning_rate": 0.0005612349362948896,
+      "loss": 1.0114,
+      "step": 21760
+    },
+    {
+      "epoch": 0.6942185656430371,
+      "grad_norm": 0.18431392312049866,
+      "learning_rate": 0.0005588850365004215,
+      "loss": 1.0059,
+      "step": 21770
+    },
+    {
+      "epoch": 0.694537453362671,
+      "grad_norm": 0.19037270545959473,
+      "learning_rate": 0.0005565449757745625,
+      "loss": 1.012,
+      "step": 21780
+    },
+    {
+      "epoch": 0.6948563410823049,
+      "grad_norm": 0.18189027905464172,
+      "learning_rate": 0.0005542147129209725,
+      "loss": 1.0081,
+      "step": 21790
+    },
+    {
+      "epoch": 0.6951752288019388,
+      "grad_norm": 0.18496230244636536,
+      "learning_rate": 0.0005518942069158012,
+      "loss": 1.0098,
+      "step": 21800
+    },
+    {
+      "epoch": 0.6954941165215728,
+      "grad_norm": 0.19114148616790771,
+      "learning_rate": 0.0005495834169069658,
+      "loss": 1.0009,
+      "step": 21810
+    },
+    {
+      "epoch": 0.6958130042412066,
+      "grad_norm": 0.19435721635818481,
+      "learning_rate": 0.0005472823022134319,
+      "loss": 1.013,
+      "step": 21820
+    },
+    {
+      "epoch": 0.6961318919608406,
+      "grad_norm": 0.18771202862262726,
+      "learning_rate": 0.000544990822324497,
+      "loss": 0.9897,
+      "step": 21830
+    },
+    {
+      "epoch": 0.6964507796804745,
+      "grad_norm": 0.18119776248931885,
+      "learning_rate": 0.0005427089368990779,
+      "loss": 0.9922,
+      "step": 21840
+    },
+    {
+      "epoch": 0.6967696674001084,
+      "grad_norm": 0.1919165849685669,
+      "learning_rate": 0.0005404366057649998,
+      "loss": 1.0073,
+      "step": 21850
+    },
+    {
+      "epoch": 0.6970885551197423,
+      "grad_norm": 0.19585365056991577,
+      "learning_rate": 0.00053817378891829,
+      "loss": 0.9996,
+      "step": 21860
+    },
+    {
+      "epoch": 0.6974074428393763,
+      "grad_norm": 0.1853964477777481,
+      "learning_rate": 0.0005359204465224725,
+      "loss": 1.0088,
+      "step": 21870
+    },
+    {
+      "epoch": 0.6977263305590101,
+      "grad_norm": 0.1896389275789261,
+      "learning_rate": 0.0005336765389078676,
+      "loss": 0.9985,
+      "step": 21880
+    },
+    {
+      "epoch": 0.6980452182786441,
+      "grad_norm": 0.18373239040374756,
+      "learning_rate": 0.000531442026570893,
+      "loss": 1.0067,
+      "step": 21890
+    },
+    {
+      "epoch": 0.698364105998278,
+      "grad_norm": 0.18755769729614258,
+      "learning_rate": 0.0005292168701733688,
+      "loss": 1.0027,
+      "step": 21900
+    },
+    {
+      "epoch": 0.6986829937179119,
+      "grad_norm": 0.18125109374523163,
+      "learning_rate": 0.0005270010305418245,
+      "loss": 0.9857,
+      "step": 21910
+    },
+    {
+      "epoch": 0.6990018814375458,
+      "grad_norm": 0.18633505702018738,
+      "learning_rate": 0.0005247944686668097,
+      "loss": 0.9839,
+      "step": 21920
+    },
+    {
+      "epoch": 0.6993207691571798,
+      "grad_norm": 0.19030959904193878,
+      "learning_rate": 0.0005225971457022069,
+      "loss": 0.9982,
+      "step": 21930
+    },
+    {
+      "epoch": 0.6996396568768136,
+      "grad_norm": 0.1889505535364151,
+      "learning_rate": 0.0005204090229645483,
+      "loss": 0.9807,
+      "step": 21940
+    },
+    {
+      "epoch": 0.6999585445964476,
+      "grad_norm": 0.18198321759700775,
+      "learning_rate": 0.0005182300619323341,
+      "loss": 0.99,
+      "step": 21950
+    },
+    {
+      "epoch": 0.7002774323160815,
+      "grad_norm": 0.19034498929977417,
+      "learning_rate": 0.0005160602242453551,
+      "loss": 1.0019,
+      "step": 21960
+    },
+    {
+      "epoch": 0.7005963200357154,
+      "grad_norm": 0.1891014575958252,
+      "learning_rate": 0.0005138994717040161,
+      "loss": 0.9735,
+      "step": 21970
+    },
+    {
+      "epoch": 0.7009152077553493,
+      "grad_norm": 0.18174348771572113,
+      "learning_rate": 0.0005117477662686652,
+      "loss": 0.9999,
+      "step": 21980
+    },
+    {
+      "epoch": 0.7012340954749833,
+      "grad_norm": 0.1839553713798523,
+      "learning_rate": 0.0005096050700589225,
+      "loss": 1.0032,
+      "step": 21990
+    },
+    {
+      "epoch": 0.7015529831946172,
+      "grad_norm": 0.18982721865177155,
+      "learning_rate": 0.0005074713453530141,
+      "loss": 0.991,
+      "step": 22000
+    },
+    {
+      "epoch": 0.7018718709142511,
+      "grad_norm": 0.18580938875675201,
+      "learning_rate": 0.0005053465545871075,
+      "loss": 1.0031,
+      "step": 22010
+    },
+    {
+      "epoch": 0.702190758633885,
+      "grad_norm": 0.19052889943122864,
+      "learning_rate": 0.0005032306603546511,
+      "loss": 0.9878,
+      "step": 22020
+    },
+    {
+      "epoch": 0.702509646353519,
+      "grad_norm": 0.18959671258926392,
+      "learning_rate": 0.0005011236254057146,
+      "loss": 1.0018,
+      "step": 22030
+    },
+    {
+      "epoch": 0.7028285340731528,
+      "grad_norm": 0.18057048320770264,
+      "learning_rate": 0.0004990254126463343,
+      "loss": 0.9908,
+      "step": 22040
+    },
+    {
+      "epoch": 0.7031474217927868,
+      "grad_norm": 0.19141310453414917,
+      "learning_rate": 0.0004969359851378588,
+      "loss": 0.9963,
+      "step": 22050
+    },
+    {
+      "epoch": 0.7034663095124207,
+      "grad_norm": 0.18355531990528107,
+      "learning_rate": 0.0004948553060963001,
+      "loss": 0.9711,
+      "step": 22060
+    },
+    {
+      "epoch": 0.7037851972320546,
+      "grad_norm": 0.19848114252090454,
+      "learning_rate": 0.0004927833388916852,
+      "loss": 1.0064,
+      "step": 22070
+    },
+    {
+      "epoch": 0.7041040849516885,
+      "grad_norm": 0.18456107378005981,
+      "learning_rate": 0.0004907200470474113,
+      "loss": 0.993,
+      "step": 22080
+    },
+    {
+      "epoch": 0.7044229726713225,
+      "grad_norm": 0.18483372032642365,
+      "learning_rate": 0.0004886653942396035,
+      "loss": 0.9807,
+      "step": 22090
+    },
+    {
+      "epoch": 0.7047418603909563,
+      "grad_norm": 0.18378837406635284,
+      "learning_rate": 0.00048661934429647597,
+      "loss": 0.9992,
+      "step": 22100
+    },
+    {
+      "epoch": 0.7050607481105903,
+      "grad_norm": 0.1917727291584015,
+      "learning_rate": 0.0004845818611976946,
+      "loss": 0.9897,
+      "step": 22110
+    },
+    {
+      "epoch": 0.7053796358302242,
+      "grad_norm": 0.18209923803806305,
+      "learning_rate": 0.0004825529090737429,
+      "loss": 0.9792,
+      "step": 22120
+    },
+    {
+      "epoch": 0.7056985235498581,
+      "grad_norm": 0.17974483966827393,
+      "learning_rate": 0.0004805324522052906,
+      "loss": 0.9956,
+      "step": 22130
+    },
+    {
+      "epoch": 0.706017411269492,
+      "grad_norm": 0.18029539287090302,
+      "learning_rate": 0.000478520455022565,
+      "loss": 0.9818,
+      "step": 22140
+    },
+    {
+      "epoch": 0.706336298989126,
+      "grad_norm": 0.182785302400589,
+      "learning_rate": 0.0004765168821047247,
+      "loss": 0.9831,
+      "step": 22150
+    },
+    {
+      "epoch": 0.7066551867087598,
+      "grad_norm": 0.1831427663564682,
+      "learning_rate": 0.0004745216981792355,
+      "loss": 0.9999,
+      "step": 22160
+    },
+    {
+      "epoch": 0.7069740744283938,
+      "grad_norm": 0.18406899273395538,
+      "learning_rate": 0.00047253486812125044,
+      "loss": 0.9654,
+      "step": 22170
+    },
+    {
+      "epoch": 0.7072929621480277,
+      "grad_norm": 0.19103997945785522,
+      "learning_rate": 0.0004705563569529904,
+      "loss": 0.9865,
+      "step": 22180
+    },
+    {
+      "epoch": 0.7076118498676616,
+      "grad_norm": 0.1879732608795166,
+      "learning_rate": 0.00046858612984312904,
+      "loss": 0.9963,
+      "step": 22190
+    },
+    {
+      "epoch": 0.7079307375872955,
+      "grad_norm": 0.18674372136592865,
+      "learning_rate": 0.00046662415210617933,
+      "loss": 0.9823,
+      "step": 22200
+    },
+    {
+      "epoch": 0.7082496253069295,
+      "grad_norm": 0.18366439640522003,
+      "learning_rate": 0.00046467038920188283,
+      "loss": 0.9865,
+      "step": 22210
+    },
+    {
+      "epoch": 0.7085685130265633,
+      "grad_norm": 0.18931563198566437,
+      "learning_rate": 0.0004627248067346017,
+      "loss": 0.98,
+      "step": 22220
+    },
+    {
+      "epoch": 0.7088874007461973,
+      "grad_norm": 0.18589584529399872,
+      "learning_rate": 0.0004607873704527135,
+      "loss": 1.0023,
+      "step": 22230
+    },
+    {
+      "epoch": 0.7092062884658312,
+      "grad_norm": 0.19276930391788483,
+      "learning_rate": 0.00045885804624800757,
+      "loss": 1.0027,
+      "step": 22240
+    },
+    {
+      "epoch": 0.7095251761854651,
+      "grad_norm": 0.18109150230884552,
+      "learning_rate": 0.000456936800155085,
+      "loss": 0.9737,
+      "step": 22250
+    },
+    {
+      "epoch": 0.709844063905099,
+      "grad_norm": 0.1827128678560257,
+      "learning_rate": 0.00045502359835076047,
+      "loss": 0.9768,
+      "step": 22260
+    },
+    {
+      "epoch": 0.710162951624733,
+      "grad_norm": 0.18957406282424927,
+      "learning_rate": 0.00045311840715346694,
+      "loss": 0.9796,
+      "step": 22270
+    },
+    {
+      "epoch": 0.7104818393443668,
+      "grad_norm": 0.1889813393354416,
+      "learning_rate": 0.0004512211930226627,
+      "loss": 0.9838,
+      "step": 22280
+    },
+    {
+      "epoch": 0.7108007270640008,
+      "grad_norm": 0.18386803567409515,
+      "learning_rate": 0.0004493319225582409,
+      "loss": 0.9903,
+      "step": 22290
+    },
+    {
+      "epoch": 0.7111196147836347,
+      "grad_norm": 0.18842735886573792,
+      "learning_rate": 0.00044745056249994127,
+      "loss": 0.9888,
+      "step": 22300
+    },
+    {
+      "epoch": 0.7114385025032686,
+      "grad_norm": 0.18693885207176208,
+      "learning_rate": 0.00044557707972676475,
+      "loss": 0.9828,
+      "step": 22310
+    },
+    {
+      "epoch": 0.7117573902229025,
+      "grad_norm": 0.1820717453956604,
+      "learning_rate": 0.0004437114412563908,
+      "loss": 0.9856,
+      "step": 22320
+    },
+    {
+      "epoch": 0.7120762779425365,
+      "grad_norm": 0.18761351704597473,
+      "learning_rate": 0.0004418536142445961,
+      "loss": 0.9836,
+      "step": 22330
+    },
+    {
+      "epoch": 0.7123951656621703,
+      "grad_norm": 0.1892370730638504,
+      "learning_rate": 0.0004400035659846766,
+      "loss": 0.9766,
+      "step": 22340
+    },
+    {
+      "epoch": 0.7127140533818043,
+      "grad_norm": 0.18151234090328217,
+      "learning_rate": 0.00043816126390687195,
+      "loss": 0.9651,
+      "step": 22350
+    },
+    {
+      "epoch": 0.7130329411014382,
+      "grad_norm": 0.1909625083208084,
+      "learning_rate": 0.0004363266755777918,
+      "loss": 0.9808,
+      "step": 22360
+    },
+    {
+      "epoch": 0.7133518288210721,
+      "grad_norm": 0.17754018306732178,
+      "learning_rate": 0.00043449976869984496,
+      "loss": 0.9813,
+      "step": 22370
+    },
+    {
+      "epoch": 0.713670716540706,
+      "grad_norm": 0.1815163642168045,
+      "learning_rate": 0.00043268051111067067,
+      "loss": 0.9625,
+      "step": 22380
+    },
+    {
+      "epoch": 0.71398960426034,
+      "grad_norm": 0.18049179017543793,
+      "learning_rate": 0.00043086887078257267,
+      "loss": 0.9729,
+      "step": 22390
+    },
+    {
+      "epoch": 0.7143084919799738,
+      "grad_norm": 0.18532010912895203,
+      "learning_rate": 0.00042906481582195513,
+      "loss": 0.9576,
+      "step": 22400
+    },
+    {
+      "epoch": 0.7146273796996078,
+      "grad_norm": 0.18741180002689362,
+      "learning_rate": 0.0004272683144687611,
+      "loss": 0.9909,
+      "step": 22410
+    },
+    {
+      "epoch": 0.7149462674192417,
+      "grad_norm": 0.18309523165225983,
+      "learning_rate": 0.0004254793350959138,
+      "loss": 0.9858,
+      "step": 22420
+    },
+    {
+      "epoch": 0.7152651551388756,
+      "grad_norm": 0.18443326652050018,
+      "learning_rate": 0.00042369784620875905,
+      "loss": 0.9638,
+      "step": 22430
+    },
+    {
+      "epoch": 0.7155840428585095,
+      "grad_norm": 0.18828490376472473,
+      "learning_rate": 0.00042192381644451176,
+      "loss": 0.9697,
+      "step": 22440
+    },
+    {
+      "epoch": 0.7159029305781435,
+      "grad_norm": 0.1892024129629135,
+      "learning_rate": 0.0004201572145717032,
+      "loss": 0.9741,
+      "step": 22450
+    },
+    {
+      "epoch": 0.7162218182977773,
+      "grad_norm": 0.18395423889160156,
+      "learning_rate": 0.0004183980094896312,
+      "loss": 0.9832,
+      "step": 22460
+    },
+    {
+      "epoch": 0.7165407060174113,
+      "grad_norm": 0.18173611164093018,
+      "learning_rate": 0.0004166461702278128,
+      "loss": 0.9703,
+      "step": 22470
+    },
+    {
+      "epoch": 0.7168595937370452,
+      "grad_norm": 0.1873852014541626,
+      "learning_rate": 0.0004149016659454389,
+      "loss": 0.9676,
+      "step": 22480
+    },
+    {
+      "epoch": 0.7171784814566791,
+      "grad_norm": 0.1870831847190857,
+      "learning_rate": 0.00041316446593083145,
+      "loss": 0.9739,
+      "step": 22490
+    },
+    {
+      "epoch": 0.717497369176313,
+      "grad_norm": 0.1896754801273346,
+      "learning_rate": 0.00041143453960090277,
+      "loss": 0.9775,
+      "step": 22500
+    },
+    {
+      "epoch": 0.717816256895947,
+      "grad_norm": 0.18270163238048553,
+      "learning_rate": 0.0004097118565006169,
+      "loss": 0.9752,
+      "step": 22510
+    },
+    {
+      "epoch": 0.7181351446155808,
+      "grad_norm": 0.1867200881242752,
+      "learning_rate": 0.00040799638630245356,
+      "loss": 0.9648,
+      "step": 22520
+    },
+    {
+      "epoch": 0.7184540323352148,
+      "grad_norm": 0.19233950972557068,
+      "learning_rate": 0.0004062880988058746,
+      "loss": 0.9662,
+      "step": 22530
+    },
+    {
+      "epoch": 0.7187729200548487,
+      "grad_norm": 0.18478399515151978,
+      "learning_rate": 0.0004045869639367917,
+      "loss": 0.9691,
+      "step": 22540
+    },
+    {
+      "epoch": 0.7190918077744826,
+      "grad_norm": 0.18730990588665009,
+      "learning_rate": 0.0004028929517470373,
+      "loss": 0.9673,
+      "step": 22550
+    },
+    {
+      "epoch": 0.7194106954941165,
+      "grad_norm": 0.18217439949512482,
+      "learning_rate": 0.00040120603241383754,
+      "loss": 0.965,
+      "step": 22560
+    },
+    {
+      "epoch": 0.7197295832137505,
+      "grad_norm": 0.18644104897975922,
+      "learning_rate": 0.0003995261762392866,
+      "loss": 0.9701,
+      "step": 22570
+    },
+    {
+      "epoch": 0.7200484709333843,
+      "grad_norm": 0.1922314167022705,
+      "learning_rate": 0.0003978533536498247,
+      "loss": 0.9698,
+      "step": 22580
+    },
+    {
+      "epoch": 0.7203673586530183,
+      "grad_norm": 0.18881219625473022,
+      "learning_rate": 0.0003961875351957167,
+      "loss": 0.9628,
+      "step": 22590
+    },
+    {
+      "epoch": 0.7206862463726522,
+      "grad_norm": 0.18104930222034454,
+      "learning_rate": 0.000394528691550534,
+      "loss": 0.967,
+      "step": 22600
+    },
+    {
+      "epoch": 0.7210051340922861,
+      "grad_norm": 0.1838918924331665,
+      "learning_rate": 0.0003928767935106386,
+      "loss": 0.9563,
+      "step": 22610
+    },
+    {
+      "epoch": 0.72132402181192,
+      "grad_norm": 0.19298873841762543,
+      "learning_rate": 0.0003912318119946682,
+      "loss": 0.9705,
+      "step": 22620
+    },
+    {
+      "epoch": 0.721642909531554,
+      "grad_norm": 0.1953434944152832,
+      "learning_rate": 0.0003895937180430247,
+      "loss": 0.9669,
+      "step": 22630
+    },
+    {
+      "epoch": 0.7219617972511878,
+      "grad_norm": 0.18364940583705902,
+      "learning_rate": 0.0003879624828173645,
+      "loss": 0.9552,
+      "step": 22640
+    },
+    {
+      "epoch": 0.7222806849708218,
+      "grad_norm": 0.1896495521068573,
+      "learning_rate": 0.0003863380776000905,
+      "loss": 0.97,
+      "step": 22650
+    },
+    {
+      "epoch": 0.7225995726904557,
+      "grad_norm": 0.19229236245155334,
+      "learning_rate": 0.0003847204737938466,
+      "loss": 0.9617,
+      "step": 22660
+    },
+    {
+      "epoch": 0.7229184604100896,
+      "grad_norm": 0.1945151388645172,
+      "learning_rate": 0.0003831096429210144,
+      "loss": 0.9585,
+      "step": 22670
+    },
+    {
+      "epoch": 0.7232373481297235,
+      "grad_norm": 0.18564851582050323,
+      "learning_rate": 0.00038150555662321163,
+      "loss": 0.9655,
+      "step": 22680
+    },
+    {
+      "epoch": 0.7235562358493575,
+      "grad_norm": 0.1905517876148224,
+      "learning_rate": 0.0003799081866607931,
+      "loss": 0.976,
+      "step": 22690
+    },
+    {
+      "epoch": 0.7238751235689913,
+      "grad_norm": 0.18559037148952484,
+      "learning_rate": 0.00037831750491235344,
+      "loss": 0.9715,
+      "step": 22700
+    },
+    {
+      "epoch": 0.7241940112886253,
+      "grad_norm": 0.18964865803718567,
+      "learning_rate": 0.0003767334833742322,
+      "loss": 0.9654,
+      "step": 22710
+    },
+    {
+      "epoch": 0.7245128990082592,
+      "grad_norm": 0.1926131397485733,
+      "learning_rate": 0.0003751560941600205,
+      "loss": 0.9556,
+      "step": 22720
+    },
+    {
+      "epoch": 0.7248317867278931,
+      "grad_norm": 0.18436503410339355,
+      "learning_rate": 0.0003735853095000704,
+      "loss": 0.9575,
+      "step": 22730
+    },
+    {
+      "epoch": 0.725150674447527,
+      "grad_norm": 0.1904323697090149,
+      "learning_rate": 0.0003720211017410059,
+      "loss": 0.9488,
+      "step": 22740
+    },
+    {
+      "epoch": 0.725469562167161,
+      "grad_norm": 0.1889037936925888,
+      "learning_rate": 0.0003704634433452362,
+      "loss": 0.9564,
+      "step": 22750
+    },
+    {
+      "epoch": 0.7257884498867948,
+      "grad_norm": 0.1985059231519699,
+      "learning_rate": 0.0003689123068904708,
+      "loss": 0.9664,
+      "step": 22760
+    },
+    {
+      "epoch": 0.7261073376064288,
+      "grad_norm": 0.196369007229805,
+      "learning_rate": 0.00036736766506923683,
+      "loss": 0.9646,
+      "step": 22770
+    },
+    {
+      "epoch": 0.7264262253260627,
+      "grad_norm": 0.18976521492004395,
+      "learning_rate": 0.00036582949068839814,
+      "loss": 0.9542,
+      "step": 22780
+    },
+    {
+      "epoch": 0.7267451130456966,
+      "grad_norm": 0.19328856468200684,
+      "learning_rate": 0.0003642977566686768,
+      "loss": 0.9763,
+      "step": 22790
+    },
+    {
+      "epoch": 0.7270640007653305,
+      "grad_norm": 0.1831946074962616,
+      "learning_rate": 0.00036277243604417616,
+      "loss": 0.9544,
+      "step": 22800
+    },
+    {
+      "epoch": 0.7273828884849645,
+      "grad_norm": 0.18955983221530914,
+      "learning_rate": 0.00036125350196190614,
+      "loss": 0.9699,
+      "step": 22810
+    },
+    {
+      "epoch": 0.7277017762045984,
+      "grad_norm": 0.19156791269779205,
+      "learning_rate": 0.0003597409276813109,
+      "loss": 0.9747,
+      "step": 22820
+    },
+    {
+      "epoch": 0.7280206639242323,
+      "grad_norm": 0.1873570680618286,
+      "learning_rate": 0.0003582346865737974,
+      "loss": 0.9647,
+      "step": 22830
+    },
+    {
+      "epoch": 0.7283395516438662,
+      "grad_norm": 0.1930588185787201,
+      "learning_rate": 0.0003567347521222671,
+      "loss": 0.9737,
+      "step": 22840
+    },
+    {
+      "epoch": 0.7286584393635002,
+      "grad_norm": 0.1854488104581833,
+      "learning_rate": 0.00035524109792064903,
+      "loss": 0.9753,
+      "step": 22850
+    },
+    {
+      "epoch": 0.728977327083134,
+      "grad_norm": 0.18703170120716095,
+      "learning_rate": 0.00035375369767343465,
+      "loss": 0.9513,
+      "step": 22860
+    },
+    {
+      "epoch": 0.729296214802768,
+      "grad_norm": 0.1863420605659485,
+      "learning_rate": 0.0003522725251952154,
+      "loss": 0.9618,
+      "step": 22870
+    },
+    {
+      "epoch": 0.7296151025224019,
+      "grad_norm": 0.1924019306898117,
+      "learning_rate": 0.0003507975544102212,
+      "loss": 0.9453,
+      "step": 22880
+    },
+    {
+      "epoch": 0.7299339902420358,
+      "grad_norm": 0.19052354991436005,
+      "learning_rate": 0.000349328759351862,
+      "loss": 0.9561,
+      "step": 22890
+    },
+    {
+      "epoch": 0.7302528779616697,
+      "grad_norm": 0.18800488114356995,
+      "learning_rate": 0.00034786611416226987,
+      "loss": 0.9726,
+      "step": 22900
+    },
+    {
+      "epoch": 0.7305717656813037,
+      "grad_norm": 0.18403072655200958,
+      "learning_rate": 0.0003464095930918445,
+      "loss": 0.9536,
+      "step": 22910
+    },
+    {
+      "epoch": 0.7308906534009375,
+      "grad_norm": 0.18697819113731384,
+      "learning_rate": 0.0003449591704987995,
+      "loss": 0.9511,
+      "step": 22920
+    },
+    {
+      "epoch": 0.7312095411205715,
+      "grad_norm": 0.18541944026947021,
+      "learning_rate": 0.0003435148208487109,
+      "loss": 0.9685,
+      "step": 22930
+    },
+    {
+      "epoch": 0.7315284288402054,
+      "grad_norm": 0.1891910880804062,
+      "learning_rate": 0.0003420765187140679,
+      "loss": 0.9524,
+      "step": 22940
+    },
+    {
+      "epoch": 0.7318473165598393,
+      "grad_norm": 0.18466798961162567,
+      "learning_rate": 0.00034064423877382523,
+      "loss": 0.9507,
+      "step": 22950
+    },
+    {
+      "epoch": 0.7321662042794732,
+      "grad_norm": 0.19009530544281006,
+      "learning_rate": 0.000339217955812957,
+      "loss": 0.9511,
+      "step": 22960
+    },
+    {
+      "epoch": 0.7324850919991072,
+      "grad_norm": 0.18500375747680664,
+      "learning_rate": 0.0003377976447220132,
+      "loss": 0.9443,
+      "step": 22970
+    },
+    {
+      "epoch": 0.732803979718741,
+      "grad_norm": 0.18694646656513214,
+      "learning_rate": 0.0003363832804966775,
+      "loss": 0.9492,
+      "step": 22980
+    },
+    {
+      "epoch": 0.733122867438375,
+      "grad_norm": 0.18722723424434662,
+      "learning_rate": 0.00033497483823732686,
+      "loss": 0.9415,
+      "step": 22990
+    },
+    {
+      "epoch": 0.7334417551580089,
+      "grad_norm": 0.19515295326709747,
+      "learning_rate": 0.0003335722931485937,
+      "loss": 0.9615,
+      "step": 23000
+    },
+    {
+      "epoch": 0.7337606428776428,
+      "grad_norm": 0.18494488298892975,
+      "learning_rate": 0.00033217562053892876,
+      "loss": 0.9512,
+      "step": 23010
+    },
+    {
+      "epoch": 0.7340795305972767,
+      "grad_norm": 0.19232329726219177,
+      "learning_rate": 0.00033078479582016675,
+      "loss": 0.9603,
+      "step": 23020
+    },
+    {
+      "epoch": 0.7343984183169107,
+      "grad_norm": 0.18243782222270966,
+      "learning_rate": 0.0003293997945070935,
+      "loss": 0.9707,
+      "step": 23030
+    },
+    {
+      "epoch": 0.7347173060365445,
+      "grad_norm": 0.18182645738124847,
+      "learning_rate": 0.00032802059221701484,
+      "loss": 0.9418,
+      "step": 23040
+    },
+    {
+      "epoch": 0.7350361937561785,
+      "grad_norm": 0.19260235130786896,
+      "learning_rate": 0.00032664716466932733,
+      "loss": 0.9697,
+      "step": 23050
+    },
+    {
+      "epoch": 0.7353550814758124,
+      "grad_norm": 0.18719175457954407,
+      "learning_rate": 0.0003252794876850907,
+      "loss": 0.9499,
+      "step": 23060
+    },
+    {
+      "epoch": 0.7356739691954463,
+      "grad_norm": 0.1868394911289215,
+      "learning_rate": 0.00032391753718660234,
+      "loss": 0.9502,
+      "step": 23070
+    },
+    {
+      "epoch": 0.7359928569150802,
+      "grad_norm": 0.1856638640165329,
+      "learning_rate": 0.00032256128919697346,
+      "loss": 0.9573,
+      "step": 23080
+    },
+    {
+      "epoch": 0.7363117446347142,
+      "grad_norm": 0.18470019102096558,
+      "learning_rate": 0.00032121071983970694,
+      "loss": 0.9338,
+      "step": 23090
+    },
+    {
+      "epoch": 0.736630632354348,
+      "grad_norm": 0.1904626041650772,
+      "learning_rate": 0.0003198658053382767,
+      "loss": 0.9534,
+      "step": 23100
+    },
+    {
+      "epoch": 0.736949520073982,
+      "grad_norm": 0.18753063678741455,
+      "learning_rate": 0.0003185265220157095,
+      "loss": 0.9358,
+      "step": 23110
+    },
+    {
+      "epoch": 0.7372684077936159,
+      "grad_norm": 0.18264374136924744,
+      "learning_rate": 0.00031719284629416804,
+      "loss": 0.9542,
+      "step": 23120
+    },
+    {
+      "epoch": 0.7375872955132498,
+      "grad_norm": 0.18596158921718597,
+      "learning_rate": 0.0003158647546945357,
+      "loss": 0.9445,
+      "step": 23130
+    },
+    {
+      "epoch": 0.7379061832328837,
+      "grad_norm": 0.1909426897764206,
+      "learning_rate": 0.0003145422238360034,
+      "loss": 0.9745,
+      "step": 23140
+    },
+    {
+      "epoch": 0.7382250709525177,
+      "grad_norm": 0.1785246878862381,
+      "learning_rate": 0.0003132252304356578,
+      "loss": 0.9695,
+      "step": 23150
+    },
+    {
+      "epoch": 0.7385439586721515,
+      "grad_norm": 0.18791161477565765,
+      "learning_rate": 0.00031191375130807145,
+      "loss": 0.9535,
+      "step": 23160
+    },
+    {
+      "epoch": 0.7388628463917855,
+      "grad_norm": 0.18513339757919312,
+      "learning_rate": 0.0003106077633648948,
+      "loss": 0.9566,
+      "step": 23170
+    },
+    {
+      "epoch": 0.7391817341114194,
+      "grad_norm": 0.19226591289043427,
+      "learning_rate": 0.0003093072436144496,
+      "loss": 0.947,
+      "step": 23180
+    },
+    {
+      "epoch": 0.7395006218310533,
+      "grad_norm": 0.18903833627700806,
+      "learning_rate": 0.00030801216916132403,
+      "loss": 0.9454,
+      "step": 23190
+    },
+    {
+      "epoch": 0.7398195095506872,
+      "grad_norm": 0.18524321913719177,
+      "learning_rate": 0.00030672251720596967,
+      "loss": 0.9444,
+      "step": 23200
+    },
+    {
+      "epoch": 0.7401383972703212,
+      "grad_norm": 0.1868593543767929,
+      "learning_rate": 0.0003054382650443004,
+      "loss": 0.9544,
+      "step": 23210
+    },
+    {
+      "epoch": 0.740457284989955,
+      "grad_norm": 0.1891583651304245,
+      "learning_rate": 0.0003041593900672922,
+      "loss": 0.9377,
+      "step": 23220
+    },
+    {
+      "epoch": 0.740776172709589,
+      "grad_norm": 0.1848241090774536,
+      "learning_rate": 0.00030288586976058574,
+      "loss": 0.9334,
+      "step": 23230
+    },
+    {
+      "epoch": 0.7410950604292229,
+      "grad_norm": 0.19397152960300446,
+      "learning_rate": 0.00030161768170408935,
+      "loss": 0.9403,
+      "step": 23240
+    },
+    {
+      "epoch": 0.7414139481488567,
+      "grad_norm": 0.19353078305721283,
+      "learning_rate": 0.0003003548035715848,
+      "loss": 0.9424,
+      "step": 23250
+    },
+    {
+      "epoch": 0.7417328358684907,
+      "grad_norm": 0.1882072240114212,
+      "learning_rate": 0.0002990972131303341,
+      "loss": 0.9421,
+      "step": 23260
+    },
+    {
+      "epoch": 0.7420517235881247,
+      "grad_norm": 0.1862560659646988,
+      "learning_rate": 0.0002978448882406881,
+      "loss": 0.9413,
+      "step": 23270
+    },
+    {
+      "epoch": 0.7423706113077585,
+      "grad_norm": 0.19289855659008026,
+      "learning_rate": 0.00029659780685569674,
+      "loss": 0.9371,
+      "step": 23280
+    },
+    {
+      "epoch": 0.7426894990273925,
+      "grad_norm": 0.1900213360786438,
+      "learning_rate": 0.00029535594702072087,
+      "loss": 0.9352,
+      "step": 23290
+    },
+    {
+      "epoch": 0.7430083867470264,
+      "grad_norm": 0.19510945677757263,
+      "learning_rate": 0.0002941192868730457,
+      "loss": 0.9612,
+      "step": 23300
+    },
+    {
+      "epoch": 0.7433272744666602,
+      "grad_norm": 0.186062753200531,
+      "learning_rate": 0.00029288780464149593,
+      "loss": 0.9451,
+      "step": 23310
+    },
+    {
+      "epoch": 0.7436461621862942,
+      "grad_norm": 0.19324566423892975,
+      "learning_rate": 0.00029166147864605294,
+      "loss": 0.9531,
+      "step": 23320
+    },
+    {
+      "epoch": 0.7439650499059282,
+      "grad_norm": 0.1828330159187317,
+      "learning_rate": 0.0002904402872974721,
+      "loss": 0.9505,
+      "step": 23330
+    },
+    {
+      "epoch": 0.744283937625562,
+      "grad_norm": 0.18823856115341187,
+      "learning_rate": 0.00028922420909690367,
+      "loss": 0.9478,
+      "step": 23340
+    },
+    {
+      "epoch": 0.744602825345196,
+      "grad_norm": 0.18443375825881958,
+      "learning_rate": 0.00028801322263551397,
+      "loss": 0.9438,
+      "step": 23350
+    },
+    {
+      "epoch": 0.7449217130648299,
+      "grad_norm": 0.1899966150522232,
+      "learning_rate": 0.0002868073065941083,
+      "loss": 0.9278,
+      "step": 23360
+    },
+    {
+      "epoch": 0.7452406007844637,
+      "grad_norm": 0.1841687709093094,
+      "learning_rate": 0.00028560643974275587,
+      "loss": 0.9396,
+      "step": 23370
+    },
+    {
+      "epoch": 0.7455594885040977,
+      "grad_norm": 0.18998615443706512,
+      "learning_rate": 0.00028441060094041583,
+      "loss": 0.9382,
+      "step": 23380
+    },
+    {
+      "epoch": 0.7458783762237317,
+      "grad_norm": 0.19826436042785645,
+      "learning_rate": 0.0002832197691345653,
+      "loss": 0.9448,
+      "step": 23390
+    },
+    {
+      "epoch": 0.7461972639433655,
+      "grad_norm": 0.19598931074142456,
+      "learning_rate": 0.0002820339233608287,
+      "loss": 0.945,
+      "step": 23400
+    },
+    {
+      "epoch": 0.7465161516629994,
+      "grad_norm": 0.18679118156433105,
+      "learning_rate": 0.00028085304274260857,
+      "loss": 0.9547,
+      "step": 23410
+    },
+    {
+      "epoch": 0.7468350393826334,
+      "grad_norm": 0.1902741640806198,
+      "learning_rate": 0.0002796771064907181,
+      "loss": 0.9444,
+      "step": 23420
+    },
+    {
+      "epoch": 0.7471539271022672,
+      "grad_norm": 0.18102550506591797,
+      "learning_rate": 0.0002785060939030151,
+      "loss": 0.9373,
+      "step": 23430
+    },
+    {
+      "epoch": 0.7474728148219012,
+      "grad_norm": 0.1945812702178955,
+      "learning_rate": 0.0002773399843640378,
+      "loss": 0.9259,
+      "step": 23440
+    },
+    {
+      "epoch": 0.7477917025415352,
+      "grad_norm": 0.1866329461336136,
+      "learning_rate": 0.00027617875734464145,
+      "loss": 0.931,
+      "step": 23450
+    },
+    {
+      "epoch": 0.748110590261169,
+      "grad_norm": 0.19371329247951508,
+      "learning_rate": 0.00027502239240163715,
+      "loss": 0.9546,
+      "step": 23460
+    },
+    {
+      "epoch": 0.748429477980803,
+      "grad_norm": 0.1864960938692093,
+      "learning_rate": 0.00027387086917743224,
+      "loss": 0.9432,
+      "step": 23470
+    },
+    {
+      "epoch": 0.7487483657004369,
+      "grad_norm": 0.1891416311264038,
+      "learning_rate": 0.0002727241673996714,
+      "loss": 0.9329,
+      "step": 23480
+    },
+    {
+      "epoch": 0.7490672534200707,
+      "grad_norm": 0.18572381138801575,
+      "learning_rate": 0.00027158226688088006,
+      "loss": 0.9348,
+      "step": 23490
+    },
+    {
+      "epoch": 0.7493861411397047,
+      "grad_norm": 0.1926565021276474,
+      "learning_rate": 0.0002704451475181089,
+      "loss": 0.9434,
+      "step": 23500
+    },
+    {
+      "epoch": 0.7497050288593387,
+      "grad_norm": 0.1821991503238678,
+      "learning_rate": 0.00026931278929257993,
+      "loss": 0.9317,
+      "step": 23510
+    },
+    {
+      "epoch": 0.7500239165789725,
+      "grad_norm": 0.1843806356191635,
+      "learning_rate": 0.00026818517226933437,
+      "loss": 0.9282,
+      "step": 23520
+    },
+    {
+      "epoch": 0.7503428042986064,
+      "grad_norm": 0.19848383963108063,
+      "learning_rate": 0.00026706227659688107,
+      "loss": 0.94,
+      "step": 23530
+    },
+    {
+      "epoch": 0.7506616920182404,
+      "grad_norm": 0.1843789517879486,
+      "learning_rate": 0.00026594408250684776,
+      "loss": 0.9373,
+      "step": 23540
+    },
+    {
+      "epoch": 0.7509805797378742,
+      "grad_norm": 0.1920584887266159,
+      "learning_rate": 0.00026483057031363234,
+      "loss": 0.9448,
+      "step": 23550
+    },
+    {
+      "epoch": 0.7512994674575082,
+      "grad_norm": 0.1851128190755844,
+      "learning_rate": 0.00026372172041405677,
+      "loss": 0.9338,
+      "step": 23560
+    },
+    {
+      "epoch": 0.7516183551771422,
+      "grad_norm": 0.19818131625652313,
+      "learning_rate": 0.0002626175132870219,
+      "loss": 0.9425,
+      "step": 23570
+    },
+    {
+      "epoch": 0.751937242896776,
+      "grad_norm": 0.18199047446250916,
+      "learning_rate": 0.0002615179294931637,
+      "loss": 0.9331,
+      "step": 23580
+    },
+    {
+      "epoch": 0.75225613061641,
+      "grad_norm": 0.18399380147457123,
+      "learning_rate": 0.000260422949674511,
+      "loss": 0.9283,
+      "step": 23590
+    },
+    {
+      "epoch": 0.7525750183360439,
+      "grad_norm": 0.18700499832630157,
+      "learning_rate": 0.00025933255455414493,
+      "loss": 0.9268,
+      "step": 23600
+    },
+    {
+      "epoch": 0.7528939060556777,
+      "grad_norm": 0.1888848841190338,
+      "learning_rate": 0.0002582467249358593,
+      "loss": 0.9426,
+      "step": 23610
+    },
+    {
+      "epoch": 0.7532127937753117,
+      "grad_norm": 0.18916404247283936,
+      "learning_rate": 0.0002571654417038226,
+      "loss": 0.9448,
+      "step": 23620
+    },
+    {
+      "epoch": 0.7535316814949456,
+      "grad_norm": 0.1902066469192505,
+      "learning_rate": 0.0002560886858222419,
+      "loss": 0.9503,
+      "step": 23630
+    },
+    {
+      "epoch": 0.7538505692145795,
+      "grad_norm": 0.17855492234230042,
+      "learning_rate": 0.0002550164383350272,
+      "loss": 0.9192,
+      "step": 23640
+    },
+    {
+      "epoch": 0.7541694569342134,
+      "grad_norm": 0.1947944611310959,
+      "learning_rate": 0.0002539486803654581,
+      "loss": 0.9391,
+      "step": 23650
+    },
+    {
+      "epoch": 0.7544883446538474,
+      "grad_norm": 0.19682112336158752,
+      "learning_rate": 0.0002528853931158513,
+      "loss": 0.9362,
+      "step": 23660
+    },
+    {
+      "epoch": 0.7548072323734814,
+      "grad_norm": 0.18450482189655304,
+      "learning_rate": 0.0002518265578672297,
+      "loss": 0.9295,
+      "step": 23670
+    },
+    {
+      "epoch": 0.7551261200931152,
+      "grad_norm": 0.1883128583431244,
+      "learning_rate": 0.00025077215597899276,
+      "loss": 0.9278,
+      "step": 23680
+    },
+    {
+      "epoch": 0.7554450078127491,
+      "grad_norm": 0.1992689073085785,
+      "learning_rate": 0.0002497221688885888,
+      "loss": 0.9508,
+      "step": 23690
+    },
+    {
+      "epoch": 0.7557638955323831,
+      "grad_norm": 0.19176770746707916,
+      "learning_rate": 0.0002486765781111873,
+      "loss": 0.9338,
+      "step": 23700
+    },
+    {
+      "epoch": 0.7560827832520169,
+      "grad_norm": 0.18524746596813202,
+      "learning_rate": 0.0002476353652393545,
+      "loss": 0.9386,
+      "step": 23710
+    },
+    {
+      "epoch": 0.7564016709716509,
+      "grad_norm": 0.19836051762104034,
+      "learning_rate": 0.0002465985119427286,
+      "loss": 0.9239,
+      "step": 23720
+    },
+    {
+      "epoch": 0.7567205586912849,
+      "grad_norm": 0.19612550735473633,
+      "learning_rate": 0.0002455659999676974,
+      "loss": 0.9326,
+      "step": 23730
+    },
+    {
+      "epoch": 0.7570394464109187,
+      "grad_norm": 0.18374860286712646,
+      "learning_rate": 0.0002445378111370768,
+      "loss": 0.9327,
+      "step": 23740
+    },
+    {
+      "epoch": 0.7573583341305526,
+      "grad_norm": 0.18519578874111176,
+      "learning_rate": 0.00024351392734979106,
+      "loss": 0.9249,
+      "step": 23750
+    },
+    {
+      "epoch": 0.7576772218501866,
+      "grad_norm": 0.1952565759420395,
+      "learning_rate": 0.00024249433058055368,
+      "loss": 0.9177,
+      "step": 23760
+    },
+    {
+      "epoch": 0.7579961095698204,
+      "grad_norm": 0.1942618042230606,
+      "learning_rate": 0.00024147900287955056,
+      "loss": 0.9411,
+      "step": 23770
+    },
+    {
+      "epoch": 0.7583149972894544,
+      "grad_norm": 0.1848842203617096,
+      "learning_rate": 0.0002404679263721236,
+      "loss": 0.9188,
+      "step": 23780
+    },
+    {
+      "epoch": 0.7586338850090883,
+      "grad_norm": 0.19031748175621033,
+      "learning_rate": 0.00023946108325845628,
+      "loss": 0.9368,
+      "step": 23790
+    },
+    {
+      "epoch": 0.7589527727287222,
+      "grad_norm": 0.19315947592258453,
+      "learning_rate": 0.00023845845581326017,
+      "loss": 0.9527,
+      "step": 23800
+    },
+    {
+      "epoch": 0.7592716604483561,
+      "grad_norm": 0.2014995813369751,
+      "learning_rate": 0.00023746002638546287,
+      "loss": 0.9196,
+      "step": 23810
+    },
+    {
+      "epoch": 0.7595905481679901,
+      "grad_norm": 0.19522787630558014,
+      "learning_rate": 0.00023646577739789735,
+      "loss": 0.9334,
+      "step": 23820
+    },
+    {
+      "epoch": 0.7599094358876239,
+      "grad_norm": 0.18518857657909393,
+      "learning_rate": 0.00023547569134699248,
+      "loss": 0.931,
+      "step": 23830
+    },
+    {
+      "epoch": 0.7602283236072579,
+      "grad_norm": 0.1907796561717987,
+      "learning_rate": 0.00023448975080246477,
+      "loss": 0.9305,
+      "step": 23840
+    },
+    {
+      "epoch": 0.7605472113268918,
+      "grad_norm": 0.18997465074062347,
+      "learning_rate": 0.00023350793840701175,
+      "loss": 0.9316,
+      "step": 23850
+    },
+    {
+      "epoch": 0.7608660990465257,
+      "grad_norm": 0.18743346631526947,
+      "learning_rate": 0.00023253023687600625,
+      "loss": 0.9189,
+      "step": 23860
+    },
+    {
+      "epoch": 0.7611849867661596,
+      "grad_norm": 0.18413014709949493,
+      "learning_rate": 0.00023155662899719193,
+      "loss": 0.9461,
+      "step": 23870
+    },
+    {
+      "epoch": 0.7615038744857936,
+      "grad_norm": 0.1894141137599945,
+      "learning_rate": 0.00023058709763038065,
+      "loss": 0.9311,
+      "step": 23880
+    },
+    {
+      "epoch": 0.7618227622054274,
+      "grad_norm": 0.18807397782802582,
+      "learning_rate": 0.00022962162570715043,
+      "loss": 0.9373,
+      "step": 23890
+    },
+    {
+      "epoch": 0.7621416499250614,
+      "grad_norm": 0.1933257281780243,
+      "learning_rate": 0.00022866019623054502,
+      "loss": 0.9213,
+      "step": 23900
+    },
+    {
+      "epoch": 0.7624605376446953,
+      "grad_norm": 0.19029848277568817,
+      "learning_rate": 0.00022770279227477477,
+      "loss": 0.9353,
+      "step": 23910
+    },
+    {
+      "epoch": 0.7627794253643292,
+      "grad_norm": 0.18332846462726593,
+      "learning_rate": 0.00022674939698491854,
+      "loss": 0.9226,
+      "step": 23920
+    },
+    {
+      "epoch": 0.7630983130839631,
+      "grad_norm": 0.18277508020401,
+      "learning_rate": 0.000225799993576627,
+      "loss": 0.9239,
+      "step": 23930
+    },
+    {
+      "epoch": 0.7634172008035971,
+      "grad_norm": 0.1963907927274704,
+      "learning_rate": 0.0002248545653358273,
+      "loss": 0.9371,
+      "step": 23940
+    },
+    {
+      "epoch": 0.7637360885232309,
+      "grad_norm": 0.18964403867721558,
+      "learning_rate": 0.00022391309561842852,
+      "loss": 0.9302,
+      "step": 23950
+    },
+    {
+      "epoch": 0.7640549762428649,
+      "grad_norm": 0.1915760636329651,
+      "learning_rate": 0.00022297556785002887,
+      "loss": 0.9303,
+      "step": 23960
+    },
+    {
+      "epoch": 0.7643738639624988,
+      "grad_norm": 0.18544916808605194,
+      "learning_rate": 0.0002220419655256239,
+      "loss": 0.9178,
+      "step": 23970
+    },
+    {
+      "epoch": 0.7646927516821327,
+      "grad_norm": 0.1847660392522812,
+      "learning_rate": 0.00022111227220931578,
+      "loss": 0.9324,
+      "step": 23980
+    },
+    {
+      "epoch": 0.7650116394017666,
+      "grad_norm": 0.19324614107608795,
+      "learning_rate": 0.00022018647153402418,
+      "loss": 0.9297,
+      "step": 23990
+    },
+    {
+      "epoch": 0.7653305271214006,
+      "grad_norm": 0.18945874273777008,
+      "learning_rate": 0.00021926454720119792,
+      "loss": 0.9435,
+      "step": 24000
+    },
+    {
+      "epoch": 0.7656494148410344,
+      "grad_norm": 0.18816037476062775,
+      "learning_rate": 0.00021834648298052806,
+      "loss": 0.9106,
+      "step": 24010
+    },
+    {
+      "epoch": 0.7659683025606684,
+      "grad_norm": 0.18085907399654388,
+      "learning_rate": 0.00021743226270966244,
+      "loss": 0.9252,
+      "step": 24020
+    },
+    {
+      "epoch": 0.7662871902803023,
+      "grad_norm": 0.19740988314151764,
+      "learning_rate": 0.00021652187029392068,
+      "loss": 0.941,
+      "step": 24030
+    },
+    {
+      "epoch": 0.7666060779999362,
+      "grad_norm": 0.18533362448215485,
+      "learning_rate": 0.00021561528970601124,
+      "loss": 0.9313,
+      "step": 24040
+    },
+    {
+      "epoch": 0.7669249657195701,
+      "grad_norm": 0.1974269449710846,
+      "learning_rate": 0.00021471250498574909,
+      "loss": 0.9306,
+      "step": 24050
+    },
+    {
+      "epoch": 0.7672438534392041,
+      "grad_norm": 0.1847943365573883,
+      "learning_rate": 0.0002138135002397747,
+      "loss": 0.9329,
+      "step": 24060
+    },
+    {
+      "epoch": 0.7675627411588379,
+      "grad_norm": 0.1897907704114914,
+      "learning_rate": 0.0002129182596412743,
+      "loss": 0.9456,
+      "step": 24070
+    },
+    {
+      "epoch": 0.7678816288784719,
+      "grad_norm": 0.18913713097572327,
+      "learning_rate": 0.00021202676742970136,
+      "loss": 0.9362,
+      "step": 24080
+    },
+    {
+      "epoch": 0.7682005165981058,
+      "grad_norm": 0.18905293941497803,
+      "learning_rate": 0.0002111390079104989,
+      "loss": 0.9248,
+      "step": 24090
+    },
+    {
+      "epoch": 0.7685194043177397,
+      "grad_norm": 0.18642562627792358,
+      "learning_rate": 0.00021025496545482323,
+      "loss": 0.9298,
+      "step": 24100
+    },
+    {
+      "epoch": 0.7688382920373736,
+      "grad_norm": 0.19463413953781128,
+      "learning_rate": 0.0002093746244992691,
+      "loss": 0.9348,
+      "step": 24110
+    },
+    {
+      "epoch": 0.7691571797570076,
+      "grad_norm": 0.18571799993515015,
+      "learning_rate": 0.0002084979695455954,
+      "loss": 0.9225,
+      "step": 24120
+    },
+    {
+      "epoch": 0.7694760674766414,
+      "grad_norm": 0.19087985157966614,
+      "learning_rate": 0.00020762498516045232,
+      "loss": 0.9244,
+      "step": 24130
+    },
+    {
+      "epoch": 0.7697949551962754,
+      "grad_norm": 0.18741409480571747,
+      "learning_rate": 0.00020675565597510982,
+      "loss": 0.9278,
+      "step": 24140
+    },
+    {
+      "epoch": 0.7701138429159093,
+      "grad_norm": 0.18996545672416687,
+      "learning_rate": 0.00020588996668518696,
+      "loss": 0.9291,
+      "step": 24150
+    },
+    {
+      "epoch": 0.7704327306355432,
+      "grad_norm": 0.18844963610172272,
+      "learning_rate": 0.00020502790205038248,
+      "loss": 0.9122,
+      "step": 24160
+    },
+    {
+      "epoch": 0.7707516183551771,
+      "grad_norm": 0.19069145619869232,
+      "learning_rate": 0.00020416944689420654,
+      "loss": 0.9184,
+      "step": 24170
+    },
+    {
+      "epoch": 0.7710705060748111,
+      "grad_norm": 0.19214212894439697,
+      "learning_rate": 0.00020331458610371345,
+      "loss": 0.931,
+      "step": 24180
+    },
+    {
+      "epoch": 0.7713893937944449,
+      "grad_norm": 0.19216571748256683,
+      "learning_rate": 0.00020246330462923582,
+      "loss": 0.9262,
+      "step": 24190
+    },
+    {
+      "epoch": 0.7717082815140789,
+      "grad_norm": 0.18903692066669464,
+      "learning_rate": 0.00020161558748411925,
+      "loss": 0.9149,
+      "step": 24200
+    },
+    {
+      "epoch": 0.7720271692337128,
+      "grad_norm": 0.18422691524028778,
+      "learning_rate": 0.00020077141974445887,
+      "loss": 0.9198,
+      "step": 24210
+    },
+    {
+      "epoch": 0.7723460569533467,
+      "grad_norm": 0.190536767244339,
+      "learning_rate": 0.00019993078654883636,
+      "loss": 0.9135,
+      "step": 24220
+    },
+    {
+      "epoch": 0.7726649446729806,
+      "grad_norm": 0.18851052224636078,
+      "learning_rate": 0.00019909367309805842,
+      "loss": 0.8966,
+      "step": 24230
+    },
+    {
+      "epoch": 0.7729838323926146,
+      "grad_norm": 0.1889980584383011,
+      "learning_rate": 0.0001982600646548962,
+      "loss": 0.9211,
+      "step": 24240
+    },
+    {
+      "epoch": 0.7733027201122484,
+      "grad_norm": 0.1888222098350525,
+      "learning_rate": 0.0001974299465438259,
+      "loss": 0.9144,
+      "step": 24250
+    },
+    {
+      "epoch": 0.7736216078318824,
+      "grad_norm": 0.19497574865818024,
+      "learning_rate": 0.00019660330415077035,
+      "loss": 0.9109,
+      "step": 24260
+    },
+    {
+      "epoch": 0.7739404955515163,
+      "grad_norm": 0.19014975428581238,
+      "learning_rate": 0.00019578012292284172,
+      "loss": 0.9264,
+      "step": 24270
+    },
+    {
+      "epoch": 0.7742593832711502,
+      "grad_norm": 0.19474774599075317,
+      "learning_rate": 0.0001949603883680855,
+      "loss": 0.9307,
+      "step": 24280
+    },
+    {
+      "epoch": 0.7745782709907841,
+      "grad_norm": 0.19689294695854187,
+      "learning_rate": 0.0001941440860552251,
+      "loss": 0.9297,
+      "step": 24290
+    },
+    {
+      "epoch": 0.7748971587104181,
+      "grad_norm": 0.18944482505321503,
+      "learning_rate": 0.00019333120161340792,
+      "loss": 0.9221,
+      "step": 24300
+    },
+    {
+      "epoch": 0.7752160464300519,
+      "grad_norm": 0.18974262475967407,
+      "learning_rate": 0.00019252172073195239,
+      "loss": 0.925,
+      "step": 24310
+    },
+    {
+      "epoch": 0.7755349341496859,
+      "grad_norm": 0.19915218651294708,
+      "learning_rate": 0.00019171562916009603,
+      "loss": 0.9313,
+      "step": 24320
+    },
+    {
+      "epoch": 0.7758538218693198,
+      "grad_norm": 0.1882493942975998,
+      "learning_rate": 0.00019091291270674447,
+      "loss": 0.9154,
+      "step": 24330
+    },
+    {
+      "epoch": 0.7761727095889537,
+      "grad_norm": 0.18934135138988495,
+      "learning_rate": 0.00019011355724022166,
+      "loss": 0.9269,
+      "step": 24340
+    },
+    {
+      "epoch": 0.7764915973085876,
+      "grad_norm": 0.18505986034870148,
+      "learning_rate": 0.0001893175486880212,
+      "loss": 0.9286,
+      "step": 24350
+    },
+    {
+      "epoch": 0.7768104850282216,
+      "grad_norm": 0.187198668718338,
+      "learning_rate": 0.0001885248730365585,
+      "loss": 0.9193,
+      "step": 24360
+    },
+    {
+      "epoch": 0.7771293727478554,
+      "grad_norm": 0.18569988012313843,
+      "learning_rate": 0.00018773551633092397,
+      "loss": 0.9095,
+      "step": 24370
+    },
+    {
+      "epoch": 0.7774482604674894,
+      "grad_norm": 0.1854165643453598,
+      "learning_rate": 0.00018694946467463756,
+      "loss": 0.9035,
+      "step": 24380
+    },
+    {
+      "epoch": 0.7777671481871233,
+      "grad_norm": 0.18419300019741058,
+      "learning_rate": 0.00018616670422940394,
+      "loss": 0.9207,
+      "step": 24390
+    },
+    {
+      "epoch": 0.7780860359067572,
+      "grad_norm": 0.19491253793239594,
+      "learning_rate": 0.00018538722121486895,
+      "loss": 0.9235,
+      "step": 24400
+    },
+    {
+      "epoch": 0.7784049236263911,
+      "grad_norm": 0.18213653564453125,
+      "learning_rate": 0.00018461100190837707,
+      "loss": 0.9231,
+      "step": 24410
+    },
+    {
+      "epoch": 0.7787238113460251,
+      "grad_norm": 0.18353652954101562,
+      "learning_rate": 0.0001838380326447297,
+      "loss": 0.9278,
+      "step": 24420
+    },
+    {
+      "epoch": 0.7790426990656589,
+      "grad_norm": 0.18761824071407318,
+      "learning_rate": 0.00018306829981594458,
+      "loss": 0.9237,
+      "step": 24430
+    },
+    {
+      "epoch": 0.7793615867852929,
+      "grad_norm": 0.1892223060131073,
+      "learning_rate": 0.0001823017898710165,
+      "loss": 0.9195,
+      "step": 24440
+    },
+    {
+      "epoch": 0.7796804745049268,
+      "grad_norm": 0.1860656440258026,
+      "learning_rate": 0.00018153848931567836,
+      "loss": 0.9091,
+      "step": 24450
+    },
+    {
+      "epoch": 0.7799993622245607,
+      "grad_norm": 0.19069086015224457,
+      "learning_rate": 0.00018077838471216377,
+      "loss": 0.9207,
+      "step": 24460
+    },
+    {
+      "epoch": 0.7803182499441946,
+      "grad_norm": 0.18532173335552216,
+      "learning_rate": 0.00018002146267897054,
+      "loss": 0.9138,
+      "step": 24470
+    },
+    {
+      "epoch": 0.7806371376638286,
+      "grad_norm": 0.18655841052532196,
+      "learning_rate": 0.00017926770989062511,
+      "loss": 0.9038,
+      "step": 24480
+    },
+    {
+      "epoch": 0.7809560253834625,
+      "grad_norm": 0.18871451914310455,
+      "learning_rate": 0.0001785171130774477,
+      "loss": 0.9208,
+      "step": 24490
+    },
+    {
+      "epoch": 0.7812749131030964,
+      "grad_norm": 0.1932729184627533,
+      "learning_rate": 0.00017776965902531916,
+      "loss": 0.9235,
+      "step": 24500
+    },
+    {
+      "epoch": 0.7815938008227303,
+      "grad_norm": 0.19129279255867004,
+      "learning_rate": 0.00017702533457544784,
+      "loss": 0.9015,
+      "step": 24510
+    },
+    {
+      "epoch": 0.7819126885423643,
+      "grad_norm": 0.19334463775157928,
+      "learning_rate": 0.00017628412662413823,
+      "loss": 0.9229,
+      "step": 24520
+    },
+    {
+      "epoch": 0.7822315762619981,
+      "grad_norm": 0.18650083243846893,
+      "learning_rate": 0.0001755460221225603,
+      "loss": 0.9277,
+      "step": 24530
+    },
+    {
+      "epoch": 0.7825504639816321,
+      "grad_norm": 0.18478453159332275,
+      "learning_rate": 0.00017481100807651963,
+      "loss": 0.9089,
+      "step": 24540
+    },
+    {
+      "epoch": 0.782869351701266,
+      "grad_norm": 0.19256654381752014,
+      "learning_rate": 0.00017407907154622863,
+      "loss": 0.9176,
+      "step": 24550
+    },
+    {
+      "epoch": 0.7831882394208999,
+      "grad_norm": 0.1951526403427124,
+      "learning_rate": 0.00017335019964607887,
+      "loss": 0.9103,
+      "step": 24560
+    },
+    {
+      "epoch": 0.7835071271405338,
+      "grad_norm": 0.1941751390695572,
+      "learning_rate": 0.00017262437954441417,
+      "loss": 0.9187,
+      "step": 24570
+    },
+    {
+      "epoch": 0.7838260148601678,
+      "grad_norm": 0.1897512674331665,
+      "learning_rate": 0.00017190159846330476,
+      "loss": 0.915,
+      "step": 24580
+    },
+    {
+      "epoch": 0.7841449025798016,
+      "grad_norm": 0.18099430203437805,
+      "learning_rate": 0.00017118184367832215,
+      "loss": 0.92,
+      "step": 24590
+    },
+    {
+      "epoch": 0.7844637902994356,
+      "grad_norm": 0.18526895344257355,
+      "learning_rate": 0.00017046510251831525,
+      "loss": 0.9181,
+      "step": 24600
+    },
+    {
+      "epoch": 0.7847826780190695,
+      "grad_norm": 0.18779991567134857,
+      "learning_rate": 0.0001697513623651875,
+      "loss": 0.9108,
+      "step": 24610
+    },
+    {
+      "epoch": 0.7851015657387034,
+      "grad_norm": 0.1872919648885727,
+      "learning_rate": 0.00016904061065367424,
+      "loss": 0.9005,
+      "step": 24620
+    },
+    {
+      "epoch": 0.7854204534583373,
+      "grad_norm": 0.19210565090179443,
+      "learning_rate": 0.00016833283487112187,
+      "loss": 0.9142,
+      "step": 24630
+    },
+    {
+      "epoch": 0.7857393411779713,
+      "grad_norm": 0.19088362157344818,
+      "learning_rate": 0.00016762802255726757,
+      "loss": 0.9082,
+      "step": 24640
+    },
+    {
+      "epoch": 0.7860582288976051,
+      "grad_norm": 0.19318880140781403,
+      "learning_rate": 0.0001669261613040197,
+      "loss": 0.9098,
+      "step": 24650
+    },
+    {
+      "epoch": 0.7863771166172391,
+      "grad_norm": 0.19324028491973877,
+      "learning_rate": 0.00016622723875523958,
+      "loss": 0.9167,
+      "step": 24660
+    },
+    {
+      "epoch": 0.786696004336873,
+      "grad_norm": 0.18459977209568024,
+      "learning_rate": 0.0001655312426065239,
+      "loss": 0.9013,
+      "step": 24670
+    },
+    {
+      "epoch": 0.7870148920565069,
+      "grad_norm": 0.18851350247859955,
+      "learning_rate": 0.00016483816060498802,
+      "loss": 0.9235,
+      "step": 24680
+    },
+    {
+      "epoch": 0.7873337797761408,
+      "grad_norm": 0.19676366448402405,
+      "learning_rate": 0.00016414798054905036,
+      "loss": 0.9191,
+      "step": 24690
+    },
+    {
+      "epoch": 0.7876526674957748,
+      "grad_norm": 0.18468333780765533,
+      "learning_rate": 0.0001634606902882176,
+      "loss": 0.9007,
+      "step": 24700
+    },
+    {
+      "epoch": 0.7879715552154086,
+      "grad_norm": 0.18907414376735687,
+      "learning_rate": 0.00016277627772287072,
+      "loss": 0.9181,
+      "step": 24710
+    },
+    {
+      "epoch": 0.7882904429350426,
+      "grad_norm": 0.18885044753551483,
+      "learning_rate": 0.00016209473080405187,
+      "loss": 0.8997,
+      "step": 24720
+    },
+    {
+      "epoch": 0.7886093306546765,
+      "grad_norm": 0.18907248973846436,
+      "learning_rate": 0.0001614160375332526,
+      "loss": 0.9155,
+      "step": 24730
+    },
+    {
+      "epoch": 0.7889282183743104,
+      "grad_norm": 0.19450700283050537,
+      "learning_rate": 0.00016074018596220224,
+      "loss": 0.9155,
+      "step": 24740
+    },
+    {
+      "epoch": 0.7892471060939443,
+      "grad_norm": 0.18370944261550903,
+      "learning_rate": 0.00016006716419265783,
+      "loss": 0.9325,
+      "step": 24750
+    },
+    {
+      "epoch": 0.7895659938135783,
+      "grad_norm": 0.18763361871242523,
+      "learning_rate": 0.00015939696037619444,
+      "loss": 0.9066,
+      "step": 24760
+    },
+    {
+      "epoch": 0.7898848815332121,
+      "grad_norm": 0.19110409915447235,
+      "learning_rate": 0.00015872956271399674,
+      "loss": 0.8973,
+      "step": 24770
+    },
+    {
+      "epoch": 0.7902037692528461,
+      "grad_norm": 0.18324126303195953,
+      "learning_rate": 0.00015806495945665133,
+      "loss": 0.9194,
+      "step": 24780
+    },
+    {
+      "epoch": 0.79052265697248,
+      "grad_norm": 0.19040825963020325,
+      "learning_rate": 0.00015740313890393964,
+      "loss": 0.9321,
+      "step": 24790
+    },
+    {
+      "epoch": 0.7908415446921139,
+      "grad_norm": 0.1885160505771637,
+      "learning_rate": 0.00015674408940463216,
+      "loss": 0.9003,
+      "step": 24800
+    },
+    {
+      "epoch": 0.7911604324117478,
+      "grad_norm": 0.1887117624282837,
+      "learning_rate": 0.00015608779935628333,
+      "loss": 0.9166,
+      "step": 24810
+    },
+    {
+      "epoch": 0.7914793201313818,
+      "grad_norm": 0.19734545052051544,
+      "learning_rate": 0.0001554342572050271,
+      "loss": 0.9099,
+      "step": 24820
+    },
+    {
+      "epoch": 0.7917982078510156,
+      "grad_norm": 0.19855238497257233,
+      "learning_rate": 0.00015478345144537376,
+      "loss": 0.9131,
+      "step": 24830
+    },
+    {
+      "epoch": 0.7921170955706496,
+      "grad_norm": 0.1890731155872345,
+      "learning_rate": 0.0001541353706200072,
+      "loss": 0.9044,
+      "step": 24840
+    },
+    {
+      "epoch": 0.7924359832902835,
+      "grad_norm": 0.1917041540145874,
+      "learning_rate": 0.0001534900033195833,
+      "loss": 0.9151,
+      "step": 24850
+    },
+    {
+      "epoch": 0.7927548710099174,
+      "grad_norm": 0.19247248768806458,
+      "learning_rate": 0.00015284733818252897,
+      "loss": 0.9068,
+      "step": 24860
+    },
+    {
+      "epoch": 0.7930737587295513,
+      "grad_norm": 0.18687069416046143,
+      "learning_rate": 0.00015220736389484244,
+      "loss": 0.9018,
+      "step": 24870
+    },
+    {
+      "epoch": 0.7933926464491853,
+      "grad_norm": 0.1889362782239914,
+      "learning_rate": 0.00015157006918989363,
+      "loss": 0.9096,
+      "step": 24880
+    },
+    {
+      "epoch": 0.7937115341688191,
+      "grad_norm": 0.19133806228637695,
+      "learning_rate": 0.00015093544284822607,
+      "loss": 0.9034,
+      "step": 24890
+    },
+    {
+      "epoch": 0.7940304218884531,
+      "grad_norm": 0.1814614236354828,
+      "learning_rate": 0.0001503034736973594,
+      "loss": 0.9042,
+      "step": 24900
+    },
+    {
+      "epoch": 0.794349309608087,
+      "grad_norm": 0.1918705403804779,
+      "learning_rate": 0.00014967415061159254,
+      "loss": 0.9141,
+      "step": 24910
+    },
+    {
+      "epoch": 0.7946681973277209,
+      "grad_norm": 0.18677212297916412,
+      "learning_rate": 0.00014904746251180796,
+      "loss": 0.9097,
+      "step": 24920
+    },
+    {
+      "epoch": 0.7949870850473548,
+      "grad_norm": 0.19333551824092865,
+      "learning_rate": 0.00014842339836527651,
+      "loss": 0.9008,
+      "step": 24930
+    },
+    {
+      "epoch": 0.7953059727669888,
+      "grad_norm": 0.19339174032211304,
+      "learning_rate": 0.00014780194718546334,
+      "loss": 0.9083,
+      "step": 24940
+    },
+    {
+      "epoch": 0.7956248604866226,
+      "grad_norm": 0.19391214847564697,
+      "learning_rate": 0.00014718309803183436,
+      "loss": 0.9166,
+      "step": 24950
+    },
+    {
+      "epoch": 0.7959437482062566,
+      "grad_norm": 0.19230784475803375,
+      "learning_rate": 0.00014656684000966363,
+      "loss": 0.9076,
+      "step": 24960
+    },
+    {
+      "epoch": 0.7962626359258905,
+      "grad_norm": 0.1877783089876175,
+      "learning_rate": 0.00014595316226984173,
+      "loss": 0.9016,
+      "step": 24970
+    },
+    {
+      "epoch": 0.7965815236455244,
+      "grad_norm": 0.1867537945508957,
+      "learning_rate": 0.00014534205400868448,
+      "loss": 0.9091,
+      "step": 24980
+    },
+    {
+      "epoch": 0.7969004113651583,
+      "grad_norm": 0.18787339329719543,
+      "learning_rate": 0.000144733504467743,
+      "loss": 0.8847,
+      "step": 24990
+    },
+    {
+      "epoch": 0.7972192990847923,
+      "grad_norm": 0.19018739461898804,
+      "learning_rate": 0.00014412750293361419,
+      "loss": 0.9036,
+      "step": 25000
+    },
+    {
+      "epoch": 0.7975381868044261,
+      "grad_norm": 0.19432775676250458,
+      "learning_rate": 0.00014352403873775206,
+      "loss": 0.9233,
+      "step": 25010
+    },
+    {
+      "epoch": 0.7978570745240601,
+      "grad_norm": 0.19632703065872192,
+      "learning_rate": 0.0001429231012562802,
+      "loss": 0.901,
+      "step": 25020
+    },
+    {
+      "epoch": 0.798175962243694,
+      "grad_norm": 0.1939215213060379,
+      "learning_rate": 0.00014232467990980434,
+      "loss": 0.8983,
+      "step": 25030
+    },
+    {
+      "epoch": 0.7984948499633279,
+      "grad_norm": 0.2033773511648178,
+      "learning_rate": 0.0001417287641632264,
+      "loss": 0.9169,
+      "step": 25040
+    },
+    {
+      "epoch": 0.7988137376829618,
+      "grad_norm": 0.18326032161712646,
+      "learning_rate": 0.00014113534352555893,
+      "loss": 0.9006,
+      "step": 25050
+    },
+    {
+      "epoch": 0.7991326254025958,
+      "grad_norm": 0.18715116381645203,
+      "learning_rate": 0.00014054440754974036,
+      "loss": 0.8999,
+      "step": 25060
+    },
+    {
+      "epoch": 0.7994515131222296,
+      "grad_norm": 0.19295094907283783,
+      "learning_rate": 0.00013995594583245116,
+      "loss": 0.9173,
+      "step": 25070
+    },
+    {
+      "epoch": 0.7997704008418636,
+      "grad_norm": 0.19653460383415222,
+      "learning_rate": 0.0001393699480139307,
+      "loss": 0.9094,
+      "step": 25080
+    },
+    {
+      "epoch": 0.8000892885614975,
+      "grad_norm": 0.18987542390823364,
+      "learning_rate": 0.00013878640377779487,
+      "loss": 0.9136,
+      "step": 25090
+    },
+    {
+      "epoch": 0.8004081762811314,
+      "grad_norm": 0.1913256049156189,
+      "learning_rate": 0.00013820530285085425,
+      "loss": 0.893,
+      "step": 25100
+    },
+    {
+      "epoch": 0.8007270640007653,
+      "grad_norm": 0.1863279491662979,
+      "learning_rate": 0.0001376266350029336,
+      "loss": 0.9053,
+      "step": 25110
+    },
+    {
+      "epoch": 0.8010459517203993,
+      "grad_norm": 0.1903347223997116,
+      "learning_rate": 0.0001370503900466916,
+      "loss": 0.9071,
+      "step": 25120
+    },
+    {
+      "epoch": 0.8013648394400331,
+      "grad_norm": 0.19538167119026184,
+      "learning_rate": 0.00013647655783744143,
+      "loss": 0.8983,
+      "step": 25130
+    },
+    {
+      "epoch": 0.8016837271596671,
+      "grad_norm": 0.19056464731693268,
+      "learning_rate": 0.00013590512827297215,
+      "loss": 0.9075,
+      "step": 25140
+    },
+    {
+      "epoch": 0.802002614879301,
+      "grad_norm": 0.19379113614559174,
+      "learning_rate": 0.00013533609129337112,
+      "loss": 0.8932,
+      "step": 25150
+    },
+    {
+      "epoch": 0.8023215025989349,
+      "grad_norm": 0.19236436486244202,
+      "learning_rate": 0.00013476943688084665,
+      "loss": 0.9199,
+      "step": 25160
+    },
+    {
+      "epoch": 0.8026403903185688,
+      "grad_norm": 0.19388633966445923,
+      "learning_rate": 0.00013420515505955158,
+      "loss": 0.8861,
+      "step": 25170
+    },
+    {
+      "epoch": 0.8029592780382028,
+      "grad_norm": 0.18646042048931122,
+      "learning_rate": 0.000133643235895408,
+      "loss": 0.8961,
+      "step": 25180
+    },
+    {
+      "epoch": 0.8032781657578366,
+      "grad_norm": 0.19682526588439941,
+      "learning_rate": 0.00013308366949593199,
+      "loss": 0.8948,
+      "step": 25190
+    },
+    {
+      "epoch": 0.8035970534774706,
+      "grad_norm": 0.1909864842891693,
+      "learning_rate": 0.0001325264460100597,
+      "loss": 0.9045,
+      "step": 25200
+    },
+    {
+      "epoch": 0.8039159411971045,
+      "grad_norm": 0.19391486048698425,
+      "learning_rate": 0.0001319715556279738,
+      "loss": 0.913,
+      "step": 25210
+    },
+    {
+      "epoch": 0.8042348289167384,
+      "grad_norm": 0.18908917903900146,
+      "learning_rate": 0.00013141898858093086,
+      "loss": 0.9024,
+      "step": 25220
+    },
+    {
+      "epoch": 0.8045537166363723,
+      "grad_norm": 0.18832768499851227,
+      "learning_rate": 0.00013086873514108925,
+      "loss": 0.892,
+      "step": 25230
+    },
+    {
+      "epoch": 0.8048726043560063,
+      "grad_norm": 0.19684334099292755,
+      "learning_rate": 0.00013032078562133812,
+      "loss": 0.9172,
+      "step": 25240
+    },
+    {
+      "epoch": 0.8051914920756401,
+      "grad_norm": 0.18681035935878754,
+      "learning_rate": 0.0001297751303751266,
+      "loss": 0.9105,
+      "step": 25250
+    },
+    {
+      "epoch": 0.8055103797952741,
+      "grad_norm": 0.19750331342220306,
+      "learning_rate": 0.00012923175979629407,
+      "loss": 0.8844,
+      "step": 25260
+    },
+    {
+      "epoch": 0.805829267514908,
+      "grad_norm": 0.17914694547653198,
+      "learning_rate": 0.00012869066431890117,
+      "loss": 0.8821,
+      "step": 25270
+    },
+    {
+      "epoch": 0.8061481552345419,
+      "grad_norm": 0.18732082843780518,
+      "learning_rate": 0.00012815183441706112,
+      "loss": 0.9071,
+      "step": 25280
+    },
+    {
+      "epoch": 0.8064670429541758,
+      "grad_norm": 0.18838287889957428,
+      "learning_rate": 0.0001276152606047724,
+      "loss": 0.9071,
+      "step": 25290
+    },
+    {
+      "epoch": 0.8067859306738098,
+      "grad_norm": 0.18626269698143005,
+      "learning_rate": 0.0001270809334357514,
+      "loss": 0.8998,
+      "step": 25300
+    },
+    {
+      "epoch": 0.8071048183934437,
+      "grad_norm": 0.19195237755775452,
+      "learning_rate": 0.00012654884350326619,
+      "loss": 0.8856,
+      "step": 25310
+    },
+    {
+      "epoch": 0.8074237061130776,
+      "grad_norm": 0.19287879765033722,
+      "learning_rate": 0.00012601898143997108,
+      "loss": 0.9198,
+      "step": 25320
+    },
+    {
+      "epoch": 0.8077425938327115,
+      "grad_norm": 0.18623973429203033,
+      "learning_rate": 0.00012549133791774162,
+      "loss": 0.9028,
+      "step": 25330
+    },
+    {
+      "epoch": 0.8080614815523455,
+      "grad_norm": 0.1909293383359909,
+      "learning_rate": 0.0001249659036475103,
+      "loss": 0.9023,
+      "step": 25340
+    },
+    {
+      "epoch": 0.8083803692719793,
+      "grad_norm": 0.1917041689157486,
+      "learning_rate": 0.00012444266937910312,
+      "loss": 0.9037,
+      "step": 25350
+    },
+    {
+      "epoch": 0.8086992569916133,
+      "grad_norm": 0.1956179291009903,
+      "learning_rate": 0.0001239216259010767,
+      "loss": 0.897,
+      "step": 25360
+    },
+    {
+      "epoch": 0.8090181447112472,
+      "grad_norm": 0.18946874141693115,
+      "learning_rate": 0.00012340276404055616,
+      "loss": 0.8903,
+      "step": 25370
+    },
+    {
+      "epoch": 0.8093370324308811,
+      "grad_norm": 0.19109565019607544,
+      "learning_rate": 0.00012288607466307355,
+      "loss": 0.9091,
+      "step": 25380
+    },
+    {
+      "epoch": 0.809655920150515,
+      "grad_norm": 0.19188068807125092,
+      "learning_rate": 0.0001223715486724071,
+      "loss": 0.8943,
+      "step": 25390
+    },
+    {
+      "epoch": 0.809974807870149,
+      "grad_norm": 0.1854020357131958,
+      "learning_rate": 0.00012185917701042106,
+      "loss": 0.8957,
+      "step": 25400
+    },
+    {
+      "epoch": 0.8102936955897828,
+      "grad_norm": 0.18928898870944977,
+      "learning_rate": 0.0001213489506569063,
+      "loss": 0.8974,
+      "step": 25410
+    },
+    {
+      "epoch": 0.8106125833094168,
+      "grad_norm": 0.19741034507751465,
+      "learning_rate": 0.00012084086062942134,
+      "loss": 0.9008,
+      "step": 25420
+    },
+    {
+      "epoch": 0.8109314710290507,
+      "grad_norm": 0.1944274604320526,
+      "learning_rate": 0.00012033489798313444,
+      "loss": 0.8994,
+      "step": 25430
+    },
+    {
+      "epoch": 0.8112503587486846,
+      "grad_norm": 0.1925448775291443,
+      "learning_rate": 0.00011983105381066592,
+      "loss": 0.9098,
+      "step": 25440
+    },
+    {
+      "epoch": 0.8115692464683185,
+      "grad_norm": 0.19471345841884613,
+      "learning_rate": 0.00011932931924193155,
+      "loss": 0.8981,
+      "step": 25450
+    },
+    {
+      "epoch": 0.8118881341879525,
+      "grad_norm": 0.18608881533145905,
+      "learning_rate": 0.0001188296854439862,
+      "loss": 0.8864,
+      "step": 25460
+    },
+    {
+      "epoch": 0.8122070219075863,
+      "grad_norm": 0.19888868927955627,
+      "learning_rate": 0.00011833214362086844,
+      "loss": 0.9101,
+      "step": 25470
+    },
+    {
+      "epoch": 0.8125259096272203,
+      "grad_norm": 0.18246127665042877,
+      "learning_rate": 0.00011783668501344572,
+      "loss": 0.9138,
+      "step": 25480
+    },
+    {
+      "epoch": 0.8128447973468542,
+      "grad_norm": 0.19178058207035065,
+      "learning_rate": 0.00011734330089926006,
+      "loss": 0.9087,
+      "step": 25490
+    },
+    {
+      "epoch": 0.8131636850664881,
+      "grad_norm": 0.19574995338916779,
+      "learning_rate": 0.0001168519825923746,
+      "loss": 0.9075,
+      "step": 25500
+    },
+    {
+      "epoch": 0.813482572786122,
+      "grad_norm": 0.19114628434181213,
+      "learning_rate": 0.00011636272144322059,
+      "loss": 0.9029,
+      "step": 25510
+    },
+    {
+      "epoch": 0.813801460505756,
+      "grad_norm": 0.20036236941814423,
+      "learning_rate": 0.00011587550883844523,
+      "loss": 0.8975,
+      "step": 25520
+    },
+    {
+      "epoch": 0.8141203482253898,
+      "grad_norm": 0.19069448113441467,
+      "learning_rate": 0.00011539033620075986,
+      "loss": 0.9043,
+      "step": 25530
+    },
+    {
+      "epoch": 0.8144392359450238,
+      "grad_norm": 0.1888374239206314,
+      "learning_rate": 0.00011490719498878924,
+      "loss": 0.9047,
+      "step": 25540
+    },
+    {
+      "epoch": 0.8147581236646577,
+      "grad_norm": 0.19065211713314056,
+      "learning_rate": 0.00011442607669692085,
+      "loss": 0.9017,
+      "step": 25550
+    },
+    {
+      "epoch": 0.8150770113842916,
+      "grad_norm": 0.18727807700634003,
+      "learning_rate": 0.00011394697285515537,
+      "loss": 0.8943,
+      "step": 25560
+    },
+    {
+      "epoch": 0.8153958991039255,
+      "grad_norm": 0.18859420716762543,
+      "learning_rate": 0.0001134698750289575,
+      "loss": 0.9018,
+      "step": 25570
+    },
+    {
+      "epoch": 0.8157147868235595,
+      "grad_norm": 0.18913602828979492,
+      "learning_rate": 0.00011299477481910747,
+      "loss": 0.9019,
+      "step": 25580
+    },
+    {
+      "epoch": 0.8160336745431933,
+      "grad_norm": 0.19403930008411407,
+      "learning_rate": 0.00011252166386155319,
+      "loss": 0.9129,
+      "step": 25590
+    },
+    {
+      "epoch": 0.8163525622628273,
+      "grad_norm": 0.1911238133907318,
+      "learning_rate": 0.00011205053382726299,
+      "loss": 0.8979,
+      "step": 25600
+    },
+    {
+      "epoch": 0.8166714499824612,
+      "grad_norm": 0.19056203961372375,
+      "learning_rate": 0.00011158137642207893,
+      "loss": 0.8961,
+      "step": 25610
+    },
+    {
+      "epoch": 0.8169903377020951,
+      "grad_norm": 0.1901327222585678,
+      "learning_rate": 0.00011111418338657102,
+      "loss": 0.9004,
+      "step": 25620
+    },
+    {
+      "epoch": 0.817309225421729,
+      "grad_norm": 0.19726039469242096,
+      "learning_rate": 0.00011064894649589143,
+      "loss": 0.9107,
+      "step": 25630
+    },
+    {
+      "epoch": 0.817628113141363,
+      "grad_norm": 0.1873912811279297,
+      "learning_rate": 0.00011018565755962999,
+      "loss": 0.9057,
+      "step": 25640
+    },
+    {
+      "epoch": 0.8179470008609968,
+      "grad_norm": 0.1929081529378891,
+      "learning_rate": 0.00010972430842166995,
+      "loss": 0.883,
+      "step": 25650
+    },
+    {
+      "epoch": 0.8182658885806308,
+      "grad_norm": 0.19839973747730255,
+      "learning_rate": 0.0001092648909600443,
+      "loss": 0.9024,
+      "step": 25660
+    },
+    {
+      "epoch": 0.8185847763002647,
+      "grad_norm": 0.19346867501735687,
+      "learning_rate": 0.00010880739708679283,
+      "loss": 0.8958,
+      "step": 25670
+    },
+    {
+      "epoch": 0.8189036640198986,
+      "grad_norm": 0.1984575092792511,
+      "learning_rate": 0.00010835181874781979,
+      "loss": 0.9047,
+      "step": 25680
+    },
+    {
+      "epoch": 0.8192225517395325,
+      "grad_norm": 0.1977538913488388,
+      "learning_rate": 0.00010789814792275205,
+      "loss": 0.9038,
+      "step": 25690
+    },
+    {
+      "epoch": 0.8195414394591665,
+      "grad_norm": 0.2003737837076187,
+      "learning_rate": 0.00010744637662479786,
+      "loss": 0.91,
+      "step": 25700
+    },
+    {
+      "epoch": 0.8198603271788003,
+      "grad_norm": 0.19274483621120453,
+      "learning_rate": 0.00010699649690060641,
+      "loss": 0.8969,
+      "step": 25710
+    },
+    {
+      "epoch": 0.8201792148984343,
+      "grad_norm": 0.18908193707466125,
+      "learning_rate": 0.00010654850083012758,
+      "loss": 0.8835,
+      "step": 25720
+    },
+    {
+      "epoch": 0.8204981026180682,
+      "grad_norm": 0.19787780940532684,
+      "learning_rate": 0.00010610238052647271,
+      "loss": 0.8967,
+      "step": 25730
+    },
+    {
+      "epoch": 0.8208169903377021,
+      "grad_norm": 0.1946023851633072,
+      "learning_rate": 0.00010565812813577561,
+      "loss": 0.8948,
+      "step": 25740
+    },
+    {
+      "epoch": 0.821135878057336,
+      "grad_norm": 0.19036608934402466,
+      "learning_rate": 0.00010521573583705442,
+      "loss": 0.8982,
+      "step": 25750
+    },
+    {
+      "epoch": 0.82145476577697,
+      "grad_norm": 0.19042381644248962,
+      "learning_rate": 0.00010477519584207379,
+      "loss": 0.8739,
+      "step": 25760
+    },
+    {
+      "epoch": 0.8217736534966038,
+      "grad_norm": 0.19218648970127106,
+      "learning_rate": 0.00010433650039520788,
+      "loss": 0.895,
+      "step": 25770
+    },
+    {
+      "epoch": 0.8220925412162378,
+      "grad_norm": 0.19477754831314087,
+      "learning_rate": 0.00010389964177330376,
+      "loss": 0.8941,
+      "step": 25780
+    },
+    {
+      "epoch": 0.8224114289358717,
+      "grad_norm": 0.19564403593540192,
+      "learning_rate": 0.00010346461228554552,
+      "loss": 0.9025,
+      "step": 25790
+    },
+    {
+      "epoch": 0.8227303166555056,
+      "grad_norm": 0.19064311683177948,
+      "learning_rate": 0.00010303140427331876,
+      "loss": 0.8768,
+      "step": 25800
+    },
+    {
+      "epoch": 0.8230492043751395,
+      "grad_norm": 0.18618376553058624,
+      "learning_rate": 0.00010260001011007584,
+      "loss": 0.8988,
+      "step": 25810
+    },
+    {
+      "epoch": 0.8233680920947735,
+      "grad_norm": 0.18917080760002136,
+      "learning_rate": 0.00010217042220120158,
+      "loss": 0.9023,
+      "step": 25820
+    },
+    {
+      "epoch": 0.8236869798144073,
+      "grad_norm": 0.19517607986927032,
+      "learning_rate": 0.00010174263298387963,
+      "loss": 0.8916,
+      "step": 25830
+    },
+    {
+      "epoch": 0.8240058675340413,
+      "grad_norm": 0.18709541857242584,
+      "learning_rate": 0.00010131663492695925,
+      "loss": 0.8908,
+      "step": 25840
+    },
+    {
+      "epoch": 0.8243247552536752,
+      "grad_norm": 0.19261054694652557,
+      "learning_rate": 0.00010089242053082271,
+      "loss": 0.8996,
+      "step": 25850
+    },
+    {
+      "epoch": 0.824643642973309,
+      "grad_norm": 0.19508786499500275,
+      "learning_rate": 0.00010046998232725337,
+      "loss": 0.8903,
+      "step": 25860
+    },
+    {
+      "epoch": 0.824962530692943,
+      "grad_norm": 0.1922270953655243,
+      "learning_rate": 0.00010004931287930405,
+      "loss": 0.8815,
+      "step": 25870
+    },
+    {
+      "epoch": 0.825281418412577,
+      "grad_norm": 0.19108553230762482,
+      "learning_rate": 9.96304047811663e-05,
+      "loss": 0.8864,
+      "step": 25880
+    },
+    {
+      "epoch": 0.8256003061322108,
+      "grad_norm": 0.19563859701156616,
+      "learning_rate": 9.921325065803983e-05,
+      "loss": 0.8991,
+      "step": 25890
+    },
+    {
+      "epoch": 0.8259191938518448,
+      "grad_norm": 0.19630859792232513,
+      "learning_rate": 9.879784316600278e-05,
+      "loss": 0.8808,
+      "step": 25900
+    },
+    {
+      "epoch": 0.8262380815714787,
+      "grad_norm": 0.2010757029056549,
+      "learning_rate": 9.838417499188239e-05,
+      "loss": 0.897,
+      "step": 25910
+    },
+    {
+      "epoch": 0.8265569692911126,
+      "grad_norm": 0.1868877112865448,
+      "learning_rate": 9.79722388531263e-05,
+      "loss": 0.8982,
+      "step": 25920
+    },
+    {
+      "epoch": 0.8268758570107465,
+      "grad_norm": 0.19479702413082123,
+      "learning_rate": 9.756202749767429e-05,
+      "loss": 0.8909,
+      "step": 25930
+    },
+    {
+      "epoch": 0.8271947447303805,
+      "grad_norm": 0.19186285138130188,
+      "learning_rate": 9.715353370383069e-05,
+      "loss": 0.8999,
+      "step": 25940
+    },
+    {
+      "epoch": 0.8275136324500143,
+      "grad_norm": 0.1874547004699707,
+      "learning_rate": 9.674675028013712e-05,
+      "loss": 0.8698,
+      "step": 25950
+    },
+    {
+      "epoch": 0.8278325201696483,
+      "grad_norm": 0.1925581693649292,
+      "learning_rate": 9.634167006524597e-05,
+      "loss": 0.8985,
+      "step": 25960
+    },
+    {
+      "epoch": 0.8281514078892822,
+      "grad_norm": 0.19503287971019745,
+      "learning_rate": 9.593828592779434e-05,
+      "loss": 0.8946,
+      "step": 25970
+    },
+    {
+      "epoch": 0.828470295608916,
+      "grad_norm": 0.2057391107082367,
+      "learning_rate": 9.553659076627844e-05,
+      "loss": 0.9048,
+      "step": 25980
+    },
+    {
+      "epoch": 0.82878918332855,
+      "grad_norm": 0.1989460438489914,
+      "learning_rate": 9.513657750892853e-05,
+      "loss": 0.8884,
+      "step": 25990
+    },
+    {
+      "epoch": 0.829108071048184,
+      "grad_norm": 0.19758674502372742,
+      "learning_rate": 9.473823911358465e-05,
+      "loss": 0.8884,
+      "step": 26000
+    },
+    {
+      "epoch": 0.8294269587678178,
+      "grad_norm": 0.18588489294052124,
+      "learning_rate": 9.434156856757232e-05,
+      "loss": 0.885,
+      "step": 26010
+    },
+    {
+      "epoch": 0.8297458464874518,
+      "grad_norm": 0.19282850623130798,
+      "learning_rate": 9.394655888757933e-05,
+      "loss": 0.8843,
+      "step": 26020
+    },
+    {
+      "epoch": 0.8300647342070857,
+      "grad_norm": 0.19358274340629578,
+      "learning_rate": 9.355320311953263e-05,
+      "loss": 0.9003,
+      "step": 26030
+    },
+    {
+      "epoch": 0.8303836219267196,
+      "grad_norm": 0.19735553860664368,
+      "learning_rate": 9.316149433847609e-05,
+      "loss": 0.8926,
+      "step": 26040
+    },
+    {
+      "epoch": 0.8307025096463535,
+      "grad_norm": 0.1917579174041748,
+      "learning_rate": 9.27714256484484e-05,
+      "loss": 0.9033,
+      "step": 26050
+    },
+    {
+      "epoch": 0.8310213973659875,
+      "grad_norm": 0.19253985583782196,
+      "learning_rate": 9.238299018236176e-05,
+      "loss": 0.892,
+      "step": 26060
+    },
+    {
+      "epoch": 0.8313402850856213,
+      "grad_norm": 0.19036774337291718,
+      "learning_rate": 9.199618110188106e-05,
+      "loss": 0.8794,
+      "step": 26070
+    },
+    {
+      "epoch": 0.8316591728052553,
+      "grad_norm": 0.19188916683197021,
+      "learning_rate": 9.161099159730329e-05,
+      "loss": 0.891,
+      "step": 26080
+    },
+    {
+      "epoch": 0.8319780605248892,
+      "grad_norm": 0.19725832343101501,
+      "learning_rate": 9.122741488743787e-05,
+      "loss": 0.8884,
+      "step": 26090
+    },
+    {
+      "epoch": 0.832296948244523,
+      "grad_norm": 0.19478730857372284,
+      "learning_rate": 9.084544421948714e-05,
+      "loss": 0.8941,
+      "step": 26100
+    },
+    {
+      "epoch": 0.832615835964157,
+      "grad_norm": 0.1925089806318283,
+      "learning_rate": 9.046507286892751e-05,
+      "loss": 0.9235,
+      "step": 26110
+    },
+    {
+      "epoch": 0.832934723683791,
+      "grad_norm": 0.1941222846508026,
+      "learning_rate": 9.008629413939108e-05,
+      "loss": 0.8937,
+      "step": 26120
+    },
+    {
+      "epoch": 0.8332536114034248,
+      "grad_norm": 0.19237853586673737,
+      "learning_rate": 8.970910136254777e-05,
+      "loss": 0.8875,
+      "step": 26130
+    },
+    {
+      "epoch": 0.8335724991230588,
+      "grad_norm": 0.19570504128932953,
+      "learning_rate": 8.93334878979879e-05,
+      "loss": 0.8816,
+      "step": 26140
+    },
+    {
+      "epoch": 0.8338913868426927,
+      "grad_norm": 0.19162657856941223,
+      "learning_rate": 8.895944713310525e-05,
+      "loss": 0.8913,
+      "step": 26150
+    },
+    {
+      "epoch": 0.8342102745623267,
+      "grad_norm": 0.1932561993598938,
+      "learning_rate": 8.858697248298071e-05,
+      "loss": 0.9011,
+      "step": 26160
+    },
+    {
+      "epoch": 0.8345291622819605,
+      "grad_norm": 0.193185493350029,
+      "learning_rate": 8.821605739026645e-05,
+      "loss": 0.8667,
+      "step": 26170
+    },
+    {
+      "epoch": 0.8348480500015945,
+      "grad_norm": 0.2088388353586197,
+      "learning_rate": 8.784669532507018e-05,
+      "loss": 0.9125,
+      "step": 26180
+    },
+    {
+      "epoch": 0.8351669377212284,
+      "grad_norm": 0.19375212490558624,
+      "learning_rate": 8.747887978484048e-05,
+      "loss": 0.8872,
+      "step": 26190
+    },
+    {
+      "epoch": 0.8354858254408623,
+      "grad_norm": 0.19213901460170746,
+      "learning_rate": 8.71126042942522e-05,
+      "loss": 0.8768,
+      "step": 26200
+    },
+    {
+      "epoch": 0.8358047131604962,
+      "grad_norm": 0.19676388800144196,
+      "learning_rate": 8.674786240509246e-05,
+      "loss": 0.8803,
+      "step": 26210
+    },
+    {
+      "epoch": 0.8361236008801302,
+      "grad_norm": 0.18894055485725403,
+      "learning_rate": 8.638464769614718e-05,
+      "loss": 0.8887,
+      "step": 26220
+    },
+    {
+      "epoch": 0.836442488599764,
+      "grad_norm": 0.19717377424240112,
+      "learning_rate": 8.602295377308798e-05,
+      "loss": 0.8849,
+      "step": 26230
+    },
+    {
+      "epoch": 0.836761376319398,
+      "grad_norm": 0.18463239073753357,
+      "learning_rate": 8.566277426835967e-05,
+      "loss": 0.8856,
+      "step": 26240
+    },
+    {
+      "epoch": 0.8370802640390319,
+      "grad_norm": 0.19209441542625427,
+      "learning_rate": 8.53041028410681e-05,
+      "loss": 0.8865,
+      "step": 26250
+    },
+    {
+      "epoch": 0.8373991517586657,
+      "grad_norm": 0.19038350880146027,
+      "learning_rate": 8.494693317686852e-05,
+      "loss": 0.8662,
+      "step": 26260
+    },
+    {
+      "epoch": 0.8377180394782997,
+      "grad_norm": 0.19096122682094574,
+      "learning_rate": 8.459125898785451e-05,
+      "loss": 0.9065,
+      "step": 26270
+    },
+    {
+      "epoch": 0.8380369271979337,
+      "grad_norm": 0.1978101283311844,
+      "learning_rate": 8.423707401244714e-05,
+      "loss": 0.9053,
+      "step": 26280
+    },
+    {
+      "epoch": 0.8383558149175675,
+      "grad_norm": 0.18875358998775482,
+      "learning_rate": 8.388437201528488e-05,
+      "loss": 0.8865,
+      "step": 26290
+    },
+    {
+      "epoch": 0.8386747026372015,
+      "grad_norm": 0.1894250363111496,
+      "learning_rate": 8.353314678711372e-05,
+      "loss": 0.8963,
+      "step": 26300
+    },
+    {
+      "epoch": 0.8389935903568354,
+      "grad_norm": 0.1935695856809616,
+      "learning_rate": 8.318339214467789e-05,
+      "loss": 0.8895,
+      "step": 26310
+    },
+    {
+      "epoch": 0.8393124780764692,
+      "grad_norm": 0.18998850882053375,
+      "learning_rate": 8.283510193061105e-05,
+      "loss": 0.8861,
+      "step": 26320
+    },
+    {
+      "epoch": 0.8396313657961032,
+      "grad_norm": 0.20382626354694366,
+      "learning_rate": 8.248827001332778e-05,
+      "loss": 0.8913,
+      "step": 26330
+    },
+    {
+      "epoch": 0.8399502535157372,
+      "grad_norm": 0.1898353546857834,
+      "learning_rate": 8.214289028691584e-05,
+      "loss": 0.8857,
+      "step": 26340
+    },
+    {
+      "epoch": 0.840269141235371,
+      "grad_norm": 0.19221539795398712,
+      "learning_rate": 8.179895667102844e-05,
+      "loss": 0.8751,
+      "step": 26350
+    },
+    {
+      "epoch": 0.840588028955005,
+      "grad_norm": 0.18822281062602997,
+      "learning_rate": 8.145646311077731e-05,
+      "loss": 0.8997,
+      "step": 26360
+    },
+    {
+      "epoch": 0.8409069166746389,
+      "grad_norm": 0.19643518328666687,
+      "learning_rate": 8.11154035766261e-05,
+      "loss": 0.894,
+      "step": 26370
+    },
+    {
+      "epoch": 0.8412258043942727,
+      "grad_norm": 0.19378352165222168,
+      "learning_rate": 8.077577206428427e-05,
+      "loss": 0.8879,
+      "step": 26380
+    },
+    {
+      "epoch": 0.8415446921139067,
+      "grad_norm": 0.1851210743188858,
+      "learning_rate": 8.043756259460127e-05,
+      "loss": 0.8805,
+      "step": 26390
+    },
+    {
+      "epoch": 0.8418635798335407,
+      "grad_norm": 0.19853575527668,
+      "learning_rate": 8.010076921346141e-05,
+      "loss": 0.904,
+      "step": 26400
+    },
+    {
+      "epoch": 0.8421824675531745,
+      "grad_norm": 0.192692831158638,
+      "learning_rate": 7.976538599167896e-05,
+      "loss": 0.8952,
+      "step": 26410
+    },
+    {
+      "epoch": 0.8425013552728084,
+      "grad_norm": 0.19446724653244019,
+      "learning_rate": 7.943140702489378e-05,
+      "loss": 0.8703,
+      "step": 26420
+    },
+    {
+      "epoch": 0.8428202429924424,
+      "grad_norm": 0.1938653141260147,
+      "learning_rate": 7.909882643346739e-05,
+      "loss": 0.8889,
+      "step": 26430
+    },
+    {
+      "epoch": 0.8431391307120762,
+      "grad_norm": 0.1853872686624527,
+      "learning_rate": 7.876763836237944e-05,
+      "loss": 0.8928,
+      "step": 26440
+    },
+    {
+      "epoch": 0.8434580184317102,
+      "grad_norm": 0.19568127393722534,
+      "learning_rate": 7.843783698112465e-05,
+      "loss": 0.9062,
+      "step": 26450
+    },
+    {
+      "epoch": 0.8437769061513442,
+      "grad_norm": 0.1882636547088623,
+      "learning_rate": 7.810941648361018e-05,
+      "loss": 0.8969,
+      "step": 26460
+    },
+    {
+      "epoch": 0.844095793870978,
+      "grad_norm": 0.1955377161502838,
+      "learning_rate": 7.778237108805339e-05,
+      "loss": 0.8771,
+      "step": 26470
+    },
+    {
+      "epoch": 0.844414681590612,
+      "grad_norm": 0.18335965275764465,
+      "learning_rate": 7.745669503688002e-05,
+      "loss": 0.8825,
+      "step": 26480
+    },
+    {
+      "epoch": 0.8447335693102459,
+      "grad_norm": 0.19274048507213593,
+      "learning_rate": 7.71323825966229e-05,
+      "loss": 0.8819,
+      "step": 26490
+    },
+    {
+      "epoch": 0.8450524570298797,
+      "grad_norm": 0.18612460792064667,
+      "learning_rate": 7.680942805782095e-05,
+      "loss": 0.8862,
+      "step": 26500
+    },
+    {
+      "epoch": 0.8453713447495137,
+      "grad_norm": 0.1887853741645813,
+      "learning_rate": 7.648782573491877e-05,
+      "loss": 0.8825,
+      "step": 26510
+    },
+    {
+      "epoch": 0.8456902324691477,
+      "grad_norm": 0.19096797704696655,
+      "learning_rate": 7.616756996616643e-05,
+      "loss": 0.8766,
+      "step": 26520
+    },
+    {
+      "epoch": 0.8460091201887815,
+      "grad_norm": 0.19669249653816223,
+      "learning_rate": 7.58486551135198e-05,
+      "loss": 0.8821,
+      "step": 26530
+    },
+    {
+      "epoch": 0.8463280079084154,
+      "grad_norm": 0.1886143982410431,
+      "learning_rate": 7.553107556254135e-05,
+      "loss": 0.8786,
+      "step": 26540
+    },
+    {
+      "epoch": 0.8466468956280494,
+      "grad_norm": 0.19534634053707123,
+      "learning_rate": 7.521482572230134e-05,
+      "loss": 0.8883,
+      "step": 26550
+    },
+    {
+      "epoch": 0.8469657833476832,
+      "grad_norm": 0.19123771786689758,
+      "learning_rate": 7.48999000252793e-05,
+      "loss": 0.8819,
+      "step": 26560
+    },
+    {
+      "epoch": 0.8472846710673172,
+      "grad_norm": 0.18510787189006805,
+      "learning_rate": 7.458629292726607e-05,
+      "loss": 0.8966,
+      "step": 26570
+    },
+    {
+      "epoch": 0.8476035587869512,
+      "grad_norm": 0.19328685104846954,
+      "learning_rate": 7.427399890726616e-05,
+      "loss": 0.8799,
+      "step": 26580
+    },
+    {
+      "epoch": 0.847922446506585,
+      "grad_norm": 0.19416575133800507,
+      "learning_rate": 7.396301246740063e-05,
+      "loss": 0.8805,
+      "step": 26590
+    },
+    {
+      "epoch": 0.848241334226219,
+      "grad_norm": 0.1965588927268982,
+      "learning_rate": 7.36533281328102e-05,
+      "loss": 0.8907,
+      "step": 26600
+    },
+    {
+      "epoch": 0.8485602219458529,
+      "grad_norm": 0.2026282250881195,
+      "learning_rate": 7.334494045155892e-05,
+      "loss": 0.8804,
+      "step": 26610
+    },
+    {
+      "epoch": 0.8488791096654867,
+      "grad_norm": 0.19632813334465027,
+      "learning_rate": 7.303784399453824e-05,
+      "loss": 0.8882,
+      "step": 26620
+    },
+    {
+      "epoch": 0.8491979973851207,
+      "grad_norm": 0.19422686100006104,
+      "learning_rate": 7.273203335537129e-05,
+      "loss": 0.8857,
+      "step": 26630
+    },
+    {
+      "epoch": 0.8495168851047546,
+      "grad_norm": 0.1914503127336502,
+      "learning_rate": 7.242750315031787e-05,
+      "loss": 0.9041,
+      "step": 26640
+    },
+    {
+      "epoch": 0.8498357728243885,
+      "grad_norm": 0.18675798177719116,
+      "learning_rate": 7.21242480181795e-05,
+      "loss": 0.8853,
+      "step": 26650
+    },
+    {
+      "epoch": 0.8501546605440224,
+      "grad_norm": 0.19239214062690735,
+      "learning_rate": 7.182226262020522e-05,
+      "loss": 0.8753,
+      "step": 26660
+    },
+    {
+      "epoch": 0.8504735482636564,
+      "grad_norm": 0.19588054716587067,
+      "learning_rate": 7.15215416399974e-05,
+      "loss": 0.8887,
+      "step": 26670
+    },
+    {
+      "epoch": 0.8507924359832902,
+      "grad_norm": 0.1920136660337448,
+      "learning_rate": 7.122207978341839e-05,
+      "loss": 0.889,
+      "step": 26680
+    },
+    {
+      "epoch": 0.8511113237029242,
+      "grad_norm": 0.18894952535629272,
+      "learning_rate": 7.092387177849706e-05,
+      "loss": 0.88,
+      "step": 26690
+    },
+    {
+      "epoch": 0.8514302114225581,
+      "grad_norm": 0.19077442586421967,
+      "learning_rate": 7.062691237533617e-05,
+      "loss": 0.8782,
+      "step": 26700
+    },
+    {
+      "epoch": 0.851749099142192,
+      "grad_norm": 0.1909032016992569,
+      "learning_rate": 7.033119634601985e-05,
+      "loss": 0.8887,
+      "step": 26710
+    },
+    {
+      "epoch": 0.8520679868618259,
+      "grad_norm": 0.19129188358783722,
+      "learning_rate": 7.003671848452163e-05,
+      "loss": 0.885,
+      "step": 26720
+    },
+    {
+      "epoch": 0.8523868745814599,
+      "grad_norm": 0.19594432413578033,
+      "learning_rate": 6.974347360661275e-05,
+      "loss": 0.8924,
+      "step": 26730
+    },
+    {
+      "epoch": 0.8527057623010937,
+      "grad_norm": 0.1917877495288849,
+      "learning_rate": 6.945145654977087e-05,
+      "loss": 0.8972,
+      "step": 26740
+    },
+    {
+      "epoch": 0.8530246500207277,
+      "grad_norm": 0.19224712252616882,
+      "learning_rate": 6.916066217308926e-05,
+      "loss": 0.8944,
+      "step": 26750
+    },
+    {
+      "epoch": 0.8533435377403616,
+      "grad_norm": 0.19904139637947083,
+      "learning_rate": 6.887108535718623e-05,
+      "loss": 0.8813,
+      "step": 26760
+    },
+    {
+      "epoch": 0.8536624254599955,
+      "grad_norm": 0.1943054050207138,
+      "learning_rate": 6.858272100411499e-05,
+      "loss": 0.8757,
+      "step": 26770
+    },
+    {
+      "epoch": 0.8539813131796294,
+      "grad_norm": 0.19775620102882385,
+      "learning_rate": 6.829556403727401e-05,
+      "loss": 0.8983,
+      "step": 26780
+    },
+    {
+      "epoch": 0.8543002008992634,
+      "grad_norm": 0.19544585049152374,
+      "learning_rate": 6.800960940131751e-05,
+      "loss": 0.8895,
+      "step": 26790
+    },
+    {
+      "epoch": 0.8546190886188972,
+      "grad_norm": 0.19292794167995453,
+      "learning_rate": 6.772485206206656e-05,
+      "loss": 0.8663,
+      "step": 26800
+    },
+    {
+      "epoch": 0.8549379763385312,
+      "grad_norm": 0.18773417174816132,
+      "learning_rate": 6.74412870064204e-05,
+      "loss": 0.8881,
+      "step": 26810
+    },
+    {
+      "epoch": 0.8552568640581651,
+      "grad_norm": 0.20264096558094025,
+      "learning_rate": 6.71589092422682e-05,
+      "loss": 0.8956,
+      "step": 26820
+    },
+    {
+      "epoch": 0.855575751777799,
+      "grad_norm": 0.19748295843601227,
+      "learning_rate": 6.687771379840115e-05,
+      "loss": 0.901,
+      "step": 26830
+    },
+    {
+      "epoch": 0.8558946394974329,
+      "grad_norm": 0.1929382085800171,
+      "learning_rate": 6.659769572442513e-05,
+      "loss": 0.8889,
+      "step": 26840
+    },
+    {
+      "epoch": 0.8562135272170669,
+      "grad_norm": 0.1998470276594162,
+      "learning_rate": 6.631885009067319e-05,
+      "loss": 0.8796,
+      "step": 26850
+    },
+    {
+      "epoch": 0.8565324149367007,
+      "grad_norm": 0.19323571026325226,
+      "learning_rate": 6.60411719881191e-05,
+      "loss": 0.8787,
+      "step": 26860
+    },
+    {
+      "epoch": 0.8568513026563347,
+      "grad_norm": 0.19377054274082184,
+      "learning_rate": 6.576465652829075e-05,
+      "loss": 0.8903,
+      "step": 26870
+    },
+    {
+      "epoch": 0.8571701903759686,
+      "grad_norm": 0.19681768119335175,
+      "learning_rate": 6.548929884318418e-05,
+      "loss": 0.8818,
+      "step": 26880
+    },
+    {
+      "epoch": 0.8574890780956025,
+      "grad_norm": 0.19799382984638214,
+      "learning_rate": 6.521509408517782e-05,
+      "loss": 0.8918,
+      "step": 26890
+    },
+    {
+      "epoch": 0.8578079658152364,
+      "grad_norm": 0.19595491886138916,
+      "learning_rate": 6.494203742694715e-05,
+      "loss": 0.8946,
+      "step": 26900
+    },
+    {
+      "epoch": 0.8581268535348704,
+      "grad_norm": 0.19243298470973969,
+      "learning_rate": 6.46701240613798e-05,
+      "loss": 0.8933,
+      "step": 26910
+    },
+    {
+      "epoch": 0.8584457412545042,
+      "grad_norm": 0.19631756842136383,
+      "learning_rate": 6.439934920149081e-05,
+      "loss": 0.8797,
+      "step": 26920
+    },
+    {
+      "epoch": 0.8587646289741382,
+      "grad_norm": 0.18874439597129822,
+      "learning_rate": 6.412970808033838e-05,
+      "loss": 0.9002,
+      "step": 26930
+    },
+    {
+      "epoch": 0.8590835166937721,
+      "grad_norm": 0.19086116552352905,
+      "learning_rate": 6.386119595094003e-05,
+      "loss": 0.8796,
+      "step": 26940
+    },
+    {
+      "epoch": 0.859402404413406,
+      "grad_norm": 0.1910838931798935,
+      "learning_rate": 6.359380808618895e-05,
+      "loss": 0.8777,
+      "step": 26950
+    },
+    {
+      "epoch": 0.8597212921330399,
+      "grad_norm": 0.1921229064464569,
+      "learning_rate": 6.332753977877079e-05,
+      "loss": 0.8886,
+      "step": 26960
+    },
+    {
+      "epoch": 0.8600401798526739,
+      "grad_norm": 0.19134828448295593,
+      "learning_rate": 6.306238634108082e-05,
+      "loss": 0.8638,
+      "step": 26970
+    },
+    {
+      "epoch": 0.8603590675723078,
+      "grad_norm": 0.1976134330034256,
+      "learning_rate": 6.279834310514136e-05,
+      "loss": 0.8728,
+      "step": 26980
+    },
+    {
+      "epoch": 0.8606779552919417,
+      "grad_norm": 0.19609402120113373,
+      "learning_rate": 6.253540542251968e-05,
+      "loss": 0.8825,
+      "step": 26990
+    },
+    {
+      "epoch": 0.8609968430115756,
+      "grad_norm": 0.19391323626041412,
+      "learning_rate": 6.227356866424601e-05,
+      "loss": 0.8678,
+      "step": 27000
+    },
+    {
+      "epoch": 0.8613157307312096,
+      "grad_norm": 0.19331073760986328,
+      "learning_rate": 6.201282822073233e-05,
+      "loss": 0.8655,
+      "step": 27010
+    },
+    {
+      "epoch": 0.8616346184508434,
+      "grad_norm": 0.19262798130512238,
+      "learning_rate": 6.175317950169087e-05,
+      "loss": 0.8838,
+      "step": 27020
+    },
+    {
+      "epoch": 0.8619535061704774,
+      "grad_norm": 0.1916375756263733,
+      "learning_rate": 6.149461793605354e-05,
+      "loss": 0.8883,
+      "step": 27030
+    },
+    {
+      "epoch": 0.8622723938901113,
+      "grad_norm": 0.18959954380989075,
+      "learning_rate": 6.123713897189136e-05,
+      "loss": 0.9003,
+      "step": 27040
+    },
+    {
+      "epoch": 0.8625912816097452,
+      "grad_norm": 0.19482597708702087,
+      "learning_rate": 6.0980738076334414e-05,
+      "loss": 0.9003,
+      "step": 27050
+    },
+    {
+      "epoch": 0.8629101693293791,
+      "grad_norm": 0.1904841810464859,
+      "learning_rate": 6.0725410735491895e-05,
+      "loss": 0.8917,
+      "step": 27060
+    },
+    {
+      "epoch": 0.8632290570490131,
+      "grad_norm": 0.19442743062973022,
+      "learning_rate": 6.0471152454372806e-05,
+      "loss": 0.877,
+      "step": 27070
+    },
+    {
+      "epoch": 0.8635479447686469,
+      "grad_norm": 0.18793179094791412,
+      "learning_rate": 6.02179587568067e-05,
+      "loss": 0.8927,
+      "step": 27080
+    },
+    {
+      "epoch": 0.8638668324882809,
+      "grad_norm": 0.20632310211658478,
+      "learning_rate": 5.9965825185364964e-05,
+      "loss": 0.8945,
+      "step": 27090
+    },
+    {
+      "epoch": 0.8641857202079148,
+      "grad_norm": 0.18707668781280518,
+      "learning_rate": 5.971474730128228e-05,
+      "loss": 0.8852,
+      "step": 27100
+    },
+    {
+      "epoch": 0.8645046079275487,
+      "grad_norm": 0.18785780668258667,
+      "learning_rate": 5.946472068437854e-05,
+      "loss": 0.8763,
+      "step": 27110
+    },
+    {
+      "epoch": 0.8648234956471826,
+      "grad_norm": 0.19697052240371704,
+      "learning_rate": 5.9215740932980986e-05,
+      "loss": 0.8899,
+      "step": 27120
+    },
+    {
+      "epoch": 0.8651423833668166,
+      "grad_norm": 0.19185379147529602,
+      "learning_rate": 5.8967803663846734e-05,
+      "loss": 0.8746,
+      "step": 27130
+    },
+    {
+      "epoch": 0.8654612710864504,
+      "grad_norm": 0.19396957755088806,
+      "learning_rate": 5.8720904512085626e-05,
+      "loss": 0.8683,
+      "step": 27140
+    },
+    {
+      "epoch": 0.8657801588060844,
+      "grad_norm": 0.18885491788387299,
+      "learning_rate": 5.8475039131083355e-05,
+      "loss": 0.8838,
+      "step": 27150
+    },
+    {
+      "epoch": 0.8660990465257183,
+      "grad_norm": 0.19360066950321198,
+      "learning_rate": 5.823020319242495e-05,
+      "loss": 0.8922,
+      "step": 27160
+    },
+    {
+      "epoch": 0.8664179342453522,
+      "grad_norm": 0.19725437462329865,
+      "learning_rate": 5.7986392385818584e-05,
+      "loss": 0.8765,
+      "step": 27170
+    },
+    {
+      "epoch": 0.8667368219649861,
+      "grad_norm": 0.19328470528125763,
+      "learning_rate": 5.774360241901975e-05,
+      "loss": 0.8879,
+      "step": 27180
+    },
+    {
+      "epoch": 0.8670557096846201,
+      "grad_norm": 0.19106145203113556,
+      "learning_rate": 5.7501829017755564e-05,
+      "loss": 0.8834,
+      "step": 27190
+    },
+    {
+      "epoch": 0.8673745974042539,
+      "grad_norm": 0.19127941131591797,
+      "learning_rate": 5.7261067925649635e-05,
+      "loss": 0.8797,
+      "step": 27200
+    },
+    {
+      "epoch": 0.8676934851238879,
+      "grad_norm": 0.19829584658145905,
+      "learning_rate": 5.7021314904147045e-05,
+      "loss": 0.8799,
+      "step": 27210
+    },
+    {
+      "epoch": 0.8680123728435218,
+      "grad_norm": 0.19009840488433838,
+      "learning_rate": 5.678256573243984e-05,
+      "loss": 0.8787,
+      "step": 27220
+    },
+    {
+      "epoch": 0.8683312605631557,
+      "grad_norm": 0.1915004998445511,
+      "learning_rate": 5.6544816207392587e-05,
+      "loss": 0.8743,
+      "step": 27230
+    },
+    {
+      "epoch": 0.8686501482827896,
+      "grad_norm": 0.19615386426448822,
+      "learning_rate": 5.630806214346851e-05,
+      "loss": 0.8757,
+      "step": 27240
+    },
+    {
+      "epoch": 0.8689690360024236,
+      "grad_norm": 0.19712796807289124,
+      "learning_rate": 5.6072299372655695e-05,
+      "loss": 0.8872,
+      "step": 27250
+    },
+    {
+      "epoch": 0.8692879237220574,
+      "grad_norm": 0.19148242473602295,
+      "learning_rate": 5.58375237443938e-05,
+      "loss": 0.8766,
+      "step": 27260
+    },
+    {
+      "epoch": 0.8696068114416914,
+      "grad_norm": 0.19317524135112762,
+      "learning_rate": 5.5603731125500924e-05,
+      "loss": 0.8802,
+      "step": 27270
+    },
+    {
+      "epoch": 0.8699256991613253,
+      "grad_norm": 0.19434620440006256,
+      "learning_rate": 5.537091740010087e-05,
+      "loss": 0.8769,
+      "step": 27280
+    },
+    {
+      "epoch": 0.8702445868809592,
+      "grad_norm": 0.19037990272045135,
+      "learning_rate": 5.513907846955069e-05,
+      "loss": 0.8812,
+      "step": 27290
+    },
+    {
+      "epoch": 0.8705634746005931,
+      "grad_norm": 0.19795745611190796,
+      "learning_rate": 5.490821025236851e-05,
+      "loss": 0.8913,
+      "step": 27300
+    },
+    {
+      "epoch": 0.8708823623202271,
+      "grad_norm": 0.2026415467262268,
+      "learning_rate": 5.467830868416169e-05,
+      "loss": 0.9003,
+      "step": 27310
+    },
+    {
+      "epoch": 0.8712012500398609,
+      "grad_norm": 0.18853889405727386,
+      "learning_rate": 5.444936971755526e-05,
+      "loss": 0.8757,
+      "step": 27320
+    },
+    {
+      "epoch": 0.8715201377594949,
+      "grad_norm": 0.18458408117294312,
+      "learning_rate": 5.42213893221207e-05,
+      "loss": 0.8898,
+      "step": 27330
+    },
+    {
+      "epoch": 0.8718390254791288,
+      "grad_norm": 0.1918436586856842,
+      "learning_rate": 5.399436348430491e-05,
+      "loss": 0.8927,
+      "step": 27340
+    },
+    {
+      "epoch": 0.8721579131987627,
+      "grad_norm": 0.19048848748207092,
+      "learning_rate": 5.3768288207359675e-05,
+      "loss": 0.8763,
+      "step": 27350
+    },
+    {
+      "epoch": 0.8724768009183966,
+      "grad_norm": 0.19431082904338837,
+      "learning_rate": 5.354315951127119e-05,
+      "loss": 0.8899,
+      "step": 27360
+    },
+    {
+      "epoch": 0.8727956886380306,
+      "grad_norm": 0.18930363655090332,
+      "learning_rate": 5.331897343269001e-05,
+      "loss": 0.8735,
+      "step": 27370
+    },
+    {
+      "epoch": 0.8731145763576644,
+      "grad_norm": 0.19170422852039337,
+      "learning_rate": 5.309572602486132e-05,
+      "loss": 0.8802,
+      "step": 27380
+    },
+    {
+      "epoch": 0.8734334640772984,
+      "grad_norm": 0.1955621987581253,
+      "learning_rate": 5.287341335755546e-05,
+      "loss": 0.883,
+      "step": 27390
+    },
+    {
+      "epoch": 0.8737523517969323,
+      "grad_norm": 0.19437985122203827,
+      "learning_rate": 5.265203151699865e-05,
+      "loss": 0.8785,
+      "step": 27400
+    },
+    {
+      "epoch": 0.8740712395165662,
+      "grad_norm": 0.19519606232643127,
+      "learning_rate": 5.243157660580418e-05,
+      "loss": 0.8694,
+      "step": 27410
+    },
+    {
+      "epoch": 0.8743901272362001,
+      "grad_norm": 0.19466404616832733,
+      "learning_rate": 5.221204474290376e-05,
+      "loss": 0.8751,
+      "step": 27420
+    },
+    {
+      "epoch": 0.8747090149558341,
+      "grad_norm": 0.19154104590415955,
+      "learning_rate": 5.199343206347918e-05,
+      "loss": 0.8816,
+      "step": 27430
+    },
+    {
+      "epoch": 0.8750279026754679,
+      "grad_norm": 0.1949530988931656,
+      "learning_rate": 5.1775734718894304e-05,
+      "loss": 0.8758,
+      "step": 27440
+    },
+    {
+      "epoch": 0.8753467903951019,
+      "grad_norm": 0.19437164068222046,
+      "learning_rate": 5.155894887662728e-05,
+      "loss": 0.8815,
+      "step": 27450
+    },
+    {
+      "epoch": 0.8756656781147358,
+      "grad_norm": 0.19124570488929749,
+      "learning_rate": 5.134307072020311e-05,
+      "loss": 0.878,
+      "step": 27460
+    },
+    {
+      "epoch": 0.8759845658343697,
+      "grad_norm": 0.1909313052892685,
+      "learning_rate": 5.1128096449126425e-05,
+      "loss": 0.887,
+      "step": 27470
+    },
+    {
+      "epoch": 0.8763034535540036,
+      "grad_norm": 0.19552530348300934,
+      "learning_rate": 5.091402227881458e-05,
+      "loss": 0.877,
+      "step": 27480
+    },
+    {
+      "epoch": 0.8766223412736376,
+      "grad_norm": 0.19062495231628418,
+      "learning_rate": 5.070084444053106e-05,
+      "loss": 0.871,
+      "step": 27490
+    },
+    {
+      "epoch": 0.8769412289932714,
+      "grad_norm": 0.19744792580604553,
+      "learning_rate": 5.0488559181319116e-05,
+      "loss": 0.8901,
+      "step": 27500
+    },
+    {
+      "epoch": 0.8772601167129054,
+      "grad_norm": 0.19660893082618713,
+      "learning_rate": 5.027716276393563e-05,
+      "loss": 0.8743,
+      "step": 27510
+    },
+    {
+      "epoch": 0.8775790044325393,
+      "grad_norm": 0.19473454356193542,
+      "learning_rate": 5.00666514667855e-05,
+      "loss": 0.8898,
+      "step": 27520
+    },
+    {
+      "epoch": 0.8778978921521732,
+      "grad_norm": 0.1918601542711258,
+      "learning_rate": 4.9857021583855885e-05,
+      "loss": 0.8592,
+      "step": 27530
+    },
+    {
+      "epoch": 0.8782167798718071,
+      "grad_norm": 0.18904858827590942,
+      "learning_rate": 4.9648269424651144e-05,
+      "loss": 0.8803,
+      "step": 27540
+    },
+    {
+      "epoch": 0.8785356675914411,
+      "grad_norm": 0.19310516119003296,
+      "learning_rate": 4.9440391314127776e-05,
+      "loss": 0.8841,
+      "step": 27550
+    },
+    {
+      "epoch": 0.8788545553110749,
+      "grad_norm": 0.1876547634601593,
+      "learning_rate": 4.9233383592629755e-05,
+      "loss": 0.8776,
+      "step": 27560
+    },
+    {
+      "epoch": 0.8791734430307089,
+      "grad_norm": 0.1921987533569336,
+      "learning_rate": 4.90272426158241e-05,
+      "loss": 0.8773,
+      "step": 27570
+    },
+    {
+      "epoch": 0.8794923307503428,
+      "grad_norm": 0.19707876443862915,
+      "learning_rate": 4.88219647546367e-05,
+      "loss": 0.8824,
+      "step": 27580
+    },
+    {
+      "epoch": 0.8798112184699767,
+      "grad_norm": 0.1927400529384613,
+      "learning_rate": 4.861754639518844e-05,
+      "loss": 0.8784,
+      "step": 27590
+    },
+    {
+      "epoch": 0.8801301061896106,
+      "grad_norm": 0.19577376544475555,
+      "learning_rate": 4.841398393873158e-05,
+      "loss": 0.8793,
+      "step": 27600
+    },
+    {
+      "epoch": 0.8804489939092446,
+      "grad_norm": 0.1940997689962387,
+      "learning_rate": 4.821127380158639e-05,
+      "loss": 0.8756,
+      "step": 27610
+    },
+    {
+      "epoch": 0.8807678816288784,
+      "grad_norm": 0.19283920526504517,
+      "learning_rate": 4.800941241507807e-05,
+      "loss": 0.8822,
+      "step": 27620
+    },
+    {
+      "epoch": 0.8810867693485124,
+      "grad_norm": 0.19780674576759338,
+      "learning_rate": 4.7808396225473926e-05,
+      "loss": 0.8815,
+      "step": 27630
+    },
+    {
+      "epoch": 0.8814056570681463,
+      "grad_norm": 0.19866904616355896,
+      "learning_rate": 4.760822169392078e-05,
+      "loss": 0.8766,
+      "step": 27640
+    },
+    {
+      "epoch": 0.8817245447877802,
+      "grad_norm": 0.19428886473178864,
+      "learning_rate": 4.74088852963827e-05,
+      "loss": 0.874,
+      "step": 27650
+    },
+    {
+      "epoch": 0.8820434325074141,
+      "grad_norm": 0.19498443603515625,
+      "learning_rate": 4.7210383523578954e-05,
+      "loss": 0.8894,
+      "step": 27660
+    },
+    {
+      "epoch": 0.8823623202270481,
+      "grad_norm": 0.19383135437965393,
+      "learning_rate": 4.701271288092223e-05,
+      "loss": 0.8709,
+      "step": 27670
+    },
+    {
+      "epoch": 0.8826812079466819,
+      "grad_norm": 0.19539567828178406,
+      "learning_rate": 4.681586988845707e-05,
+      "loss": 0.8806,
+      "step": 27680
+    },
+    {
+      "epoch": 0.8830000956663159,
+      "grad_norm": 0.1919242888689041,
+      "learning_rate": 4.661985108079873e-05,
+      "loss": 0.8994,
+      "step": 27690
+    },
+    {
+      "epoch": 0.8833189833859498,
+      "grad_norm": 0.19845853745937347,
+      "learning_rate": 4.6424653007072e-05,
+      "loss": 0.8786,
+      "step": 27700
+    },
+    {
+      "epoch": 0.8836378711055837,
+      "grad_norm": 0.1964893341064453,
+      "learning_rate": 4.62302722308506e-05,
+      "loss": 0.8652,
+      "step": 27710
+    },
+    {
+      "epoch": 0.8839567588252176,
+      "grad_norm": 0.1927543431520462,
+      "learning_rate": 4.603670533009658e-05,
+      "loss": 0.8723,
+      "step": 27720
+    },
+    {
+      "epoch": 0.8842756465448516,
+      "grad_norm": 0.1906014084815979,
+      "learning_rate": 4.5843948897100144e-05,
+      "loss": 0.8771,
+      "step": 27730
+    },
+    {
+      "epoch": 0.8845945342644854,
+      "grad_norm": 0.19443483650684357,
+      "learning_rate": 4.565199953841964e-05,
+      "loss": 0.8573,
+      "step": 27740
+    },
+    {
+      "epoch": 0.8849134219841194,
+      "grad_norm": 0.18828009068965912,
+      "learning_rate": 4.5460853874821776e-05,
+      "loss": 0.8908,
+      "step": 27750
+    },
+    {
+      "epoch": 0.8852323097037533,
+      "grad_norm": 0.19496281445026398,
+      "learning_rate": 4.52705085412222e-05,
+      "loss": 0.8787,
+      "step": 27760
+    },
+    {
+      "epoch": 0.8855511974233872,
+      "grad_norm": 0.1892516016960144,
+      "learning_rate": 4.50809601866262e-05,
+      "loss": 0.8804,
+      "step": 27770
+    },
+    {
+      "epoch": 0.8858700851430211,
+      "grad_norm": 0.19120721518993378,
+      "learning_rate": 4.4892205474069755e-05,
+      "loss": 0.8658,
+      "step": 27780
+    },
+    {
+      "epoch": 0.8861889728626551,
+      "grad_norm": 0.1939810812473297,
+      "learning_rate": 4.4704241080560745e-05,
+      "loss": 0.8731,
+      "step": 27790
+    },
+    {
+      "epoch": 0.886507860582289,
+      "grad_norm": 0.1969224512577057,
+      "learning_rate": 4.4517063697020485e-05,
+      "loss": 0.8849,
+      "step": 27800
+    },
+    {
+      "epoch": 0.8868267483019229,
+      "grad_norm": 0.19289876520633698,
+      "learning_rate": 4.433067002822545e-05,
+      "loss": 0.8784,
+      "step": 27810
+    },
+    {
+      "epoch": 0.8871456360215568,
+      "grad_norm": 0.1962817758321762,
+      "learning_rate": 4.414505679274928e-05,
+      "loss": 0.8611,
+      "step": 27820
+    },
+    {
+      "epoch": 0.8874645237411908,
+      "grad_norm": 0.19316478073596954,
+      "learning_rate": 4.396022072290497e-05,
+      "loss": 0.8717,
+      "step": 27830
+    },
+    {
+      "epoch": 0.8877834114608246,
+      "grad_norm": 0.19402272999286652,
+      "learning_rate": 4.3776158564687404e-05,
+      "loss": 0.8757,
+      "step": 27840
+    },
+    {
+      "epoch": 0.8881022991804586,
+      "grad_norm": 0.19799882173538208,
+      "learning_rate": 4.359286707771605e-05,
+      "loss": 0.8913,
+      "step": 27850
+    },
+    {
+      "epoch": 0.8884211869000925,
+      "grad_norm": 0.19214606285095215,
+      "learning_rate": 4.341034303517787e-05,
+      "loss": 0.8773,
+      "step": 27860
+    },
+    {
+      "epoch": 0.8887400746197264,
+      "grad_norm": 0.18697836995124817,
+      "learning_rate": 4.322858322377056e-05,
+      "loss": 0.8758,
+      "step": 27870
+    },
+    {
+      "epoch": 0.8890589623393603,
+      "grad_norm": 0.20322303473949432,
+      "learning_rate": 4.3047584443645955e-05,
+      "loss": 0.8815,
+      "step": 27880
+    },
+    {
+      "epoch": 0.8893778500589943,
+      "grad_norm": 0.19178463518619537,
+      "learning_rate": 4.2867343508353714e-05,
+      "loss": 0.8632,
+      "step": 27890
+    },
+    {
+      "epoch": 0.8896967377786281,
+      "grad_norm": 0.18942740559577942,
+      "learning_rate": 4.2687857244785214e-05,
+      "loss": 0.8832,
+      "step": 27900
+    },
+    {
+      "epoch": 0.8900156254982621,
+      "grad_norm": 0.187669038772583,
+      "learning_rate": 4.2509122493117696e-05,
+      "loss": 0.8843,
+      "step": 27910
+    },
+    {
+      "epoch": 0.890334513217896,
+      "grad_norm": 0.19848766922950745,
+      "learning_rate": 4.2331136106758614e-05,
+      "loss": 0.8919,
+      "step": 27920
+    },
+    {
+      "epoch": 0.8906534009375299,
+      "grad_norm": 0.19923046231269836,
+      "learning_rate": 4.2153894952290276e-05,
+      "loss": 0.8728,
+      "step": 27930
+    },
+    {
+      "epoch": 0.8909722886571638,
+      "grad_norm": 0.19249597191810608,
+      "learning_rate": 4.1977395909414635e-05,
+      "loss": 0.8744,
+      "step": 27940
+    },
+    {
+      "epoch": 0.8912911763767978,
+      "grad_norm": 0.19503191113471985,
+      "learning_rate": 4.180163587089841e-05,
+      "loss": 0.8794,
+      "step": 27950
+    },
+    {
+      "epoch": 0.8916100640964316,
+      "grad_norm": 0.19658716022968292,
+      "learning_rate": 4.162661174251835e-05,
+      "loss": 0.8863,
+      "step": 27960
+    },
+    {
+      "epoch": 0.8919289518160656,
+      "grad_norm": 0.20051929354667664,
+      "learning_rate": 4.1452320443006753e-05,
+      "loss": 0.8899,
+      "step": 27970
+    },
+    {
+      "epoch": 0.8922478395356995,
+      "grad_norm": 0.19176195561885834,
+      "learning_rate": 4.1278758903997235e-05,
+      "loss": 0.8968,
+      "step": 27980
+    },
+    {
+      "epoch": 0.8925667272553334,
+      "grad_norm": 0.19284586608409882,
+      "learning_rate": 4.110592406997074e-05,
+      "loss": 0.8748,
+      "step": 27990
+    },
+    {
+      "epoch": 0.8928856149749673,
+      "grad_norm": 0.18788652122020721,
+      "learning_rate": 4.0933812898201676e-05,
+      "loss": 0.8764,
+      "step": 28000
+    },
+    {
+      "epoch": 0.8932045026946013,
+      "grad_norm": 0.19523321092128754,
+      "learning_rate": 4.0762422358704414e-05,
+      "loss": 0.8765,
+      "step": 28010
+    },
+    {
+      "epoch": 0.8935233904142351,
+      "grad_norm": 0.18856965005397797,
+      "learning_rate": 4.059174943417997e-05,
+      "loss": 0.8778,
+      "step": 28020
+    },
+    {
+      "epoch": 0.8938422781338691,
+      "grad_norm": 0.188978374004364,
+      "learning_rate": 4.0421791119962764e-05,
+      "loss": 0.8822,
+      "step": 28030
+    },
+    {
+      "epoch": 0.894161165853503,
+      "grad_norm": 0.18726089596748352,
+      "learning_rate": 4.0252544423967835e-05,
+      "loss": 0.8723,
+      "step": 28040
+    },
+    {
+      "epoch": 0.8944800535731369,
+      "grad_norm": 0.19660618901252747,
+      "learning_rate": 4.008400636663814e-05,
+      "loss": 0.8619,
+      "step": 28050
+    },
+    {
+      "epoch": 0.8947989412927708,
+      "grad_norm": 0.1990547776222229,
+      "learning_rate": 3.991617398089205e-05,
+      "loss": 0.8734,
+      "step": 28060
+    },
+    {
+      "epoch": 0.8951178290124048,
+      "grad_norm": 0.1974078267812729,
+      "learning_rate": 3.974904431207121e-05,
+      "loss": 0.8741,
+      "step": 28070
+    },
+    {
+      "epoch": 0.8954367167320386,
+      "grad_norm": 0.1837972104549408,
+      "learning_rate": 3.958261441788843e-05,
+      "loss": 0.8747,
+      "step": 28080
+    },
+    {
+      "epoch": 0.8957556044516726,
+      "grad_norm": 0.20282982289791107,
+      "learning_rate": 3.941688136837593e-05,
+      "loss": 0.871,
+      "step": 28090
+    },
+    {
+      "epoch": 0.8960744921713065,
+      "grad_norm": 0.19532689452171326,
+      "learning_rate": 3.925184224583376e-05,
+      "loss": 0.8797,
+      "step": 28100
+    },
+    {
+      "epoch": 0.8963933798909404,
+      "grad_norm": 0.19317756593227386,
+      "learning_rate": 3.908749414477842e-05,
+      "loss": 0.8684,
+      "step": 28110
+    },
+    {
+      "epoch": 0.8967122676105743,
+      "grad_norm": 0.19880421459674835,
+      "learning_rate": 3.8923834171891724e-05,
+      "loss": 0.8912,
+      "step": 28120
+    },
+    {
+      "epoch": 0.8970311553302083,
+      "grad_norm": 0.1922730654478073,
+      "learning_rate": 3.876085944596985e-05,
+      "loss": 0.8719,
+      "step": 28130
+    },
+    {
+      "epoch": 0.8973500430498421,
+      "grad_norm": 0.20053960382938385,
+      "learning_rate": 3.8598567097872634e-05,
+      "loss": 0.8934,
+      "step": 28140
+    },
+    {
+      "epoch": 0.8976689307694761,
+      "grad_norm": 0.1984051764011383,
+      "learning_rate": 3.843695427047303e-05,
+      "loss": 0.8794,
+      "step": 28150
+    },
+    {
+      "epoch": 0.89798781848911,
+      "grad_norm": 0.19294370710849762,
+      "learning_rate": 3.8276018118606846e-05,
+      "loss": 0.8787,
+      "step": 28160
+    },
+    {
+      "epoch": 0.8983067062087439,
+      "grad_norm": 0.19863317906856537,
+      "learning_rate": 3.8115755809022624e-05,
+      "loss": 0.8681,
+      "step": 28170
+    },
+    {
+      "epoch": 0.8986255939283778,
+      "grad_norm": 0.1922876238822937,
+      "learning_rate": 3.7956164520331766e-05,
+      "loss": 0.8711,
+      "step": 28180
+    },
+    {
+      "epoch": 0.8989444816480118,
+      "grad_norm": 0.1919499635696411,
+      "learning_rate": 3.7797241442958946e-05,
+      "loss": 0.8707,
+      "step": 28190
+    },
+    {
+      "epoch": 0.8992633693676456,
+      "grad_norm": 0.19040320813655853,
+      "learning_rate": 3.7638983779092486e-05,
+      "loss": 0.8747,
+      "step": 28200
+    },
+    {
+      "epoch": 0.8995822570872796,
+      "grad_norm": 0.19564911723136902,
+      "learning_rate": 3.748138874263523e-05,
+      "loss": 0.8796,
+      "step": 28210
+    },
+    {
+      "epoch": 0.8999011448069135,
+      "grad_norm": 0.19297778606414795,
+      "learning_rate": 3.732445355915546e-05,
+      "loss": 0.8868,
+      "step": 28220
+    },
+    {
+      "epoch": 0.9002200325265474,
+      "grad_norm": 0.18817225098609924,
+      "learning_rate": 3.716817546583803e-05,
+      "loss": 0.8635,
+      "step": 28230
+    },
+    {
+      "epoch": 0.9005389202461813,
+      "grad_norm": 0.19400273263454437,
+      "learning_rate": 3.7012551711435764e-05,
+      "loss": 0.884,
+      "step": 28240
+    },
+    {
+      "epoch": 0.9008578079658153,
+      "grad_norm": 0.19113945960998535,
+      "learning_rate": 3.6857579556221e-05,
+      "loss": 0.8822,
+      "step": 28250
+    },
+    {
+      "epoch": 0.9011766956854491,
+      "grad_norm": 0.19377689063549042,
+      "learning_rate": 3.670325627193734e-05,
+      "loss": 0.8593,
+      "step": 28260
+    },
+    {
+      "epoch": 0.9014955834050831,
+      "grad_norm": 0.19520173966884613,
+      "learning_rate": 3.654957914175167e-05,
+      "loss": 0.882,
+      "step": 28270
+    },
+    {
+      "epoch": 0.901814471124717,
+      "grad_norm": 0.1930304616689682,
+      "learning_rate": 3.639654546020629e-05,
+      "loss": 0.8676,
+      "step": 28280
+    },
+    {
+      "epoch": 0.9021333588443509,
+      "grad_norm": 0.18980911374092102,
+      "learning_rate": 3.6244152533171275e-05,
+      "loss": 0.8825,
+      "step": 28290
+    },
+    {
+      "epoch": 0.9024522465639848,
+      "grad_norm": 0.20174168050289154,
+      "learning_rate": 3.609239767779709e-05,
+      "loss": 0.902,
+      "step": 28300
+    },
+    {
+      "epoch": 0.9027711342836188,
+      "grad_norm": 0.18596351146697998,
+      "learning_rate": 3.594127822246733e-05,
+      "loss": 0.8835,
+      "step": 28310
+    },
+    {
+      "epoch": 0.9030900220032526,
+      "grad_norm": 0.1928795576095581,
+      "learning_rate": 3.579079150675168e-05,
+      "loss": 0.8707,
+      "step": 28320
+    },
+    {
+      "epoch": 0.9034089097228866,
+      "grad_norm": 0.18944554030895233,
+      "learning_rate": 3.564093488135911e-05,
+      "loss": 0.8735,
+      "step": 28330
+    },
+    {
+      "epoch": 0.9037277974425205,
+      "grad_norm": 0.18926158547401428,
+      "learning_rate": 3.54917057080912e-05,
+      "loss": 0.8601,
+      "step": 28340
+    },
+    {
+      "epoch": 0.9040466851621544,
+      "grad_norm": 0.19643589854240417,
+      "learning_rate": 3.534310135979569e-05,
+      "loss": 0.882,
+      "step": 28350
+    },
+    {
+      "epoch": 0.9043655728817883,
+      "grad_norm": 0.19318707287311554,
+      "learning_rate": 3.51951192203203e-05,
+      "loss": 0.8793,
+      "step": 28360
+    },
+    {
+      "epoch": 0.9046844606014223,
+      "grad_norm": 0.1926054209470749,
+      "learning_rate": 3.504775668446659e-05,
+      "loss": 0.8793,
+      "step": 28370
+    },
+    {
+      "epoch": 0.9050033483210561,
+      "grad_norm": 0.19345009326934814,
+      "learning_rate": 3.4901011157944126e-05,
+      "loss": 0.867,
+      "step": 28380
+    },
+    {
+      "epoch": 0.9053222360406901,
+      "grad_norm": 0.19208772480487823,
+      "learning_rate": 3.475488005732481e-05,
+      "loss": 0.8662,
+      "step": 28390
+    },
+    {
+      "epoch": 0.905641123760324,
+      "grad_norm": 0.19586001336574554,
+      "learning_rate": 3.460936080999741e-05,
+      "loss": 0.8716,
+      "step": 28400
+    },
+    {
+      "epoch": 0.9059600114799579,
+      "grad_norm": 0.20053620636463165,
+      "learning_rate": 3.4464450854122266e-05,
+      "loss": 0.8779,
+      "step": 28410
+    },
+    {
+      "epoch": 0.9062788991995918,
+      "grad_norm": 0.19501394033432007,
+      "learning_rate": 3.4320147638586154e-05,
+      "loss": 0.8713,
+      "step": 28420
+    },
+    {
+      "epoch": 0.9065977869192258,
+      "grad_norm": 0.19111500680446625,
+      "learning_rate": 3.4176448622957427e-05,
+      "loss": 0.8946,
+      "step": 28430
+    },
+    {
+      "epoch": 0.9069166746388596,
+      "grad_norm": 0.19476771354675293,
+      "learning_rate": 3.4033351277441255e-05,
+      "loss": 0.8855,
+      "step": 28440
+    },
+    {
+      "epoch": 0.9072355623584936,
+      "grad_norm": 0.19051986932754517,
+      "learning_rate": 3.38908530828351e-05,
+      "loss": 0.8852,
+      "step": 28450
+    },
+    {
+      "epoch": 0.9075544500781275,
+      "grad_norm": 0.1968327760696411,
+      "learning_rate": 3.374895153048438e-05,
+      "loss": 0.87,
+      "step": 28460
+    },
+    {
+      "epoch": 0.9078733377977614,
+      "grad_norm": 0.1960689276456833,
+      "learning_rate": 3.360764412223827e-05,
+      "loss": 0.8775,
+      "step": 28470
+    },
+    {
+      "epoch": 0.9081922255173953,
+      "grad_norm": 0.1949181854724884,
+      "learning_rate": 3.3466928370405766e-05,
+      "loss": 0.8774,
+      "step": 28480
+    },
+    {
+      "epoch": 0.9085111132370293,
+      "grad_norm": 0.20106466114521027,
+      "learning_rate": 3.332680179771184e-05,
+      "loss": 0.882,
+      "step": 28490
+    },
+    {
+      "epoch": 0.9088300009566631,
+      "grad_norm": 0.1960950791835785,
+      "learning_rate": 3.318726193725388e-05,
+      "loss": 0.8785,
+      "step": 28500
+    },
+    {
+      "epoch": 0.9091488886762971,
+      "grad_norm": 0.18996301293373108,
+      "learning_rate": 3.304830633245822e-05,
+      "loss": 0.8744,
+      "step": 28510
+    },
+    {
+      "epoch": 0.909467776395931,
+      "grad_norm": 0.18958501517772675,
+      "learning_rate": 3.290993253703689e-05,
+      "loss": 0.8837,
+      "step": 28520
+    },
+    {
+      "epoch": 0.9097866641155649,
+      "grad_norm": 0.1934012621641159,
+      "learning_rate": 3.277213811494463e-05,
+      "loss": 0.85,
+      "step": 28530
+    },
+    {
+      "epoch": 0.9101055518351988,
+      "grad_norm": 0.18933558464050293,
+      "learning_rate": 3.263492064033587e-05,
+      "loss": 0.8757,
+      "step": 28540
+    },
+    {
+      "epoch": 0.9104244395548328,
+      "grad_norm": 0.1904170960187912,
+      "learning_rate": 3.249827769752215e-05,
+      "loss": 0.8675,
+      "step": 28550
+    },
+    {
+      "epoch": 0.9107433272744666,
+      "grad_norm": 0.1956118792295456,
+      "learning_rate": 3.2362206880929494e-05,
+      "loss": 0.8857,
+      "step": 28560
+    },
+    {
+      "epoch": 0.9110622149941006,
+      "grad_norm": 0.1950976848602295,
+      "learning_rate": 3.222670579505612e-05,
+      "loss": 0.8793,
+      "step": 28570
+    },
+    {
+      "epoch": 0.9113811027137345,
+      "grad_norm": 0.19186055660247803,
+      "learning_rate": 3.2091772054430256e-05,
+      "loss": 0.88,
+      "step": 28580
+    },
+    {
+      "epoch": 0.9116999904333684,
+      "grad_norm": 0.1933923065662384,
+      "learning_rate": 3.195740328356815e-05,
+      "loss": 0.8882,
+      "step": 28590
+    },
+    {
+      "epoch": 0.9120188781530023,
+      "grad_norm": 0.20145314931869507,
+      "learning_rate": 3.182359711693219e-05,
+      "loss": 0.8918,
+      "step": 28600
+    },
+    {
+      "epoch": 0.9123377658726363,
+      "grad_norm": 0.20328111946582794,
+      "learning_rate": 3.169035119888938e-05,
+      "loss": 0.8868,
+      "step": 28610
+    },
+    {
+      "epoch": 0.9126566535922701,
+      "grad_norm": 0.18951085209846497,
+      "learning_rate": 3.155766318366973e-05,
+      "loss": 0.8737,
+      "step": 28620
+    },
+    {
+      "epoch": 0.9129755413119041,
+      "grad_norm": 0.18694986402988434,
+      "learning_rate": 3.142553073532508e-05,
+      "loss": 0.8656,
+      "step": 28630
+    },
+    {
+      "epoch": 0.913294429031538,
+      "grad_norm": 0.19038134813308716,
+      "learning_rate": 3.129395152768789e-05,
+      "loss": 0.8748,
+      "step": 28640
+    },
+    {
+      "epoch": 0.913613316751172,
+      "grad_norm": 0.19913242757320404,
+      "learning_rate": 3.116292324433033e-05,
+      "loss": 0.8684,
+      "step": 28650
+    },
+    {
+      "epoch": 0.9139322044708058,
+      "grad_norm": 0.20065714418888092,
+      "learning_rate": 3.103244357852349e-05,
+      "loss": 0.8734,
+      "step": 28660
+    },
+    {
+      "epoch": 0.9142510921904398,
+      "grad_norm": 0.19351083040237427,
+      "learning_rate": 3.090251023319679e-05,
+      "loss": 0.8794,
+      "step": 28670
+    },
+    {
+      "epoch": 0.9145699799100737,
+      "grad_norm": 0.1949431151151657,
+      "learning_rate": 3.0773120920897486e-05,
+      "loss": 0.8843,
+      "step": 28680
+    },
+    {
+      "epoch": 0.9148888676297076,
+      "grad_norm": 0.19440433382987976,
+      "learning_rate": 3.064427336375045e-05,
+      "loss": 0.8854,
+      "step": 28690
+    },
+    {
+      "epoch": 0.9152077553493415,
+      "grad_norm": 0.19945016503334045,
+      "learning_rate": 3.0515965293418096e-05,
+      "loss": 0.8745,
+      "step": 28700
+    },
+    {
+      "epoch": 0.9155266430689755,
+      "grad_norm": 0.19426259398460388,
+      "learning_rate": 3.038819445106034e-05,
+      "loss": 0.877,
+      "step": 28710
+    },
+    {
+      "epoch": 0.9158455307886093,
+      "grad_norm": 0.19654859602451324,
+      "learning_rate": 3.0260958587294935e-05,
+      "loss": 0.8796,
+      "step": 28720
+    },
+    {
+      "epoch": 0.9161644185082433,
+      "grad_norm": 0.19408033788204193,
+      "learning_rate": 3.013425546215782e-05,
+      "loss": 0.8922,
+      "step": 28730
+    },
+    {
+      "epoch": 0.9164833062278772,
+      "grad_norm": 0.2011575996875763,
+      "learning_rate": 3.0008082845063706e-05,
+      "loss": 0.8958,
+      "step": 28740
+    },
+    {
+      "epoch": 0.9168021939475111,
+      "grad_norm": 0.19144727289676666,
+      "learning_rate": 2.9882438514766802e-05,
+      "loss": 0.885,
+      "step": 28750
+    },
+    {
+      "epoch": 0.917121081667145,
+      "grad_norm": 0.191970556974411,
+      "learning_rate": 2.9757320259321707e-05,
+      "loss": 0.8791,
+      "step": 28760
+    },
+    {
+      "epoch": 0.917439969386779,
+      "grad_norm": 0.20373089611530304,
+      "learning_rate": 2.963272587604447e-05,
+      "loss": 0.8728,
+      "step": 28770
+    },
+    {
+      "epoch": 0.9177588571064128,
+      "grad_norm": 0.202840656042099,
+      "learning_rate": 2.9508653171473813e-05,
+      "loss": 0.8794,
+      "step": 28780
+    },
+    {
+      "epoch": 0.9180777448260468,
+      "grad_norm": 0.19466151297092438,
+      "learning_rate": 2.9385099961332523e-05,
+      "loss": 0.8728,
+      "step": 28790
+    },
+    {
+      "epoch": 0.9183966325456807,
+      "grad_norm": 0.19009439647197723,
+      "learning_rate": 2.926206407048898e-05,
+      "loss": 0.8707,
+      "step": 28800
+    },
+    {
+      "epoch": 0.9187155202653146,
+      "grad_norm": 0.19433505833148956,
+      "learning_rate": 2.913954333291889e-05,
+      "loss": 0.8721,
+      "step": 28810
+    },
+    {
+      "epoch": 0.9190344079849485,
+      "grad_norm": 0.19531868398189545,
+      "learning_rate": 2.901753559166712e-05,
+      "loss": 0.8554,
+      "step": 28820
+    },
+    {
+      "epoch": 0.9193532957045825,
+      "grad_norm": 0.1936861276626587,
+      "learning_rate": 2.8896038698809757e-05,
+      "loss": 0.8692,
+      "step": 28830
+    },
+    {
+      "epoch": 0.9196721834242163,
+      "grad_norm": 0.19250236451625824,
+      "learning_rate": 2.8775050515416274e-05,
+      "loss": 0.8772,
+      "step": 28840
+    },
+    {
+      "epoch": 0.9199910711438503,
+      "grad_norm": 0.18581518530845642,
+      "learning_rate": 2.865456891151188e-05,
+      "loss": 0.8771,
+      "step": 28850
+    },
+    {
+      "epoch": 0.9203099588634842,
+      "grad_norm": 0.1961974948644638,
+      "learning_rate": 2.8534591766040024e-05,
+      "loss": 0.8798,
+      "step": 28860
+    },
+    {
+      "epoch": 0.920628846583118,
+      "grad_norm": 0.19266340136528015,
+      "learning_rate": 2.8415116966825072e-05,
+      "loss": 0.8634,
+      "step": 28870
+    },
+    {
+      "epoch": 0.920947734302752,
+      "grad_norm": 0.19667044281959534,
+      "learning_rate": 2.8296142410535074e-05,
+      "loss": 0.8706,
+      "step": 28880
+    },
+    {
+      "epoch": 0.921266622022386,
+      "grad_norm": 0.19851966202259064,
+      "learning_rate": 2.8177666002644782e-05,
+      "loss": 0.8848,
+      "step": 28890
+    },
+    {
+      "epoch": 0.9215855097420198,
+      "grad_norm": 0.1962861716747284,
+      "learning_rate": 2.8059685657398748e-05,
+      "loss": 0.8833,
+      "step": 28900
+    },
+    {
+      "epoch": 0.9219043974616538,
+      "grad_norm": 0.19509483873844147,
+      "learning_rate": 2.7942199297774628e-05,
+      "loss": 0.8732,
+      "step": 28910
+    },
+    {
+      "epoch": 0.9222232851812877,
+      "grad_norm": 0.19322551786899567,
+      "learning_rate": 2.7825204855446604e-05,
+      "loss": 0.8804,
+      "step": 28920
+    },
+    {
+      "epoch": 0.9225421729009216,
+      "grad_norm": 0.19780372083187103,
+      "learning_rate": 2.7708700270748953e-05,
+      "loss": 0.884,
+      "step": 28930
+    },
+    {
+      "epoch": 0.9228610606205555,
+      "grad_norm": 0.19941799342632294,
+      "learning_rate": 2.759268349263983e-05,
+      "loss": 0.8738,
+      "step": 28940
+    },
+    {
+      "epoch": 0.9231799483401895,
+      "grad_norm": 0.18909300863742828,
+      "learning_rate": 2.7477152478665126e-05,
+      "loss": 0.8751,
+      "step": 28950
+    },
+    {
+      "epoch": 0.9234988360598233,
+      "grad_norm": 0.18900838494300842,
+      "learning_rate": 2.736210519492252e-05,
+      "loss": 0.8734,
+      "step": 28960
+    },
+    {
+      "epoch": 0.9238177237794573,
+      "grad_norm": 0.19971856474876404,
+      "learning_rate": 2.7247539616025663e-05,
+      "loss": 0.8702,
+      "step": 28970
+    },
+    {
+      "epoch": 0.9241366114990912,
+      "grad_norm": 0.20252341032028198,
+      "learning_rate": 2.7133453725068553e-05,
+      "loss": 0.8648,
+      "step": 28980
+    },
+    {
+      "epoch": 0.924455499218725,
+      "grad_norm": 0.19686760008335114,
+      "learning_rate": 2.701984551358999e-05,
+      "loss": 0.8739,
+      "step": 28990
+    },
+    {
+      "epoch": 0.924774386938359,
+      "grad_norm": 0.20540061593055725,
+      "learning_rate": 2.6906712981538236e-05,
+      "loss": 0.865,
+      "step": 29000
+    },
+    {
+      "epoch": 0.925093274657993,
+      "grad_norm": 0.1936776340007782,
+      "learning_rate": 2.6794054137235806e-05,
+      "loss": 0.8734,
+      "step": 29010
+    },
+    {
+      "epoch": 0.9254121623776268,
+      "grad_norm": 0.1934029906988144,
+      "learning_rate": 2.668186699734439e-05,
+      "loss": 0.8753,
+      "step": 29020
+    },
+    {
+      "epoch": 0.9257310500972608,
+      "grad_norm": 0.1935998946428299,
+      "learning_rate": 2.657014958682998e-05,
+      "loss": 0.8686,
+      "step": 29030
+    },
+    {
+      "epoch": 0.9260499378168947,
+      "grad_norm": 0.19146475195884705,
+      "learning_rate": 2.6458899938928035e-05,
+      "loss": 0.8644,
+      "step": 29040
+    },
+    {
+      "epoch": 0.9263688255365286,
+      "grad_norm": 0.19215962290763855,
+      "learning_rate": 2.634811609510889e-05,
+      "loss": 0.8693,
+      "step": 29050
+    },
+    {
+      "epoch": 0.9266877132561625,
+      "grad_norm": 0.19085708260536194,
+      "learning_rate": 2.6237796105043294e-05,
+      "loss": 0.8823,
+      "step": 29060
+    },
+    {
+      "epoch": 0.9270066009757965,
+      "grad_norm": 0.19523738324642181,
+      "learning_rate": 2.6127938026568044e-05,
+      "loss": 0.872,
+      "step": 29070
+    },
+    {
+      "epoch": 0.9273254886954303,
+      "grad_norm": 0.19307196140289307,
+      "learning_rate": 2.6018539925651806e-05,
+      "loss": 0.8551,
+      "step": 29080
+    },
+    {
+      "epoch": 0.9276443764150643,
+      "grad_norm": 0.19289293885231018,
+      "learning_rate": 2.5909599876361067e-05,
+      "loss": 0.8643,
+      "step": 29090
+    },
+    {
+      "epoch": 0.9279632641346982,
+      "grad_norm": 0.1898387372493744,
+      "learning_rate": 2.5801115960826234e-05,
+      "loss": 0.8827,
+      "step": 29100
+    },
+    {
+      "epoch": 0.928282151854332,
+      "grad_norm": 0.20208875834941864,
+      "learning_rate": 2.5693086269207865e-05,
+      "loss": 0.8995,
+      "step": 29110
+    },
+    {
+      "epoch": 0.928601039573966,
+      "grad_norm": 0.19447356462478638,
+      "learning_rate": 2.5585508899663037e-05,
+      "loss": 0.8855,
+      "step": 29120
+    },
+    {
+      "epoch": 0.9289199272936,
+      "grad_norm": 0.19338631629943848,
+      "learning_rate": 2.5478381958311885e-05,
+      "loss": 0.8704,
+      "step": 29130
+    },
+    {
+      "epoch": 0.9292388150132338,
+      "grad_norm": 0.19808538258075714,
+      "learning_rate": 2.5371703559204248e-05,
+      "loss": 0.8744,
+      "step": 29140
+    },
+    {
+      "epoch": 0.9295577027328678,
+      "grad_norm": 0.1872604638338089,
+      "learning_rate": 2.526547182428646e-05,
+      "loss": 0.8598,
+      "step": 29150
+    },
+    {
+      "epoch": 0.9298765904525017,
+      "grad_norm": 0.190872460603714,
+      "learning_rate": 2.5159684883368308e-05,
+      "loss": 0.856,
+      "step": 29160
+    },
+    {
+      "epoch": 0.9301954781721355,
+      "grad_norm": 0.19496072828769684,
+      "learning_rate": 2.505434087409009e-05,
+      "loss": 0.88,
+      "step": 29170
+    },
+    {
+      "epoch": 0.9305143658917695,
+      "grad_norm": 0.19592823088169098,
+      "learning_rate": 2.4949437941889832e-05,
+      "loss": 0.8721,
+      "step": 29180
+    },
+    {
+      "epoch": 0.9308332536114035,
+      "grad_norm": 0.19178549945354462,
+      "learning_rate": 2.4844974239970637e-05,
+      "loss": 0.8512,
+      "step": 29190
+    },
+    {
+      "epoch": 0.9311521413310373,
+      "grad_norm": 0.19600598514080048,
+      "learning_rate": 2.4740947929268202e-05,
+      "loss": 0.8758,
+      "step": 29200
+    },
+    {
+      "epoch": 0.9314710290506713,
+      "grad_norm": 0.1950349062681198,
+      "learning_rate": 2.4637357178418395e-05,
+      "loss": 0.8914,
+      "step": 29210
+    },
+    {
+      "epoch": 0.9317899167703052,
+      "grad_norm": 0.19837765395641327,
+      "learning_rate": 2.4534200163725028e-05,
+      "loss": 0.8767,
+      "step": 29220
+    },
+    {
+      "epoch": 0.932108804489939,
+      "grad_norm": 0.1905105859041214,
+      "learning_rate": 2.443147506912777e-05,
+      "loss": 0.8675,
+      "step": 29230
+    },
+    {
+      "epoch": 0.932427692209573,
+      "grad_norm": 0.20346423983573914,
+      "learning_rate": 2.4329180086170172e-05,
+      "loss": 0.865,
+      "step": 29240
+    },
+    {
+      "epoch": 0.932746579929207,
+      "grad_norm": 0.19541527330875397,
+      "learning_rate": 2.4227313413967803e-05,
+      "loss": 0.8722,
+      "step": 29250
+    },
+    {
+      "epoch": 0.9330654676488408,
+      "grad_norm": 0.19942022860050201,
+      "learning_rate": 2.412587325917658e-05,
+      "loss": 0.872,
+      "step": 29260
+    },
+    {
+      "epoch": 0.9333843553684747,
+      "grad_norm": 0.20535457134246826,
+      "learning_rate": 2.4024857835961166e-05,
+      "loss": 0.866,
+      "step": 29270
+    },
+    {
+      "epoch": 0.9337032430881087,
+      "grad_norm": 0.19743776321411133,
+      "learning_rate": 2.392426536596356e-05,
+      "loss": 0.8674,
+      "step": 29280
+    },
+    {
+      "epoch": 0.9340221308077425,
+      "grad_norm": 0.20449943840503693,
+      "learning_rate": 2.3824094078271775e-05,
+      "loss": 0.8671,
+      "step": 29290
+    },
+    {
+      "epoch": 0.9343410185273765,
+      "grad_norm": 0.18968933820724487,
+      "learning_rate": 2.3724342209388646e-05,
+      "loss": 0.8635,
+      "step": 29300
+    },
+    {
+      "epoch": 0.9346599062470105,
+      "grad_norm": 0.18928585946559906,
+      "learning_rate": 2.3625008003200812e-05,
+      "loss": 0.8791,
+      "step": 29310
+    },
+    {
+      "epoch": 0.9349787939666443,
+      "grad_norm": 0.18719816207885742,
+      "learning_rate": 2.352608971094778e-05,
+      "loss": 0.874,
+      "step": 29320
+    },
+    {
+      "epoch": 0.9352976816862782,
+      "grad_norm": 0.1976853758096695,
+      "learning_rate": 2.3427585591191153e-05,
+      "loss": 0.8902,
+      "step": 29330
+    },
+    {
+      "epoch": 0.9356165694059122,
+      "grad_norm": 0.19273672997951508,
+      "learning_rate": 2.3329493909783962e-05,
+      "loss": 0.8614,
+      "step": 29340
+    },
+    {
+      "epoch": 0.935935457125546,
+      "grad_norm": 0.196641743183136,
+      "learning_rate": 2.3231812939840138e-05,
+      "loss": 0.8632,
+      "step": 29350
+    },
+    {
+      "epoch": 0.93625434484518,
+      "grad_norm": 0.19882845878601074,
+      "learning_rate": 2.3134540961704098e-05,
+      "loss": 0.8514,
+      "step": 29360
+    },
+    {
+      "epoch": 0.936573232564814,
+      "grad_norm": 0.1976517140865326,
+      "learning_rate": 2.3037676262920523e-05,
+      "loss": 0.8798,
+      "step": 29370
+    },
+    {
+      "epoch": 0.9368921202844478,
+      "grad_norm": 0.19035455584526062,
+      "learning_rate": 2.2941217138204138e-05,
+      "loss": 0.8758,
+      "step": 29380
+    },
+    {
+      "epoch": 0.9372110080040817,
+      "grad_norm": 0.19680151343345642,
+      "learning_rate": 2.2845161889409744e-05,
+      "loss": 0.8729,
+      "step": 29390
+    },
+    {
+      "epoch": 0.9375298957237157,
+      "grad_norm": 0.18864302337169647,
+      "learning_rate": 2.2749508825502283e-05,
+      "loss": 0.8776,
+      "step": 29400
+    },
+    {
+      "epoch": 0.9378487834433495,
+      "grad_norm": 0.1876661777496338,
+      "learning_rate": 2.2654256262527105e-05,
+      "loss": 0.8727,
+      "step": 29410
+    },
+    {
+      "epoch": 0.9381676711629835,
+      "grad_norm": 0.1983334869146347,
+      "learning_rate": 2.2559402523580303e-05,
+      "loss": 0.8783,
+      "step": 29420
+    },
+    {
+      "epoch": 0.9384865588826174,
+      "grad_norm": 0.2038796991109848,
+      "learning_rate": 2.2464945938779194e-05,
+      "loss": 0.8678,
+      "step": 29430
+    },
+    {
+      "epoch": 0.9388054466022513,
+      "grad_norm": 0.19537989795207977,
+      "learning_rate": 2.2370884845232913e-05,
+      "loss": 0.8618,
+      "step": 29440
+    },
+    {
+      "epoch": 0.9391243343218852,
+      "grad_norm": 0.19477014243602753,
+      "learning_rate": 2.2277217587013156e-05,
+      "loss": 0.8654,
+      "step": 29450
+    },
+    {
+      "epoch": 0.9394432220415192,
+      "grad_norm": 0.19163036346435547,
+      "learning_rate": 2.2183942515125016e-05,
+      "loss": 0.877,
+      "step": 29460
+    },
+    {
+      "epoch": 0.9397621097611532,
+      "grad_norm": 0.19399482011795044,
+      "learning_rate": 2.2091057987477944e-05,
+      "loss": 0.8717,
+      "step": 29470
+    },
+    {
+      "epoch": 0.940080997480787,
+      "grad_norm": 0.19094112515449524,
+      "learning_rate": 2.1998562368856864e-05,
+      "loss": 0.8701,
+      "step": 29480
+    },
+    {
+      "epoch": 0.940399885200421,
+      "grad_norm": 0.193534255027771,
+      "learning_rate": 2.190645403089337e-05,
+      "loss": 0.8704,
+      "step": 29490
+    },
+    {
+      "epoch": 0.9407187729200549,
+      "grad_norm": 0.1996348351240158,
+      "learning_rate": 2.181473135203705e-05,
+      "loss": 0.8753,
+      "step": 29500
+    },
+    {
+      "epoch": 0.9410376606396887,
+      "grad_norm": 0.19400392472743988,
+      "learning_rate": 2.172339271752697e-05,
+      "loss": 0.8673,
+      "step": 29510
+    },
+    {
+      "epoch": 0.9413565483593227,
+      "grad_norm": 0.20478451251983643,
+      "learning_rate": 2.1632436519363203e-05,
+      "loss": 0.8657,
+      "step": 29520
+    },
+    {
+      "epoch": 0.9416754360789567,
+      "grad_norm": 0.20005236566066742,
+      "learning_rate": 2.1541861156278552e-05,
+      "loss": 0.8521,
+      "step": 29530
+    },
+    {
+      "epoch": 0.9419943237985905,
+      "grad_norm": 0.19527098536491394,
+      "learning_rate": 2.145166503371037e-05,
+      "loss": 0.8808,
+      "step": 29540
+    },
+    {
+      "epoch": 0.9423132115182244,
+      "grad_norm": 0.20347745716571808,
+      "learning_rate": 2.1361846563772446e-05,
+      "loss": 0.8769,
+      "step": 29550
+    },
+    {
+      "epoch": 0.9426320992378584,
+      "grad_norm": 0.19539164006710052,
+      "learning_rate": 2.127240416522709e-05,
+      "loss": 0.8669,
+      "step": 29560
+    },
+    {
+      "epoch": 0.9429509869574922,
+      "grad_norm": 0.19707153737545013,
+      "learning_rate": 2.1183336263457258e-05,
+      "loss": 0.8782,
+      "step": 29570
+    },
+    {
+      "epoch": 0.9432698746771262,
+      "grad_norm": 0.1947970688343048,
+      "learning_rate": 2.109464129043888e-05,
+      "loss": 0.8599,
+      "step": 29580
+    },
+    {
+      "epoch": 0.9435887623967602,
+      "grad_norm": 0.20084436237812042,
+      "learning_rate": 2.100631768471321e-05,
+      "loss": 0.8637,
+      "step": 29590
+    },
+    {
+      "epoch": 0.943907650116394,
+      "grad_norm": 0.19857023656368256,
+      "learning_rate": 2.0918363891359365e-05,
+      "loss": 0.8783,
+      "step": 29600
+    },
+    {
+      "epoch": 0.944226537836028,
+      "grad_norm": 0.19687621295452118,
+      "learning_rate": 2.0830778361966933e-05,
+      "loss": 0.8798,
+      "step": 29610
+    },
+    {
+      "epoch": 0.9445454255556619,
+      "grad_norm": 0.18758653104305267,
+      "learning_rate": 2.0743559554608734e-05,
+      "loss": 0.8792,
+      "step": 29620
+    },
+    {
+      "epoch": 0.9448643132752957,
+      "grad_norm": 0.2017526775598526,
+      "learning_rate": 2.0656705933813646e-05,
+      "loss": 0.8811,
+      "step": 29630
+    },
+    {
+      "epoch": 0.9451832009949297,
+      "grad_norm": 0.20141223073005676,
+      "learning_rate": 2.05702159705396e-05,
+      "loss": 0.8659,
+      "step": 29640
+    },
+    {
+      "epoch": 0.9455020887145636,
+      "grad_norm": 0.19546176493167877,
+      "learning_rate": 2.0484088142146646e-05,
+      "loss": 0.8589,
+      "step": 29650
+    },
+    {
+      "epoch": 0.9458209764341975,
+      "grad_norm": 0.19249585270881653,
+      "learning_rate": 2.039832093237016e-05,
+      "loss": 0.8678,
+      "step": 29660
+    },
+    {
+      "epoch": 0.9461398641538314,
+      "grad_norm": 0.19663062691688538,
+      "learning_rate": 2.0312912831294133e-05,
+      "loss": 0.8846,
+      "step": 29670
+    },
+    {
+      "epoch": 0.9464587518734654,
+      "grad_norm": 0.1930922567844391,
+      "learning_rate": 2.022786233532461e-05,
+      "loss": 0.8756,
+      "step": 29680
+    },
+    {
+      "epoch": 0.9467776395930992,
+      "grad_norm": 0.19463899731636047,
+      "learning_rate": 2.014316794716319e-05,
+      "loss": 0.8715,
+      "step": 29690
+    },
+    {
+      "epoch": 0.9470965273127332,
+      "grad_norm": 0.20353105664253235,
+      "learning_rate": 2.0058828175780694e-05,
+      "loss": 0.8674,
+      "step": 29700
+    },
+    {
+      "epoch": 0.9474154150323671,
+      "grad_norm": 0.19156238436698914,
+      "learning_rate": 1.9974841536390925e-05,
+      "loss": 0.8515,
+      "step": 29710
+    },
+    {
+      "epoch": 0.947734302752001,
+      "grad_norm": 0.19340883195400238,
+      "learning_rate": 1.989120655042449e-05,
+      "loss": 0.8643,
+      "step": 29720
+    },
+    {
+      "epoch": 0.9480531904716349,
+      "grad_norm": 0.2009674459695816,
+      "learning_rate": 1.9807921745502785e-05,
+      "loss": 0.8846,
+      "step": 29730
+    },
+    {
+      "epoch": 0.9483720781912689,
+      "grad_norm": 0.1952427476644516,
+      "learning_rate": 1.972498565541209e-05,
+      "loss": 0.8587,
+      "step": 29740
+    },
+    {
+      "epoch": 0.9486909659109027,
+      "grad_norm": 0.1934279054403305,
+      "learning_rate": 1.964239682007775e-05,
+      "loss": 0.8716,
+      "step": 29750
+    },
+    {
+      "epoch": 0.9490098536305367,
+      "grad_norm": 0.19285337626934052,
+      "learning_rate": 1.956015378553845e-05,
+      "loss": 0.8741,
+      "step": 29760
+    },
+    {
+      "epoch": 0.9493287413501706,
+      "grad_norm": 0.20002107322216034,
+      "learning_rate": 1.947825510392065e-05,
+      "loss": 0.8645,
+      "step": 29770
+    },
+    {
+      "epoch": 0.9496476290698045,
+      "grad_norm": 0.1990961879491806,
+      "learning_rate": 1.939669933341307e-05,
+      "loss": 0.8674,
+      "step": 29780
+    },
+    {
+      "epoch": 0.9499665167894384,
+      "grad_norm": 0.19558599591255188,
+      "learning_rate": 1.9315485038241333e-05,
+      "loss": 0.873,
+      "step": 29790
+    },
+    {
+      "epoch": 0.9502854045090724,
+      "grad_norm": 0.19908514618873596,
+      "learning_rate": 1.9234610788642647e-05,
+      "loss": 0.8719,
+      "step": 29800
+    },
+    {
+      "epoch": 0.9506042922287062,
+      "grad_norm": 0.19451043009757996,
+      "learning_rate": 1.9154075160840683e-05,
+      "loss": 0.8565,
+      "step": 29810
+    },
+    {
+      "epoch": 0.9509231799483402,
+      "grad_norm": 0.19632062315940857,
+      "learning_rate": 1.907387673702047e-05,
+      "loss": 0.8798,
+      "step": 29820
+    },
+    {
+      "epoch": 0.9512420676679741,
+      "grad_norm": 0.19675464928150177,
+      "learning_rate": 1.8994014105303468e-05,
+      "loss": 0.8812,
+      "step": 29830
+    },
+    {
+      "epoch": 0.951560955387608,
+      "grad_norm": 0.19777949154376984,
+      "learning_rate": 1.8914485859722682e-05,
+      "loss": 0.8797,
+      "step": 29840
+    },
+    {
+      "epoch": 0.9518798431072419,
+      "grad_norm": 0.1946035623550415,
+      "learning_rate": 1.8835290600197926e-05,
+      "loss": 0.8758,
+      "step": 29850
+    },
+    {
+      "epoch": 0.9521987308268759,
+      "grad_norm": 0.19789068400859833,
+      "learning_rate": 1.8756426932511182e-05,
+      "loss": 0.8627,
+      "step": 29860
+    },
+    {
+      "epoch": 0.9525176185465097,
+      "grad_norm": 0.19965241849422455,
+      "learning_rate": 1.8677893468282023e-05,
+      "loss": 0.8783,
+      "step": 29870
+    },
+    {
+      "epoch": 0.9528365062661437,
+      "grad_norm": 0.18935656547546387,
+      "learning_rate": 1.8599688824943227e-05,
+      "loss": 0.8654,
+      "step": 29880
+    },
+    {
+      "epoch": 0.9531553939857776,
+      "grad_norm": 0.20136713981628418,
+      "learning_rate": 1.852181162571638e-05,
+      "loss": 0.863,
+      "step": 29890
+    },
+    {
+      "epoch": 0.9534742817054115,
+      "grad_norm": 0.19391171634197235,
+      "learning_rate": 1.8444260499587665e-05,
+      "loss": 0.863,
+      "step": 29900
+    },
+    {
+      "epoch": 0.9537931694250454,
+      "grad_norm": 0.19194775819778442,
+      "learning_rate": 1.8367034081283716e-05,
+      "loss": 0.8753,
+      "step": 29910
+    },
+    {
+      "epoch": 0.9541120571446794,
+      "grad_norm": 0.19135677814483643,
+      "learning_rate": 1.829013101124761e-05,
+      "loss": 0.8741,
+      "step": 29920
+    },
+    {
+      "epoch": 0.9544309448643132,
+      "grad_norm": 0.19014357030391693,
+      "learning_rate": 1.8213549935614886e-05,
+      "loss": 0.865,
+      "step": 29930
+    },
+    {
+      "epoch": 0.9547498325839472,
+      "grad_norm": 0.19929145276546478,
+      "learning_rate": 1.8137289506189752e-05,
+      "loss": 0.8698,
+      "step": 29940
+    },
+    {
+      "epoch": 0.9550687203035811,
+      "grad_norm": 0.1992056667804718,
+      "learning_rate": 1.806134838042133e-05,
+      "loss": 0.8966,
+      "step": 29950
+    },
+    {
+      "epoch": 0.955387608023215,
+      "grad_norm": 0.1995675414800644,
+      "learning_rate": 1.7985725221380017e-05,
+      "loss": 0.8682,
+      "step": 29960
+    },
+    {
+      "epoch": 0.9557064957428489,
+      "grad_norm": 0.19337597489356995,
+      "learning_rate": 1.7910418697733964e-05,
+      "loss": 0.8775,
+      "step": 29970
+    },
+    {
+      "epoch": 0.9560253834624829,
+      "grad_norm": 0.19523291289806366,
+      "learning_rate": 1.7835427483725635e-05,
+      "loss": 0.8676,
+      "step": 29980
+    },
+    {
+      "epoch": 0.9563442711821167,
+      "grad_norm": 0.20105189085006714,
+      "learning_rate": 1.7760750259148444e-05,
+      "loss": 0.8863,
+      "step": 29990
+    },
+    {
+      "epoch": 0.9566631589017507,
+      "grad_norm": 0.20133532583713531,
+      "learning_rate": 1.7686385709323553e-05,
+      "loss": 0.8821,
+      "step": 30000
+    },
+    {
+      "epoch": 0.9569820466213846,
+      "grad_norm": 0.19689635932445526,
+      "learning_rate": 1.7612332525076688e-05,
+      "loss": 0.8852,
+      "step": 30010
+    },
+    {
+      "epoch": 0.9573009343410185,
+      "grad_norm": 0.1972358226776123,
+      "learning_rate": 1.7538589402715118e-05,
+      "loss": 0.8579,
+      "step": 30020
+    },
+    {
+      "epoch": 0.9576198220606524,
+      "grad_norm": 0.19419169425964355,
+      "learning_rate": 1.7465155044004687e-05,
+      "loss": 0.8679,
+      "step": 30030
+    },
+    {
+      "epoch": 0.9579387097802864,
+      "grad_norm": 0.1934502273797989,
+      "learning_rate": 1.7392028156146985e-05,
+      "loss": 0.8688,
+      "step": 30040
+    },
+    {
+      "epoch": 0.9582575974999202,
+      "grad_norm": 0.1983090192079544,
+      "learning_rate": 1.7319207451756553e-05,
+      "loss": 0.8622,
+      "step": 30050
+    },
+    {
+      "epoch": 0.9585764852195542,
+      "grad_norm": 0.19536088407039642,
+      "learning_rate": 1.724669164883824e-05,
+      "loss": 0.8875,
+      "step": 30060
+    },
+    {
+      "epoch": 0.9588953729391881,
+      "grad_norm": 0.19346052408218384,
+      "learning_rate": 1.7174479470764624e-05,
+      "loss": 0.8837,
+      "step": 30070
+    },
+    {
+      "epoch": 0.959214260658822,
+      "grad_norm": 0.19292378425598145,
+      "learning_rate": 1.7102569646253555e-05,
+      "loss": 0.8687,
+      "step": 30080
+    },
+    {
+      "epoch": 0.9595331483784559,
+      "grad_norm": 0.1938813477754593,
+      "learning_rate": 1.7030960909345757e-05,
+      "loss": 0.8707,
+      "step": 30090
+    },
+    {
+      "epoch": 0.9598520360980899,
+      "grad_norm": 0.19189944863319397,
+      "learning_rate": 1.6959651999382535e-05,
+      "loss": 0.8792,
+      "step": 30100
+    },
+    {
+      "epoch": 0.9601709238177237,
+      "grad_norm": 0.19455473124980927,
+      "learning_rate": 1.688864166098361e-05,
+      "loss": 0.876,
+      "step": 30110
+    },
+    {
+      "epoch": 0.9604898115373577,
+      "grad_norm": 0.19673947989940643,
+      "learning_rate": 1.6817928644024988e-05,
+      "loss": 0.8685,
+      "step": 30120
+    },
+    {
+      "epoch": 0.9608086992569916,
+      "grad_norm": 0.19511686265468597,
+      "learning_rate": 1.6747511703616973e-05,
+      "loss": 0.8592,
+      "step": 30130
+    },
+    {
+      "epoch": 0.9611275869766255,
+      "grad_norm": 0.19796845316886902,
+      "learning_rate": 1.6677389600082237e-05,
+      "loss": 0.8745,
+      "step": 30140
+    },
+    {
+      "epoch": 0.9614464746962594,
+      "grad_norm": 0.19610048830509186,
+      "learning_rate": 1.6607561098934e-05,
+      "loss": 0.8725,
+      "step": 30150
+    },
+    {
+      "epoch": 0.9617653624158934,
+      "grad_norm": 0.19613753259181976,
+      "learning_rate": 1.6538024970854306e-05,
+      "loss": 0.8635,
+      "step": 30160
+    },
+    {
+      "epoch": 0.9620842501355272,
+      "grad_norm": 0.19956007599830627,
+      "learning_rate": 1.646877999167236e-05,
+      "loss": 0.8612,
+      "step": 30170
+    },
+    {
+      "epoch": 0.9624031378551612,
+      "grad_norm": 0.19366030395030975,
+      "learning_rate": 1.639982494234301e-05,
+      "loss": 0.8742,
+      "step": 30180
+    },
+    {
+      "epoch": 0.9627220255747951,
+      "grad_norm": 0.1933685839176178,
+      "learning_rate": 1.633115860892524e-05,
+      "loss": 0.8786,
+      "step": 30190
+    },
+    {
+      "epoch": 0.963040913294429,
+      "grad_norm": 0.1933773159980774,
+      "learning_rate": 1.6262779782560837e-05,
+      "loss": 0.8671,
+      "step": 30200
+    },
+    {
+      "epoch": 0.9633598010140629,
+      "grad_norm": 0.1973232924938202,
+      "learning_rate": 1.6194687259453118e-05,
+      "loss": 0.8751,
+      "step": 30210
+    },
+    {
+      "epoch": 0.9636786887336969,
+      "grad_norm": 0.1968729943037033,
+      "learning_rate": 1.6126879840845686e-05,
+      "loss": 0.8847,
+      "step": 30220
+    },
+    {
+      "epoch": 0.9639975764533307,
+      "grad_norm": 0.20425496995449066,
+      "learning_rate": 1.6059356333001364e-05,
+      "loss": 0.8802,
+      "step": 30230
+    },
+    {
+      "epoch": 0.9643164641729647,
+      "grad_norm": 0.19971290230751038,
+      "learning_rate": 1.599211554718118e-05,
+      "loss": 0.8764,
+      "step": 30240
+    },
+    {
+      "epoch": 0.9646353518925986,
+      "grad_norm": 0.19143065810203552,
+      "learning_rate": 1.5925156299623422e-05,
+      "loss": 0.8513,
+      "step": 30250
+    },
+    {
+      "epoch": 0.9649542396122325,
+      "grad_norm": 0.20351962745189667,
+      "learning_rate": 1.5858477411522817e-05,
+      "loss": 0.8607,
+      "step": 30260
+    },
+    {
+      "epoch": 0.9652731273318664,
+      "grad_norm": 0.1943381279706955,
+      "learning_rate": 1.5792077709009763e-05,
+      "loss": 0.8688,
+      "step": 30270
+    },
+    {
+      "epoch": 0.9655920150515004,
+      "grad_norm": 0.20330402255058289,
+      "learning_rate": 1.5725956023129666e-05,
+      "loss": 0.8571,
+      "step": 30280
+    },
+    {
+      "epoch": 0.9659109027711343,
+      "grad_norm": 0.19449760019779205,
+      "learning_rate": 1.566011118982237e-05,
+      "loss": 0.8664,
+      "step": 30290
+    },
+    {
+      "epoch": 0.9662297904907682,
+      "grad_norm": 0.19076880812644958,
+      "learning_rate": 1.559454204990166e-05,
+      "loss": 0.8774,
+      "step": 30300
+    },
+    {
+      "epoch": 0.9665486782104021,
+      "grad_norm": 0.19695602357387543,
+      "learning_rate": 1.5529247449034844e-05,
+      "loss": 0.8672,
+      "step": 30310
+    },
+    {
+      "epoch": 0.9668675659300361,
+      "grad_norm": 0.1943412721157074,
+      "learning_rate": 1.5464226237722444e-05,
+      "loss": 0.8717,
+      "step": 30320
+    },
+    {
+      "epoch": 0.9671864536496699,
+      "grad_norm": 0.1951615959405899,
+      "learning_rate": 1.5399477271277955e-05,
+      "loss": 0.868,
+      "step": 30330
+    },
+    {
+      "epoch": 0.9675053413693039,
+      "grad_norm": 0.19740445911884308,
+      "learning_rate": 1.5334999409807695e-05,
+      "loss": 0.8923,
+      "step": 30340
+    },
+    {
+      "epoch": 0.9678242290889378,
+      "grad_norm": 0.2107999175786972,
+      "learning_rate": 1.5270791518190734e-05,
+      "loss": 0.8857,
+      "step": 30350
+    },
+    {
+      "epoch": 0.9681431168085717,
+      "grad_norm": 0.19357022643089294,
+      "learning_rate": 1.5206852466058905e-05,
+      "loss": 0.8742,
+      "step": 30360
+    },
+    {
+      "epoch": 0.9684620045282056,
+      "grad_norm": 0.19716483354568481,
+      "learning_rate": 1.5143181127776916e-05,
+      "loss": 0.8889,
+      "step": 30370
+    },
+    {
+      "epoch": 0.9687808922478396,
+      "grad_norm": 0.19570153951644897,
+      "learning_rate": 1.5079776382422542e-05,
+      "loss": 0.8634,
+      "step": 30380
+    },
+    {
+      "epoch": 0.9690997799674734,
+      "grad_norm": 0.19478130340576172,
+      "learning_rate": 1.5016637113766854e-05,
+      "loss": 0.8679,
+      "step": 30390
+    },
+    {
+      "epoch": 0.9694186676871074,
+      "grad_norm": 0.19353069365024567,
+      "learning_rate": 1.495376221025461e-05,
+      "loss": 0.8798,
+      "step": 30400
+    },
+    {
+      "epoch": 0.9697375554067413,
+      "grad_norm": 0.19400358200073242,
+      "learning_rate": 1.4891150564984654e-05,
+      "loss": 0.862,
+      "step": 30410
+    },
+    {
+      "epoch": 0.9700564431263752,
+      "grad_norm": 0.20103727281093597,
+      "learning_rate": 1.482880107569045e-05,
+      "loss": 0.8758,
+      "step": 30420
+    },
+    {
+      "epoch": 0.9703753308460091,
+      "grad_norm": 0.19329263269901276,
+      "learning_rate": 1.4766712644720671e-05,
+      "loss": 0.8734,
+      "step": 30430
+    },
+    {
+      "epoch": 0.9706942185656431,
+      "grad_norm": 0.19491326808929443,
+      "learning_rate": 1.4704884179019873e-05,
+      "loss": 0.8584,
+      "step": 30440
+    },
+    {
+      "epoch": 0.9710131062852769,
+      "grad_norm": 0.18893815577030182,
+      "learning_rate": 1.464331459010925e-05,
+      "loss": 0.8752,
+      "step": 30450
+    },
+    {
+      "epoch": 0.9713319940049109,
+      "grad_norm": 0.19456712901592255,
+      "learning_rate": 1.4582002794067476e-05,
+      "loss": 0.8795,
+      "step": 30460
+    },
+    {
+      "epoch": 0.9716508817245448,
+      "grad_norm": 0.19400033354759216,
+      "learning_rate": 1.4520947711511627e-05,
+      "loss": 0.8845,
+      "step": 30470
+    },
+    {
+      "epoch": 0.9719697694441787,
+      "grad_norm": 0.19873221218585968,
+      "learning_rate": 1.446014826757816e-05,
+      "loss": 0.8835,
+      "step": 30480
+    },
+    {
+      "epoch": 0.9722886571638126,
+      "grad_norm": 0.19608688354492188,
+      "learning_rate": 1.4399603391904017e-05,
+      "loss": 0.8795,
+      "step": 30490
+    },
+    {
+      "epoch": 0.9726075448834466,
+      "grad_norm": 0.18778324127197266,
+      "learning_rate": 1.4339312018607758e-05,
+      "loss": 0.8729,
+      "step": 30500
+    },
+    {
+      "epoch": 0.9729264326030804,
+      "grad_norm": 0.19760362803936005,
+      "learning_rate": 1.4279273086270809e-05,
+      "loss": 0.8744,
+      "step": 30510
+    },
+    {
+      "epoch": 0.9732453203227144,
+      "grad_norm": 0.19934231042861938,
+      "learning_rate": 1.4219485537918775e-05,
+      "loss": 0.8793,
+      "step": 30520
+    },
+    {
+      "epoch": 0.9735642080423483,
+      "grad_norm": 0.19571653008460999,
+      "learning_rate": 1.4159948321002827e-05,
+      "loss": 0.8726,
+      "step": 30530
+    },
+    {
+      "epoch": 0.9738830957619822,
+      "grad_norm": 0.19574718177318573,
+      "learning_rate": 1.4100660387381168e-05,
+      "loss": 0.8679,
+      "step": 30540
+    },
+    {
+      "epoch": 0.9742019834816161,
+      "grad_norm": 0.19988127052783966,
+      "learning_rate": 1.4041620693300609e-05,
+      "loss": 0.8799,
+      "step": 30550
+    },
+    {
+      "epoch": 0.9745208712012501,
+      "grad_norm": 0.19439704716205597,
+      "learning_rate": 1.3982828199378157e-05,
+      "loss": 0.8787,
+      "step": 30560
+    },
+    {
+      "epoch": 0.9748397589208839,
+      "grad_norm": 0.195739284157753,
+      "learning_rate": 1.3924281870582728e-05,
+      "loss": 0.8751,
+      "step": 30570
+    },
+    {
+      "epoch": 0.9751586466405179,
+      "grad_norm": 0.19495446979999542,
+      "learning_rate": 1.3865980676216942e-05,
+      "loss": 0.871,
+      "step": 30580
+    },
+    {
+      "epoch": 0.9754775343601518,
+      "grad_norm": 0.1931915432214737,
+      "learning_rate": 1.3807923589898957e-05,
+      "loss": 0.8798,
+      "step": 30590
+    },
+    {
+      "epoch": 0.9757964220797857,
+      "grad_norm": 0.2003222554922104,
+      "learning_rate": 1.3750109589544415e-05,
+      "loss": 0.8967,
+      "step": 30600
+    },
+    {
+      "epoch": 0.9761153097994196,
+      "grad_norm": 0.18771512806415558,
+      "learning_rate": 1.3692537657348436e-05,
+      "loss": 0.8726,
+      "step": 30610
+    },
+    {
+      "epoch": 0.9764341975190536,
+      "grad_norm": 0.19001594185829163,
+      "learning_rate": 1.3635206779767707e-05,
+      "loss": 0.8707,
+      "step": 30620
+    },
+    {
+      "epoch": 0.9767530852386874,
+      "grad_norm": 0.19214805960655212,
+      "learning_rate": 1.3578115947502643e-05,
+      "loss": 0.8654,
+      "step": 30630
+    },
+    {
+      "epoch": 0.9770719729583214,
+      "grad_norm": 0.20203489065170288,
+      "learning_rate": 1.3521264155479603e-05,
+      "loss": 0.8795,
+      "step": 30640
+    },
+    {
+      "epoch": 0.9773908606779553,
+      "grad_norm": 0.19535228610038757,
+      "learning_rate": 1.3464650402833215e-05,
+      "loss": 0.8786,
+      "step": 30650
+    },
+    {
+      "epoch": 0.9777097483975892,
+      "grad_norm": 0.19746150076389313,
+      "learning_rate": 1.3408273692888739e-05,
+      "loss": 0.8657,
+      "step": 30660
+    },
+    {
+      "epoch": 0.9780286361172231,
+      "grad_norm": 0.19501303136348724,
+      "learning_rate": 1.3352133033144537e-05,
+      "loss": 0.8683,
+      "step": 30670
+    },
+    {
+      "epoch": 0.9783475238368571,
+      "grad_norm": 0.19745995104312897,
+      "learning_rate": 1.3296227435254582e-05,
+      "loss": 0.8707,
+      "step": 30680
+    },
+    {
+      "epoch": 0.9786664115564909,
+      "grad_norm": 0.19865138828754425,
+      "learning_rate": 1.3240555915011072e-05,
+      "loss": 0.878,
+      "step": 30690
+    },
+    {
+      "epoch": 0.9789852992761249,
+      "grad_norm": 0.192633718252182,
+      "learning_rate": 1.31851174923271e-05,
+      "loss": 0.8738,
+      "step": 30700
+    },
+    {
+      "epoch": 0.9793041869957588,
+      "grad_norm": 0.1950562447309494,
+      "learning_rate": 1.3129911191219392e-05,
+      "loss": 0.8836,
+      "step": 30710
+    },
+    {
+      "epoch": 0.9796230747153927,
+      "grad_norm": 0.19444888830184937,
+      "learning_rate": 1.307493603979115e-05,
+      "loss": 0.866,
+      "step": 30720
+    },
+    {
+      "epoch": 0.9799419624350266,
+      "grad_norm": 0.19572365283966064,
+      "learning_rate": 1.3020191070214908e-05,
+      "loss": 0.8579,
+      "step": 30730
+    },
+    {
+      "epoch": 0.9802608501546606,
+      "grad_norm": 0.19500453770160675,
+      "learning_rate": 1.2965675318715508e-05,
+      "loss": 0.8549,
+      "step": 30740
+    },
+    {
+      "epoch": 0.9805797378742944,
+      "grad_norm": 0.19590529799461365,
+      "learning_rate": 1.2911387825553141e-05,
+      "loss": 0.8733,
+      "step": 30750
+    },
+    {
+      "epoch": 0.9808986255939284,
+      "grad_norm": 0.1961866170167923,
+      "learning_rate": 1.2857327635006445e-05,
+      "loss": 0.8761,
+      "step": 30760
+    },
+    {
+      "epoch": 0.9812175133135623,
+      "grad_norm": 0.19800586998462677,
+      "learning_rate": 1.2803493795355673e-05,
+      "loss": 0.8644,
+      "step": 30770
+    },
+    {
+      "epoch": 0.9815364010331962,
+      "grad_norm": 0.19354160130023956,
+      "learning_rate": 1.2749885358865947e-05,
+      "loss": 0.8724,
+      "step": 30780
+    },
+    {
+      "epoch": 0.9818552887528301,
+      "grad_norm": 0.20072521269321442,
+      "learning_rate": 1.269650138177057e-05,
+      "loss": 0.8646,
+      "step": 30790
+    },
+    {
+      "epoch": 0.9821741764724641,
+      "grad_norm": 0.19742512702941895,
+      "learning_rate": 1.2643340924254416e-05,
+      "loss": 0.8601,
+      "step": 30800
+    },
+    {
+      "epoch": 0.9824930641920979,
+      "grad_norm": 0.1960420161485672,
+      "learning_rate": 1.2590403050437372e-05,
+      "loss": 0.8592,
+      "step": 30810
+    },
+    {
+      "epoch": 0.9828119519117319,
+      "grad_norm": 0.1924024224281311,
+      "learning_rate": 1.2537686828357875e-05,
+      "loss": 0.8603,
+      "step": 30820
+    },
+    {
+      "epoch": 0.9831308396313658,
+      "grad_norm": 0.1885261833667755,
+      "learning_rate": 1.2485191329956501e-05,
+      "loss": 0.8623,
+      "step": 30830
+    },
+    {
+      "epoch": 0.9834497273509997,
+      "grad_norm": 0.1895126849412918,
+      "learning_rate": 1.2432915631059624e-05,
+      "loss": 0.8588,
+      "step": 30840
+    },
+    {
+      "epoch": 0.9837686150706336,
+      "grad_norm": 0.19329601526260376,
+      "learning_rate": 1.2380858811363149e-05,
+      "loss": 0.8686,
+      "step": 30850
+    },
+    {
+      "epoch": 0.9840875027902676,
+      "grad_norm": 0.20007088780403137,
+      "learning_rate": 1.2329019954416307e-05,
+      "loss": 0.865,
+      "step": 30860
+    },
+    {
+      "epoch": 0.9844063905099014,
+      "grad_norm": 0.19103839993476868,
+      "learning_rate": 1.2277398147605525e-05,
+      "loss": 0.8638,
+      "step": 30870
+    },
+    {
+      "epoch": 0.9847252782295354,
+      "grad_norm": 0.1993708461523056,
+      "learning_rate": 1.2225992482138354e-05,
+      "loss": 0.8641,
+      "step": 30880
+    },
+    {
+      "epoch": 0.9850441659491693,
+      "grad_norm": 0.1940104365348816,
+      "learning_rate": 1.2174802053027486e-05,
+      "loss": 0.8703,
+      "step": 30890
+    },
+    {
+      "epoch": 0.9853630536688032,
+      "grad_norm": 0.1930721402168274,
+      "learning_rate": 1.2123825959074799e-05,
+      "loss": 0.8716,
+      "step": 30900
+    },
+    {
+      "epoch": 0.9856819413884371,
+      "grad_norm": 0.1939040571451187,
+      "learning_rate": 1.2073063302855502e-05,
+      "loss": 0.8594,
+      "step": 30910
+    },
+    {
+      "epoch": 0.9860008291080711,
+      "grad_norm": 0.19301022589206696,
+      "learning_rate": 1.2022513190702337e-05,
+      "loss": 0.8726,
+      "step": 30920
+    },
+    {
+      "epoch": 0.9863197168277049,
+      "grad_norm": 0.19931401312351227,
+      "learning_rate": 1.1972174732689848e-05,
+      "loss": 0.8657,
+      "step": 30930
+    },
+    {
+      "epoch": 0.9866386045473389,
+      "grad_norm": 0.1907082498073578,
+      "learning_rate": 1.1922047042618712e-05,
+      "loss": 0.8565,
+      "step": 30940
+    },
+    {
+      "epoch": 0.9869574922669728,
+      "grad_norm": 0.19912324845790863,
+      "learning_rate": 1.1872129238000134e-05,
+      "loss": 0.8668,
+      "step": 30950
+    },
+    {
+      "epoch": 0.9872763799866067,
+      "grad_norm": 0.19937056303024292,
+      "learning_rate": 1.1822420440040315e-05,
+      "loss": 0.8742,
+      "step": 30960
+    },
+    {
+      "epoch": 0.9875952677062406,
+      "grad_norm": 0.19725482165813446,
+      "learning_rate": 1.1772919773624978e-05,
+      "loss": 0.8697,
+      "step": 30970
+    },
+    {
+      "epoch": 0.9879141554258746,
+      "grad_norm": 0.20150548219680786,
+      "learning_rate": 1.1723626367303966e-05,
+      "loss": 0.8748,
+      "step": 30980
+    },
+    {
+      "epoch": 0.9882330431455084,
+      "grad_norm": 0.200142964720726,
+      "learning_rate": 1.1674539353275899e-05,
+      "loss": 0.8869,
+      "step": 30990
+    },
+    {
+      "epoch": 0.9885519308651424,
+      "grad_norm": 0.18772722780704498,
+      "learning_rate": 1.1625657867372895e-05,
+      "loss": 0.8625,
+      "step": 31000
+    },
+    {
+      "epoch": 0.9888708185847763,
+      "grad_norm": 0.2008049041032791,
+      "learning_rate": 1.1576981049045353e-05,
+      "loss": 0.8983,
+      "step": 31010
+    },
+    {
+      "epoch": 0.9891897063044102,
+      "grad_norm": 0.1990879476070404,
+      "learning_rate": 1.1528508041346812e-05,
+      "loss": 0.868,
+      "step": 31020
+    },
+    {
+      "epoch": 0.9895085940240441,
+      "grad_norm": 0.19177481532096863,
+      "learning_rate": 1.1480237990918854e-05,
+      "loss": 0.869,
+      "step": 31030
+    },
+    {
+      "epoch": 0.9898274817436781,
+      "grad_norm": 0.19938461482524872,
+      "learning_rate": 1.1432170047976097e-05,
+      "loss": 0.8819,
+      "step": 31040
+    },
+    {
+      "epoch": 0.9901463694633119,
+      "grad_norm": 0.20128324627876282,
+      "learning_rate": 1.1384303366291208e-05,
+      "loss": 0.8721,
+      "step": 31050
+    },
+    {
+      "epoch": 0.9904652571829459,
+      "grad_norm": 0.19463203847408295,
+      "learning_rate": 1.1336637103180043e-05,
+      "loss": 0.863,
+      "step": 31060
+    },
+    {
+      "epoch": 0.9907841449025798,
+      "grad_norm": 0.20132602751255035,
+      "learning_rate": 1.1289170419486774e-05,
+      "loss": 0.8694,
+      "step": 31070
+    },
+    {
+      "epoch": 0.9911030326222137,
+      "grad_norm": 0.19993562996387482,
+      "learning_rate": 1.1241902479569133e-05,
+      "loss": 0.8651,
+      "step": 31080
+    },
+    {
+      "epoch": 0.9914219203418476,
+      "grad_norm": 0.20050421357154846,
+      "learning_rate": 1.1194832451283707e-05,
+      "loss": 0.8667,
+      "step": 31090
+    },
+    {
+      "epoch": 0.9917408080614816,
+      "grad_norm": 0.19323430955410004,
+      "learning_rate": 1.1147959505971274e-05,
+      "loss": 0.8682,
+      "step": 31100
+    },
+    {
+      "epoch": 0.9920596957811154,
+      "grad_norm": 0.19004856050014496,
+      "learning_rate": 1.110128281844223e-05,
+      "loss": 0.8526,
+      "step": 31110
+    },
+    {
+      "epoch": 0.9923785835007494,
+      "grad_norm": 0.19814559817314148,
+      "learning_rate": 1.1054801566962041e-05,
+      "loss": 0.8697,
+      "step": 31120
+    },
+    {
+      "epoch": 0.9926974712203833,
+      "grad_norm": 0.1958136260509491,
+      "learning_rate": 1.1008514933236803e-05,
+      "loss": 0.8659,
+      "step": 31130
+    },
+    {
+      "epoch": 0.9930163589400173,
+      "grad_norm": 0.19568458199501038,
+      "learning_rate": 1.096242210239881e-05,
+      "loss": 0.8782,
+      "step": 31140
+    },
+    {
+      "epoch": 0.9933352466596511,
+      "grad_norm": 0.20434001088142395,
+      "learning_rate": 1.0916522262992225e-05,
+      "loss": 0.868,
+      "step": 31150
+    },
+    {
+      "epoch": 0.9936541343792851,
+      "grad_norm": 0.19457891583442688,
+      "learning_rate": 1.0870814606958793e-05,
+      "loss": 0.8653,
+      "step": 31160
+    },
+    {
+      "epoch": 0.993973022098919,
+      "grad_norm": 0.19215627014636993,
+      "learning_rate": 1.0825298329623609e-05,
+      "loss": 0.8612,
+      "step": 31170
+    },
+    {
+      "epoch": 0.9942919098185529,
+      "grad_norm": 0.19397249817848206,
+      "learning_rate": 1.0779972629680952e-05,
+      "loss": 0.8594,
+      "step": 31180
+    },
+    {
+      "epoch": 0.9946107975381868,
+      "grad_norm": 0.19824834167957306,
+      "learning_rate": 1.0734836709180184e-05,
+      "loss": 0.8608,
+      "step": 31190
+    },
+    {
+      "epoch": 0.9949296852578208,
+      "grad_norm": 0.20180141925811768,
+      "learning_rate": 1.0689889773511703e-05,
+      "loss": 0.8641,
+      "step": 31200
+    },
+    {
+      "epoch": 0.9952485729774546,
+      "grad_norm": 0.2057270109653473,
+      "learning_rate": 1.0645131031392937e-05,
+      "loss": 0.8876,
+      "step": 31210
+    },
+    {
+      "epoch": 0.9955674606970886,
+      "grad_norm": 0.19196845591068268,
+      "learning_rate": 1.060055969485445e-05,
+      "loss": 0.8536,
+      "step": 31220
+    },
+    {
+      "epoch": 0.9958863484167225,
+      "grad_norm": 0.18629442155361176,
+      "learning_rate": 1.0556174979226025e-05,
+      "loss": 0.8596,
+      "step": 31230
+    },
+    {
+      "epoch": 0.9962052361363564,
+      "grad_norm": 0.1941034346818924,
+      "learning_rate": 1.0511976103122883e-05,
+      "loss": 0.856,
+      "step": 31240
+    },
+    {
+      "epoch": 0.9965241238559903,
+      "grad_norm": 0.20251859724521637,
+      "learning_rate": 1.0467962288431913e-05,
+      "loss": 0.8726,
+      "step": 31250
+    },
+    {
+      "epoch": 0.9968430115756243,
+      "grad_norm": 0.19077450037002563,
+      "learning_rate": 1.0424132760297974e-05,
+      "loss": 0.8549,
+      "step": 31260
+    },
+    {
+      "epoch": 0.9971618992952581,
+      "grad_norm": 0.2010996788740158,
+      "learning_rate": 1.0380486747110261e-05,
+      "loss": 0.8776,
+      "step": 31270
+    },
+    {
+      "epoch": 0.9974807870148921,
+      "grad_norm": 0.2025349736213684,
+      "learning_rate": 1.033702348048871e-05,
+      "loss": 0.8617,
+      "step": 31280
+    },
+    {
+      "epoch": 0.997799674734526,
+      "grad_norm": 0.19157281517982483,
+      "learning_rate": 1.0293742195270484e-05,
+      "loss": 0.8576,
+      "step": 31290
+    },
+    {
+      "epoch": 0.9981185624541599,
+      "grad_norm": 0.20192226767539978,
+      "learning_rate": 1.0250642129496489e-05,
+      "loss": 0.8726,
+      "step": 31300
+    },
+    {
+      "epoch": 0.9984374501737938,
+      "grad_norm": 0.19540460407733917,
+      "learning_rate": 1.0207722524397968e-05,
+      "loss": 0.8525,
+      "step": 31310
+    },
+    {
+      "epoch": 0.9987563378934278,
+      "grad_norm": 0.19221796095371246,
+      "learning_rate": 1.0164982624383143e-05,
+      "loss": 0.8681,
+      "step": 31320
+    },
+    {
+      "epoch": 0.9990752256130616,
+      "grad_norm": 0.1946989744901657,
+      "learning_rate": 1.0122421677023911e-05,
+      "loss": 0.8784,
+      "step": 31330
+    },
+    {
+      "epoch": 0.9993941133326956,
+      "grad_norm": 0.1920672059059143,
+      "learning_rate": 1.0080038933042593e-05,
+      "loss": 0.8539,
+      "step": 31340
+    },
+    {
+      "epoch": 0.9997130010523295,
+      "grad_norm": 0.19885636866092682,
+      "learning_rate": 1.003783364629875e-05,
+      "loss": 0.8791,
+      "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": 1.1770770611838413e+18,
+  "train_batch_size": 512,
+  "trial_name": null,
+  "trial_params": null
+}