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": 3.2215495109558105,
+      "learning_rate": 3.1888771963391685e-05,
+      "loss": 9.4793,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006377754392678338,
+      "grad_norm": 1.5021514892578125,
+      "learning_rate": 6.377754392678337e-05,
+      "loss": 8.1296,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009566631589017507,
+      "grad_norm": 0.8992422819137573,
+      "learning_rate": 9.566631589017505e-05,
+      "loss": 7.1242,
+      "step": 30
+    },
+    {
+      "epoch": 0.0012755508785356677,
+      "grad_norm": 0.7726672887802124,
+      "learning_rate": 0.00012755508785356674,
+      "loss": 6.4833,
+      "step": 40
+    },
+    {
+      "epoch": 0.0015944385981695845,
+      "grad_norm": 0.9172486066818237,
+      "learning_rate": 0.00015944385981695843,
+      "loss": 6.1522,
+      "step": 50
+    },
+    {
+      "epoch": 0.0019133263178035013,
+      "grad_norm": 0.9134557247161865,
+      "learning_rate": 0.0001913326317803501,
+      "loss": 5.9715,
+      "step": 60
+    },
+    {
+      "epoch": 0.0022322140374374183,
+      "grad_norm": 0.7963905334472656,
+      "learning_rate": 0.00022322140374374182,
+      "loss": 5.792,
+      "step": 70
+    },
+    {
+      "epoch": 0.0025511017570713354,
+      "grad_norm": 1.257215142250061,
+      "learning_rate": 0.0002551101757071335,
+      "loss": 5.6415,
+      "step": 80
+    },
+    {
+      "epoch": 0.002869989476705252,
+      "grad_norm": 0.7022666931152344,
+      "learning_rate": 0.0002869989476705252,
+      "loss": 5.4826,
+      "step": 90
+    },
+    {
+      "epoch": 0.003188877196339169,
+      "grad_norm": 0.8765685558319092,
+      "learning_rate": 0.00031888771963391687,
+      "loss": 5.2978,
+      "step": 100
+    },
+    {
+      "epoch": 0.003507764915973086,
+      "grad_norm": 0.8249574303627014,
+      "learning_rate": 0.00035077649159730856,
+      "loss": 5.126,
+      "step": 110
+    },
+    {
+      "epoch": 0.0038266526356070026,
+      "grad_norm": 1.0378992557525635,
+      "learning_rate": 0.0003826652635607002,
+      "loss": 4.9622,
+      "step": 120
+    },
+    {
+      "epoch": 0.00414554035524092,
+      "grad_norm": 0.8625181913375854,
+      "learning_rate": 0.00041455403552409194,
+      "loss": 4.8241,
+      "step": 130
+    },
+    {
+      "epoch": 0.004464428074874837,
+      "grad_norm": 0.7947406768798828,
+      "learning_rate": 0.00044644280748748364,
+      "loss": 4.65,
+      "step": 140
+    },
+    {
+      "epoch": 0.004783315794508754,
+      "grad_norm": 0.7589473128318787,
+      "learning_rate": 0.0004783315794508753,
+      "loss": 4.5083,
+      "step": 150
+    },
+    {
+      "epoch": 0.005102203514142671,
+      "grad_norm": 0.8085631728172302,
+      "learning_rate": 0.000510220351414267,
+      "loss": 4.3782,
+      "step": 160
+    },
+    {
+      "epoch": 0.005421091233776587,
+      "grad_norm": 0.7931132912635803,
+      "learning_rate": 0.0005421091233776587,
+      "loss": 4.3206,
+      "step": 170
+    },
+    {
+      "epoch": 0.005739978953410504,
+      "grad_norm": 0.9039464592933655,
+      "learning_rate": 0.0005739978953410504,
+      "loss": 4.2164,
+      "step": 180
+    },
+    {
+      "epoch": 0.006058866673044421,
+      "grad_norm": 0.9057785272598267,
+      "learning_rate": 0.0006058866673044421,
+      "loss": 4.1272,
+      "step": 190
+    },
+    {
+      "epoch": 0.006377754392678338,
+      "grad_norm": 0.9266021847724915,
+      "learning_rate": 0.0006377754392678337,
+      "loss": 4.0183,
+      "step": 200
+    },
+    {
+      "epoch": 0.006696642112312255,
+      "grad_norm": 0.724240243434906,
+      "learning_rate": 0.0006696642112312254,
+      "loss": 3.9509,
+      "step": 210
+    },
+    {
+      "epoch": 0.007015529831946172,
+      "grad_norm": 0.7587215900421143,
+      "learning_rate": 0.0007015529831946171,
+      "loss": 3.8491,
+      "step": 220
+    },
+    {
+      "epoch": 0.007334417551580088,
+      "grad_norm": 0.8444303870201111,
+      "learning_rate": 0.0007334417551580089,
+      "loss": 3.7938,
+      "step": 230
+    },
+    {
+      "epoch": 0.007653305271214005,
+      "grad_norm": 0.7100092172622681,
+      "learning_rate": 0.0007653305271214004,
+      "loss": 3.7144,
+      "step": 240
+    },
+    {
+      "epoch": 0.007972192990847922,
+      "grad_norm": 0.7530874609947205,
+      "learning_rate": 0.0007972192990847921,
+      "loss": 3.6929,
+      "step": 250
+    },
+    {
+      "epoch": 0.00829108071048184,
+      "grad_norm": 0.8656921982765198,
+      "learning_rate": 0.0008291080710481839,
+      "loss": 3.6039,
+      "step": 260
+    },
+    {
+      "epoch": 0.008609968430115756,
+      "grad_norm": 0.6925029158592224,
+      "learning_rate": 0.0008609968430115756,
+      "loss": 3.5808,
+      "step": 270
+    },
+    {
+      "epoch": 0.008928856149749673,
+      "grad_norm": 0.6730265021324158,
+      "learning_rate": 0.0008928856149749673,
+      "loss": 3.5033,
+      "step": 280
+    },
+    {
+      "epoch": 0.00924774386938359,
+      "grad_norm": 0.8135679960250854,
+      "learning_rate": 0.0009247743869383588,
+      "loss": 3.4702,
+      "step": 290
+    },
+    {
+      "epoch": 0.009566631589017507,
+      "grad_norm": 0.7290470600128174,
+      "learning_rate": 0.0009566631589017505,
+      "loss": 3.4341,
+      "step": 300
+    },
+    {
+      "epoch": 0.009885519308651424,
+      "grad_norm": 0.7947729229927063,
+      "learning_rate": 0.0009885519308651422,
+      "loss": 3.3841,
+      "step": 310
+    },
+    {
+      "epoch": 0.010204407028285341,
+      "grad_norm": 0.8279941082000732,
+      "learning_rate": 0.001,
+      "loss": 3.3531,
+      "step": 320
+    },
+    {
+      "epoch": 0.010523294747919257,
+      "grad_norm": 0.8101890683174133,
+      "learning_rate": 0.001,
+      "loss": 3.2957,
+      "step": 330
+    },
+    {
+      "epoch": 0.010842182467553174,
+      "grad_norm": 0.6927778720855713,
+      "learning_rate": 0.001,
+      "loss": 3.272,
+      "step": 340
+    },
+    {
+      "epoch": 0.01116107018718709,
+      "grad_norm": 0.6906934976577759,
+      "learning_rate": 0.001,
+      "loss": 3.2044,
+      "step": 350
+    },
+    {
+      "epoch": 0.011479957906821008,
+      "grad_norm": 0.8388308882713318,
+      "learning_rate": 0.001,
+      "loss": 3.1869,
+      "step": 360
+    },
+    {
+      "epoch": 0.011798845626454925,
+      "grad_norm": 0.719241201877594,
+      "learning_rate": 0.001,
+      "loss": 3.1254,
+      "step": 370
+    },
+    {
+      "epoch": 0.012117733346088842,
+      "grad_norm": 0.7356446981430054,
+      "learning_rate": 0.001,
+      "loss": 3.107,
+      "step": 380
+    },
+    {
+      "epoch": 0.012436621065722759,
+      "grad_norm": 0.6843870878219604,
+      "learning_rate": 0.001,
+      "loss": 3.079,
+      "step": 390
+    },
+    {
+      "epoch": 0.012755508785356676,
+      "grad_norm": 0.6850771307945251,
+      "learning_rate": 0.001,
+      "loss": 3.04,
+      "step": 400
+    },
+    {
+      "epoch": 0.013074396504990593,
+      "grad_norm": 0.7865222096443176,
+      "learning_rate": 0.001,
+      "loss": 2.9971,
+      "step": 410
+    },
+    {
+      "epoch": 0.01339328422462451,
+      "grad_norm": 0.6876953840255737,
+      "learning_rate": 0.001,
+      "loss": 2.973,
+      "step": 420
+    },
+    {
+      "epoch": 0.013712171944258427,
+      "grad_norm": 0.7146596908569336,
+      "learning_rate": 0.001,
+      "loss": 2.9296,
+      "step": 430
+    },
+    {
+      "epoch": 0.014031059663892344,
+      "grad_norm": 0.6826304197311401,
+      "learning_rate": 0.001,
+      "loss": 2.8941,
+      "step": 440
+    },
+    {
+      "epoch": 0.014349947383526261,
+      "grad_norm": 0.8065870404243469,
+      "learning_rate": 0.001,
+      "loss": 2.8534,
+      "step": 450
+    },
+    {
+      "epoch": 0.014668835103160176,
+      "grad_norm": 0.6614107489585876,
+      "learning_rate": 0.001,
+      "loss": 2.8442,
+      "step": 460
+    },
+    {
+      "epoch": 0.014987722822794093,
+      "grad_norm": 0.6620937585830688,
+      "learning_rate": 0.001,
+      "loss": 2.8236,
+      "step": 470
+    },
+    {
+      "epoch": 0.01530661054242801,
+      "grad_norm": 0.7093802094459534,
+      "learning_rate": 0.001,
+      "loss": 2.8044,
+      "step": 480
+    },
+    {
+      "epoch": 0.01562549826206193,
+      "grad_norm": 0.6854017376899719,
+      "learning_rate": 0.001,
+      "loss": 2.7472,
+      "step": 490
+    },
+    {
+      "epoch": 0.015944385981695845,
+      "grad_norm": 0.6393734216690063,
+      "learning_rate": 0.001,
+      "loss": 2.7519,
+      "step": 500
+    },
+    {
+      "epoch": 0.016263273701329763,
+      "grad_norm": 0.6611132025718689,
+      "learning_rate": 0.001,
+      "loss": 2.7141,
+      "step": 510
+    },
+    {
+      "epoch": 0.01658216142096368,
+      "grad_norm": 0.68845534324646,
+      "learning_rate": 0.001,
+      "loss": 2.6999,
+      "step": 520
+    },
+    {
+      "epoch": 0.016901049140597594,
+      "grad_norm": 0.7871103882789612,
+      "learning_rate": 0.001,
+      "loss": 2.7055,
+      "step": 530
+    },
+    {
+      "epoch": 0.017219936860231513,
+      "grad_norm": 0.6550216674804688,
+      "learning_rate": 0.001,
+      "loss": 2.6843,
+      "step": 540
+    },
+    {
+      "epoch": 0.017538824579865428,
+      "grad_norm": 0.7184970378875732,
+      "learning_rate": 0.001,
+      "loss": 2.638,
+      "step": 550
+    },
+    {
+      "epoch": 0.017857712299499347,
+      "grad_norm": 0.6269710659980774,
+      "learning_rate": 0.001,
+      "loss": 2.6259,
+      "step": 560
+    },
+    {
+      "epoch": 0.018176600019133262,
+      "grad_norm": 0.622220516204834,
+      "learning_rate": 0.001,
+      "loss": 2.6048,
+      "step": 570
+    },
+    {
+      "epoch": 0.01849548773876718,
+      "grad_norm": 0.6675232648849487,
+      "learning_rate": 0.001,
+      "loss": 2.6026,
+      "step": 580
+    },
+    {
+      "epoch": 0.018814375458401096,
+      "grad_norm": 0.6323223114013672,
+      "learning_rate": 0.001,
+      "loss": 2.5598,
+      "step": 590
+    },
+    {
+      "epoch": 0.019133263178035015,
+      "grad_norm": 0.6446636915206909,
+      "learning_rate": 0.001,
+      "loss": 2.5686,
+      "step": 600
+    },
+    {
+      "epoch": 0.01945215089766893,
+      "grad_norm": 0.6598224639892578,
+      "learning_rate": 0.001,
+      "loss": 2.5535,
+      "step": 610
+    },
+    {
+      "epoch": 0.01977103861730285,
+      "grad_norm": 0.6299172639846802,
+      "learning_rate": 0.001,
+      "loss": 2.4965,
+      "step": 620
+    },
+    {
+      "epoch": 0.020089926336936764,
+      "grad_norm": 0.6265530586242676,
+      "learning_rate": 0.001,
+      "loss": 2.5023,
+      "step": 630
+    },
+    {
+      "epoch": 0.020408814056570683,
+      "grad_norm": 0.6296660304069519,
+      "learning_rate": 0.001,
+      "loss": 2.5133,
+      "step": 640
+    },
+    {
+      "epoch": 0.0207277017762046,
+      "grad_norm": 0.6240696907043457,
+      "learning_rate": 0.001,
+      "loss": 2.4724,
+      "step": 650
+    },
+    {
+      "epoch": 0.021046589495838514,
+      "grad_norm": 0.6711713671684265,
+      "learning_rate": 0.001,
+      "loss": 2.4841,
+      "step": 660
+    },
+    {
+      "epoch": 0.021365477215472432,
+      "grad_norm": 0.6260756254196167,
+      "learning_rate": 0.001,
+      "loss": 2.4608,
+      "step": 670
+    },
+    {
+      "epoch": 0.021684364935106348,
+      "grad_norm": 0.6255728602409363,
+      "learning_rate": 0.001,
+      "loss": 2.4568,
+      "step": 680
+    },
+    {
+      "epoch": 0.022003252654740266,
+      "grad_norm": 0.6062188148498535,
+      "learning_rate": 0.001,
+      "loss": 2.4329,
+      "step": 690
+    },
+    {
+      "epoch": 0.02232214037437418,
+      "grad_norm": 0.6365958452224731,
+      "learning_rate": 0.001,
+      "loss": 2.4095,
+      "step": 700
+    },
+    {
+      "epoch": 0.0226410280940081,
+      "grad_norm": 0.6736142635345459,
+      "learning_rate": 0.001,
+      "loss": 2.4445,
+      "step": 710
+    },
+    {
+      "epoch": 0.022959915813642016,
+      "grad_norm": 0.6197222471237183,
+      "learning_rate": 0.001,
+      "loss": 2.3984,
+      "step": 720
+    },
+    {
+      "epoch": 0.023278803533275935,
+      "grad_norm": 0.6317213177680969,
+      "learning_rate": 0.001,
+      "loss": 2.403,
+      "step": 730
+    },
+    {
+      "epoch": 0.02359769125290985,
+      "grad_norm": 0.6519914865493774,
+      "learning_rate": 0.001,
+      "loss": 2.3768,
+      "step": 740
+    },
+    {
+      "epoch": 0.02391657897254377,
+      "grad_norm": 0.749313473701477,
+      "learning_rate": 0.001,
+      "loss": 2.4057,
+      "step": 750
+    },
+    {
+      "epoch": 0.024235466692177684,
+      "grad_norm": 0.6350657939910889,
+      "learning_rate": 0.001,
+      "loss": 2.4037,
+      "step": 760
+    },
+    {
+      "epoch": 0.024554354411811603,
+      "grad_norm": 0.6400778889656067,
+      "learning_rate": 0.001,
+      "loss": 2.3994,
+      "step": 770
+    },
+    {
+      "epoch": 0.024873242131445518,
+      "grad_norm": 0.621253252029419,
+      "learning_rate": 0.001,
+      "loss": 2.3554,
+      "step": 780
+    },
+    {
+      "epoch": 0.025192129851079433,
+      "grad_norm": 0.6236314177513123,
+      "learning_rate": 0.001,
+      "loss": 2.3783,
+      "step": 790
+    },
+    {
+      "epoch": 0.025511017570713352,
+      "grad_norm": 0.6254419684410095,
+      "learning_rate": 0.001,
+      "loss": 2.3943,
+      "step": 800
+    },
+    {
+      "epoch": 0.025829905290347267,
+      "grad_norm": 0.6776648759841919,
+      "learning_rate": 0.001,
+      "loss": 2.3347,
+      "step": 810
+    },
+    {
+      "epoch": 0.026148793009981186,
+      "grad_norm": 0.6007510423660278,
+      "learning_rate": 0.001,
+      "loss": 2.372,
+      "step": 820
+    },
+    {
+      "epoch": 0.0264676807296151,
+      "grad_norm": 0.5976624488830566,
+      "learning_rate": 0.001,
+      "loss": 2.33,
+      "step": 830
+    },
+    {
+      "epoch": 0.02678656844924902,
+      "grad_norm": 0.6124452352523804,
+      "learning_rate": 0.001,
+      "loss": 2.327,
+      "step": 840
+    },
+    {
+      "epoch": 0.027105456168882935,
+      "grad_norm": 0.5857394337654114,
+      "learning_rate": 0.001,
+      "loss": 2.3208,
+      "step": 850
+    },
+    {
+      "epoch": 0.027424343888516854,
+      "grad_norm": 0.6345723271369934,
+      "learning_rate": 0.001,
+      "loss": 2.317,
+      "step": 860
+    },
+    {
+      "epoch": 0.02774323160815077,
+      "grad_norm": 0.6067624688148499,
+      "learning_rate": 0.001,
+      "loss": 2.2959,
+      "step": 870
+    },
+    {
+      "epoch": 0.028062119327784688,
+      "grad_norm": 0.5995084047317505,
+      "learning_rate": 0.001,
+      "loss": 2.3031,
+      "step": 880
+    },
+    {
+      "epoch": 0.028381007047418604,
+      "grad_norm": 0.6133849024772644,
+      "learning_rate": 0.001,
+      "loss": 2.283,
+      "step": 890
+    },
+    {
+      "epoch": 0.028699894767052522,
+      "grad_norm": 0.6345779895782471,
+      "learning_rate": 0.001,
+      "loss": 2.2874,
+      "step": 900
+    },
+    {
+      "epoch": 0.029018782486686438,
+      "grad_norm": 0.5908248424530029,
+      "learning_rate": 0.001,
+      "loss": 2.2586,
+      "step": 910
+    },
+    {
+      "epoch": 0.029337670206320353,
+      "grad_norm": 0.5945861339569092,
+      "learning_rate": 0.001,
+      "loss": 2.2623,
+      "step": 920
+    },
+    {
+      "epoch": 0.02965655792595427,
+      "grad_norm": 0.6086012721061707,
+      "learning_rate": 0.001,
+      "loss": 2.2776,
+      "step": 930
+    },
+    {
+      "epoch": 0.029975445645588187,
+      "grad_norm": 0.596208930015564,
+      "learning_rate": 0.001,
+      "loss": 2.2784,
+      "step": 940
+    },
+    {
+      "epoch": 0.030294333365222106,
+      "grad_norm": 0.618234395980835,
+      "learning_rate": 0.001,
+      "loss": 2.2536,
+      "step": 950
+    },
+    {
+      "epoch": 0.03061322108485602,
+      "grad_norm": 0.6146029829978943,
+      "learning_rate": 0.001,
+      "loss": 2.2574,
+      "step": 960
+    },
+    {
+      "epoch": 0.03093210880448994,
+      "grad_norm": 0.6115110516548157,
+      "learning_rate": 0.001,
+      "loss": 2.2441,
+      "step": 970
+    },
+    {
+      "epoch": 0.03125099652412386,
+      "grad_norm": 0.6049351692199707,
+      "learning_rate": 0.001,
+      "loss": 2.2694,
+      "step": 980
+    },
+    {
+      "epoch": 0.03156988424375777,
+      "grad_norm": 0.5909845232963562,
+      "learning_rate": 0.001,
+      "loss": 2.2532,
+      "step": 990
+    },
+    {
+      "epoch": 0.03188877196339169,
+      "grad_norm": 0.5601927638053894,
+      "learning_rate": 0.001,
+      "loss": 2.2379,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03220765968302561,
+      "grad_norm": 0.6240832209587097,
+      "learning_rate": 0.001,
+      "loss": 2.2355,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03252654740265953,
+      "grad_norm": 0.6171403527259827,
+      "learning_rate": 0.001,
+      "loss": 2.252,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03284543512229344,
+      "grad_norm": 0.587386429309845,
+      "learning_rate": 0.001,
+      "loss": 2.2236,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03316432284192736,
+      "grad_norm": 0.5910850167274475,
+      "learning_rate": 0.001,
+      "loss": 2.1752,
+      "step": 1040
+    },
+    {
+      "epoch": 0.033483210561561276,
+      "grad_norm": 0.5837855339050293,
+      "learning_rate": 0.001,
+      "loss": 2.2151,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03380209828119519,
+      "grad_norm": 0.5770293474197388,
+      "learning_rate": 0.001,
+      "loss": 2.1735,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03412098600082911,
+      "grad_norm": 0.6157304644584656,
+      "learning_rate": 0.001,
+      "loss": 2.2111,
+      "step": 1070
+    },
+    {
+      "epoch": 0.034439873720463025,
+      "grad_norm": 0.6692126989364624,
+      "learning_rate": 0.001,
+      "loss": 2.2099,
+      "step": 1080
+    },
+    {
+      "epoch": 0.034758761440096944,
+      "grad_norm": 0.6117860078811646,
+      "learning_rate": 0.001,
+      "loss": 2.2038,
+      "step": 1090
+    },
+    {
+      "epoch": 0.035077649159730856,
+      "grad_norm": 0.6830983757972717,
+      "learning_rate": 0.001,
+      "loss": 2.2173,
+      "step": 1100
+    },
+    {
+      "epoch": 0.035396536879364775,
+      "grad_norm": 0.5630941987037659,
+      "learning_rate": 0.001,
+      "loss": 2.1761,
+      "step": 1110
+    },
+    {
+      "epoch": 0.035715424598998693,
+      "grad_norm": 0.5978312492370605,
+      "learning_rate": 0.001,
+      "loss": 2.1767,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03603431231863261,
+      "grad_norm": 0.6004764437675476,
+      "learning_rate": 0.001,
+      "loss": 2.1733,
+      "step": 1130
+    },
+    {
+      "epoch": 0.036353200038266524,
+      "grad_norm": 0.619663655757904,
+      "learning_rate": 0.001,
+      "loss": 2.1967,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03667208775790044,
+      "grad_norm": 0.653950035572052,
+      "learning_rate": 0.001,
+      "loss": 2.1843,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03699097547753436,
+      "grad_norm": 0.6021509766578674,
+      "learning_rate": 0.001,
+      "loss": 2.1771,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03730986319716828,
+      "grad_norm": 0.5993155837059021,
+      "learning_rate": 0.001,
+      "loss": 2.136,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03762875091680219,
+      "grad_norm": 0.5950784683227539,
+      "learning_rate": 0.001,
+      "loss": 2.1163,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03794763863643611,
+      "grad_norm": 0.5831239819526672,
+      "learning_rate": 0.001,
+      "loss": 2.1326,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03826652635607003,
+      "grad_norm": 0.5934585332870483,
+      "learning_rate": 0.001,
+      "loss": 2.1463,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03858541407570394,
+      "grad_norm": 0.5983244776725769,
+      "learning_rate": 0.001,
+      "loss": 2.1268,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03890430179533786,
+      "grad_norm": 0.6157175302505493,
+      "learning_rate": 0.001,
+      "loss": 2.145,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03922318951497178,
+      "grad_norm": 0.5686699151992798,
+      "learning_rate": 0.001,
+      "loss": 2.1292,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0395420772346057,
+      "grad_norm": 0.6384000182151794,
+      "learning_rate": 0.001,
+      "loss": 2.1362,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03986096495423961,
+      "grad_norm": 0.5988278985023499,
+      "learning_rate": 0.001,
+      "loss": 2.1187,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04017985267387353,
+      "grad_norm": 0.5760336518287659,
+      "learning_rate": 0.001,
+      "loss": 2.1132,
+      "step": 1260
+    },
+    {
+      "epoch": 0.04049874039350745,
+      "grad_norm": 0.5890830159187317,
+      "learning_rate": 0.001,
+      "loss": 2.1298,
+      "step": 1270
+    },
+    {
+      "epoch": 0.040817628113141366,
+      "grad_norm": 0.5832691788673401,
+      "learning_rate": 0.001,
+      "loss": 2.1087,
+      "step": 1280
+    },
+    {
+      "epoch": 0.04113651583277528,
+      "grad_norm": 0.5678040981292725,
+      "learning_rate": 0.001,
+      "loss": 2.1192,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0414554035524092,
+      "grad_norm": 0.5775576829910278,
+      "learning_rate": 0.001,
+      "loss": 2.0936,
+      "step": 1300
+    },
+    {
+      "epoch": 0.041774291272043115,
+      "grad_norm": 0.6111060380935669,
+      "learning_rate": 0.001,
+      "loss": 2.0832,
+      "step": 1310
+    },
+    {
+      "epoch": 0.04209317899167703,
+      "grad_norm": 0.5604788661003113,
+      "learning_rate": 0.001,
+      "loss": 2.1251,
+      "step": 1320
+    },
+    {
+      "epoch": 0.042412066711310946,
+      "grad_norm": 0.6098892688751221,
+      "learning_rate": 0.001,
+      "loss": 2.1046,
+      "step": 1330
+    },
+    {
+      "epoch": 0.042730954430944865,
+      "grad_norm": 0.5982756614685059,
+      "learning_rate": 0.001,
+      "loss": 2.0743,
+      "step": 1340
+    },
+    {
+      "epoch": 0.04304984215057878,
+      "grad_norm": 0.5638964176177979,
+      "learning_rate": 0.001,
+      "loss": 2.1103,
+      "step": 1350
+    },
+    {
+      "epoch": 0.043368729870212695,
+      "grad_norm": 0.5656088590621948,
+      "learning_rate": 0.001,
+      "loss": 2.0873,
+      "step": 1360
+    },
+    {
+      "epoch": 0.043687617589846614,
+      "grad_norm": 0.5705325603485107,
+      "learning_rate": 0.001,
+      "loss": 2.0958,
+      "step": 1370
+    },
+    {
+      "epoch": 0.04400650530948053,
+      "grad_norm": 0.5603035688400269,
+      "learning_rate": 0.001,
+      "loss": 2.1022,
+      "step": 1380
+    },
+    {
+      "epoch": 0.04432539302911445,
+      "grad_norm": 0.56059330701828,
+      "learning_rate": 0.001,
+      "loss": 2.0511,
+      "step": 1390
+    },
+    {
+      "epoch": 0.04464428074874836,
+      "grad_norm": 0.5691202878952026,
+      "learning_rate": 0.001,
+      "loss": 2.0845,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04496316846838228,
+      "grad_norm": 0.5622802376747131,
+      "learning_rate": 0.001,
+      "loss": 2.0817,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0452820561880162,
+      "grad_norm": 0.5683780908584595,
+      "learning_rate": 0.001,
+      "loss": 2.05,
+      "step": 1420
+    },
+    {
+      "epoch": 0.04560094390765012,
+      "grad_norm": 0.5761978626251221,
+      "learning_rate": 0.001,
+      "loss": 2.0766,
+      "step": 1430
+    },
+    {
+      "epoch": 0.04591983162728403,
+      "grad_norm": 0.5588021278381348,
+      "learning_rate": 0.001,
+      "loss": 2.0496,
+      "step": 1440
+    },
+    {
+      "epoch": 0.04623871934691795,
+      "grad_norm": 0.5767732858657837,
+      "learning_rate": 0.001,
+      "loss": 2.0992,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04655760706655187,
+      "grad_norm": 0.5720701217651367,
+      "learning_rate": 0.001,
+      "loss": 2.0471,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04687649478618578,
+      "grad_norm": 0.6206384301185608,
+      "learning_rate": 0.001,
+      "loss": 2.0519,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0471953825058197,
+      "grad_norm": 0.5901126861572266,
+      "learning_rate": 0.001,
+      "loss": 2.0665,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04751427022545362,
+      "grad_norm": 0.5666248202323914,
+      "learning_rate": 0.001,
+      "loss": 2.0524,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04783315794508754,
+      "grad_norm": 0.5244306325912476,
+      "learning_rate": 0.001,
+      "loss": 2.0374,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04815204566472145,
+      "grad_norm": 0.5552513599395752,
+      "learning_rate": 0.001,
+      "loss": 2.0222,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04847093338435537,
+      "grad_norm": 0.5625154376029968,
+      "learning_rate": 0.001,
+      "loss": 2.0465,
+      "step": 1520
+    },
+    {
+      "epoch": 0.048789821103989287,
+      "grad_norm": 0.5688402056694031,
+      "learning_rate": 0.001,
+      "loss": 2.0549,
+      "step": 1530
+    },
+    {
+      "epoch": 0.049108708823623205,
+      "grad_norm": 0.5560811758041382,
+      "learning_rate": 0.001,
+      "loss": 2.0122,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04942759654325712,
+      "grad_norm": 0.5479153394699097,
+      "learning_rate": 0.001,
+      "loss": 2.0345,
+      "step": 1550
+    },
+    {
+      "epoch": 0.049746484262891036,
+      "grad_norm": 0.5255228877067566,
+      "learning_rate": 0.001,
+      "loss": 2.0217,
+      "step": 1560
+    },
+    {
+      "epoch": 0.050065371982524955,
+      "grad_norm": 0.582529604434967,
+      "learning_rate": 0.001,
+      "loss": 2.0241,
+      "step": 1570
+    },
+    {
+      "epoch": 0.050384259702158866,
+      "grad_norm": 0.53663569688797,
+      "learning_rate": 0.001,
+      "loss": 2.0001,
+      "step": 1580
+    },
+    {
+      "epoch": 0.050703147421792785,
+      "grad_norm": 0.550887405872345,
+      "learning_rate": 0.001,
+      "loss": 2.0114,
+      "step": 1590
+    },
+    {
+      "epoch": 0.051022035141426704,
+      "grad_norm": 0.5510720610618591,
+      "learning_rate": 0.001,
+      "loss": 2.0124,
+      "step": 1600
+    },
+    {
+      "epoch": 0.05134092286106062,
+      "grad_norm": 0.5833602547645569,
+      "learning_rate": 0.001,
+      "loss": 2.0178,
+      "step": 1610
+    },
+    {
+      "epoch": 0.051659810580694535,
+      "grad_norm": 0.5617836713790894,
+      "learning_rate": 0.001,
+      "loss": 2.0381,
+      "step": 1620
+    },
+    {
+      "epoch": 0.05197869830032845,
+      "grad_norm": 0.5440929532051086,
+      "learning_rate": 0.001,
+      "loss": 1.9956,
+      "step": 1630
+    },
+    {
+      "epoch": 0.05229758601996237,
+      "grad_norm": 0.5375956892967224,
+      "learning_rate": 0.001,
+      "loss": 2.0045,
+      "step": 1640
+    },
+    {
+      "epoch": 0.05261647373959629,
+      "grad_norm": 0.5563559532165527,
+      "learning_rate": 0.001,
+      "loss": 2.0267,
+      "step": 1650
+    },
+    {
+      "epoch": 0.0529353614592302,
+      "grad_norm": 0.5130823850631714,
+      "learning_rate": 0.001,
+      "loss": 2.007,
+      "step": 1660
+    },
+    {
+      "epoch": 0.05325424917886412,
+      "grad_norm": 0.5466340780258179,
+      "learning_rate": 0.001,
+      "loss": 2.0031,
+      "step": 1670
+    },
+    {
+      "epoch": 0.05357313689849804,
+      "grad_norm": 0.5090089440345764,
+      "learning_rate": 0.001,
+      "loss": 1.964,
+      "step": 1680
+    },
+    {
+      "epoch": 0.05389202461813196,
+      "grad_norm": 0.5429556369781494,
+      "learning_rate": 0.001,
+      "loss": 1.9897,
+      "step": 1690
+    },
+    {
+      "epoch": 0.05421091233776587,
+      "grad_norm": 0.5265209078788757,
+      "learning_rate": 0.001,
+      "loss": 1.9896,
+      "step": 1700
+    },
+    {
+      "epoch": 0.05452980005739979,
+      "grad_norm": 0.5308935046195984,
+      "learning_rate": 0.001,
+      "loss": 1.9797,
+      "step": 1710
+    },
+    {
+      "epoch": 0.05484868777703371,
+      "grad_norm": 0.52045077085495,
+      "learning_rate": 0.001,
+      "loss": 1.9848,
+      "step": 1720
+    },
+    {
+      "epoch": 0.05516757549666762,
+      "grad_norm": 0.5526434183120728,
+      "learning_rate": 0.001,
+      "loss": 1.9823,
+      "step": 1730
+    },
+    {
+      "epoch": 0.05548646321630154,
+      "grad_norm": 0.5421900749206543,
+      "learning_rate": 0.001,
+      "loss": 1.9867,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05580535093593546,
+      "grad_norm": 0.5268239974975586,
+      "learning_rate": 0.001,
+      "loss": 1.973,
+      "step": 1750
+    },
+    {
+      "epoch": 0.056124238655569376,
+      "grad_norm": 0.5331604480743408,
+      "learning_rate": 0.001,
+      "loss": 1.969,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05644312637520329,
+      "grad_norm": 0.5327712297439575,
+      "learning_rate": 0.001,
+      "loss": 1.966,
+      "step": 1770
+    },
+    {
+      "epoch": 0.05676201409483721,
+      "grad_norm": 0.526186466217041,
+      "learning_rate": 0.001,
+      "loss": 1.9931,
+      "step": 1780
+    },
+    {
+      "epoch": 0.057080901814471126,
+      "grad_norm": 0.5733531713485718,
+      "learning_rate": 0.001,
+      "loss": 1.9699,
+      "step": 1790
+    },
+    {
+      "epoch": 0.057399789534105045,
+      "grad_norm": 0.5519174933433533,
+      "learning_rate": 0.001,
+      "loss": 1.9623,
+      "step": 1800
+    },
+    {
+      "epoch": 0.057718677253738956,
+      "grad_norm": 0.6039563417434692,
+      "learning_rate": 0.001,
+      "loss": 1.9908,
+      "step": 1810
+    },
+    {
+      "epoch": 0.058037564973372875,
+      "grad_norm": 0.5261481404304504,
+      "learning_rate": 0.001,
+      "loss": 1.9698,
+      "step": 1820
+    },
+    {
+      "epoch": 0.058356452693006794,
+      "grad_norm": 0.5257373452186584,
+      "learning_rate": 0.001,
+      "loss": 1.9554,
+      "step": 1830
+    },
+    {
+      "epoch": 0.058675340412640706,
+      "grad_norm": 0.5442422032356262,
+      "learning_rate": 0.001,
+      "loss": 1.956,
+      "step": 1840
+    },
+    {
+      "epoch": 0.058994228132274625,
+      "grad_norm": 0.5888342261314392,
+      "learning_rate": 0.001,
+      "loss": 1.9254,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05931311585190854,
+      "grad_norm": 0.5307484865188599,
+      "learning_rate": 0.001,
+      "loss": 1.9524,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05963200357154246,
+      "grad_norm": 0.5235530734062195,
+      "learning_rate": 0.001,
+      "loss": 1.9542,
+      "step": 1870
+    },
+    {
+      "epoch": 0.059950891291176374,
+      "grad_norm": 0.4977993071079254,
+      "learning_rate": 0.001,
+      "loss": 1.9636,
+      "step": 1880
+    },
+    {
+      "epoch": 0.06026977901081029,
+      "grad_norm": 0.5077757239341736,
+      "learning_rate": 0.001,
+      "loss": 1.9524,
+      "step": 1890
+    },
+    {
+      "epoch": 0.06058866673044421,
+      "grad_norm": 0.5321727395057678,
+      "learning_rate": 0.001,
+      "loss": 1.9548,
+      "step": 1900
+    },
+    {
+      "epoch": 0.06090755445007813,
+      "grad_norm": 0.49839040637016296,
+      "learning_rate": 0.001,
+      "loss": 1.9353,
+      "step": 1910
+    },
+    {
+      "epoch": 0.06122644216971204,
+      "grad_norm": 0.5668623447418213,
+      "learning_rate": 0.001,
+      "loss": 1.9226,
+      "step": 1920
+    },
+    {
+      "epoch": 0.06154532988934596,
+      "grad_norm": 0.4942353665828705,
+      "learning_rate": 0.001,
+      "loss": 1.9532,
+      "step": 1930
+    },
+    {
+      "epoch": 0.06186421760897988,
+      "grad_norm": 0.4908944070339203,
+      "learning_rate": 0.001,
+      "loss": 1.9236,
+      "step": 1940
+    },
+    {
+      "epoch": 0.0621831053286138,
+      "grad_norm": 0.5142762660980225,
+      "learning_rate": 0.001,
+      "loss": 1.9319,
+      "step": 1950
+    },
+    {
+      "epoch": 0.06250199304824772,
+      "grad_norm": 0.5228368043899536,
+      "learning_rate": 0.001,
+      "loss": 1.9281,
+      "step": 1960
+    },
+    {
+      "epoch": 0.06282088076788163,
+      "grad_norm": 0.5202351212501526,
+      "learning_rate": 0.001,
+      "loss": 1.9435,
+      "step": 1970
+    },
+    {
+      "epoch": 0.06313976848751554,
+      "grad_norm": 0.5125612616539001,
+      "learning_rate": 0.001,
+      "loss": 1.9153,
+      "step": 1980
+    },
+    {
+      "epoch": 0.06345865620714947,
+      "grad_norm": 0.4858326315879822,
+      "learning_rate": 0.001,
+      "loss": 1.9081,
+      "step": 1990
+    },
+    {
+      "epoch": 0.06377754392678338,
+      "grad_norm": 0.48831015825271606,
+      "learning_rate": 0.001,
+      "loss": 1.9123,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06409643164641729,
+      "grad_norm": 0.5106445550918579,
+      "learning_rate": 0.001,
+      "loss": 1.9144,
+      "step": 2010
+    },
+    {
+      "epoch": 0.06441531936605122,
+      "grad_norm": 0.5125754475593567,
+      "learning_rate": 0.001,
+      "loss": 1.9336,
+      "step": 2020
+    },
+    {
+      "epoch": 0.06473420708568513,
+      "grad_norm": 0.5162148475646973,
+      "learning_rate": 0.001,
+      "loss": 1.9143,
+      "step": 2030
+    },
+    {
+      "epoch": 0.06505309480531905,
+      "grad_norm": 0.5132099986076355,
+      "learning_rate": 0.001,
+      "loss": 1.9116,
+      "step": 2040
+    },
+    {
+      "epoch": 0.06537198252495297,
+      "grad_norm": 0.4962635338306427,
+      "learning_rate": 0.001,
+      "loss": 1.8872,
+      "step": 2050
+    },
+    {
+      "epoch": 0.06569087024458688,
+      "grad_norm": 0.5262729525566101,
+      "learning_rate": 0.001,
+      "loss": 1.8783,
+      "step": 2060
+    },
+    {
+      "epoch": 0.0660097579642208,
+      "grad_norm": 0.49188458919525146,
+      "learning_rate": 0.001,
+      "loss": 1.91,
+      "step": 2070
+    },
+    {
+      "epoch": 0.06632864568385471,
+      "grad_norm": 0.4812409579753876,
+      "learning_rate": 0.001,
+      "loss": 1.8985,
+      "step": 2080
+    },
+    {
+      "epoch": 0.06664753340348863,
+      "grad_norm": 0.48991620540618896,
+      "learning_rate": 0.001,
+      "loss": 1.9143,
+      "step": 2090
+    },
+    {
+      "epoch": 0.06696642112312255,
+      "grad_norm": 0.481096088886261,
+      "learning_rate": 0.001,
+      "loss": 1.9079,
+      "step": 2100
+    },
+    {
+      "epoch": 0.06728530884275646,
+      "grad_norm": 0.4740917384624481,
+      "learning_rate": 0.001,
+      "loss": 1.8818,
+      "step": 2110
+    },
+    {
+      "epoch": 0.06760419656239038,
+      "grad_norm": 0.5008795261383057,
+      "learning_rate": 0.001,
+      "loss": 1.8831,
+      "step": 2120
+    },
+    {
+      "epoch": 0.0679230842820243,
+      "grad_norm": 0.49072882533073425,
+      "learning_rate": 0.001,
+      "loss": 1.8938,
+      "step": 2130
+    },
+    {
+      "epoch": 0.06824197200165821,
+      "grad_norm": 0.47333481907844543,
+      "learning_rate": 0.001,
+      "loss": 1.8753,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06856085972129214,
+      "grad_norm": 0.4727560579776764,
+      "learning_rate": 0.001,
+      "loss": 1.8894,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06887974744092605,
+      "grad_norm": 0.4693233072757721,
+      "learning_rate": 0.001,
+      "loss": 1.8974,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06919863516055996,
+      "grad_norm": 0.4827735722064972,
+      "learning_rate": 0.001,
+      "loss": 1.88,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06951752288019389,
+      "grad_norm": 0.5058079361915588,
+      "learning_rate": 0.001,
+      "loss": 1.8769,
+      "step": 2180
+    },
+    {
+      "epoch": 0.0698364105998278,
+      "grad_norm": 0.4859229624271393,
+      "learning_rate": 0.001,
+      "loss": 1.8586,
+      "step": 2190
+    },
+    {
+      "epoch": 0.07015529831946171,
+      "grad_norm": 0.49526458978652954,
+      "learning_rate": 0.001,
+      "loss": 1.8846,
+      "step": 2200
+    },
+    {
+      "epoch": 0.07047418603909564,
+      "grad_norm": 0.4747912585735321,
+      "learning_rate": 0.001,
+      "loss": 1.8608,
+      "step": 2210
+    },
+    {
+      "epoch": 0.07079307375872955,
+      "grad_norm": 0.5108914375305176,
+      "learning_rate": 0.001,
+      "loss": 1.8728,
+      "step": 2220
+    },
+    {
+      "epoch": 0.07111196147836348,
+      "grad_norm": 0.44781816005706787,
+      "learning_rate": 0.001,
+      "loss": 1.8597,
+      "step": 2230
+    },
+    {
+      "epoch": 0.07143084919799739,
+      "grad_norm": 0.4641876518726349,
+      "learning_rate": 0.001,
+      "loss": 1.8608,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0717497369176313,
+      "grad_norm": 0.47702521085739136,
+      "learning_rate": 0.001,
+      "loss": 1.838,
+      "step": 2250
+    },
+    {
+      "epoch": 0.07206862463726522,
+      "grad_norm": 0.46187707781791687,
+      "learning_rate": 0.001,
+      "loss": 1.8257,
+      "step": 2260
+    },
+    {
+      "epoch": 0.07238751235689914,
+      "grad_norm": 0.4767541289329529,
+      "learning_rate": 0.001,
+      "loss": 1.8606,
+      "step": 2270
+    },
+    {
+      "epoch": 0.07270640007653305,
+      "grad_norm": 0.47086700797080994,
+      "learning_rate": 0.001,
+      "loss": 1.8422,
+      "step": 2280
+    },
+    {
+      "epoch": 0.07302528779616697,
+      "grad_norm": 0.4712027311325073,
+      "learning_rate": 0.001,
+      "loss": 1.8214,
+      "step": 2290
+    },
+    {
+      "epoch": 0.07334417551580089,
+      "grad_norm": 0.4686843156814575,
+      "learning_rate": 0.001,
+      "loss": 1.8307,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0736630632354348,
+      "grad_norm": 0.4426477253437042,
+      "learning_rate": 0.001,
+      "loss": 1.8205,
+      "step": 2310
+    },
+    {
+      "epoch": 0.07398195095506872,
+      "grad_norm": 0.465758353471756,
+      "learning_rate": 0.001,
+      "loss": 1.8373,
+      "step": 2320
+    },
+    {
+      "epoch": 0.07430083867470264,
+      "grad_norm": 0.45455947518348694,
+      "learning_rate": 0.001,
+      "loss": 1.837,
+      "step": 2330
+    },
+    {
+      "epoch": 0.07461972639433656,
+      "grad_norm": 0.4629955291748047,
+      "learning_rate": 0.001,
+      "loss": 1.8196,
+      "step": 2340
+    },
+    {
+      "epoch": 0.07493861411397047,
+      "grad_norm": 0.4526189863681793,
+      "learning_rate": 0.001,
+      "loss": 1.8441,
+      "step": 2350
+    },
+    {
+      "epoch": 0.07525750183360438,
+      "grad_norm": 0.4585621953010559,
+      "learning_rate": 0.001,
+      "loss": 1.8329,
+      "step": 2360
+    },
+    {
+      "epoch": 0.07557638955323831,
+      "grad_norm": 0.4587751626968384,
+      "learning_rate": 0.001,
+      "loss": 1.7881,
+      "step": 2370
+    },
+    {
+      "epoch": 0.07589527727287222,
+      "grad_norm": 0.4355972111225128,
+      "learning_rate": 0.001,
+      "loss": 1.796,
+      "step": 2380
+    },
+    {
+      "epoch": 0.07621416499250613,
+      "grad_norm": 0.4731498956680298,
+      "learning_rate": 0.001,
+      "loss": 1.8371,
+      "step": 2390
+    },
+    {
+      "epoch": 0.07653305271214006,
+      "grad_norm": 0.45605430006980896,
+      "learning_rate": 0.001,
+      "loss": 1.8212,
+      "step": 2400
+    },
+    {
+      "epoch": 0.07685194043177397,
+      "grad_norm": 0.4283037483692169,
+      "learning_rate": 0.001,
+      "loss": 1.8333,
+      "step": 2410
+    },
+    {
+      "epoch": 0.07717082815140788,
+      "grad_norm": 0.4527207911014557,
+      "learning_rate": 0.001,
+      "loss": 1.8028,
+      "step": 2420
+    },
+    {
+      "epoch": 0.07748971587104181,
+      "grad_norm": 0.44645947217941284,
+      "learning_rate": 0.001,
+      "loss": 1.8036,
+      "step": 2430
+    },
+    {
+      "epoch": 0.07780860359067572,
+      "grad_norm": 0.43729937076568604,
+      "learning_rate": 0.001,
+      "loss": 1.7946,
+      "step": 2440
+    },
+    {
+      "epoch": 0.07812749131030965,
+      "grad_norm": 0.4428195655345917,
+      "learning_rate": 0.001,
+      "loss": 1.7917,
+      "step": 2450
+    },
+    {
+      "epoch": 0.07844637902994356,
+      "grad_norm": 0.45311400294303894,
+      "learning_rate": 0.001,
+      "loss": 1.8078,
+      "step": 2460
+    },
+    {
+      "epoch": 0.07876526674957747,
+      "grad_norm": 0.43074876070022583,
+      "learning_rate": 0.001,
+      "loss": 1.8011,
+      "step": 2470
+    },
+    {
+      "epoch": 0.0790841544692114,
+      "grad_norm": 0.43784239888191223,
+      "learning_rate": 0.001,
+      "loss": 1.7801,
+      "step": 2480
+    },
+    {
+      "epoch": 0.07940304218884531,
+      "grad_norm": 0.42712098360061646,
+      "learning_rate": 0.001,
+      "loss": 1.8244,
+      "step": 2490
+    },
+    {
+      "epoch": 0.07972192990847922,
+      "grad_norm": 0.45883429050445557,
+      "learning_rate": 0.001,
+      "loss": 1.8038,
+      "step": 2500
+    },
+    {
+      "epoch": 0.08004081762811315,
+      "grad_norm": 0.4601421654224396,
+      "learning_rate": 0.001,
+      "loss": 1.7772,
+      "step": 2510
+    },
+    {
+      "epoch": 0.08035970534774706,
+      "grad_norm": 0.47060567140579224,
+      "learning_rate": 0.001,
+      "loss": 1.8024,
+      "step": 2520
+    },
+    {
+      "epoch": 0.08067859306738097,
+      "grad_norm": 0.42333802580833435,
+      "learning_rate": 0.001,
+      "loss": 1.7989,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0809974807870149,
+      "grad_norm": 0.414415180683136,
+      "learning_rate": 0.001,
+      "loss": 1.792,
+      "step": 2540
+    },
+    {
+      "epoch": 0.0813163685066488,
+      "grad_norm": 0.41989174485206604,
+      "learning_rate": 0.001,
+      "loss": 1.7762,
+      "step": 2550
+    },
+    {
+      "epoch": 0.08163525622628273,
+      "grad_norm": 0.4492782950401306,
+      "learning_rate": 0.001,
+      "loss": 1.7852,
+      "step": 2560
+    },
+    {
+      "epoch": 0.08195414394591664,
+      "grad_norm": 0.43293607234954834,
+      "learning_rate": 0.001,
+      "loss": 1.7578,
+      "step": 2570
+    },
+    {
+      "epoch": 0.08227303166555056,
+      "grad_norm": 0.4404749572277069,
+      "learning_rate": 0.001,
+      "loss": 1.7949,
+      "step": 2580
+    },
+    {
+      "epoch": 0.08259191938518448,
+      "grad_norm": 0.482251912355423,
+      "learning_rate": 0.001,
+      "loss": 1.7615,
+      "step": 2590
+    },
+    {
+      "epoch": 0.0829108071048184,
+      "grad_norm": 0.441723495721817,
+      "learning_rate": 0.001,
+      "loss": 1.7755,
+      "step": 2600
+    },
+    {
+      "epoch": 0.0832296948244523,
+      "grad_norm": 0.4477382302284241,
+      "learning_rate": 0.001,
+      "loss": 1.7761,
+      "step": 2610
+    },
+    {
+      "epoch": 0.08354858254408623,
+      "grad_norm": 0.4355809688568115,
+      "learning_rate": 0.001,
+      "loss": 1.7598,
+      "step": 2620
+    },
+    {
+      "epoch": 0.08386747026372014,
+      "grad_norm": 0.4181293547153473,
+      "learning_rate": 0.001,
+      "loss": 1.7601,
+      "step": 2630
+    },
+    {
+      "epoch": 0.08418635798335405,
+      "grad_norm": 0.4101440906524658,
+      "learning_rate": 0.001,
+      "loss": 1.7626,
+      "step": 2640
+    },
+    {
+      "epoch": 0.08450524570298798,
+      "grad_norm": 0.41453298926353455,
+      "learning_rate": 0.001,
+      "loss": 1.7624,
+      "step": 2650
+    },
+    {
+      "epoch": 0.08482413342262189,
+      "grad_norm": 0.4500748813152313,
+      "learning_rate": 0.001,
+      "loss": 1.7717,
+      "step": 2660
+    },
+    {
+      "epoch": 0.08514302114225582,
+      "grad_norm": 0.45811501145362854,
+      "learning_rate": 0.001,
+      "loss": 1.7451,
+      "step": 2670
+    },
+    {
+      "epoch": 0.08546190886188973,
+      "grad_norm": 0.40631115436553955,
+      "learning_rate": 0.001,
+      "loss": 1.7792,
+      "step": 2680
+    },
+    {
+      "epoch": 0.08578079658152364,
+      "grad_norm": 0.4175332188606262,
+      "learning_rate": 0.001,
+      "loss": 1.762,
+      "step": 2690
+    },
+    {
+      "epoch": 0.08609968430115757,
+      "grad_norm": 0.4147797226905823,
+      "learning_rate": 0.001,
+      "loss": 1.758,
+      "step": 2700
+    },
+    {
+      "epoch": 0.08641857202079148,
+      "grad_norm": 0.4185612201690674,
+      "learning_rate": 0.001,
+      "loss": 1.7508,
+      "step": 2710
+    },
+    {
+      "epoch": 0.08673745974042539,
+      "grad_norm": 0.4215739965438843,
+      "learning_rate": 0.001,
+      "loss": 1.7676,
+      "step": 2720
+    },
+    {
+      "epoch": 0.08705634746005932,
+      "grad_norm": 0.4120407998561859,
+      "learning_rate": 0.001,
+      "loss": 1.7702,
+      "step": 2730
+    },
+    {
+      "epoch": 0.08737523517969323,
+      "grad_norm": 0.4168885350227356,
+      "learning_rate": 0.001,
+      "loss": 1.7794,
+      "step": 2740
+    },
+    {
+      "epoch": 0.08769412289932715,
+      "grad_norm": 0.420948326587677,
+      "learning_rate": 0.001,
+      "loss": 1.7321,
+      "step": 2750
+    },
+    {
+      "epoch": 0.08801301061896107,
+      "grad_norm": 0.4142162501811981,
+      "learning_rate": 0.001,
+      "loss": 1.7386,
+      "step": 2760
+    },
+    {
+      "epoch": 0.08833189833859498,
+      "grad_norm": 0.4113253951072693,
+      "learning_rate": 0.001,
+      "loss": 1.7362,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0886507860582289,
+      "grad_norm": 0.41032281517982483,
+      "learning_rate": 0.001,
+      "loss": 1.76,
+      "step": 2780
+    },
+    {
+      "epoch": 0.08896967377786281,
+      "grad_norm": 0.3968314230442047,
+      "learning_rate": 0.001,
+      "loss": 1.7167,
+      "step": 2790
+    },
+    {
+      "epoch": 0.08928856149749673,
+      "grad_norm": 0.4045749306678772,
+      "learning_rate": 0.001,
+      "loss": 1.7521,
+      "step": 2800
+    },
+    {
+      "epoch": 0.08960744921713065,
+      "grad_norm": 0.38197824358940125,
+      "learning_rate": 0.001,
+      "loss": 1.71,
+      "step": 2810
+    },
+    {
+      "epoch": 0.08992633693676456,
+      "grad_norm": 0.40112364292144775,
+      "learning_rate": 0.001,
+      "loss": 1.7327,
+      "step": 2820
+    },
+    {
+      "epoch": 0.09024522465639848,
+      "grad_norm": 0.40666356682777405,
+      "learning_rate": 0.001,
+      "loss": 1.7358,
+      "step": 2830
+    },
+    {
+      "epoch": 0.0905641123760324,
+      "grad_norm": 0.40545156598091125,
+      "learning_rate": 0.001,
+      "loss": 1.7315,
+      "step": 2840
+    },
+    {
+      "epoch": 0.09088300009566631,
+      "grad_norm": 0.4269545376300812,
+      "learning_rate": 0.001,
+      "loss": 1.7073,
+      "step": 2850
+    },
+    {
+      "epoch": 0.09120188781530024,
+      "grad_norm": 0.4080691635608673,
+      "learning_rate": 0.001,
+      "loss": 1.7141,
+      "step": 2860
+    },
+    {
+      "epoch": 0.09152077553493415,
+      "grad_norm": 0.41239744424819946,
+      "learning_rate": 0.001,
+      "loss": 1.7126,
+      "step": 2870
+    },
+    {
+      "epoch": 0.09183966325456806,
+      "grad_norm": 0.3845699727535248,
+      "learning_rate": 0.001,
+      "loss": 1.7263,
+      "step": 2880
+    },
+    {
+      "epoch": 0.09215855097420199,
+      "grad_norm": 0.38979536294937134,
+      "learning_rate": 0.001,
+      "loss": 1.7173,
+      "step": 2890
+    },
+    {
+      "epoch": 0.0924774386938359,
+      "grad_norm": 0.3904477655887604,
+      "learning_rate": 0.001,
+      "loss": 1.7167,
+      "step": 2900
+    },
+    {
+      "epoch": 0.09279632641346981,
+      "grad_norm": 0.4068198502063751,
+      "learning_rate": 0.001,
+      "loss": 1.723,
+      "step": 2910
+    },
+    {
+      "epoch": 0.09311521413310374,
+      "grad_norm": 0.37980249524116516,
+      "learning_rate": 0.001,
+      "loss": 1.724,
+      "step": 2920
+    },
+    {
+      "epoch": 0.09343410185273765,
+      "grad_norm": 0.3883666396141052,
+      "learning_rate": 0.001,
+      "loss": 1.7292,
+      "step": 2930
+    },
+    {
+      "epoch": 0.09375298957237156,
+      "grad_norm": 0.38029372692108154,
+      "learning_rate": 0.001,
+      "loss": 1.7312,
+      "step": 2940
+    },
+    {
+      "epoch": 0.09407187729200549,
+      "grad_norm": 0.40660858154296875,
+      "learning_rate": 0.001,
+      "loss": 1.7042,
+      "step": 2950
+    },
+    {
+      "epoch": 0.0943907650116394,
+      "grad_norm": 0.40074729919433594,
+      "learning_rate": 0.001,
+      "loss": 1.6959,
+      "step": 2960
+    },
+    {
+      "epoch": 0.09470965273127332,
+      "grad_norm": 0.38984328508377075,
+      "learning_rate": 0.001,
+      "loss": 1.7042,
+      "step": 2970
+    },
+    {
+      "epoch": 0.09502854045090724,
+      "grad_norm": 0.36554253101348877,
+      "learning_rate": 0.001,
+      "loss": 1.6994,
+      "step": 2980
+    },
+    {
+      "epoch": 0.09534742817054115,
+      "grad_norm": 0.3839636445045471,
+      "learning_rate": 0.001,
+      "loss": 1.7045,
+      "step": 2990
+    },
+    {
+      "epoch": 0.09566631589017507,
+      "grad_norm": 0.3920130729675293,
+      "learning_rate": 0.001,
+      "loss": 1.7151,
+      "step": 3000
+    },
+    {
+      "epoch": 0.09598520360980899,
+      "grad_norm": 0.37574824690818787,
+      "learning_rate": 0.001,
+      "loss": 1.6945,
+      "step": 3010
+    },
+    {
+      "epoch": 0.0963040913294429,
+      "grad_norm": 0.37784722447395325,
+      "learning_rate": 0.001,
+      "loss": 1.6825,
+      "step": 3020
+    },
+    {
+      "epoch": 0.09662297904907682,
+      "grad_norm": 0.37999963760375977,
+      "learning_rate": 0.001,
+      "loss": 1.6928,
+      "step": 3030
+    },
+    {
+      "epoch": 0.09694186676871074,
+      "grad_norm": 0.3637866675853729,
+      "learning_rate": 0.001,
+      "loss": 1.6893,
+      "step": 3040
+    },
+    {
+      "epoch": 0.09726075448834465,
+      "grad_norm": 0.37975192070007324,
+      "learning_rate": 0.001,
+      "loss": 1.7077,
+      "step": 3050
+    },
+    {
+      "epoch": 0.09757964220797857,
+      "grad_norm": 0.3667464852333069,
+      "learning_rate": 0.001,
+      "loss": 1.6983,
+      "step": 3060
+    },
+    {
+      "epoch": 0.09789852992761248,
+      "grad_norm": 0.37685081362724304,
+      "learning_rate": 0.001,
+      "loss": 1.676,
+      "step": 3070
+    },
+    {
+      "epoch": 0.09821741764724641,
+      "grad_norm": 0.35984039306640625,
+      "learning_rate": 0.001,
+      "loss": 1.6843,
+      "step": 3080
+    },
+    {
+      "epoch": 0.09853630536688032,
+      "grad_norm": 0.3787494897842407,
+      "learning_rate": 0.001,
+      "loss": 1.6745,
+      "step": 3090
+    },
+    {
+      "epoch": 0.09885519308651423,
+      "grad_norm": 0.3618012070655823,
+      "learning_rate": 0.001,
+      "loss": 1.666,
+      "step": 3100
+    },
+    {
+      "epoch": 0.09917408080614816,
+      "grad_norm": 0.37266436219215393,
+      "learning_rate": 0.001,
+      "loss": 1.6618,
+      "step": 3110
+    },
+    {
+      "epoch": 0.09949296852578207,
+      "grad_norm": 0.36171668767929077,
+      "learning_rate": 0.001,
+      "loss": 1.6866,
+      "step": 3120
+    },
+    {
+      "epoch": 0.09981185624541598,
+      "grad_norm": 0.3792877495288849,
+      "learning_rate": 0.001,
+      "loss": 1.6932,
+      "step": 3130
+    },
+    {
+      "epoch": 0.10013074396504991,
+      "grad_norm": 0.3717607259750366,
+      "learning_rate": 0.001,
+      "loss": 1.6773,
+      "step": 3140
+    },
+    {
+      "epoch": 0.10044963168468382,
+      "grad_norm": 0.3575504422187805,
+      "learning_rate": 0.001,
+      "loss": 1.6799,
+      "step": 3150
+    },
+    {
+      "epoch": 0.10076851940431773,
+      "grad_norm": 0.37053510546684265,
+      "learning_rate": 0.001,
+      "loss": 1.6593,
+      "step": 3160
+    },
+    {
+      "epoch": 0.10108740712395166,
+      "grad_norm": 0.35849571228027344,
+      "learning_rate": 0.001,
+      "loss": 1.6635,
+      "step": 3170
+    },
+    {
+      "epoch": 0.10140629484358557,
+      "grad_norm": 0.3712310194969177,
+      "learning_rate": 0.001,
+      "loss": 1.6642,
+      "step": 3180
+    },
+    {
+      "epoch": 0.1017251825632195,
+      "grad_norm": 0.36214131116867065,
+      "learning_rate": 0.001,
+      "loss": 1.6422,
+      "step": 3190
+    },
+    {
+      "epoch": 0.10204407028285341,
+      "grad_norm": 0.3629904091358185,
+      "learning_rate": 0.001,
+      "loss": 1.6492,
+      "step": 3200
+    },
+    {
+      "epoch": 0.10236295800248732,
+      "grad_norm": 0.34454378485679626,
+      "learning_rate": 0.001,
+      "loss": 1.68,
+      "step": 3210
+    },
+    {
+      "epoch": 0.10268184572212125,
+      "grad_norm": 0.356357604265213,
+      "learning_rate": 0.001,
+      "loss": 1.6717,
+      "step": 3220
+    },
+    {
+      "epoch": 0.10300073344175516,
+      "grad_norm": 0.35266950726509094,
+      "learning_rate": 0.001,
+      "loss": 1.6358,
+      "step": 3230
+    },
+    {
+      "epoch": 0.10331962116138907,
+      "grad_norm": 0.3726760149002075,
+      "learning_rate": 0.001,
+      "loss": 1.6207,
+      "step": 3240
+    },
+    {
+      "epoch": 0.103638508881023,
+      "grad_norm": 0.35605040192604065,
+      "learning_rate": 0.001,
+      "loss": 1.6467,
+      "step": 3250
+    },
+    {
+      "epoch": 0.1039573966006569,
+      "grad_norm": 0.36946994066238403,
+      "learning_rate": 0.001,
+      "loss": 1.663,
+      "step": 3260
+    },
+    {
+      "epoch": 0.10427628432029083,
+      "grad_norm": 0.3503727614879608,
+      "learning_rate": 0.001,
+      "loss": 1.6157,
+      "step": 3270
+    },
+    {
+      "epoch": 0.10459517203992474,
+      "grad_norm": 0.3665166199207306,
+      "learning_rate": 0.001,
+      "loss": 1.6582,
+      "step": 3280
+    },
+    {
+      "epoch": 0.10491405975955866,
+      "grad_norm": 0.3531331717967987,
+      "learning_rate": 0.001,
+      "loss": 1.6184,
+      "step": 3290
+    },
+    {
+      "epoch": 0.10523294747919258,
+      "grad_norm": 0.3589783012866974,
+      "learning_rate": 0.001,
+      "loss": 1.6514,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1055518351988265,
+      "grad_norm": 0.3521636724472046,
+      "learning_rate": 0.001,
+      "loss": 1.6394,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1058707229184604,
+      "grad_norm": 0.3591137230396271,
+      "learning_rate": 0.001,
+      "loss": 1.6222,
+      "step": 3320
+    },
+    {
+      "epoch": 0.10618961063809433,
+      "grad_norm": 0.34527188539505005,
+      "learning_rate": 0.001,
+      "loss": 1.6849,
+      "step": 3330
+    },
+    {
+      "epoch": 0.10650849835772824,
+      "grad_norm": 0.33738335967063904,
+      "learning_rate": 0.001,
+      "loss": 1.6509,
+      "step": 3340
+    },
+    {
+      "epoch": 0.10682738607736215,
+      "grad_norm": 0.342528373003006,
+      "learning_rate": 0.001,
+      "loss": 1.6568,
+      "step": 3350
+    },
+    {
+      "epoch": 0.10714627379699608,
+      "grad_norm": 0.3343750536441803,
+      "learning_rate": 0.001,
+      "loss": 1.6315,
+      "step": 3360
+    },
+    {
+      "epoch": 0.10746516151662999,
+      "grad_norm": 0.33615440130233765,
+      "learning_rate": 0.001,
+      "loss": 1.6399,
+      "step": 3370
+    },
+    {
+      "epoch": 0.10778404923626392,
+      "grad_norm": 0.3438841700553894,
+      "learning_rate": 0.001,
+      "loss": 1.6302,
+      "step": 3380
+    },
+    {
+      "epoch": 0.10810293695589783,
+      "grad_norm": 0.3580395579338074,
+      "learning_rate": 0.001,
+      "loss": 1.5996,
+      "step": 3390
+    },
+    {
+      "epoch": 0.10842182467553174,
+      "grad_norm": 0.34119513630867004,
+      "learning_rate": 0.001,
+      "loss": 1.6179,
+      "step": 3400
+    },
+    {
+      "epoch": 0.10874071239516567,
+      "grad_norm": 0.33958378434181213,
+      "learning_rate": 0.001,
+      "loss": 1.6254,
+      "step": 3410
+    },
+    {
+      "epoch": 0.10905960011479958,
+      "grad_norm": 0.3339976668357849,
+      "learning_rate": 0.001,
+      "loss": 1.6334,
+      "step": 3420
+    },
+    {
+      "epoch": 0.10937848783443349,
+      "grad_norm": 0.3434707522392273,
+      "learning_rate": 0.001,
+      "loss": 1.6307,
+      "step": 3430
+    },
+    {
+      "epoch": 0.10969737555406742,
+      "grad_norm": 0.33472973108291626,
+      "learning_rate": 0.001,
+      "loss": 1.6306,
+      "step": 3440
+    },
+    {
+      "epoch": 0.11001626327370133,
+      "grad_norm": 0.3320460915565491,
+      "learning_rate": 0.001,
+      "loss": 1.6366,
+      "step": 3450
+    },
+    {
+      "epoch": 0.11033515099333524,
+      "grad_norm": 0.3379921317100525,
+      "learning_rate": 0.001,
+      "loss": 1.5962,
+      "step": 3460
+    },
+    {
+      "epoch": 0.11065403871296917,
+      "grad_norm": 0.3337703049182892,
+      "learning_rate": 0.001,
+      "loss": 1.624,
+      "step": 3470
+    },
+    {
+      "epoch": 0.11097292643260308,
+      "grad_norm": 0.3293701410293579,
+      "learning_rate": 0.001,
+      "loss": 1.6056,
+      "step": 3480
+    },
+    {
+      "epoch": 0.111291814152237,
+      "grad_norm": 0.3398142158985138,
+      "learning_rate": 0.001,
+      "loss": 1.593,
+      "step": 3490
+    },
+    {
+      "epoch": 0.11161070187187092,
+      "grad_norm": 0.33788856863975525,
+      "learning_rate": 0.001,
+      "loss": 1.6093,
+      "step": 3500
+    },
+    {
+      "epoch": 0.11192958959150483,
+      "grad_norm": 0.3267028033733368,
+      "learning_rate": 0.001,
+      "loss": 1.6069,
+      "step": 3510
+    },
+    {
+      "epoch": 0.11224847731113875,
+      "grad_norm": 0.32580670714378357,
+      "learning_rate": 0.001,
+      "loss": 1.5999,
+      "step": 3520
+    },
+    {
+      "epoch": 0.11256736503077266,
+      "grad_norm": 0.31678780913352966,
+      "learning_rate": 0.001,
+      "loss": 1.5919,
+      "step": 3530
+    },
+    {
+      "epoch": 0.11288625275040658,
+      "grad_norm": 0.31694918870925903,
+      "learning_rate": 0.001,
+      "loss": 1.6112,
+      "step": 3540
+    },
+    {
+      "epoch": 0.1132051404700405,
+      "grad_norm": 0.31398412585258484,
+      "learning_rate": 0.001,
+      "loss": 1.6043,
+      "step": 3550
+    },
+    {
+      "epoch": 0.11352402818967441,
+      "grad_norm": 0.3309810757637024,
+      "learning_rate": 0.001,
+      "loss": 1.5972,
+      "step": 3560
+    },
+    {
+      "epoch": 0.11384291590930833,
+      "grad_norm": 0.33123892545700073,
+      "learning_rate": 0.001,
+      "loss": 1.5764,
+      "step": 3570
+    },
+    {
+      "epoch": 0.11416180362894225,
+      "grad_norm": 0.3245631158351898,
+      "learning_rate": 0.001,
+      "loss": 1.5751,
+      "step": 3580
+    },
+    {
+      "epoch": 0.11448069134857616,
+      "grad_norm": 0.32239916920661926,
+      "learning_rate": 0.001,
+      "loss": 1.6094,
+      "step": 3590
+    },
+    {
+      "epoch": 0.11479957906821009,
+      "grad_norm": 0.3406887948513031,
+      "learning_rate": 0.001,
+      "loss": 1.6002,
+      "step": 3600
+    },
+    {
+      "epoch": 0.115118466787844,
+      "grad_norm": 0.3163871467113495,
+      "learning_rate": 0.001,
+      "loss": 1.5643,
+      "step": 3610
+    },
+    {
+      "epoch": 0.11543735450747791,
+      "grad_norm": 0.3176119327545166,
+      "learning_rate": 0.001,
+      "loss": 1.5611,
+      "step": 3620
+    },
+    {
+      "epoch": 0.11575624222711184,
+      "grad_norm": 0.32567960023880005,
+      "learning_rate": 0.001,
+      "loss": 1.5995,
+      "step": 3630
+    },
+    {
+      "epoch": 0.11607512994674575,
+      "grad_norm": 0.3347359895706177,
+      "learning_rate": 0.001,
+      "loss": 1.6058,
+      "step": 3640
+    },
+    {
+      "epoch": 0.11639401766637966,
+      "grad_norm": 0.3217657506465912,
+      "learning_rate": 0.001,
+      "loss": 1.5745,
+      "step": 3650
+    },
+    {
+      "epoch": 0.11671290538601359,
+      "grad_norm": 0.31946665048599243,
+      "learning_rate": 0.001,
+      "loss": 1.559,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1170317931056475,
+      "grad_norm": 0.3089522421360016,
+      "learning_rate": 0.001,
+      "loss": 1.5884,
+      "step": 3670
+    },
+    {
+      "epoch": 0.11735068082528141,
+      "grad_norm": 0.31650635600090027,
+      "learning_rate": 0.001,
+      "loss": 1.6132,
+      "step": 3680
+    },
+    {
+      "epoch": 0.11766956854491534,
+      "grad_norm": 0.3170093595981598,
+      "learning_rate": 0.001,
+      "loss": 1.5718,
+      "step": 3690
+    },
+    {
+      "epoch": 0.11798845626454925,
+      "grad_norm": 0.30170735716819763,
+      "learning_rate": 0.001,
+      "loss": 1.5786,
+      "step": 3700
+    },
+    {
+      "epoch": 0.11830734398418317,
+      "grad_norm": 0.3147510290145874,
+      "learning_rate": 0.001,
+      "loss": 1.5805,
+      "step": 3710
+    },
+    {
+      "epoch": 0.11862623170381709,
+      "grad_norm": 0.30160295963287354,
+      "learning_rate": 0.001,
+      "loss": 1.5567,
+      "step": 3720
+    },
+    {
+      "epoch": 0.118945119423451,
+      "grad_norm": 0.32837751507759094,
+      "learning_rate": 0.001,
+      "loss": 1.5771,
+      "step": 3730
+    },
+    {
+      "epoch": 0.11926400714308492,
+      "grad_norm": 0.31761300563812256,
+      "learning_rate": 0.001,
+      "loss": 1.5816,
+      "step": 3740
+    },
+    {
+      "epoch": 0.11958289486271884,
+      "grad_norm": 0.3200889527797699,
+      "learning_rate": 0.001,
+      "loss": 1.5536,
+      "step": 3750
+    },
+    {
+      "epoch": 0.11990178258235275,
+      "grad_norm": 0.30934789776802063,
+      "learning_rate": 0.001,
+      "loss": 1.5577,
+      "step": 3760
+    },
+    {
+      "epoch": 0.12022067030198667,
+      "grad_norm": 0.30753806233406067,
+      "learning_rate": 0.001,
+      "loss": 1.5784,
+      "step": 3770
+    },
+    {
+      "epoch": 0.12053955802162059,
+      "grad_norm": 0.3071571886539459,
+      "learning_rate": 0.001,
+      "loss": 1.5894,
+      "step": 3780
+    },
+    {
+      "epoch": 0.12085844574125451,
+      "grad_norm": 0.31278368830680847,
+      "learning_rate": 0.001,
+      "loss": 1.5458,
+      "step": 3790
+    },
+    {
+      "epoch": 0.12117733346088842,
+      "grad_norm": 0.3064911365509033,
+      "learning_rate": 0.001,
+      "loss": 1.5434,
+      "step": 3800
+    },
+    {
+      "epoch": 0.12149622118052233,
+      "grad_norm": 0.3021445572376251,
+      "learning_rate": 0.001,
+      "loss": 1.5439,
+      "step": 3810
+    },
+    {
+      "epoch": 0.12181510890015626,
+      "grad_norm": 0.30052146315574646,
+      "learning_rate": 0.001,
+      "loss": 1.5531,
+      "step": 3820
+    },
+    {
+      "epoch": 0.12213399661979017,
+      "grad_norm": 0.3149283826351166,
+      "learning_rate": 0.001,
+      "loss": 1.5463,
+      "step": 3830
+    },
+    {
+      "epoch": 0.12245288433942408,
+      "grad_norm": 0.30645984411239624,
+      "learning_rate": 0.001,
+      "loss": 1.561,
+      "step": 3840
+    },
+    {
+      "epoch": 0.12277177205905801,
+      "grad_norm": 0.3132129907608032,
+      "learning_rate": 0.001,
+      "loss": 1.5622,
+      "step": 3850
+    },
+    {
+      "epoch": 0.12309065977869192,
+      "grad_norm": 0.31083235144615173,
+      "learning_rate": 0.001,
+      "loss": 1.5573,
+      "step": 3860
+    },
+    {
+      "epoch": 0.12340954749832583,
+      "grad_norm": 0.3180761933326721,
+      "learning_rate": 0.001,
+      "loss": 1.5747,
+      "step": 3870
+    },
+    {
+      "epoch": 0.12372843521795976,
+      "grad_norm": 0.3081989288330078,
+      "learning_rate": 0.001,
+      "loss": 1.536,
+      "step": 3880
+    },
+    {
+      "epoch": 0.12404732293759367,
+      "grad_norm": 0.30181053280830383,
+      "learning_rate": 0.001,
+      "loss": 1.5566,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1243662106572276,
+      "grad_norm": 0.3038397431373596,
+      "learning_rate": 0.001,
+      "loss": 1.5501,
+      "step": 3900
+    },
+    {
+      "epoch": 0.12468509837686151,
+      "grad_norm": 0.2994053363800049,
+      "learning_rate": 0.001,
+      "loss": 1.5416,
+      "step": 3910
+    },
+    {
+      "epoch": 0.12500398609649543,
+      "grad_norm": 0.29628464579582214,
+      "learning_rate": 0.001,
+      "loss": 1.5511,
+      "step": 3920
+    },
+    {
+      "epoch": 0.12532287381612933,
+      "grad_norm": 0.2959844768047333,
+      "learning_rate": 0.001,
+      "loss": 1.5207,
+      "step": 3930
+    },
+    {
+      "epoch": 0.12564176153576326,
+      "grad_norm": 0.294956773519516,
+      "learning_rate": 0.001,
+      "loss": 1.5164,
+      "step": 3940
+    },
+    {
+      "epoch": 0.12596064925539718,
+      "grad_norm": 0.28598493337631226,
+      "learning_rate": 0.001,
+      "loss": 1.5311,
+      "step": 3950
+    },
+    {
+      "epoch": 0.12627953697503108,
+      "grad_norm": 0.2978171706199646,
+      "learning_rate": 0.001,
+      "loss": 1.5241,
+      "step": 3960
+    },
+    {
+      "epoch": 0.126598424694665,
+      "grad_norm": 0.2851006090641022,
+      "learning_rate": 0.001,
+      "loss": 1.5414,
+      "step": 3970
+    },
+    {
+      "epoch": 0.12691731241429893,
+      "grad_norm": 0.2982429265975952,
+      "learning_rate": 0.001,
+      "loss": 1.5179,
+      "step": 3980
+    },
+    {
+      "epoch": 0.12723620013393283,
+      "grad_norm": 0.28805026412010193,
+      "learning_rate": 0.001,
+      "loss": 1.5252,
+      "step": 3990
+    },
+    {
+      "epoch": 0.12755508785356676,
+      "grad_norm": 0.3060665726661682,
+      "learning_rate": 0.001,
+      "loss": 1.5211,
+      "step": 4000
+    },
+    {
+      "epoch": 0.12787397557320068,
+      "grad_norm": 0.2944425642490387,
+      "learning_rate": 0.001,
+      "loss": 1.5364,
+      "step": 4010
+    },
+    {
+      "epoch": 0.12819286329283458,
+      "grad_norm": 0.29006776213645935,
+      "learning_rate": 0.001,
+      "loss": 1.5241,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1285117510124685,
+      "grad_norm": 0.29584676027297974,
+      "learning_rate": 0.001,
+      "loss": 1.5295,
+      "step": 4030
+    },
+    {
+      "epoch": 0.12883063873210243,
+      "grad_norm": 0.28195831179618835,
+      "learning_rate": 0.001,
+      "loss": 1.525,
+      "step": 4040
+    },
+    {
+      "epoch": 0.12914952645173636,
+      "grad_norm": 0.2912329137325287,
+      "learning_rate": 0.001,
+      "loss": 1.5306,
+      "step": 4050
+    },
+    {
+      "epoch": 0.12946841417137026,
+      "grad_norm": 0.28421998023986816,
+      "learning_rate": 0.001,
+      "loss": 1.5337,
+      "step": 4060
+    },
+    {
+      "epoch": 0.12978730189100418,
+      "grad_norm": 0.2915787994861603,
+      "learning_rate": 0.001,
+      "loss": 1.5267,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1301061896106381,
+      "grad_norm": 0.28192031383514404,
+      "learning_rate": 0.001,
+      "loss": 1.5003,
+      "step": 4080
+    },
+    {
+      "epoch": 0.130425077330272,
+      "grad_norm": 0.28582581877708435,
+      "learning_rate": 0.001,
+      "loss": 1.518,
+      "step": 4090
+    },
+    {
+      "epoch": 0.13074396504990593,
+      "grad_norm": 0.28264012932777405,
+      "learning_rate": 0.001,
+      "loss": 1.5238,
+      "step": 4100
+    },
+    {
+      "epoch": 0.13106285276953986,
+      "grad_norm": 0.2747161388397217,
+      "learning_rate": 0.001,
+      "loss": 1.5093,
+      "step": 4110
+    },
+    {
+      "epoch": 0.13138174048917375,
+      "grad_norm": 0.2909848690032959,
+      "learning_rate": 0.001,
+      "loss": 1.4931,
+      "step": 4120
+    },
+    {
+      "epoch": 0.13170062820880768,
+      "grad_norm": 0.2762302756309509,
+      "learning_rate": 0.001,
+      "loss": 1.5131,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1320195159284416,
+      "grad_norm": 0.29010942578315735,
+      "learning_rate": 0.001,
+      "loss": 1.5016,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1323384036480755,
+      "grad_norm": 0.28639352321624756,
+      "learning_rate": 0.001,
+      "loss": 1.4924,
+      "step": 4150
+    },
+    {
+      "epoch": 0.13265729136770943,
+      "grad_norm": 0.27551382780075073,
+      "learning_rate": 0.001,
+      "loss": 1.4986,
+      "step": 4160
+    },
+    {
+      "epoch": 0.13297617908734335,
+      "grad_norm": 0.2695043087005615,
+      "learning_rate": 0.001,
+      "loss": 1.5002,
+      "step": 4170
+    },
+    {
+      "epoch": 0.13329506680697725,
+      "grad_norm": 0.287056028842926,
+      "learning_rate": 0.001,
+      "loss": 1.4837,
+      "step": 4180
+    },
+    {
+      "epoch": 0.13361395452661118,
+      "grad_norm": 0.27863118052482605,
+      "learning_rate": 0.001,
+      "loss": 1.4895,
+      "step": 4190
+    },
+    {
+      "epoch": 0.1339328422462451,
+      "grad_norm": 0.2812846004962921,
+      "learning_rate": 0.001,
+      "loss": 1.4992,
+      "step": 4200
+    },
+    {
+      "epoch": 0.134251729965879,
+      "grad_norm": 0.27025651931762695,
+      "learning_rate": 0.001,
+      "loss": 1.4919,
+      "step": 4210
+    },
+    {
+      "epoch": 0.13457061768551293,
+      "grad_norm": 0.26951390504837036,
+      "learning_rate": 0.001,
+      "loss": 1.4928,
+      "step": 4220
+    },
+    {
+      "epoch": 0.13488950540514685,
+      "grad_norm": 0.27416202425956726,
+      "learning_rate": 0.001,
+      "loss": 1.4852,
+      "step": 4230
+    },
+    {
+      "epoch": 0.13520839312478075,
+      "grad_norm": 0.27375853061676025,
+      "learning_rate": 0.001,
+      "loss": 1.5213,
+      "step": 4240
+    },
+    {
+      "epoch": 0.13552728084441468,
+      "grad_norm": 0.26884832978248596,
+      "learning_rate": 0.001,
+      "loss": 1.4888,
+      "step": 4250
+    },
+    {
+      "epoch": 0.1358461685640486,
+      "grad_norm": 0.27149176597595215,
+      "learning_rate": 0.001,
+      "loss": 1.5023,
+      "step": 4260
+    },
+    {
+      "epoch": 0.13616505628368253,
+      "grad_norm": 0.27007535099983215,
+      "learning_rate": 0.001,
+      "loss": 1.4405,
+      "step": 4270
+    },
+    {
+      "epoch": 0.13648394400331643,
+      "grad_norm": 0.26930558681488037,
+      "learning_rate": 0.001,
+      "loss": 1.4861,
+      "step": 4280
+    },
+    {
+      "epoch": 0.13680283172295035,
+      "grad_norm": 0.2684692144393921,
+      "learning_rate": 0.001,
+      "loss": 1.4629,
+      "step": 4290
+    },
+    {
+      "epoch": 0.13712171944258428,
+      "grad_norm": 0.2631552815437317,
+      "learning_rate": 0.001,
+      "loss": 1.4726,
+      "step": 4300
+    },
+    {
+      "epoch": 0.13744060716221818,
+      "grad_norm": 0.2739587426185608,
+      "learning_rate": 0.001,
+      "loss": 1.488,
+      "step": 4310
+    },
+    {
+      "epoch": 0.1377594948818521,
+      "grad_norm": 0.2854085862636566,
+      "learning_rate": 0.001,
+      "loss": 1.4637,
+      "step": 4320
+    },
+    {
+      "epoch": 0.13807838260148603,
+      "grad_norm": 0.2662985622882843,
+      "learning_rate": 0.001,
+      "loss": 1.4772,
+      "step": 4330
+    },
+    {
+      "epoch": 0.13839727032111993,
+      "grad_norm": 0.27411964535713196,
+      "learning_rate": 0.001,
+      "loss": 1.4772,
+      "step": 4340
+    },
+    {
+      "epoch": 0.13871615804075385,
+      "grad_norm": 0.2799747884273529,
+      "learning_rate": 0.001,
+      "loss": 1.4814,
+      "step": 4350
+    },
+    {
+      "epoch": 0.13903504576038778,
+      "grad_norm": 0.266664057970047,
+      "learning_rate": 0.001,
+      "loss": 1.4773,
+      "step": 4360
+    },
+    {
+      "epoch": 0.13935393348002167,
+      "grad_norm": 0.26244714856147766,
+      "learning_rate": 0.001,
+      "loss": 1.4922,
+      "step": 4370
+    },
+    {
+      "epoch": 0.1396728211996556,
+      "grad_norm": 0.268928200006485,
+      "learning_rate": 0.001,
+      "loss": 1.4656,
+      "step": 4380
+    },
+    {
+      "epoch": 0.13999170891928953,
+      "grad_norm": 0.26305675506591797,
+      "learning_rate": 0.001,
+      "loss": 1.4728,
+      "step": 4390
+    },
+    {
+      "epoch": 0.14031059663892342,
+      "grad_norm": 0.26536253094673157,
+      "learning_rate": 0.001,
+      "loss": 1.4829,
+      "step": 4400
+    },
+    {
+      "epoch": 0.14062948435855735,
+      "grad_norm": 0.2680259048938751,
+      "learning_rate": 0.001,
+      "loss": 1.4879,
+      "step": 4410
+    },
+    {
+      "epoch": 0.14094837207819128,
+      "grad_norm": 0.2655434012413025,
+      "learning_rate": 0.001,
+      "loss": 1.4493,
+      "step": 4420
+    },
+    {
+      "epoch": 0.14126725979782517,
+      "grad_norm": 0.2688651978969574,
+      "learning_rate": 0.001,
+      "loss": 1.4648,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1415861475174591,
+      "grad_norm": 0.2561871409416199,
+      "learning_rate": 0.001,
+      "loss": 1.4479,
+      "step": 4440
+    },
+    {
+      "epoch": 0.14190503523709302,
+      "grad_norm": 0.2676304876804352,
+      "learning_rate": 0.001,
+      "loss": 1.4353,
+      "step": 4450
+    },
+    {
+      "epoch": 0.14222392295672695,
+      "grad_norm": 0.26640716195106506,
+      "learning_rate": 0.001,
+      "loss": 1.4368,
+      "step": 4460
+    },
+    {
+      "epoch": 0.14254281067636085,
+      "grad_norm": 0.26717209815979004,
+      "learning_rate": 0.001,
+      "loss": 1.4525,
+      "step": 4470
+    },
+    {
+      "epoch": 0.14286169839599477,
+      "grad_norm": 0.253572016954422,
+      "learning_rate": 0.001,
+      "loss": 1.4607,
+      "step": 4480
+    },
+    {
+      "epoch": 0.1431805861156287,
+      "grad_norm": 0.2572033107280731,
+      "learning_rate": 0.001,
+      "loss": 1.4488,
+      "step": 4490
+    },
+    {
+      "epoch": 0.1434994738352626,
+      "grad_norm": 0.2625178098678589,
+      "learning_rate": 0.001,
+      "loss": 1.4813,
+      "step": 4500
+    },
+    {
+      "epoch": 0.14381836155489652,
+      "grad_norm": 0.26461777091026306,
+      "learning_rate": 0.001,
+      "loss": 1.4848,
+      "step": 4510
+    },
+    {
+      "epoch": 0.14413724927453045,
+      "grad_norm": 0.2646142840385437,
+      "learning_rate": 0.001,
+      "loss": 1.4555,
+      "step": 4520
+    },
+    {
+      "epoch": 0.14445613699416435,
+      "grad_norm": 0.2517695128917694,
+      "learning_rate": 0.001,
+      "loss": 1.4482,
+      "step": 4530
+    },
+    {
+      "epoch": 0.14477502471379827,
+      "grad_norm": 0.26152729988098145,
+      "learning_rate": 0.001,
+      "loss": 1.4518,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1450939124334322,
+      "grad_norm": 0.25882112979888916,
+      "learning_rate": 0.001,
+      "loss": 1.45,
+      "step": 4550
+    },
+    {
+      "epoch": 0.1454128001530661,
+      "grad_norm": 0.25552093982696533,
+      "learning_rate": 0.001,
+      "loss": 1.4729,
+      "step": 4560
+    },
+    {
+      "epoch": 0.14573168787270002,
+      "grad_norm": 0.25193673372268677,
+      "learning_rate": 0.001,
+      "loss": 1.4483,
+      "step": 4570
+    },
+    {
+      "epoch": 0.14605057559233395,
+      "grad_norm": 0.24570788443088531,
+      "learning_rate": 0.001,
+      "loss": 1.435,
+      "step": 4580
+    },
+    {
+      "epoch": 0.14636946331196785,
+      "grad_norm": 0.2533837854862213,
+      "learning_rate": 0.001,
+      "loss": 1.4343,
+      "step": 4590
+    },
+    {
+      "epoch": 0.14668835103160177,
+      "grad_norm": 0.25800344347953796,
+      "learning_rate": 0.001,
+      "loss": 1.4592,
+      "step": 4600
+    },
+    {
+      "epoch": 0.1470072387512357,
+      "grad_norm": 0.2562967836856842,
+      "learning_rate": 0.001,
+      "loss": 1.4477,
+      "step": 4610
+    },
+    {
+      "epoch": 0.1473261264708696,
+      "grad_norm": 0.2644691467285156,
+      "learning_rate": 0.001,
+      "loss": 1.452,
+      "step": 4620
+    },
+    {
+      "epoch": 0.14764501419050352,
+      "grad_norm": 0.2588248550891876,
+      "learning_rate": 0.001,
+      "loss": 1.4492,
+      "step": 4630
+    },
+    {
+      "epoch": 0.14796390191013745,
+      "grad_norm": 0.24891693890094757,
+      "learning_rate": 0.001,
+      "loss": 1.4514,
+      "step": 4640
+    },
+    {
+      "epoch": 0.14828278962977134,
+      "grad_norm": 0.2602750360965729,
+      "learning_rate": 0.001,
+      "loss": 1.4502,
+      "step": 4650
+    },
+    {
+      "epoch": 0.14860167734940527,
+      "grad_norm": 0.2690906226634979,
+      "learning_rate": 0.001,
+      "loss": 1.447,
+      "step": 4660
+    },
+    {
+      "epoch": 0.1489205650690392,
+      "grad_norm": 0.2503431439399719,
+      "learning_rate": 0.001,
+      "loss": 1.4279,
+      "step": 4670
+    },
+    {
+      "epoch": 0.14923945278867312,
+      "grad_norm": 0.2497878074645996,
+      "learning_rate": 0.001,
+      "loss": 1.4413,
+      "step": 4680
+    },
+    {
+      "epoch": 0.14955834050830702,
+      "grad_norm": 0.25570768117904663,
+      "learning_rate": 0.001,
+      "loss": 1.4402,
+      "step": 4690
+    },
+    {
+      "epoch": 0.14987722822794095,
+      "grad_norm": 0.2501462697982788,
+      "learning_rate": 0.001,
+      "loss": 1.4119,
+      "step": 4700
+    },
+    {
+      "epoch": 0.15019611594757487,
+      "grad_norm": 0.2574600577354431,
+      "learning_rate": 0.001,
+      "loss": 1.4297,
+      "step": 4710
+    },
+    {
+      "epoch": 0.15051500366720877,
+      "grad_norm": 0.26934248208999634,
+      "learning_rate": 0.001,
+      "loss": 1.4183,
+      "step": 4720
+    },
+    {
+      "epoch": 0.1508338913868427,
+      "grad_norm": 0.24917541444301605,
+      "learning_rate": 0.001,
+      "loss": 1.4301,
+      "step": 4730
+    },
+    {
+      "epoch": 0.15115277910647662,
+      "grad_norm": 0.25259947776794434,
+      "learning_rate": 0.001,
+      "loss": 1.4254,
+      "step": 4740
+    },
+    {
+      "epoch": 0.15147166682611052,
+      "grad_norm": 0.2534835636615753,
+      "learning_rate": 0.001,
+      "loss": 1.413,
+      "step": 4750
+    },
+    {
+      "epoch": 0.15179055454574444,
+      "grad_norm": 0.237448051571846,
+      "learning_rate": 0.001,
+      "loss": 1.4096,
+      "step": 4760
+    },
+    {
+      "epoch": 0.15210944226537837,
+      "grad_norm": 0.2515486478805542,
+      "learning_rate": 0.001,
+      "loss": 1.4226,
+      "step": 4770
+    },
+    {
+      "epoch": 0.15242832998501227,
+      "grad_norm": 0.24898935854434967,
+      "learning_rate": 0.001,
+      "loss": 1.4077,
+      "step": 4780
+    },
+    {
+      "epoch": 0.1527472177046462,
+      "grad_norm": 0.2587830424308777,
+      "learning_rate": 0.001,
+      "loss": 1.4354,
+      "step": 4790
+    },
+    {
+      "epoch": 0.15306610542428012,
+      "grad_norm": 0.2498152256011963,
+      "learning_rate": 0.001,
+      "loss": 1.4091,
+      "step": 4800
+    },
+    {
+      "epoch": 0.15338499314391402,
+      "grad_norm": 0.2435469776391983,
+      "learning_rate": 0.001,
+      "loss": 1.4354,
+      "step": 4810
+    },
+    {
+      "epoch": 0.15370388086354794,
+      "grad_norm": 0.2446700930595398,
+      "learning_rate": 0.001,
+      "loss": 1.4227,
+      "step": 4820
+    },
+    {
+      "epoch": 0.15402276858318187,
+      "grad_norm": 0.23855316638946533,
+      "learning_rate": 0.001,
+      "loss": 1.4085,
+      "step": 4830
+    },
+    {
+      "epoch": 0.15434165630281577,
+      "grad_norm": 0.23332339525222778,
+      "learning_rate": 0.001,
+      "loss": 1.425,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1546605440224497,
+      "grad_norm": 0.25121748447418213,
+      "learning_rate": 0.001,
+      "loss": 1.4188,
+      "step": 4850
+    },
+    {
+      "epoch": 0.15497943174208362,
+      "grad_norm": 0.23875190317630768,
+      "learning_rate": 0.001,
+      "loss": 1.4166,
+      "step": 4860
+    },
+    {
+      "epoch": 0.15529831946171752,
+      "grad_norm": 0.2450498342514038,
+      "learning_rate": 0.001,
+      "loss": 1.4002,
+      "step": 4870
+    },
+    {
+      "epoch": 0.15561720718135144,
+      "grad_norm": 0.24624621868133545,
+      "learning_rate": 0.001,
+      "loss": 1.3945,
+      "step": 4880
+    },
+    {
+      "epoch": 0.15593609490098537,
+      "grad_norm": 0.24817205965518951,
+      "learning_rate": 0.001,
+      "loss": 1.4173,
+      "step": 4890
+    },
+    {
+      "epoch": 0.1562549826206193,
+      "grad_norm": 0.2515687346458435,
+      "learning_rate": 0.001,
+      "loss": 1.4208,
+      "step": 4900
+    },
+    {
+      "epoch": 0.1565738703402532,
+      "grad_norm": 0.2485036849975586,
+      "learning_rate": 0.001,
+      "loss": 1.4264,
+      "step": 4910
+    },
+    {
+      "epoch": 0.15689275805988712,
+      "grad_norm": 0.24037910997867584,
+      "learning_rate": 0.001,
+      "loss": 1.411,
+      "step": 4920
+    },
+    {
+      "epoch": 0.15721164577952104,
+      "grad_norm": 0.2485659420490265,
+      "learning_rate": 0.001,
+      "loss": 1.4192,
+      "step": 4930
+    },
+    {
+      "epoch": 0.15753053349915494,
+      "grad_norm": 0.23820529878139496,
+      "learning_rate": 0.001,
+      "loss": 1.3934,
+      "step": 4940
+    },
+    {
+      "epoch": 0.15784942121878887,
+      "grad_norm": 0.2456134855747223,
+      "learning_rate": 0.001,
+      "loss": 1.3899,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1581683089384228,
+      "grad_norm": 0.24716608226299286,
+      "learning_rate": 0.001,
+      "loss": 1.4188,
+      "step": 4960
+    },
+    {
+      "epoch": 0.1584871966580567,
+      "grad_norm": 0.23966428637504578,
+      "learning_rate": 0.001,
+      "loss": 1.413,
+      "step": 4970
+    },
+    {
+      "epoch": 0.15880608437769062,
+      "grad_norm": 0.2461225539445877,
+      "learning_rate": 0.001,
+      "loss": 1.3964,
+      "step": 4980
+    },
+    {
+      "epoch": 0.15912497209732454,
+      "grad_norm": 0.2454925775527954,
+      "learning_rate": 0.001,
+      "loss": 1.3956,
+      "step": 4990
+    },
+    {
+      "epoch": 0.15944385981695844,
+      "grad_norm": 0.24597546458244324,
+      "learning_rate": 0.001,
+      "loss": 1.3761,
+      "step": 5000
+    },
+    {
+      "epoch": 0.15976274753659236,
+      "grad_norm": 0.24570181965827942,
+      "learning_rate": 0.001,
+      "loss": 1.4079,
+      "step": 5010
+    },
+    {
+      "epoch": 0.1600816352562263,
+      "grad_norm": 0.23773568868637085,
+      "learning_rate": 0.001,
+      "loss": 1.3889,
+      "step": 5020
+    },
+    {
+      "epoch": 0.1604005229758602,
+      "grad_norm": 0.2613687217235565,
+      "learning_rate": 0.001,
+      "loss": 1.428,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1607194106954941,
+      "grad_norm": 0.24893403053283691,
+      "learning_rate": 0.001,
+      "loss": 1.4064,
+      "step": 5040
+    },
+    {
+      "epoch": 0.16103829841512804,
+      "grad_norm": 0.24659830331802368,
+      "learning_rate": 0.001,
+      "loss": 1.3935,
+      "step": 5050
+    },
+    {
+      "epoch": 0.16135718613476194,
+      "grad_norm": 0.25039970874786377,
+      "learning_rate": 0.001,
+      "loss": 1.4005,
+      "step": 5060
+    },
+    {
+      "epoch": 0.16167607385439586,
+      "grad_norm": 0.23986603319644928,
+      "learning_rate": 0.001,
+      "loss": 1.3882,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1619949615740298,
+      "grad_norm": 0.2418118715286255,
+      "learning_rate": 0.001,
+      "loss": 1.3963,
+      "step": 5080
+    },
+    {
+      "epoch": 0.16231384929366371,
+      "grad_norm": 0.23808790743350983,
+      "learning_rate": 0.001,
+      "loss": 1.4192,
+      "step": 5090
+    },
+    {
+      "epoch": 0.1626327370132976,
+      "grad_norm": 0.2416272610425949,
+      "learning_rate": 0.001,
+      "loss": 1.4061,
+      "step": 5100
+    },
+    {
+      "epoch": 0.16295162473293154,
+      "grad_norm": 0.24399322271347046,
+      "learning_rate": 0.001,
+      "loss": 1.3958,
+      "step": 5110
+    },
+    {
+      "epoch": 0.16327051245256546,
+      "grad_norm": 0.23533989489078522,
+      "learning_rate": 0.001,
+      "loss": 1.3766,
+      "step": 5120
+    },
+    {
+      "epoch": 0.16358940017219936,
+      "grad_norm": 0.23567648231983185,
+      "learning_rate": 0.001,
+      "loss": 1.394,
+      "step": 5130
+    },
+    {
+      "epoch": 0.1639082878918333,
+      "grad_norm": 0.23541750013828278,
+      "learning_rate": 0.001,
+      "loss": 1.3877,
+      "step": 5140
+    },
+    {
+      "epoch": 0.1642271756114672,
+      "grad_norm": 0.2420421689748764,
+      "learning_rate": 0.001,
+      "loss": 1.3941,
+      "step": 5150
+    },
+    {
+      "epoch": 0.1645460633311011,
+      "grad_norm": 0.24234576523303986,
+      "learning_rate": 0.001,
+      "loss": 1.3893,
+      "step": 5160
+    },
+    {
+      "epoch": 0.16486495105073504,
+      "grad_norm": 0.23756305873394012,
+      "learning_rate": 0.001,
+      "loss": 1.3961,
+      "step": 5170
+    },
+    {
+      "epoch": 0.16518383877036896,
+      "grad_norm": 0.23691706359386444,
+      "learning_rate": 0.001,
+      "loss": 1.4036,
+      "step": 5180
+    },
+    {
+      "epoch": 0.16550272649000286,
+      "grad_norm": 0.2374681830406189,
+      "learning_rate": 0.001,
+      "loss": 1.4124,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1658216142096368,
+      "grad_norm": 0.2333029806613922,
+      "learning_rate": 0.001,
+      "loss": 1.373,
+      "step": 5200
+    },
+    {
+      "epoch": 0.1661405019292707,
+      "grad_norm": 0.23446527123451233,
+      "learning_rate": 0.001,
+      "loss": 1.3737,
+      "step": 5210
+    },
+    {
+      "epoch": 0.1664593896489046,
+      "grad_norm": 0.24269339442253113,
+      "learning_rate": 0.001,
+      "loss": 1.3741,
+      "step": 5220
+    },
+    {
+      "epoch": 0.16677827736853854,
+      "grad_norm": 0.25331172347068787,
+      "learning_rate": 0.001,
+      "loss": 1.3733,
+      "step": 5230
+    },
+    {
+      "epoch": 0.16709716508817246,
+      "grad_norm": 0.24028655886650085,
+      "learning_rate": 0.001,
+      "loss": 1.3879,
+      "step": 5240
+    },
+    {
+      "epoch": 0.16741605280780636,
+      "grad_norm": 0.2346467673778534,
+      "learning_rate": 0.001,
+      "loss": 1.3834,
+      "step": 5250
+    },
+    {
+      "epoch": 0.16773494052744028,
+      "grad_norm": 0.2357403039932251,
+      "learning_rate": 0.001,
+      "loss": 1.3681,
+      "step": 5260
+    },
+    {
+      "epoch": 0.1680538282470742,
+      "grad_norm": 0.2432994395494461,
+      "learning_rate": 0.001,
+      "loss": 1.3637,
+      "step": 5270
+    },
+    {
+      "epoch": 0.1683727159667081,
+      "grad_norm": 0.23171168565750122,
+      "learning_rate": 0.001,
+      "loss": 1.3785,
+      "step": 5280
+    },
+    {
+      "epoch": 0.16869160368634203,
+      "grad_norm": 0.2409990429878235,
+      "learning_rate": 0.001,
+      "loss": 1.3765,
+      "step": 5290
+    },
+    {
+      "epoch": 0.16901049140597596,
+      "grad_norm": 0.23982109129428864,
+      "learning_rate": 0.001,
+      "loss": 1.3738,
+      "step": 5300
+    },
+    {
+      "epoch": 0.16932937912560989,
+      "grad_norm": 0.23728245496749878,
+      "learning_rate": 0.001,
+      "loss": 1.3802,
+      "step": 5310
+    },
+    {
+      "epoch": 0.16964826684524378,
+      "grad_norm": 0.23985517024993896,
+      "learning_rate": 0.001,
+      "loss": 1.3708,
+      "step": 5320
+    },
+    {
+      "epoch": 0.1699671545648777,
+      "grad_norm": 0.2339295744895935,
+      "learning_rate": 0.001,
+      "loss": 1.3887,
+      "step": 5330
+    },
+    {
+      "epoch": 0.17028604228451164,
+      "grad_norm": 0.23314034938812256,
+      "learning_rate": 0.001,
+      "loss": 1.3959,
+      "step": 5340
+    },
+    {
+      "epoch": 0.17060493000414553,
+      "grad_norm": 0.2369535267353058,
+      "learning_rate": 0.001,
+      "loss": 1.3737,
+      "step": 5350
+    },
+    {
+      "epoch": 0.17092381772377946,
+      "grad_norm": 0.2344839721918106,
+      "learning_rate": 0.001,
+      "loss": 1.3763,
+      "step": 5360
+    },
+    {
+      "epoch": 0.17124270544341338,
+      "grad_norm": 0.23449836671352386,
+      "learning_rate": 0.001,
+      "loss": 1.3603,
+      "step": 5370
+    },
+    {
+      "epoch": 0.17156159316304728,
+      "grad_norm": 0.2421630322933197,
+      "learning_rate": 0.001,
+      "loss": 1.3503,
+      "step": 5380
+    },
+    {
+      "epoch": 0.1718804808826812,
+      "grad_norm": 0.23793990910053253,
+      "learning_rate": 0.001,
+      "loss": 1.3711,
+      "step": 5390
+    },
+    {
+      "epoch": 0.17219936860231513,
+      "grad_norm": 0.24048961699008942,
+      "learning_rate": 0.001,
+      "loss": 1.398,
+      "step": 5400
+    },
+    {
+      "epoch": 0.17251825632194903,
+      "grad_norm": 0.22494950890541077,
+      "learning_rate": 0.001,
+      "loss": 1.3571,
+      "step": 5410
+    },
+    {
+      "epoch": 0.17283714404158296,
+      "grad_norm": 0.22547166049480438,
+      "learning_rate": 0.001,
+      "loss": 1.3788,
+      "step": 5420
+    },
+    {
+      "epoch": 0.17315603176121688,
+      "grad_norm": 0.23615238070487976,
+      "learning_rate": 0.001,
+      "loss": 1.37,
+      "step": 5430
+    },
+    {
+      "epoch": 0.17347491948085078,
+      "grad_norm": 0.23652249574661255,
+      "learning_rate": 0.001,
+      "loss": 1.3757,
+      "step": 5440
+    },
+    {
+      "epoch": 0.1737938072004847,
+      "grad_norm": 0.24254925549030304,
+      "learning_rate": 0.001,
+      "loss": 1.3745,
+      "step": 5450
+    },
+    {
+      "epoch": 0.17411269492011863,
+      "grad_norm": 0.2400273233652115,
+      "learning_rate": 0.001,
+      "loss": 1.3562,
+      "step": 5460
+    },
+    {
+      "epoch": 0.17443158263975253,
+      "grad_norm": 0.23082616925239563,
+      "learning_rate": 0.001,
+      "loss": 1.3822,
+      "step": 5470
+    },
+    {
+      "epoch": 0.17475047035938646,
+      "grad_norm": 0.2507484257221222,
+      "learning_rate": 0.001,
+      "loss": 1.3813,
+      "step": 5480
+    },
+    {
+      "epoch": 0.17506935807902038,
+      "grad_norm": 0.23675039410591125,
+      "learning_rate": 0.001,
+      "loss": 1.3426,
+      "step": 5490
+    },
+    {
+      "epoch": 0.1753882457986543,
+      "grad_norm": 0.2322949916124344,
+      "learning_rate": 0.001,
+      "loss": 1.3429,
+      "step": 5500
+    },
+    {
+      "epoch": 0.1757071335182882,
+      "grad_norm": 0.23295463621616364,
+      "learning_rate": 0.001,
+      "loss": 1.3746,
+      "step": 5510
+    },
+    {
+      "epoch": 0.17602602123792213,
+      "grad_norm": 0.2300776094198227,
+      "learning_rate": 0.001,
+      "loss": 1.3767,
+      "step": 5520
+    },
+    {
+      "epoch": 0.17634490895755606,
+      "grad_norm": 0.23304656147956848,
+      "learning_rate": 0.001,
+      "loss": 1.3554,
+      "step": 5530
+    },
+    {
+      "epoch": 0.17666379667718995,
+      "grad_norm": 0.23406460881233215,
+      "learning_rate": 0.001,
+      "loss": 1.3605,
+      "step": 5540
+    },
+    {
+      "epoch": 0.17698268439682388,
+      "grad_norm": 0.2351730912923813,
+      "learning_rate": 0.001,
+      "loss": 1.3747,
+      "step": 5550
+    },
+    {
+      "epoch": 0.1773015721164578,
+      "grad_norm": 0.2326095998287201,
+      "learning_rate": 0.001,
+      "loss": 1.3662,
+      "step": 5560
+    },
+    {
+      "epoch": 0.1776204598360917,
+      "grad_norm": 0.23386625945568085,
+      "learning_rate": 0.001,
+      "loss": 1.3551,
+      "step": 5570
+    },
+    {
+      "epoch": 0.17793934755572563,
+      "grad_norm": 0.2271582931280136,
+      "learning_rate": 0.001,
+      "loss": 1.3441,
+      "step": 5580
+    },
+    {
+      "epoch": 0.17825823527535956,
+      "grad_norm": 0.23613621294498444,
+      "learning_rate": 0.001,
+      "loss": 1.3816,
+      "step": 5590
+    },
+    {
+      "epoch": 0.17857712299499345,
+      "grad_norm": 0.23003219068050385,
+      "learning_rate": 0.001,
+      "loss": 1.3676,
+      "step": 5600
+    },
+    {
+      "epoch": 0.17889601071462738,
+      "grad_norm": 0.23304754495620728,
+      "learning_rate": 0.001,
+      "loss": 1.365,
+      "step": 5610
+    },
+    {
+      "epoch": 0.1792148984342613,
+      "grad_norm": 0.23864856362342834,
+      "learning_rate": 0.001,
+      "loss": 1.3803,
+      "step": 5620
+    },
+    {
+      "epoch": 0.1795337861538952,
+      "grad_norm": 0.23164138197898865,
+      "learning_rate": 0.001,
+      "loss": 1.3666,
+      "step": 5630
+    },
+    {
+      "epoch": 0.17985267387352913,
+      "grad_norm": 0.22914890944957733,
+      "learning_rate": 0.001,
+      "loss": 1.3537,
+      "step": 5640
+    },
+    {
+      "epoch": 0.18017156159316305,
+      "grad_norm": 0.23517832159996033,
+      "learning_rate": 0.001,
+      "loss": 1.3742,
+      "step": 5650
+    },
+    {
+      "epoch": 0.18049044931279695,
+      "grad_norm": 0.23220737278461456,
+      "learning_rate": 0.001,
+      "loss": 1.3631,
+      "step": 5660
+    },
+    {
+      "epoch": 0.18080933703243088,
+      "grad_norm": 0.23347978293895721,
+      "learning_rate": 0.001,
+      "loss": 1.3663,
+      "step": 5670
+    },
+    {
+      "epoch": 0.1811282247520648,
+      "grad_norm": 0.2269410789012909,
+      "learning_rate": 0.001,
+      "loss": 1.3452,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1814471124716987,
+      "grad_norm": 0.23430763185024261,
+      "learning_rate": 0.001,
+      "loss": 1.3517,
+      "step": 5690
+    },
+    {
+      "epoch": 0.18176600019133263,
+      "grad_norm": 0.22661398351192474,
+      "learning_rate": 0.001,
+      "loss": 1.357,
+      "step": 5700
+    },
+    {
+      "epoch": 0.18208488791096655,
+      "grad_norm": 0.23189307749271393,
+      "learning_rate": 0.001,
+      "loss": 1.3619,
+      "step": 5710
+    },
+    {
+      "epoch": 0.18240377563060048,
+      "grad_norm": 0.22509999573230743,
+      "learning_rate": 0.001,
+      "loss": 1.3737,
+      "step": 5720
+    },
+    {
+      "epoch": 0.18272266335023438,
+      "grad_norm": 0.23490670323371887,
+      "learning_rate": 0.001,
+      "loss": 1.3397,
+      "step": 5730
+    },
+    {
+      "epoch": 0.1830415510698683,
+      "grad_norm": 0.22804059088230133,
+      "learning_rate": 0.001,
+      "loss": 1.3344,
+      "step": 5740
+    },
+    {
+      "epoch": 0.18336043878950223,
+      "grad_norm": 0.22229887545108795,
+      "learning_rate": 0.001,
+      "loss": 1.3693,
+      "step": 5750
+    },
+    {
+      "epoch": 0.18367932650913613,
+      "grad_norm": 0.23096612095832825,
+      "learning_rate": 0.001,
+      "loss": 1.3536,
+      "step": 5760
+    },
+    {
+      "epoch": 0.18399821422877005,
+      "grad_norm": 0.23599880933761597,
+      "learning_rate": 0.001,
+      "loss": 1.3496,
+      "step": 5770
+    },
+    {
+      "epoch": 0.18431710194840398,
+      "grad_norm": 0.22269390523433685,
+      "learning_rate": 0.001,
+      "loss": 1.3378,
+      "step": 5780
+    },
+    {
+      "epoch": 0.18463598966803788,
+      "grad_norm": 0.22350311279296875,
+      "learning_rate": 0.001,
+      "loss": 1.34,
+      "step": 5790
+    },
+    {
+      "epoch": 0.1849548773876718,
+      "grad_norm": 0.23092792928218842,
+      "learning_rate": 0.001,
+      "loss": 1.3526,
+      "step": 5800
+    },
+    {
+      "epoch": 0.18527376510730573,
+      "grad_norm": 0.22100470960140228,
+      "learning_rate": 0.001,
+      "loss": 1.3526,
+      "step": 5810
+    },
+    {
+      "epoch": 0.18559265282693962,
+      "grad_norm": 0.2348584532737732,
+      "learning_rate": 0.001,
+      "loss": 1.3433,
+      "step": 5820
+    },
+    {
+      "epoch": 0.18591154054657355,
+      "grad_norm": 0.22047287225723267,
+      "learning_rate": 0.001,
+      "loss": 1.3441,
+      "step": 5830
+    },
+    {
+      "epoch": 0.18623042826620748,
+      "grad_norm": 0.22145791351795197,
+      "learning_rate": 0.001,
+      "loss": 1.3441,
+      "step": 5840
+    },
+    {
+      "epoch": 0.18654931598584137,
+      "grad_norm": 0.22203807532787323,
+      "learning_rate": 0.001,
+      "loss": 1.3493,
+      "step": 5850
+    },
+    {
+      "epoch": 0.1868682037054753,
+      "grad_norm": 0.22726795077323914,
+      "learning_rate": 0.001,
+      "loss": 1.3482,
+      "step": 5860
+    },
+    {
+      "epoch": 0.18718709142510923,
+      "grad_norm": 0.22857330739498138,
+      "learning_rate": 0.001,
+      "loss": 1.34,
+      "step": 5870
+    },
+    {
+      "epoch": 0.18750597914474312,
+      "grad_norm": 0.22861173748970032,
+      "learning_rate": 0.001,
+      "loss": 1.3598,
+      "step": 5880
+    },
+    {
+      "epoch": 0.18782486686437705,
+      "grad_norm": 0.22939689457416534,
+      "learning_rate": 0.001,
+      "loss": 1.3443,
+      "step": 5890
+    },
+    {
+      "epoch": 0.18814375458401097,
+      "grad_norm": 0.23213569819927216,
+      "learning_rate": 0.001,
+      "loss": 1.3412,
+      "step": 5900
+    },
+    {
+      "epoch": 0.18846264230364487,
+      "grad_norm": 0.2435399442911148,
+      "learning_rate": 0.001,
+      "loss": 1.3379,
+      "step": 5910
+    },
+    {
+      "epoch": 0.1887815300232788,
+      "grad_norm": 0.22397400438785553,
+      "learning_rate": 0.001,
+      "loss": 1.3425,
+      "step": 5920
+    },
+    {
+      "epoch": 0.18910041774291272,
+      "grad_norm": 0.23101122677326202,
+      "learning_rate": 0.001,
+      "loss": 1.341,
+      "step": 5930
+    },
+    {
+      "epoch": 0.18941930546254665,
+      "grad_norm": 0.22613437473773956,
+      "learning_rate": 0.001,
+      "loss": 1.3503,
+      "step": 5940
+    },
+    {
+      "epoch": 0.18973819318218055,
+      "grad_norm": 0.23326362669467926,
+      "learning_rate": 0.001,
+      "loss": 1.3355,
+      "step": 5950
+    },
+    {
+      "epoch": 0.19005708090181447,
+      "grad_norm": 0.22755029797554016,
+      "learning_rate": 0.001,
+      "loss": 1.3639,
+      "step": 5960
+    },
+    {
+      "epoch": 0.1903759686214484,
+      "grad_norm": 0.23369061946868896,
+      "learning_rate": 0.001,
+      "loss": 1.3495,
+      "step": 5970
+    },
+    {
+      "epoch": 0.1906948563410823,
+      "grad_norm": 0.2289803922176361,
+      "learning_rate": 0.001,
+      "loss": 1.3372,
+      "step": 5980
+    },
+    {
+      "epoch": 0.19101374406071622,
+      "grad_norm": 0.22431151568889618,
+      "learning_rate": 0.001,
+      "loss": 1.3442,
+      "step": 5990
+    },
+    {
+      "epoch": 0.19133263178035015,
+      "grad_norm": 0.22803178429603577,
+      "learning_rate": 0.001,
+      "loss": 1.325,
+      "step": 6000
+    },
+    {
+      "epoch": 0.19165151949998405,
+      "grad_norm": 0.2185739278793335,
+      "learning_rate": 0.001,
+      "loss": 1.3482,
+      "step": 6010
+    },
+    {
+      "epoch": 0.19197040721961797,
+      "grad_norm": 0.2274511605501175,
+      "learning_rate": 0.001,
+      "loss": 1.3283,
+      "step": 6020
+    },
+    {
+      "epoch": 0.1922892949392519,
+      "grad_norm": 0.22346645593643188,
+      "learning_rate": 0.001,
+      "loss": 1.3133,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1926081826588858,
+      "grad_norm": 0.21699689328670502,
+      "learning_rate": 0.001,
+      "loss": 1.3313,
+      "step": 6040
+    },
+    {
+      "epoch": 0.19292707037851972,
+      "grad_norm": 0.22448177635669708,
+      "learning_rate": 0.001,
+      "loss": 1.3327,
+      "step": 6050
+    },
+    {
+      "epoch": 0.19324595809815365,
+      "grad_norm": 0.2269693911075592,
+      "learning_rate": 0.001,
+      "loss": 1.325,
+      "step": 6060
+    },
+    {
+      "epoch": 0.19356484581778755,
+      "grad_norm": 0.22963395714759827,
+      "learning_rate": 0.001,
+      "loss": 1.3337,
+      "step": 6070
+    },
+    {
+      "epoch": 0.19388373353742147,
+      "grad_norm": 0.22270570695400238,
+      "learning_rate": 0.001,
+      "loss": 1.3263,
+      "step": 6080
+    },
+    {
+      "epoch": 0.1942026212570554,
+      "grad_norm": 0.22263237833976746,
+      "learning_rate": 0.001,
+      "loss": 1.3291,
+      "step": 6090
+    },
+    {
+      "epoch": 0.1945215089766893,
+      "grad_norm": 0.21834489703178406,
+      "learning_rate": 0.001,
+      "loss": 1.3361,
+      "step": 6100
+    },
+    {
+      "epoch": 0.19484039669632322,
+      "grad_norm": 0.22711575031280518,
+      "learning_rate": 0.001,
+      "loss": 1.3428,
+      "step": 6110
+    },
+    {
+      "epoch": 0.19515928441595715,
+      "grad_norm": 0.23172850906848907,
+      "learning_rate": 0.001,
+      "loss": 1.3353,
+      "step": 6120
+    },
+    {
+      "epoch": 0.19547817213559107,
+      "grad_norm": 0.22978724539279938,
+      "learning_rate": 0.001,
+      "loss": 1.3139,
+      "step": 6130
+    },
+    {
+      "epoch": 0.19579705985522497,
+      "grad_norm": 0.22300496697425842,
+      "learning_rate": 0.001,
+      "loss": 1.3414,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1961159475748589,
+      "grad_norm": 0.22236841917037964,
+      "learning_rate": 0.001,
+      "loss": 1.3035,
+      "step": 6150
+    },
+    {
+      "epoch": 0.19643483529449282,
+      "grad_norm": 0.22211438417434692,
+      "learning_rate": 0.001,
+      "loss": 1.3141,
+      "step": 6160
+    },
+    {
+      "epoch": 0.19675372301412672,
+      "grad_norm": 0.21687041223049164,
+      "learning_rate": 0.001,
+      "loss": 1.3444,
+      "step": 6170
+    },
+    {
+      "epoch": 0.19707261073376064,
+      "grad_norm": 0.2247399538755417,
+      "learning_rate": 0.001,
+      "loss": 1.3274,
+      "step": 6180
+    },
+    {
+      "epoch": 0.19739149845339457,
+      "grad_norm": 0.22150497138500214,
+      "learning_rate": 0.001,
+      "loss": 1.3467,
+      "step": 6190
+    },
+    {
+      "epoch": 0.19771038617302847,
+      "grad_norm": 0.22610719501972198,
+      "learning_rate": 0.001,
+      "loss": 1.331,
+      "step": 6200
+    },
+    {
+      "epoch": 0.1980292738926624,
+      "grad_norm": 0.21927721798419952,
+      "learning_rate": 0.001,
+      "loss": 1.3341,
+      "step": 6210
+    },
+    {
+      "epoch": 0.19834816161229632,
+      "grad_norm": 0.22373303771018982,
+      "learning_rate": 0.001,
+      "loss": 1.3066,
+      "step": 6220
+    },
+    {
+      "epoch": 0.19866704933193022,
+      "grad_norm": 0.22470548748970032,
+      "learning_rate": 0.001,
+      "loss": 1.3102,
+      "step": 6230
+    },
+    {
+      "epoch": 0.19898593705156414,
+      "grad_norm": 0.22980283200740814,
+      "learning_rate": 0.001,
+      "loss": 1.326,
+      "step": 6240
+    },
+    {
+      "epoch": 0.19930482477119807,
+      "grad_norm": 0.2177285999059677,
+      "learning_rate": 0.001,
+      "loss": 1.3367,
+      "step": 6250
+    },
+    {
+      "epoch": 0.19962371249083197,
+      "grad_norm": 0.23009584844112396,
+      "learning_rate": 0.001,
+      "loss": 1.3157,
+      "step": 6260
+    },
+    {
+      "epoch": 0.1999426002104659,
+      "grad_norm": 0.22571367025375366,
+      "learning_rate": 0.001,
+      "loss": 1.3271,
+      "step": 6270
+    },
+    {
+      "epoch": 0.20026148793009982,
+      "grad_norm": 0.22114919126033783,
+      "learning_rate": 0.001,
+      "loss": 1.3243,
+      "step": 6280
+    },
+    {
+      "epoch": 0.20058037564973372,
+      "grad_norm": 0.22167930006980896,
+      "learning_rate": 0.001,
+      "loss": 1.3355,
+      "step": 6290
+    },
+    {
+      "epoch": 0.20089926336936764,
+      "grad_norm": 0.22725333273410797,
+      "learning_rate": 0.001,
+      "loss": 1.3277,
+      "step": 6300
+    },
+    {
+      "epoch": 0.20121815108900157,
+      "grad_norm": 0.22026440501213074,
+      "learning_rate": 0.001,
+      "loss": 1.3198,
+      "step": 6310
+    },
+    {
+      "epoch": 0.20153703880863547,
+      "grad_norm": 0.22313068807125092,
+      "learning_rate": 0.001,
+      "loss": 1.321,
+      "step": 6320
+    },
+    {
+      "epoch": 0.2018559265282694,
+      "grad_norm": 0.22668123245239258,
+      "learning_rate": 0.001,
+      "loss": 1.3187,
+      "step": 6330
+    },
+    {
+      "epoch": 0.20217481424790332,
+      "grad_norm": 0.22383753955364227,
+      "learning_rate": 0.001,
+      "loss": 1.3243,
+      "step": 6340
+    },
+    {
+      "epoch": 0.20249370196753724,
+      "grad_norm": 0.22911064326763153,
+      "learning_rate": 0.001,
+      "loss": 1.3146,
+      "step": 6350
+    },
+    {
+      "epoch": 0.20281258968717114,
+      "grad_norm": 0.21768306195735931,
+      "learning_rate": 0.001,
+      "loss": 1.3181,
+      "step": 6360
+    },
+    {
+      "epoch": 0.20313147740680507,
+      "grad_norm": 0.21696394681930542,
+      "learning_rate": 0.001,
+      "loss": 1.3349,
+      "step": 6370
+    },
+    {
+      "epoch": 0.203450365126439,
+      "grad_norm": 0.21784517168998718,
+      "learning_rate": 0.001,
+      "loss": 1.3312,
+      "step": 6380
+    },
+    {
+      "epoch": 0.2037692528460729,
+      "grad_norm": 0.22547577321529388,
+      "learning_rate": 0.001,
+      "loss": 1.3171,
+      "step": 6390
+    },
+    {
+      "epoch": 0.20408814056570682,
+      "grad_norm": 0.22969777882099152,
+      "learning_rate": 0.001,
+      "loss": 1.3022,
+      "step": 6400
+    },
+    {
+      "epoch": 0.20440702828534074,
+      "grad_norm": 0.23056884109973907,
+      "learning_rate": 0.001,
+      "loss": 1.3302,
+      "step": 6410
+    },
+    {
+      "epoch": 0.20472591600497464,
+      "grad_norm": 0.21913301944732666,
+      "learning_rate": 0.001,
+      "loss": 1.3091,
+      "step": 6420
+    },
+    {
+      "epoch": 0.20504480372460857,
+      "grad_norm": 0.2276025414466858,
+      "learning_rate": 0.001,
+      "loss": 1.3374,
+      "step": 6430
+    },
+    {
+      "epoch": 0.2053636914442425,
+      "grad_norm": 0.21844200789928436,
+      "learning_rate": 0.001,
+      "loss": 1.3138,
+      "step": 6440
+    },
+    {
+      "epoch": 0.2056825791638764,
+      "grad_norm": 0.21239309012889862,
+      "learning_rate": 0.001,
+      "loss": 1.3079,
+      "step": 6450
+    },
+    {
+      "epoch": 0.20600146688351031,
+      "grad_norm": 0.21648620069026947,
+      "learning_rate": 0.001,
+      "loss": 1.3069,
+      "step": 6460
+    },
+    {
+      "epoch": 0.20632035460314424,
+      "grad_norm": 0.2209746241569519,
+      "learning_rate": 0.001,
+      "loss": 1.3235,
+      "step": 6470
+    },
+    {
+      "epoch": 0.20663924232277814,
+      "grad_norm": 0.22533218562602997,
+      "learning_rate": 0.001,
+      "loss": 1.3189,
+      "step": 6480
+    },
+    {
+      "epoch": 0.20695813004241206,
+      "grad_norm": 0.22373533248901367,
+      "learning_rate": 0.001,
+      "loss": 1.321,
+      "step": 6490
+    },
+    {
+      "epoch": 0.207277017762046,
+      "grad_norm": 0.22977867722511292,
+      "learning_rate": 0.001,
+      "loss": 1.3111,
+      "step": 6500
+    },
+    {
+      "epoch": 0.2075959054816799,
+      "grad_norm": 0.2225407063961029,
+      "learning_rate": 0.001,
+      "loss": 1.3111,
+      "step": 6510
+    },
+    {
+      "epoch": 0.2079147932013138,
+      "grad_norm": 0.22135379910469055,
+      "learning_rate": 0.001,
+      "loss": 1.3111,
+      "step": 6520
+    },
+    {
+      "epoch": 0.20823368092094774,
+      "grad_norm": 0.2236153930425644,
+      "learning_rate": 0.001,
+      "loss": 1.3216,
+      "step": 6530
+    },
+    {
+      "epoch": 0.20855256864058166,
+      "grad_norm": 0.21490934491157532,
+      "learning_rate": 0.001,
+      "loss": 1.3331,
+      "step": 6540
+    },
+    {
+      "epoch": 0.20887145636021556,
+      "grad_norm": 0.22031015157699585,
+      "learning_rate": 0.001,
+      "loss": 1.3187,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2091903440798495,
+      "grad_norm": 0.21435341238975525,
+      "learning_rate": 0.001,
+      "loss": 1.3103,
+      "step": 6560
+    },
+    {
+      "epoch": 0.20950923179948341,
+      "grad_norm": 0.21968773007392883,
+      "learning_rate": 0.001,
+      "loss": 1.316,
+      "step": 6570
+    },
+    {
+      "epoch": 0.2098281195191173,
+      "grad_norm": 0.21836096048355103,
+      "learning_rate": 0.001,
+      "loss": 1.316,
+      "step": 6580
+    },
+    {
+      "epoch": 0.21014700723875124,
+      "grad_norm": 0.21528609097003937,
+      "learning_rate": 0.001,
+      "loss": 1.3201,
+      "step": 6590
+    },
+    {
+      "epoch": 0.21046589495838516,
+      "grad_norm": 0.22226691246032715,
+      "learning_rate": 0.001,
+      "loss": 1.2988,
+      "step": 6600
+    },
+    {
+      "epoch": 0.21078478267801906,
+      "grad_norm": 0.21093668043613434,
+      "learning_rate": 0.001,
+      "loss": 1.3211,
+      "step": 6610
+    },
+    {
+      "epoch": 0.211103670397653,
+      "grad_norm": 0.21473108232021332,
+      "learning_rate": 0.001,
+      "loss": 1.307,
+      "step": 6620
+    },
+    {
+      "epoch": 0.2114225581172869,
+      "grad_norm": 0.21987295150756836,
+      "learning_rate": 0.001,
+      "loss": 1.3161,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2117414458369208,
+      "grad_norm": 0.2141866534948349,
+      "learning_rate": 0.001,
+      "loss": 1.3273,
+      "step": 6640
+    },
+    {
+      "epoch": 0.21206033355655474,
+      "grad_norm": 0.211161270737648,
+      "learning_rate": 0.001,
+      "loss": 1.3012,
+      "step": 6650
+    },
+    {
+      "epoch": 0.21237922127618866,
+      "grad_norm": 0.22050651907920837,
+      "learning_rate": 0.001,
+      "loss": 1.3133,
+      "step": 6660
+    },
+    {
+      "epoch": 0.21269810899582256,
+      "grad_norm": 0.22973541915416718,
+      "learning_rate": 0.001,
+      "loss": 1.3085,
+      "step": 6670
+    },
+    {
+      "epoch": 0.21301699671545649,
+      "grad_norm": 0.21897663176059723,
+      "learning_rate": 0.001,
+      "loss": 1.3053,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2133358844350904,
+      "grad_norm": 0.22126558423042297,
+      "learning_rate": 0.001,
+      "loss": 1.3155,
+      "step": 6690
+    },
+    {
+      "epoch": 0.2136547721547243,
+      "grad_norm": 0.2211180031299591,
+      "learning_rate": 0.001,
+      "loss": 1.3106,
+      "step": 6700
+    },
+    {
+      "epoch": 0.21397365987435824,
+      "grad_norm": 0.2270718365907669,
+      "learning_rate": 0.001,
+      "loss": 1.2997,
+      "step": 6710
+    },
+    {
+      "epoch": 0.21429254759399216,
+      "grad_norm": 0.22596475481987,
+      "learning_rate": 0.001,
+      "loss": 1.3124,
+      "step": 6720
+    },
+    {
+      "epoch": 0.21461143531362606,
+      "grad_norm": 0.21958962082862854,
+      "learning_rate": 0.001,
+      "loss": 1.3033,
+      "step": 6730
+    },
+    {
+      "epoch": 0.21493032303325998,
+      "grad_norm": 0.21273846924304962,
+      "learning_rate": 0.001,
+      "loss": 1.3066,
+      "step": 6740
+    },
+    {
+      "epoch": 0.2152492107528939,
+      "grad_norm": 0.21829353272914886,
+      "learning_rate": 0.001,
+      "loss": 1.321,
+      "step": 6750
+    },
+    {
+      "epoch": 0.21556809847252784,
+      "grad_norm": 0.21527628600597382,
+      "learning_rate": 0.001,
+      "loss": 1.3192,
+      "step": 6760
+    },
+    {
+      "epoch": 0.21588698619216173,
+      "grad_norm": 0.2100159376859665,
+      "learning_rate": 0.001,
+      "loss": 1.301,
+      "step": 6770
+    },
+    {
+      "epoch": 0.21620587391179566,
+      "grad_norm": 0.2193724364042282,
+      "learning_rate": 0.001,
+      "loss": 1.2945,
+      "step": 6780
+    },
+    {
+      "epoch": 0.21652476163142959,
+      "grad_norm": 0.22077114880084991,
+      "learning_rate": 0.001,
+      "loss": 1.2939,
+      "step": 6790
+    },
+    {
+      "epoch": 0.21684364935106348,
+      "grad_norm": 0.2293231189250946,
+      "learning_rate": 0.001,
+      "loss": 1.2818,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2171625370706974,
+      "grad_norm": 0.21325518190860748,
+      "learning_rate": 0.001,
+      "loss": 1.3176,
+      "step": 6810
+    },
+    {
+      "epoch": 0.21748142479033133,
+      "grad_norm": 0.2264716476202011,
+      "learning_rate": 0.001,
+      "loss": 1.3017,
+      "step": 6820
+    },
+    {
+      "epoch": 0.21780031250996523,
+      "grad_norm": 0.21110965311527252,
+      "learning_rate": 0.001,
+      "loss": 1.3018,
+      "step": 6830
+    },
+    {
+      "epoch": 0.21811920022959916,
+      "grad_norm": 0.21436037123203278,
+      "learning_rate": 0.001,
+      "loss": 1.2995,
+      "step": 6840
+    },
+    {
+      "epoch": 0.21843808794923308,
+      "grad_norm": 0.2117929756641388,
+      "learning_rate": 0.001,
+      "loss": 1.282,
+      "step": 6850
+    },
+    {
+      "epoch": 0.21875697566886698,
+      "grad_norm": 0.21568013727664948,
+      "learning_rate": 0.001,
+      "loss": 1.2839,
+      "step": 6860
+    },
+    {
+      "epoch": 0.2190758633885009,
+      "grad_norm": 0.21337729692459106,
+      "learning_rate": 0.001,
+      "loss": 1.2981,
+      "step": 6870
+    },
+    {
+      "epoch": 0.21939475110813483,
+      "grad_norm": 0.21572987735271454,
+      "learning_rate": 0.001,
+      "loss": 1.3014,
+      "step": 6880
+    },
+    {
+      "epoch": 0.21971363882776873,
+      "grad_norm": 0.2122887820005417,
+      "learning_rate": 0.001,
+      "loss": 1.2972,
+      "step": 6890
+    },
+    {
+      "epoch": 0.22003252654740266,
+      "grad_norm": 0.21946419775485992,
+      "learning_rate": 0.001,
+      "loss": 1.3096,
+      "step": 6900
+    },
+    {
+      "epoch": 0.22035141426703658,
+      "grad_norm": 0.21935173869132996,
+      "learning_rate": 0.001,
+      "loss": 1.3027,
+      "step": 6910
+    },
+    {
+      "epoch": 0.22067030198667048,
+      "grad_norm": 0.2146780639886856,
+      "learning_rate": 0.001,
+      "loss": 1.3165,
+      "step": 6920
+    },
+    {
+      "epoch": 0.2209891897063044,
+      "grad_norm": 0.21132074296474457,
+      "learning_rate": 0.001,
+      "loss": 1.3073,
+      "step": 6930
+    },
+    {
+      "epoch": 0.22130807742593833,
+      "grad_norm": 0.21391667425632477,
+      "learning_rate": 0.001,
+      "loss": 1.2909,
+      "step": 6940
+    },
+    {
+      "epoch": 0.22162696514557226,
+      "grad_norm": 0.21686720848083496,
+      "learning_rate": 0.001,
+      "loss": 1.2877,
+      "step": 6950
+    },
+    {
+      "epoch": 0.22194585286520616,
+      "grad_norm": 0.2125672698020935,
+      "learning_rate": 0.001,
+      "loss": 1.2829,
+      "step": 6960
+    },
+    {
+      "epoch": 0.22226474058484008,
+      "grad_norm": 0.20903590321540833,
+      "learning_rate": 0.001,
+      "loss": 1.2921,
+      "step": 6970
+    },
+    {
+      "epoch": 0.222583628304474,
+      "grad_norm": 0.21638572216033936,
+      "learning_rate": 0.001,
+      "loss": 1.2918,
+      "step": 6980
+    },
+    {
+      "epoch": 0.2229025160241079,
+      "grad_norm": 0.2114180624485016,
+      "learning_rate": 0.001,
+      "loss": 1.3036,
+      "step": 6990
+    },
+    {
+      "epoch": 0.22322140374374183,
+      "grad_norm": 0.21430033445358276,
+      "learning_rate": 0.001,
+      "loss": 1.3194,
+      "step": 7000
+    },
+    {
+      "epoch": 0.22354029146337576,
+      "grad_norm": 0.21226227283477783,
+      "learning_rate": 0.001,
+      "loss": 1.2825,
+      "step": 7010
+    },
+    {
+      "epoch": 0.22385917918300965,
+      "grad_norm": 0.21256470680236816,
+      "learning_rate": 0.001,
+      "loss": 1.3053,
+      "step": 7020
+    },
+    {
+      "epoch": 0.22417806690264358,
+      "grad_norm": 0.22182166576385498,
+      "learning_rate": 0.001,
+      "loss": 1.2713,
+      "step": 7030
+    },
+    {
+      "epoch": 0.2244969546222775,
+      "grad_norm": 0.21688029170036316,
+      "learning_rate": 0.001,
+      "loss": 1.2743,
+      "step": 7040
+    },
+    {
+      "epoch": 0.2248158423419114,
+      "grad_norm": 0.2168598622083664,
+      "learning_rate": 0.001,
+      "loss": 1.2936,
+      "step": 7050
+    },
+    {
+      "epoch": 0.22513473006154533,
+      "grad_norm": 0.2037907838821411,
+      "learning_rate": 0.001,
+      "loss": 1.2862,
+      "step": 7060
+    },
+    {
+      "epoch": 0.22545361778117926,
+      "grad_norm": 0.21954897046089172,
+      "learning_rate": 0.001,
+      "loss": 1.2856,
+      "step": 7070
+    },
+    {
+      "epoch": 0.22577250550081315,
+      "grad_norm": 0.21183601021766663,
+      "learning_rate": 0.001,
+      "loss": 1.3106,
+      "step": 7080
+    },
+    {
+      "epoch": 0.22609139322044708,
+      "grad_norm": 0.21412093937397003,
+      "learning_rate": 0.001,
+      "loss": 1.3166,
+      "step": 7090
+    },
+    {
+      "epoch": 0.226410280940081,
+      "grad_norm": 0.2191409319639206,
+      "learning_rate": 0.001,
+      "loss": 1.2924,
+      "step": 7100
+    },
+    {
+      "epoch": 0.2267291686597149,
+      "grad_norm": 0.21817272901535034,
+      "learning_rate": 0.001,
+      "loss": 1.3071,
+      "step": 7110
+    },
+    {
+      "epoch": 0.22704805637934883,
+      "grad_norm": 0.20995283126831055,
+      "learning_rate": 0.001,
+      "loss": 1.2837,
+      "step": 7120
+    },
+    {
+      "epoch": 0.22736694409898275,
+      "grad_norm": 0.21756869554519653,
+      "learning_rate": 0.001,
+      "loss": 1.3008,
+      "step": 7130
+    },
+    {
+      "epoch": 0.22768583181861665,
+      "grad_norm": 0.21312449872493744,
+      "learning_rate": 0.001,
+      "loss": 1.2956,
+      "step": 7140
+    },
+    {
+      "epoch": 0.22800471953825058,
+      "grad_norm": 0.2182151824235916,
+      "learning_rate": 0.001,
+      "loss": 1.2939,
+      "step": 7150
+    },
+    {
+      "epoch": 0.2283236072578845,
+      "grad_norm": 0.2198774218559265,
+      "learning_rate": 0.001,
+      "loss": 1.2938,
+      "step": 7160
+    },
+    {
+      "epoch": 0.22864249497751843,
+      "grad_norm": 0.2158440351486206,
+      "learning_rate": 0.001,
+      "loss": 1.3113,
+      "step": 7170
+    },
+    {
+      "epoch": 0.22896138269715233,
+      "grad_norm": 0.21088720858097076,
+      "learning_rate": 0.001,
+      "loss": 1.2764,
+      "step": 7180
+    },
+    {
+      "epoch": 0.22928027041678625,
+      "grad_norm": 0.20957180857658386,
+      "learning_rate": 0.001,
+      "loss": 1.2852,
+      "step": 7190
+    },
+    {
+      "epoch": 0.22959915813642018,
+      "grad_norm": 0.2130306363105774,
+      "learning_rate": 0.001,
+      "loss": 1.2853,
+      "step": 7200
+    },
+    {
+      "epoch": 0.22991804585605408,
+      "grad_norm": 0.21509139239788055,
+      "learning_rate": 0.001,
+      "loss": 1.2817,
+      "step": 7210
+    },
+    {
+      "epoch": 0.230236933575688,
+      "grad_norm": 0.2113594263792038,
+      "learning_rate": 0.001,
+      "loss": 1.261,
+      "step": 7220
+    },
+    {
+      "epoch": 0.23055582129532193,
+      "grad_norm": 0.20813211798667908,
+      "learning_rate": 0.001,
+      "loss": 1.2936,
+      "step": 7230
+    },
+    {
+      "epoch": 0.23087470901495583,
+      "grad_norm": 0.21231849491596222,
+      "learning_rate": 0.001,
+      "loss": 1.2991,
+      "step": 7240
+    },
+    {
+      "epoch": 0.23119359673458975,
+      "grad_norm": 0.21319466829299927,
+      "learning_rate": 0.001,
+      "loss": 1.2865,
+      "step": 7250
+    },
+    {
+      "epoch": 0.23151248445422368,
+      "grad_norm": 0.2093285769224167,
+      "learning_rate": 0.001,
+      "loss": 1.2739,
+      "step": 7260
+    },
+    {
+      "epoch": 0.23183137217385758,
+      "grad_norm": 0.21632854640483856,
+      "learning_rate": 0.001,
+      "loss": 1.274,
+      "step": 7270
+    },
+    {
+      "epoch": 0.2321502598934915,
+      "grad_norm": 0.2091766893863678,
+      "learning_rate": 0.001,
+      "loss": 1.3008,
+      "step": 7280
+    },
+    {
+      "epoch": 0.23246914761312543,
+      "grad_norm": 0.20856739580631256,
+      "learning_rate": 0.001,
+      "loss": 1.2885,
+      "step": 7290
+    },
+    {
+      "epoch": 0.23278803533275932,
+      "grad_norm": 0.2127932757139206,
+      "learning_rate": 0.001,
+      "loss": 1.2793,
+      "step": 7300
+    },
+    {
+      "epoch": 0.23310692305239325,
+      "grad_norm": 0.2079467624425888,
+      "learning_rate": 0.001,
+      "loss": 1.2964,
+      "step": 7310
+    },
+    {
+      "epoch": 0.23342581077202718,
+      "grad_norm": 0.21427300572395325,
+      "learning_rate": 0.001,
+      "loss": 1.2781,
+      "step": 7320
+    },
+    {
+      "epoch": 0.23374469849166107,
+      "grad_norm": 0.20755432546138763,
+      "learning_rate": 0.001,
+      "loss": 1.2683,
+      "step": 7330
+    },
+    {
+      "epoch": 0.234063586211295,
+      "grad_norm": 0.20725442469120026,
+      "learning_rate": 0.001,
+      "loss": 1.2912,
+      "step": 7340
+    },
+    {
+      "epoch": 0.23438247393092893,
+      "grad_norm": 0.20349465310573578,
+      "learning_rate": 0.001,
+      "loss": 1.2808,
+      "step": 7350
+    },
+    {
+      "epoch": 0.23470136165056282,
+      "grad_norm": 0.20813782513141632,
+      "learning_rate": 0.001,
+      "loss": 1.2753,
+      "step": 7360
+    },
+    {
+      "epoch": 0.23502024937019675,
+      "grad_norm": 0.2123478800058365,
+      "learning_rate": 0.001,
+      "loss": 1.2779,
+      "step": 7370
+    },
+    {
+      "epoch": 0.23533913708983067,
+      "grad_norm": 0.21471963822841644,
+      "learning_rate": 0.001,
+      "loss": 1.2853,
+      "step": 7380
+    },
+    {
+      "epoch": 0.2356580248094646,
+      "grad_norm": 0.2094661146402359,
+      "learning_rate": 0.001,
+      "loss": 1.2864,
+      "step": 7390
+    },
+    {
+      "epoch": 0.2359769125290985,
+      "grad_norm": 0.2113139033317566,
+      "learning_rate": 0.001,
+      "loss": 1.2727,
+      "step": 7400
+    },
+    {
+      "epoch": 0.23629580024873242,
+      "grad_norm": 0.2110980600118637,
+      "learning_rate": 0.001,
+      "loss": 1.2748,
+      "step": 7410
+    },
+    {
+      "epoch": 0.23661468796836635,
+      "grad_norm": 0.22186870872974396,
+      "learning_rate": 0.001,
+      "loss": 1.2975,
+      "step": 7420
+    },
+    {
+      "epoch": 0.23693357568800025,
+      "grad_norm": 0.2114681750535965,
+      "learning_rate": 0.001,
+      "loss": 1.2569,
+      "step": 7430
+    },
+    {
+      "epoch": 0.23725246340763417,
+      "grad_norm": 0.21068887412548065,
+      "learning_rate": 0.001,
+      "loss": 1.2703,
+      "step": 7440
+    },
+    {
+      "epoch": 0.2375713511272681,
+      "grad_norm": 0.21688032150268555,
+      "learning_rate": 0.001,
+      "loss": 1.294,
+      "step": 7450
+    },
+    {
+      "epoch": 0.237890238846902,
+      "grad_norm": 0.20603680610656738,
+      "learning_rate": 0.001,
+      "loss": 1.2963,
+      "step": 7460
+    },
+    {
+      "epoch": 0.23820912656653592,
+      "grad_norm": 0.2126203328371048,
+      "learning_rate": 0.001,
+      "loss": 1.2791,
+      "step": 7470
+    },
+    {
+      "epoch": 0.23852801428616985,
+      "grad_norm": 0.20721890032291412,
+      "learning_rate": 0.001,
+      "loss": 1.2941,
+      "step": 7480
+    },
+    {
+      "epoch": 0.23884690200580375,
+      "grad_norm": 0.20675908029079437,
+      "learning_rate": 0.001,
+      "loss": 1.2644,
+      "step": 7490
+    },
+    {
+      "epoch": 0.23916578972543767,
+      "grad_norm": 0.2099614143371582,
+      "learning_rate": 0.001,
+      "loss": 1.2948,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2394846774450716,
+      "grad_norm": 0.2123858779668808,
+      "learning_rate": 0.001,
+      "loss": 1.2781,
+      "step": 7510
+    },
+    {
+      "epoch": 0.2398035651647055,
+      "grad_norm": 0.20939020812511444,
+      "learning_rate": 0.001,
+      "loss": 1.2867,
+      "step": 7520
+    },
+    {
+      "epoch": 0.24012245288433942,
+      "grad_norm": 0.1992987096309662,
+      "learning_rate": 0.001,
+      "loss": 1.2857,
+      "step": 7530
+    },
+    {
+      "epoch": 0.24044134060397335,
+      "grad_norm": 0.20575982332229614,
+      "learning_rate": 0.001,
+      "loss": 1.2829,
+      "step": 7540
+    },
+    {
+      "epoch": 0.24076022832360724,
+      "grad_norm": 0.2120957225561142,
+      "learning_rate": 0.001,
+      "loss": 1.2595,
+      "step": 7550
+    },
+    {
+      "epoch": 0.24107911604324117,
+      "grad_norm": 0.2008555829524994,
+      "learning_rate": 0.001,
+      "loss": 1.2722,
+      "step": 7560
+    },
+    {
+      "epoch": 0.2413980037628751,
+      "grad_norm": 0.2073550969362259,
+      "learning_rate": 0.001,
+      "loss": 1.2973,
+      "step": 7570
+    },
+    {
+      "epoch": 0.24171689148250902,
+      "grad_norm": 0.21187391877174377,
+      "learning_rate": 0.001,
+      "loss": 1.2726,
+      "step": 7580
+    },
+    {
+      "epoch": 0.24203577920214292,
+      "grad_norm": 0.20674507319927216,
+      "learning_rate": 0.001,
+      "loss": 1.2773,
+      "step": 7590
+    },
+    {
+      "epoch": 0.24235466692177685,
+      "grad_norm": 0.20273102819919586,
+      "learning_rate": 0.001,
+      "loss": 1.2873,
+      "step": 7600
+    },
+    {
+      "epoch": 0.24267355464141077,
+      "grad_norm": 0.2100614309310913,
+      "learning_rate": 0.001,
+      "loss": 1.253,
+      "step": 7610
+    },
+    {
+      "epoch": 0.24299244236104467,
+      "grad_norm": 0.22176362574100494,
+      "learning_rate": 0.001,
+      "loss": 1.2605,
+      "step": 7620
+    },
+    {
+      "epoch": 0.2433113300806786,
+      "grad_norm": 0.2098623365163803,
+      "learning_rate": 0.001,
+      "loss": 1.2823,
+      "step": 7630
+    },
+    {
+      "epoch": 0.24363021780031252,
+      "grad_norm": 0.2050837129354477,
+      "learning_rate": 0.001,
+      "loss": 1.3001,
+      "step": 7640
+    },
+    {
+      "epoch": 0.24394910551994642,
+      "grad_norm": 0.20297372341156006,
+      "learning_rate": 0.001,
+      "loss": 1.2672,
+      "step": 7650
+    },
+    {
+      "epoch": 0.24426799323958034,
+      "grad_norm": 0.19953450560569763,
+      "learning_rate": 0.001,
+      "loss": 1.2662,
+      "step": 7660
+    },
+    {
+      "epoch": 0.24458688095921427,
+      "grad_norm": 0.20888438820838928,
+      "learning_rate": 0.001,
+      "loss": 1.2693,
+      "step": 7670
+    },
+    {
+      "epoch": 0.24490576867884817,
+      "grad_norm": 0.20096510648727417,
+      "learning_rate": 0.001,
+      "loss": 1.2601,
+      "step": 7680
+    },
+    {
+      "epoch": 0.2452246563984821,
+      "grad_norm": 0.21057695150375366,
+      "learning_rate": 0.001,
+      "loss": 1.2813,
+      "step": 7690
+    },
+    {
+      "epoch": 0.24554354411811602,
+      "grad_norm": 0.2141493558883667,
+      "learning_rate": 0.001,
+      "loss": 1.2634,
+      "step": 7700
+    },
+    {
+      "epoch": 0.24586243183774992,
+      "grad_norm": 0.20686107873916626,
+      "learning_rate": 0.001,
+      "loss": 1.2446,
+      "step": 7710
+    },
+    {
+      "epoch": 0.24618131955738384,
+      "grad_norm": 0.20684966444969177,
+      "learning_rate": 0.001,
+      "loss": 1.2648,
+      "step": 7720
+    },
+    {
+      "epoch": 0.24650020727701777,
+      "grad_norm": 0.2081177532672882,
+      "learning_rate": 0.001,
+      "loss": 1.2761,
+      "step": 7730
+    },
+    {
+      "epoch": 0.24681909499665167,
+      "grad_norm": 0.2092471718788147,
+      "learning_rate": 0.001,
+      "loss": 1.2679,
+      "step": 7740
+    },
+    {
+      "epoch": 0.2471379827162856,
+      "grad_norm": 0.20629584789276123,
+      "learning_rate": 0.001,
+      "loss": 1.2924,
+      "step": 7750
+    },
+    {
+      "epoch": 0.24745687043591952,
+      "grad_norm": 0.20295438170433044,
+      "learning_rate": 0.001,
+      "loss": 1.262,
+      "step": 7760
+    },
+    {
+      "epoch": 0.24777575815555342,
+      "grad_norm": 0.1990422010421753,
+      "learning_rate": 0.001,
+      "loss": 1.2745,
+      "step": 7770
+    },
+    {
+      "epoch": 0.24809464587518734,
+      "grad_norm": 0.2118285447359085,
+      "learning_rate": 0.001,
+      "loss": 1.2472,
+      "step": 7780
+    },
+    {
+      "epoch": 0.24841353359482127,
+      "grad_norm": 0.2059514820575714,
+      "learning_rate": 0.001,
+      "loss": 1.281,
+      "step": 7790
+    },
+    {
+      "epoch": 0.2487324213144552,
+      "grad_norm": 0.20811574161052704,
+      "learning_rate": 0.001,
+      "loss": 1.2628,
+      "step": 7800
+    },
+    {
+      "epoch": 0.2490513090340891,
+      "grad_norm": 0.21515397727489471,
+      "learning_rate": 0.001,
+      "loss": 1.2752,
+      "step": 7810
+    },
+    {
+      "epoch": 0.24937019675372302,
+      "grad_norm": 0.20975211262702942,
+      "learning_rate": 0.001,
+      "loss": 1.2606,
+      "step": 7820
+    },
+    {
+      "epoch": 0.24968908447335694,
+      "grad_norm": 0.19748273491859436,
+      "learning_rate": 0.001,
+      "loss": 1.2481,
+      "step": 7830
+    },
+    {
+      "epoch": 0.25000797219299087,
+      "grad_norm": 0.20883125066757202,
+      "learning_rate": 0.001,
+      "loss": 1.252,
+      "step": 7840
+    },
+    {
+      "epoch": 0.25032685991262477,
+      "grad_norm": 0.1997639238834381,
+      "learning_rate": 0.001,
+      "loss": 1.2382,
+      "step": 7850
+    },
+    {
+      "epoch": 0.25064574763225866,
+      "grad_norm": 0.20731328427791595,
+      "learning_rate": 0.001,
+      "loss": 1.2407,
+      "step": 7860
+    },
+    {
+      "epoch": 0.2509646353518926,
+      "grad_norm": 0.20873534679412842,
+      "learning_rate": 0.001,
+      "loss": 1.2475,
+      "step": 7870
+    },
+    {
+      "epoch": 0.2512835230715265,
+      "grad_norm": 0.20659437775611877,
+      "learning_rate": 0.001,
+      "loss": 1.2505,
+      "step": 7880
+    },
+    {
+      "epoch": 0.2516024107911604,
+      "grad_norm": 0.20546947419643402,
+      "learning_rate": 0.001,
+      "loss": 1.2891,
+      "step": 7890
+    },
+    {
+      "epoch": 0.25192129851079437,
+      "grad_norm": 0.20524942874908447,
+      "learning_rate": 0.001,
+      "loss": 1.2644,
+      "step": 7900
+    },
+    {
+      "epoch": 0.25224018623042826,
+      "grad_norm": 0.2121630758047104,
+      "learning_rate": 0.001,
+      "loss": 1.2607,
+      "step": 7910
+    },
+    {
+      "epoch": 0.25255907395006216,
+      "grad_norm": 0.20716558396816254,
+      "learning_rate": 0.001,
+      "loss": 1.2671,
+      "step": 7920
+    },
+    {
+      "epoch": 0.2528779616696961,
+      "grad_norm": 0.21428154408931732,
+      "learning_rate": 0.001,
+      "loss": 1.2506,
+      "step": 7930
+    },
+    {
+      "epoch": 0.25319684938933,
+      "grad_norm": 0.202268585562706,
+      "learning_rate": 0.001,
+      "loss": 1.2561,
+      "step": 7940
+    },
+    {
+      "epoch": 0.2535157371089639,
+      "grad_norm": 0.20032866299152374,
+      "learning_rate": 0.001,
+      "loss": 1.2575,
+      "step": 7950
+    },
+    {
+      "epoch": 0.25383462482859787,
+      "grad_norm": 0.2026292085647583,
+      "learning_rate": 0.001,
+      "loss": 1.2714,
+      "step": 7960
+    },
+    {
+      "epoch": 0.25415351254823176,
+      "grad_norm": 0.20408962666988373,
+      "learning_rate": 0.001,
+      "loss": 1.2591,
+      "step": 7970
+    },
+    {
+      "epoch": 0.25447240026786566,
+      "grad_norm": 0.2172502875328064,
+      "learning_rate": 0.001,
+      "loss": 1.2891,
+      "step": 7980
+    },
+    {
+      "epoch": 0.2547912879874996,
+      "grad_norm": 0.20545944571495056,
+      "learning_rate": 0.001,
+      "loss": 1.2786,
+      "step": 7990
+    },
+    {
+      "epoch": 0.2551101757071335,
+      "grad_norm": 0.20460359752178192,
+      "learning_rate": 0.001,
+      "loss": 1.2502,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2554290634267674,
+      "grad_norm": 0.20056135952472687,
+      "learning_rate": 0.001,
+      "loss": 1.2456,
+      "step": 8010
+    },
+    {
+      "epoch": 0.25574795114640136,
+      "grad_norm": 0.20083419978618622,
+      "learning_rate": 0.001,
+      "loss": 1.2616,
+      "step": 8020
+    },
+    {
+      "epoch": 0.25606683886603526,
+      "grad_norm": 0.19361791014671326,
+      "learning_rate": 0.001,
+      "loss": 1.2442,
+      "step": 8030
+    },
+    {
+      "epoch": 0.25638572658566916,
+      "grad_norm": 0.21003025770187378,
+      "learning_rate": 0.001,
+      "loss": 1.2615,
+      "step": 8040
+    },
+    {
+      "epoch": 0.2567046143053031,
+      "grad_norm": 0.20616671442985535,
+      "learning_rate": 0.001,
+      "loss": 1.2771,
+      "step": 8050
+    },
+    {
+      "epoch": 0.257023502024937,
+      "grad_norm": 0.20462070405483246,
+      "learning_rate": 0.001,
+      "loss": 1.2612,
+      "step": 8060
+    },
+    {
+      "epoch": 0.2573423897445709,
+      "grad_norm": 0.2028227299451828,
+      "learning_rate": 0.001,
+      "loss": 1.2701,
+      "step": 8070
+    },
+    {
+      "epoch": 0.25766127746420486,
+      "grad_norm": 0.20757098495960236,
+      "learning_rate": 0.001,
+      "loss": 1.2567,
+      "step": 8080
+    },
+    {
+      "epoch": 0.25798016518383876,
+      "grad_norm": 0.20266075432300568,
+      "learning_rate": 0.001,
+      "loss": 1.2876,
+      "step": 8090
+    },
+    {
+      "epoch": 0.2582990529034727,
+      "grad_norm": 0.20029297471046448,
+      "learning_rate": 0.001,
+      "loss": 1.2747,
+      "step": 8100
+    },
+    {
+      "epoch": 0.2586179406231066,
+      "grad_norm": 0.2076231986284256,
+      "learning_rate": 0.001,
+      "loss": 1.2667,
+      "step": 8110
+    },
+    {
+      "epoch": 0.2589368283427405,
+      "grad_norm": 0.20260873436927795,
+      "learning_rate": 0.001,
+      "loss": 1.2404,
+      "step": 8120
+    },
+    {
+      "epoch": 0.25925571606237446,
+      "grad_norm": 0.20188632607460022,
+      "learning_rate": 0.001,
+      "loss": 1.2461,
+      "step": 8130
+    },
+    {
+      "epoch": 0.25957460378200836,
+      "grad_norm": 0.20284780859947205,
+      "learning_rate": 0.001,
+      "loss": 1.2333,
+      "step": 8140
+    },
+    {
+      "epoch": 0.25989349150164226,
+      "grad_norm": 0.20577070116996765,
+      "learning_rate": 0.001,
+      "loss": 1.2544,
+      "step": 8150
+    },
+    {
+      "epoch": 0.2602123792212762,
+      "grad_norm": 0.20342661440372467,
+      "learning_rate": 0.001,
+      "loss": 1.2508,
+      "step": 8160
+    },
+    {
+      "epoch": 0.2605312669409101,
+      "grad_norm": 0.20312359929084778,
+      "learning_rate": 0.001,
+      "loss": 1.2402,
+      "step": 8170
+    },
+    {
+      "epoch": 0.260850154660544,
+      "grad_norm": 0.20663923025131226,
+      "learning_rate": 0.001,
+      "loss": 1.2609,
+      "step": 8180
+    },
+    {
+      "epoch": 0.26116904238017796,
+      "grad_norm": 0.20582140982151031,
+      "learning_rate": 0.001,
+      "loss": 1.2448,
+      "step": 8190
+    },
+    {
+      "epoch": 0.26148793009981186,
+      "grad_norm": 0.20321714878082275,
+      "learning_rate": 0.001,
+      "loss": 1.2531,
+      "step": 8200
+    },
+    {
+      "epoch": 0.26180681781944576,
+      "grad_norm": 0.1960032731294632,
+      "learning_rate": 0.001,
+      "loss": 1.273,
+      "step": 8210
+    },
+    {
+      "epoch": 0.2621257055390797,
+      "grad_norm": 0.20806966722011566,
+      "learning_rate": 0.001,
+      "loss": 1.2685,
+      "step": 8220
+    },
+    {
+      "epoch": 0.2624445932587136,
+      "grad_norm": 0.1973854899406433,
+      "learning_rate": 0.001,
+      "loss": 1.2501,
+      "step": 8230
+    },
+    {
+      "epoch": 0.2627634809783475,
+      "grad_norm": 0.19955401122570038,
+      "learning_rate": 0.001,
+      "loss": 1.2412,
+      "step": 8240
+    },
+    {
+      "epoch": 0.26308236869798146,
+      "grad_norm": 0.19847828149795532,
+      "learning_rate": 0.001,
+      "loss": 1.2345,
+      "step": 8250
+    },
+    {
+      "epoch": 0.26340125641761536,
+      "grad_norm": 0.20189815759658813,
+      "learning_rate": 0.001,
+      "loss": 1.2618,
+      "step": 8260
+    },
+    {
+      "epoch": 0.26372014413724926,
+      "grad_norm": 0.20722807943820953,
+      "learning_rate": 0.001,
+      "loss": 1.2596,
+      "step": 8270
+    },
+    {
+      "epoch": 0.2640390318568832,
+      "grad_norm": 0.20148128271102905,
+      "learning_rate": 0.001,
+      "loss": 1.2566,
+      "step": 8280
+    },
+    {
+      "epoch": 0.2643579195765171,
+      "grad_norm": 0.21168474853038788,
+      "learning_rate": 0.001,
+      "loss": 1.2536,
+      "step": 8290
+    },
+    {
+      "epoch": 0.264676807296151,
+      "grad_norm": 0.2093587964773178,
+      "learning_rate": 0.001,
+      "loss": 1.2445,
+      "step": 8300
+    },
+    {
+      "epoch": 0.26499569501578496,
+      "grad_norm": 0.19885730743408203,
+      "learning_rate": 0.001,
+      "loss": 1.2429,
+      "step": 8310
+    },
+    {
+      "epoch": 0.26531458273541886,
+      "grad_norm": 0.18978652358055115,
+      "learning_rate": 0.001,
+      "loss": 1.2362,
+      "step": 8320
+    },
+    {
+      "epoch": 0.26563347045505276,
+      "grad_norm": 0.20467509329319,
+      "learning_rate": 0.001,
+      "loss": 1.2345,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2659523581746867,
+      "grad_norm": 0.20757165551185608,
+      "learning_rate": 0.001,
+      "loss": 1.2598,
+      "step": 8340
+    },
+    {
+      "epoch": 0.2662712458943206,
+      "grad_norm": 0.21299956738948822,
+      "learning_rate": 0.001,
+      "loss": 1.2489,
+      "step": 8350
+    },
+    {
+      "epoch": 0.2665901336139545,
+      "grad_norm": 0.19811475276947021,
+      "learning_rate": 0.001,
+      "loss": 1.2773,
+      "step": 8360
+    },
+    {
+      "epoch": 0.26690902133358846,
+      "grad_norm": 0.2071559578180313,
+      "learning_rate": 0.001,
+      "loss": 1.2648,
+      "step": 8370
+    },
+    {
+      "epoch": 0.26722790905322236,
+      "grad_norm": 0.19420473277568817,
+      "learning_rate": 0.001,
+      "loss": 1.2542,
+      "step": 8380
+    },
+    {
+      "epoch": 0.26754679677285625,
+      "grad_norm": 0.20581373572349548,
+      "learning_rate": 0.001,
+      "loss": 1.2541,
+      "step": 8390
+    },
+    {
+      "epoch": 0.2678656844924902,
+      "grad_norm": 0.20274776220321655,
+      "learning_rate": 0.001,
+      "loss": 1.2484,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2681845722121241,
+      "grad_norm": 0.19483987987041473,
+      "learning_rate": 0.001,
+      "loss": 1.2419,
+      "step": 8410
+    },
+    {
+      "epoch": 0.268503459931758,
+      "grad_norm": 0.19840216636657715,
+      "learning_rate": 0.001,
+      "loss": 1.2419,
+      "step": 8420
+    },
+    {
+      "epoch": 0.26882234765139196,
+      "grad_norm": 0.19197873771190643,
+      "learning_rate": 0.001,
+      "loss": 1.2376,
+      "step": 8430
+    },
+    {
+      "epoch": 0.26914123537102586,
+      "grad_norm": 0.21160933375358582,
+      "learning_rate": 0.001,
+      "loss": 1.237,
+      "step": 8440
+    },
+    {
+      "epoch": 0.26946012309065975,
+      "grad_norm": 0.1986304521560669,
+      "learning_rate": 0.001,
+      "loss": 1.2527,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2697790108102937,
+      "grad_norm": 0.19879919290542603,
+      "learning_rate": 0.001,
+      "loss": 1.2396,
+      "step": 8460
+    },
+    {
+      "epoch": 0.2700978985299276,
+      "grad_norm": 0.20429867506027222,
+      "learning_rate": 0.001,
+      "loss": 1.2508,
+      "step": 8470
+    },
+    {
+      "epoch": 0.2704167862495615,
+      "grad_norm": 0.19134077429771423,
+      "learning_rate": 0.001,
+      "loss": 1.2362,
+      "step": 8480
+    },
+    {
+      "epoch": 0.27073567396919546,
+      "grad_norm": 0.20987531542778015,
+      "learning_rate": 0.001,
+      "loss": 1.2356,
+      "step": 8490
+    },
+    {
+      "epoch": 0.27105456168882935,
+      "grad_norm": 0.19630587100982666,
+      "learning_rate": 0.001,
+      "loss": 1.2669,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2713734494084633,
+      "grad_norm": 0.19786277413368225,
+      "learning_rate": 0.001,
+      "loss": 1.2348,
+      "step": 8510
+    },
+    {
+      "epoch": 0.2716923371280972,
+      "grad_norm": 0.19404275715351105,
+      "learning_rate": 0.001,
+      "loss": 1.2329,
+      "step": 8520
+    },
+    {
+      "epoch": 0.2720112248477311,
+      "grad_norm": 0.19640901684761047,
+      "learning_rate": 0.001,
+      "loss": 1.2343,
+      "step": 8530
+    },
+    {
+      "epoch": 0.27233011256736506,
+      "grad_norm": 0.19867222011089325,
+      "learning_rate": 0.001,
+      "loss": 1.2477,
+      "step": 8540
+    },
+    {
+      "epoch": 0.27264900028699895,
+      "grad_norm": 0.2037269026041031,
+      "learning_rate": 0.001,
+      "loss": 1.2681,
+      "step": 8550
+    },
+    {
+      "epoch": 0.27296788800663285,
+      "grad_norm": 0.20036159455776215,
+      "learning_rate": 0.001,
+      "loss": 1.253,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2732867757262668,
+      "grad_norm": 0.19691398739814758,
+      "learning_rate": 0.001,
+      "loss": 1.2617,
+      "step": 8570
+    },
+    {
+      "epoch": 0.2736056634459007,
+      "grad_norm": 0.19771870970726013,
+      "learning_rate": 0.001,
+      "loss": 1.2303,
+      "step": 8580
+    },
+    {
+      "epoch": 0.2739245511655346,
+      "grad_norm": 0.19885499775409698,
+      "learning_rate": 0.001,
+      "loss": 1.2328,
+      "step": 8590
+    },
+    {
+      "epoch": 0.27424343888516856,
+      "grad_norm": 0.19315415620803833,
+      "learning_rate": 0.001,
+      "loss": 1.2546,
+      "step": 8600
+    },
+    {
+      "epoch": 0.27456232660480245,
+      "grad_norm": 0.1995774805545807,
+      "learning_rate": 0.001,
+      "loss": 1.2558,
+      "step": 8610
+    },
+    {
+      "epoch": 0.27488121432443635,
+      "grad_norm": 0.1985071748495102,
+      "learning_rate": 0.001,
+      "loss": 1.2273,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2752001020440703,
+      "grad_norm": 0.1915522962808609,
+      "learning_rate": 0.001,
+      "loss": 1.234,
+      "step": 8630
+    },
+    {
+      "epoch": 0.2755189897637042,
+      "grad_norm": 0.19988441467285156,
+      "learning_rate": 0.001,
+      "loss": 1.2308,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2758378774833381,
+      "grad_norm": 0.19376257061958313,
+      "learning_rate": 0.001,
+      "loss": 1.2494,
+      "step": 8650
+    },
+    {
+      "epoch": 0.27615676520297205,
+      "grad_norm": 0.19480600953102112,
+      "learning_rate": 0.001,
+      "loss": 1.2374,
+      "step": 8660
+    },
+    {
+      "epoch": 0.27647565292260595,
+      "grad_norm": 0.2003336101770401,
+      "learning_rate": 0.001,
+      "loss": 1.24,
+      "step": 8670
+    },
+    {
+      "epoch": 0.27679454064223985,
+      "grad_norm": 0.19479194283485413,
+      "learning_rate": 0.001,
+      "loss": 1.2266,
+      "step": 8680
+    },
+    {
+      "epoch": 0.2771134283618738,
+      "grad_norm": 0.1989787369966507,
+      "learning_rate": 0.001,
+      "loss": 1.2187,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2774323160815077,
+      "grad_norm": 0.19739316403865814,
+      "learning_rate": 0.001,
+      "loss": 1.2395,
+      "step": 8700
+    },
+    {
+      "epoch": 0.2777512038011416,
+      "grad_norm": 0.19161522388458252,
+      "learning_rate": 0.001,
+      "loss": 1.243,
+      "step": 8710
+    },
+    {
+      "epoch": 0.27807009152077555,
+      "grad_norm": 0.1980825811624527,
+      "learning_rate": 0.001,
+      "loss": 1.2383,
+      "step": 8720
+    },
+    {
+      "epoch": 0.27838897924040945,
+      "grad_norm": 0.19861190021038055,
+      "learning_rate": 0.001,
+      "loss": 1.2142,
+      "step": 8730
+    },
+    {
+      "epoch": 0.27870786696004335,
+      "grad_norm": 0.19875994324684143,
+      "learning_rate": 0.001,
+      "loss": 1.2325,
+      "step": 8740
+    },
+    {
+      "epoch": 0.2790267546796773,
+      "grad_norm": 0.2007092535495758,
+      "learning_rate": 0.001,
+      "loss": 1.2412,
+      "step": 8750
+    },
+    {
+      "epoch": 0.2793456423993112,
+      "grad_norm": 0.19943733513355255,
+      "learning_rate": 0.001,
+      "loss": 1.2317,
+      "step": 8760
+    },
+    {
+      "epoch": 0.2796645301189451,
+      "grad_norm": 0.19409018754959106,
+      "learning_rate": 0.001,
+      "loss": 1.2405,
+      "step": 8770
+    },
+    {
+      "epoch": 0.27998341783857905,
+      "grad_norm": 0.20169158279895782,
+      "learning_rate": 0.001,
+      "loss": 1.2454,
+      "step": 8780
+    },
+    {
+      "epoch": 0.28030230555821295,
+      "grad_norm": 0.19915443658828735,
+      "learning_rate": 0.001,
+      "loss": 1.2327,
+      "step": 8790
+    },
+    {
+      "epoch": 0.28062119327784685,
+      "grad_norm": 0.19803552329540253,
+      "learning_rate": 0.001,
+      "loss": 1.2259,
+      "step": 8800
+    },
+    {
+      "epoch": 0.2809400809974808,
+      "grad_norm": 0.19060640037059784,
+      "learning_rate": 0.001,
+      "loss": 1.2495,
+      "step": 8810
+    },
+    {
+      "epoch": 0.2812589687171147,
+      "grad_norm": 0.19586661458015442,
+      "learning_rate": 0.001,
+      "loss": 1.2341,
+      "step": 8820
+    },
+    {
+      "epoch": 0.2815778564367486,
+      "grad_norm": 0.2034222036600113,
+      "learning_rate": 0.001,
+      "loss": 1.2595,
+      "step": 8830
+    },
+    {
+      "epoch": 0.28189674415638255,
+      "grad_norm": 0.1960514336824417,
+      "learning_rate": 0.001,
+      "loss": 1.2333,
+      "step": 8840
+    },
+    {
+      "epoch": 0.28221563187601645,
+      "grad_norm": 0.2002130150794983,
+      "learning_rate": 0.001,
+      "loss": 1.2387,
+      "step": 8850
+    },
+    {
+      "epoch": 0.28253451959565035,
+      "grad_norm": 0.1940252184867859,
+      "learning_rate": 0.001,
+      "loss": 1.2403,
+      "step": 8860
+    },
+    {
+      "epoch": 0.2828534073152843,
+      "grad_norm": 0.1985214501619339,
+      "learning_rate": 0.001,
+      "loss": 1.2211,
+      "step": 8870
+    },
+    {
+      "epoch": 0.2831722950349182,
+      "grad_norm": 0.1938486248254776,
+      "learning_rate": 0.001,
+      "loss": 1.2421,
+      "step": 8880
+    },
+    {
+      "epoch": 0.2834911827545521,
+      "grad_norm": 0.1903192102909088,
+      "learning_rate": 0.001,
+      "loss": 1.2221,
+      "step": 8890
+    },
+    {
+      "epoch": 0.28381007047418605,
+      "grad_norm": 0.19713927805423737,
+      "learning_rate": 0.001,
+      "loss": 1.245,
+      "step": 8900
+    },
+    {
+      "epoch": 0.28412895819381995,
+      "grad_norm": 0.2070610523223877,
+      "learning_rate": 0.001,
+      "loss": 1.2296,
+      "step": 8910
+    },
+    {
+      "epoch": 0.2844478459134539,
+      "grad_norm": 0.19895747303962708,
+      "learning_rate": 0.001,
+      "loss": 1.2276,
+      "step": 8920
+    },
+    {
+      "epoch": 0.2847667336330878,
+      "grad_norm": 0.1950913965702057,
+      "learning_rate": 0.001,
+      "loss": 1.239,
+      "step": 8930
+    },
+    {
+      "epoch": 0.2850856213527217,
+      "grad_norm": 0.20076586306095123,
+      "learning_rate": 0.001,
+      "loss": 1.2338,
+      "step": 8940
+    },
+    {
+      "epoch": 0.28540450907235565,
+      "grad_norm": 0.1983807235956192,
+      "learning_rate": 0.001,
+      "loss": 1.2431,
+      "step": 8950
+    },
+    {
+      "epoch": 0.28572339679198955,
+      "grad_norm": 0.1993577629327774,
+      "learning_rate": 0.001,
+      "loss": 1.2283,
+      "step": 8960
+    },
+    {
+      "epoch": 0.28604228451162345,
+      "grad_norm": 0.19628196954727173,
+      "learning_rate": 0.001,
+      "loss": 1.2281,
+      "step": 8970
+    },
+    {
+      "epoch": 0.2863611722312574,
+      "grad_norm": 0.19559252262115479,
+      "learning_rate": 0.001,
+      "loss": 1.2316,
+      "step": 8980
+    },
+    {
+      "epoch": 0.2866800599508913,
+      "grad_norm": 0.1995292752981186,
+      "learning_rate": 0.001,
+      "loss": 1.2321,
+      "step": 8990
+    },
+    {
+      "epoch": 0.2869989476705252,
+      "grad_norm": 0.19853276014328003,
+      "learning_rate": 0.001,
+      "loss": 1.2576,
+      "step": 9000
+    },
+    {
+      "epoch": 0.28731783539015915,
+      "grad_norm": 0.19437505304813385,
+      "learning_rate": 0.001,
+      "loss": 1.2182,
+      "step": 9010
+    },
+    {
+      "epoch": 0.28763672310979305,
+      "grad_norm": 0.19659703969955444,
+      "learning_rate": 0.001,
+      "loss": 1.2313,
+      "step": 9020
+    },
+    {
+      "epoch": 0.28795561082942694,
+      "grad_norm": 0.19425907731056213,
+      "learning_rate": 0.001,
+      "loss": 1.2316,
+      "step": 9030
+    },
+    {
+      "epoch": 0.2882744985490609,
+      "grad_norm": 0.19886291027069092,
+      "learning_rate": 0.001,
+      "loss": 1.2373,
+      "step": 9040
+    },
+    {
+      "epoch": 0.2885933862686948,
+      "grad_norm": 0.18857437372207642,
+      "learning_rate": 0.001,
+      "loss": 1.2226,
+      "step": 9050
+    },
+    {
+      "epoch": 0.2889122739883287,
+      "grad_norm": 0.1934089958667755,
+      "learning_rate": 0.001,
+      "loss": 1.2191,
+      "step": 9060
+    },
+    {
+      "epoch": 0.28923116170796265,
+      "grad_norm": 0.19871141016483307,
+      "learning_rate": 0.001,
+      "loss": 1.2246,
+      "step": 9070
+    },
+    {
+      "epoch": 0.28955004942759655,
+      "grad_norm": 0.19774867594242096,
+      "learning_rate": 0.001,
+      "loss": 1.2335,
+      "step": 9080
+    },
+    {
+      "epoch": 0.28986893714723044,
+      "grad_norm": 0.19685259461402893,
+      "learning_rate": 0.001,
+      "loss": 1.2537,
+      "step": 9090
+    },
+    {
+      "epoch": 0.2901878248668644,
+      "grad_norm": 0.19641003012657166,
+      "learning_rate": 0.001,
+      "loss": 1.2202,
+      "step": 9100
+    },
+    {
+      "epoch": 0.2905067125864983,
+      "grad_norm": 0.19249022006988525,
+      "learning_rate": 0.001,
+      "loss": 1.2282,
+      "step": 9110
+    },
+    {
+      "epoch": 0.2908256003061322,
+      "grad_norm": 0.20958197116851807,
+      "learning_rate": 0.001,
+      "loss": 1.2346,
+      "step": 9120
+    },
+    {
+      "epoch": 0.29114448802576615,
+      "grad_norm": 0.1979653686285019,
+      "learning_rate": 0.001,
+      "loss": 1.2356,
+      "step": 9130
+    },
+    {
+      "epoch": 0.29146337574540004,
+      "grad_norm": 0.20007383823394775,
+      "learning_rate": 0.001,
+      "loss": 1.2219,
+      "step": 9140
+    },
+    {
+      "epoch": 0.29178226346503394,
+      "grad_norm": 0.1912803053855896,
+      "learning_rate": 0.001,
+      "loss": 1.2164,
+      "step": 9150
+    },
+    {
+      "epoch": 0.2921011511846679,
+      "grad_norm": 0.2007441520690918,
+      "learning_rate": 0.001,
+      "loss": 1.234,
+      "step": 9160
+    },
+    {
+      "epoch": 0.2924200389043018,
+      "grad_norm": 0.1962765008211136,
+      "learning_rate": 0.001,
+      "loss": 1.2397,
+      "step": 9170
+    },
+    {
+      "epoch": 0.2927389266239357,
+      "grad_norm": 0.19007021188735962,
+      "learning_rate": 0.001,
+      "loss": 1.2328,
+      "step": 9180
+    },
+    {
+      "epoch": 0.29305781434356964,
+      "grad_norm": 0.19233642518520355,
+      "learning_rate": 0.001,
+      "loss": 1.2173,
+      "step": 9190
+    },
+    {
+      "epoch": 0.29337670206320354,
+      "grad_norm": 0.19066978991031647,
+      "learning_rate": 0.001,
+      "loss": 1.2226,
+      "step": 9200
+    },
+    {
+      "epoch": 0.29369558978283744,
+      "grad_norm": 0.19215360283851624,
+      "learning_rate": 0.001,
+      "loss": 1.2153,
+      "step": 9210
+    },
+    {
+      "epoch": 0.2940144775024714,
+      "grad_norm": 0.1964012086391449,
+      "learning_rate": 0.001,
+      "loss": 1.2221,
+      "step": 9220
+    },
+    {
+      "epoch": 0.2943333652221053,
+      "grad_norm": 0.196467325091362,
+      "learning_rate": 0.001,
+      "loss": 1.2316,
+      "step": 9230
+    },
+    {
+      "epoch": 0.2946522529417392,
+      "grad_norm": 0.19289752840995789,
+      "learning_rate": 0.001,
+      "loss": 1.21,
+      "step": 9240
+    },
+    {
+      "epoch": 0.29497114066137314,
+      "grad_norm": 0.19339169561862946,
+      "learning_rate": 0.001,
+      "loss": 1.2243,
+      "step": 9250
+    },
+    {
+      "epoch": 0.29529002838100704,
+      "grad_norm": 0.19633044302463531,
+      "learning_rate": 0.001,
+      "loss": 1.2165,
+      "step": 9260
+    },
+    {
+      "epoch": 0.29560891610064094,
+      "grad_norm": 0.1902557909488678,
+      "learning_rate": 0.001,
+      "loss": 1.2105,
+      "step": 9270
+    },
+    {
+      "epoch": 0.2959278038202749,
+      "grad_norm": 0.1978464275598526,
+      "learning_rate": 0.001,
+      "loss": 1.221,
+      "step": 9280
+    },
+    {
+      "epoch": 0.2962466915399088,
+      "grad_norm": 0.1890365481376648,
+      "learning_rate": 0.001,
+      "loss": 1.2059,
+      "step": 9290
+    },
+    {
+      "epoch": 0.2965655792595427,
+      "grad_norm": 0.1897841989994049,
+      "learning_rate": 0.001,
+      "loss": 1.2232,
+      "step": 9300
+    },
+    {
+      "epoch": 0.29688446697917664,
+      "grad_norm": 0.19626909494400024,
+      "learning_rate": 0.001,
+      "loss": 1.2306,
+      "step": 9310
+    },
+    {
+      "epoch": 0.29720335469881054,
+      "grad_norm": 0.19002099335193634,
+      "learning_rate": 0.001,
+      "loss": 1.2019,
+      "step": 9320
+    },
+    {
+      "epoch": 0.29752224241844444,
+      "grad_norm": 0.18954624235630035,
+      "learning_rate": 0.001,
+      "loss": 1.2302,
+      "step": 9330
+    },
+    {
+      "epoch": 0.2978411301380784,
+      "grad_norm": 0.1903507560491562,
+      "learning_rate": 0.001,
+      "loss": 1.2088,
+      "step": 9340
+    },
+    {
+      "epoch": 0.2981600178577123,
+      "grad_norm": 0.1922921985387802,
+      "learning_rate": 0.001,
+      "loss": 1.215,
+      "step": 9350
+    },
+    {
+      "epoch": 0.29847890557734624,
+      "grad_norm": 0.20056477189064026,
+      "learning_rate": 0.001,
+      "loss": 1.2221,
+      "step": 9360
+    },
+    {
+      "epoch": 0.29879779329698014,
+      "grad_norm": 0.18921448290348053,
+      "learning_rate": 0.001,
+      "loss": 1.2233,
+      "step": 9370
+    },
+    {
+      "epoch": 0.29911668101661404,
+      "grad_norm": 0.1903228610754013,
+      "learning_rate": 0.001,
+      "loss": 1.2194,
+      "step": 9380
+    },
+    {
+      "epoch": 0.299435568736248,
+      "grad_norm": 0.19145965576171875,
+      "learning_rate": 0.001,
+      "loss": 1.2065,
+      "step": 9390
+    },
+    {
+      "epoch": 0.2997544564558819,
+      "grad_norm": 0.1889374703168869,
+      "learning_rate": 0.001,
+      "loss": 1.219,
+      "step": 9400
+    },
+    {
+      "epoch": 0.3000733441755158,
+      "grad_norm": 0.18882133066654205,
+      "learning_rate": 0.001,
+      "loss": 1.2229,
+      "step": 9410
+    },
+    {
+      "epoch": 0.30039223189514974,
+      "grad_norm": 0.19275881350040436,
+      "learning_rate": 0.001,
+      "loss": 1.2212,
+      "step": 9420
+    },
+    {
+      "epoch": 0.30071111961478364,
+      "grad_norm": 0.19594255089759827,
+      "learning_rate": 0.001,
+      "loss": 1.211,
+      "step": 9430
+    },
+    {
+      "epoch": 0.30103000733441754,
+      "grad_norm": 0.19568689167499542,
+      "learning_rate": 0.001,
+      "loss": 1.2228,
+      "step": 9440
+    },
+    {
+      "epoch": 0.3013488950540515,
+      "grad_norm": 0.18956463038921356,
+      "learning_rate": 0.001,
+      "loss": 1.2271,
+      "step": 9450
+    },
+    {
+      "epoch": 0.3016677827736854,
+      "grad_norm": 0.19660969078540802,
+      "learning_rate": 0.001,
+      "loss": 1.2256,
+      "step": 9460
+    },
+    {
+      "epoch": 0.3019866704933193,
+      "grad_norm": 0.1920958012342453,
+      "learning_rate": 0.001,
+      "loss": 1.2402,
+      "step": 9470
+    },
+    {
+      "epoch": 0.30230555821295324,
+      "grad_norm": 0.19480574131011963,
+      "learning_rate": 0.001,
+      "loss": 1.2192,
+      "step": 9480
+    },
+    {
+      "epoch": 0.30262444593258714,
+      "grad_norm": 0.1968194842338562,
+      "learning_rate": 0.001,
+      "loss": 1.2208,
+      "step": 9490
+    },
+    {
+      "epoch": 0.30294333365222104,
+      "grad_norm": 0.19631661474704742,
+      "learning_rate": 0.001,
+      "loss": 1.2202,
+      "step": 9500
+    },
+    {
+      "epoch": 0.303262221371855,
+      "grad_norm": 0.19253483414649963,
+      "learning_rate": 0.001,
+      "loss": 1.208,
+      "step": 9510
+    },
+    {
+      "epoch": 0.3035811090914889,
+      "grad_norm": 0.19759908318519592,
+      "learning_rate": 0.001,
+      "loss": 1.2177,
+      "step": 9520
+    },
+    {
+      "epoch": 0.3038999968111228,
+      "grad_norm": 0.1910729557275772,
+      "learning_rate": 0.001,
+      "loss": 1.1982,
+      "step": 9530
+    },
+    {
+      "epoch": 0.30421888453075674,
+      "grad_norm": 0.19392292201519012,
+      "learning_rate": 0.001,
+      "loss": 1.2137,
+      "step": 9540
+    },
+    {
+      "epoch": 0.30453777225039064,
+      "grad_norm": 0.19059373438358307,
+      "learning_rate": 0.001,
+      "loss": 1.1914,
+      "step": 9550
+    },
+    {
+      "epoch": 0.30485665997002453,
+      "grad_norm": 0.18858584761619568,
+      "learning_rate": 0.001,
+      "loss": 1.2147,
+      "step": 9560
+    },
+    {
+      "epoch": 0.3051755476896585,
+      "grad_norm": 0.19062539935112,
+      "learning_rate": 0.001,
+      "loss": 1.2092,
+      "step": 9570
+    },
+    {
+      "epoch": 0.3054944354092924,
+      "grad_norm": 0.19087275862693787,
+      "learning_rate": 0.001,
+      "loss": 1.2288,
+      "step": 9580
+    },
+    {
+      "epoch": 0.3058133231289263,
+      "grad_norm": 0.18619580566883087,
+      "learning_rate": 0.001,
+      "loss": 1.2138,
+      "step": 9590
+    },
+    {
+      "epoch": 0.30613221084856024,
+      "grad_norm": 0.1980937123298645,
+      "learning_rate": 0.001,
+      "loss": 1.2036,
+      "step": 9600
+    },
+    {
+      "epoch": 0.30645109856819414,
+      "grad_norm": 0.1878875494003296,
+      "learning_rate": 0.001,
+      "loss": 1.2163,
+      "step": 9610
+    },
+    {
+      "epoch": 0.30676998628782803,
+      "grad_norm": 0.1916397362947464,
+      "learning_rate": 0.001,
+      "loss": 1.2165,
+      "step": 9620
+    },
+    {
+      "epoch": 0.307088874007462,
+      "grad_norm": 0.19258850812911987,
+      "learning_rate": 0.001,
+      "loss": 1.1858,
+      "step": 9630
+    },
+    {
+      "epoch": 0.3074077617270959,
+      "grad_norm": 0.18846213817596436,
+      "learning_rate": 0.001,
+      "loss": 1.2184,
+      "step": 9640
+    },
+    {
+      "epoch": 0.3077266494467298,
+      "grad_norm": 0.18937738239765167,
+      "learning_rate": 0.001,
+      "loss": 1.2119,
+      "step": 9650
+    },
+    {
+      "epoch": 0.30804553716636374,
+      "grad_norm": 0.18987585604190826,
+      "learning_rate": 0.001,
+      "loss": 1.2016,
+      "step": 9660
+    },
+    {
+      "epoch": 0.30836442488599763,
+      "grad_norm": 0.19264313578605652,
+      "learning_rate": 0.001,
+      "loss": 1.2025,
+      "step": 9670
+    },
+    {
+      "epoch": 0.30868331260563153,
+      "grad_norm": 0.1826832890510559,
+      "learning_rate": 0.001,
+      "loss": 1.2143,
+      "step": 9680
+    },
+    {
+      "epoch": 0.3090022003252655,
+      "grad_norm": 0.1982695609331131,
+      "learning_rate": 0.001,
+      "loss": 1.2236,
+      "step": 9690
+    },
+    {
+      "epoch": 0.3093210880448994,
+      "grad_norm": 0.1873609572649002,
+      "learning_rate": 0.001,
+      "loss": 1.2129,
+      "step": 9700
+    },
+    {
+      "epoch": 0.3096399757645333,
+      "grad_norm": 0.1930837631225586,
+      "learning_rate": 0.001,
+      "loss": 1.2252,
+      "step": 9710
+    },
+    {
+      "epoch": 0.30995886348416724,
+      "grad_norm": 0.19828754663467407,
+      "learning_rate": 0.001,
+      "loss": 1.2227,
+      "step": 9720
+    },
+    {
+      "epoch": 0.31027775120380113,
+      "grad_norm": 0.19810664653778076,
+      "learning_rate": 0.001,
+      "loss": 1.2245,
+      "step": 9730
+    },
+    {
+      "epoch": 0.31059663892343503,
+      "grad_norm": 0.19223099946975708,
+      "learning_rate": 0.001,
+      "loss": 1.2154,
+      "step": 9740
+    },
+    {
+      "epoch": 0.310915526643069,
+      "grad_norm": 0.19165821373462677,
+      "learning_rate": 0.001,
+      "loss": 1.2004,
+      "step": 9750
+    },
+    {
+      "epoch": 0.3112344143627029,
+      "grad_norm": 0.1895214170217514,
+      "learning_rate": 0.001,
+      "loss": 1.2117,
+      "step": 9760
+    },
+    {
+      "epoch": 0.31155330208233684,
+      "grad_norm": 0.19103483855724335,
+      "learning_rate": 0.001,
+      "loss": 1.2319,
+      "step": 9770
+    },
+    {
+      "epoch": 0.31187218980197073,
+      "grad_norm": 0.18547318875789642,
+      "learning_rate": 0.001,
+      "loss": 1.1935,
+      "step": 9780
+    },
+    {
+      "epoch": 0.31219107752160463,
+      "grad_norm": 0.18510322272777557,
+      "learning_rate": 0.001,
+      "loss": 1.2051,
+      "step": 9790
+    },
+    {
+      "epoch": 0.3125099652412386,
+      "grad_norm": 0.18715843558311462,
+      "learning_rate": 0.001,
+      "loss": 1.2166,
+      "step": 9800
+    },
+    {
+      "epoch": 0.3128288529608725,
+      "grad_norm": 0.18825583159923553,
+      "learning_rate": 0.001,
+      "loss": 1.2168,
+      "step": 9810
+    },
+    {
+      "epoch": 0.3131477406805064,
+      "grad_norm": 0.1959000676870346,
+      "learning_rate": 0.001,
+      "loss": 1.2143,
+      "step": 9820
+    },
+    {
+      "epoch": 0.31346662840014033,
+      "grad_norm": 0.1898072212934494,
+      "learning_rate": 0.001,
+      "loss": 1.2043,
+      "step": 9830
+    },
+    {
+      "epoch": 0.31378551611977423,
+      "grad_norm": 0.18969126045703888,
+      "learning_rate": 0.001,
+      "loss": 1.2081,
+      "step": 9840
+    },
+    {
+      "epoch": 0.31410440383940813,
+      "grad_norm": 0.1923128217458725,
+      "learning_rate": 0.001,
+      "loss": 1.1894,
+      "step": 9850
+    },
+    {
+      "epoch": 0.3144232915590421,
+      "grad_norm": 0.18702685832977295,
+      "learning_rate": 0.001,
+      "loss": 1.2191,
+      "step": 9860
+    },
+    {
+      "epoch": 0.314742179278676,
+      "grad_norm": 0.18185706436634064,
+      "learning_rate": 0.001,
+      "loss": 1.2183,
+      "step": 9870
+    },
+    {
+      "epoch": 0.3150610669983099,
+      "grad_norm": 0.1877908557653427,
+      "learning_rate": 0.001,
+      "loss": 1.2166,
+      "step": 9880
+    },
+    {
+      "epoch": 0.31537995471794383,
+      "grad_norm": 0.19115397334098816,
+      "learning_rate": 0.001,
+      "loss": 1.1977,
+      "step": 9890
+    },
+    {
+      "epoch": 0.31569884243757773,
+      "grad_norm": 0.18303260207176208,
+      "learning_rate": 0.001,
+      "loss": 1.2028,
+      "step": 9900
+    },
+    {
+      "epoch": 0.31601773015721163,
+      "grad_norm": 0.18337474763393402,
+      "learning_rate": 0.001,
+      "loss": 1.1928,
+      "step": 9910
+    },
+    {
+      "epoch": 0.3163366178768456,
+      "grad_norm": 0.19313004612922668,
+      "learning_rate": 0.001,
+      "loss": 1.1989,
+      "step": 9920
+    },
+    {
+      "epoch": 0.3166555055964795,
+      "grad_norm": 0.19004032015800476,
+      "learning_rate": 0.001,
+      "loss": 1.2133,
+      "step": 9930
+    },
+    {
+      "epoch": 0.3169743933161134,
+      "grad_norm": 0.1894432008266449,
+      "learning_rate": 0.001,
+      "loss": 1.2166,
+      "step": 9940
+    },
+    {
+      "epoch": 0.31729328103574733,
+      "grad_norm": 0.19445645809173584,
+      "learning_rate": 0.001,
+      "loss": 1.2175,
+      "step": 9950
+    },
+    {
+      "epoch": 0.31761216875538123,
+      "grad_norm": 0.18664146959781647,
+      "learning_rate": 0.001,
+      "loss": 1.2049,
+      "step": 9960
+    },
+    {
+      "epoch": 0.31793105647501513,
+      "grad_norm": 0.1890193223953247,
+      "learning_rate": 0.001,
+      "loss": 1.223,
+      "step": 9970
+    },
+    {
+      "epoch": 0.3182499441946491,
+      "grad_norm": 0.18643978238105774,
+      "learning_rate": 0.001,
+      "loss": 1.2026,
+      "step": 9980
+    },
+    {
+      "epoch": 0.318568831914283,
+      "grad_norm": 0.18719391524791718,
+      "learning_rate": 0.001,
+      "loss": 1.2092,
+      "step": 9990
+    },
+    {
+      "epoch": 0.3188877196339169,
+      "grad_norm": 0.18894146382808685,
+      "learning_rate": 0.001,
+      "loss": 1.211,
+      "step": 10000
+    },
+    {
+      "epoch": 0.31920660735355083,
+      "grad_norm": 0.19064651429653168,
+      "learning_rate": 0.001,
+      "loss": 1.2126,
+      "step": 10010
+    },
+    {
+      "epoch": 0.31952549507318473,
+      "grad_norm": 0.19338785111904144,
+      "learning_rate": 0.001,
+      "loss": 1.1982,
+      "step": 10020
+    },
+    {
+      "epoch": 0.3198443827928186,
+      "grad_norm": 0.19196036458015442,
+      "learning_rate": 0.001,
+      "loss": 1.2046,
+      "step": 10030
+    },
+    {
+      "epoch": 0.3201632705124526,
+      "grad_norm": 0.1986386626958847,
+      "learning_rate": 0.001,
+      "loss": 1.2066,
+      "step": 10040
+    },
+    {
+      "epoch": 0.3204821582320865,
+      "grad_norm": 0.20869721472263336,
+      "learning_rate": 0.001,
+      "loss": 1.226,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3208010459517204,
+      "grad_norm": 0.18814514577388763,
+      "learning_rate": 0.001,
+      "loss": 1.209,
+      "step": 10060
+    },
+    {
+      "epoch": 0.32111993367135433,
+      "grad_norm": 0.19529877603054047,
+      "learning_rate": 0.001,
+      "loss": 1.1918,
+      "step": 10070
+    },
+    {
+      "epoch": 0.3214388213909882,
+      "grad_norm": 0.1876852661371231,
+      "learning_rate": 0.001,
+      "loss": 1.2105,
+      "step": 10080
+    },
+    {
+      "epoch": 0.3217577091106221,
+      "grad_norm": 0.18959972262382507,
+      "learning_rate": 0.001,
+      "loss": 1.205,
+      "step": 10090
+    },
+    {
+      "epoch": 0.3220765968302561,
+      "grad_norm": 0.19066168367862701,
+      "learning_rate": 0.001,
+      "loss": 1.2123,
+      "step": 10100
+    },
+    {
+      "epoch": 0.32239548454989,
+      "grad_norm": 0.18713495135307312,
+      "learning_rate": 0.001,
+      "loss": 1.2129,
+      "step": 10110
+    },
+    {
+      "epoch": 0.3227143722695239,
+      "grad_norm": 0.18978673219680786,
+      "learning_rate": 0.001,
+      "loss": 1.185,
+      "step": 10120
+    },
+    {
+      "epoch": 0.32303325998915783,
+      "grad_norm": 0.19576674699783325,
+      "learning_rate": 0.001,
+      "loss": 1.1905,
+      "step": 10130
+    },
+    {
+      "epoch": 0.3233521477087917,
+      "grad_norm": 0.18363942205905914,
+      "learning_rate": 0.001,
+      "loss": 1.1981,
+      "step": 10140
+    },
+    {
+      "epoch": 0.3236710354284256,
+      "grad_norm": 0.19066651165485382,
+      "learning_rate": 0.001,
+      "loss": 1.2113,
+      "step": 10150
+    },
+    {
+      "epoch": 0.3239899231480596,
+      "grad_norm": 0.1898638755083084,
+      "learning_rate": 0.001,
+      "loss": 1.1983,
+      "step": 10160
+    },
+    {
+      "epoch": 0.3243088108676935,
+      "grad_norm": 0.18936438858509064,
+      "learning_rate": 0.001,
+      "loss": 1.1957,
+      "step": 10170
+    },
+    {
+      "epoch": 0.32462769858732743,
+      "grad_norm": 0.18695956468582153,
+      "learning_rate": 0.001,
+      "loss": 1.1932,
+      "step": 10180
+    },
+    {
+      "epoch": 0.3249465863069613,
+      "grad_norm": 0.19916784763336182,
+      "learning_rate": 0.001,
+      "loss": 1.188,
+      "step": 10190
+    },
+    {
+      "epoch": 0.3252654740265952,
+      "grad_norm": 0.1864575296640396,
+      "learning_rate": 0.001,
+      "loss": 1.1994,
+      "step": 10200
+    },
+    {
+      "epoch": 0.3255843617462292,
+      "grad_norm": 0.18736214935779572,
+      "learning_rate": 0.001,
+      "loss": 1.197,
+      "step": 10210
+    },
+    {
+      "epoch": 0.3259032494658631,
+      "grad_norm": 0.18586505949497223,
+      "learning_rate": 0.001,
+      "loss": 1.2005,
+      "step": 10220
+    },
+    {
+      "epoch": 0.326222137185497,
+      "grad_norm": 0.18664199113845825,
+      "learning_rate": 0.001,
+      "loss": 1.2323,
+      "step": 10230
+    },
+    {
+      "epoch": 0.3265410249051309,
+      "grad_norm": 0.18509739637374878,
+      "learning_rate": 0.001,
+      "loss": 1.2228,
+      "step": 10240
+    },
+    {
+      "epoch": 0.3268599126247648,
+      "grad_norm": 0.18879742920398712,
+      "learning_rate": 0.001,
+      "loss": 1.18,
+      "step": 10250
+    },
+    {
+      "epoch": 0.3271788003443987,
+      "grad_norm": 0.19517406821250916,
+      "learning_rate": 0.001,
+      "loss": 1.2042,
+      "step": 10260
+    },
+    {
+      "epoch": 0.3274976880640327,
+      "grad_norm": 0.18386130034923553,
+      "learning_rate": 0.001,
+      "loss": 1.2029,
+      "step": 10270
+    },
+    {
+      "epoch": 0.3278165757836666,
+      "grad_norm": 0.19367419183254242,
+      "learning_rate": 0.001,
+      "loss": 1.2048,
+      "step": 10280
+    },
+    {
+      "epoch": 0.3281354635033005,
+      "grad_norm": 0.18701566755771637,
+      "learning_rate": 0.001,
+      "loss": 1.2081,
+      "step": 10290
+    },
+    {
+      "epoch": 0.3284543512229344,
+      "grad_norm": 0.18644075095653534,
+      "learning_rate": 0.001,
+      "loss": 1.1982,
+      "step": 10300
+    },
+    {
+      "epoch": 0.3287732389425683,
+      "grad_norm": 0.1888340711593628,
+      "learning_rate": 0.001,
+      "loss": 1.2174,
+      "step": 10310
+    },
+    {
+      "epoch": 0.3290921266622022,
+      "grad_norm": 0.1851629763841629,
+      "learning_rate": 0.001,
+      "loss": 1.1892,
+      "step": 10320
+    },
+    {
+      "epoch": 0.3294110143818362,
+      "grad_norm": 0.18563096225261688,
+      "learning_rate": 0.001,
+      "loss": 1.225,
+      "step": 10330
+    },
+    {
+      "epoch": 0.3297299021014701,
+      "grad_norm": 0.18963338434696198,
+      "learning_rate": 0.001,
+      "loss": 1.2111,
+      "step": 10340
+    },
+    {
+      "epoch": 0.33004878982110397,
+      "grad_norm": 0.19024492800235748,
+      "learning_rate": 0.001,
+      "loss": 1.1871,
+      "step": 10350
+    },
+    {
+      "epoch": 0.3303676775407379,
+      "grad_norm": 0.19344636797904968,
+      "learning_rate": 0.001,
+      "loss": 1.1991,
+      "step": 10360
+    },
+    {
+      "epoch": 0.3306865652603718,
+      "grad_norm": 0.18379245698451996,
+      "learning_rate": 0.001,
+      "loss": 1.2176,
+      "step": 10370
+    },
+    {
+      "epoch": 0.3310054529800057,
+      "grad_norm": 0.18924225866794586,
+      "learning_rate": 0.001,
+      "loss": 1.1919,
+      "step": 10380
+    },
+    {
+      "epoch": 0.3313243406996397,
+      "grad_norm": 0.18927711248397827,
+      "learning_rate": 0.001,
+      "loss": 1.1853,
+      "step": 10390
+    },
+    {
+      "epoch": 0.3316432284192736,
+      "grad_norm": 0.18071341514587402,
+      "learning_rate": 0.001,
+      "loss": 1.2049,
+      "step": 10400
+    },
+    {
+      "epoch": 0.33196211613890747,
+      "grad_norm": 0.18121443688869476,
+      "learning_rate": 0.001,
+      "loss": 1.185,
+      "step": 10410
+    },
+    {
+      "epoch": 0.3322810038585414,
+      "grad_norm": 0.18824678659439087,
+      "learning_rate": 0.001,
+      "loss": 1.196,
+      "step": 10420
+    },
+    {
+      "epoch": 0.3325998915781753,
+      "grad_norm": 0.181623637676239,
+      "learning_rate": 0.001,
+      "loss": 1.1843,
+      "step": 10430
+    },
+    {
+      "epoch": 0.3329187792978092,
+      "grad_norm": 0.18745820224285126,
+      "learning_rate": 0.001,
+      "loss": 1.2181,
+      "step": 10440
+    },
+    {
+      "epoch": 0.3332376670174432,
+      "grad_norm": 0.18517442047595978,
+      "learning_rate": 0.001,
+      "loss": 1.1942,
+      "step": 10450
+    },
+    {
+      "epoch": 0.33355655473707707,
+      "grad_norm": 0.1849382370710373,
+      "learning_rate": 0.001,
+      "loss": 1.1822,
+      "step": 10460
+    },
+    {
+      "epoch": 0.33387544245671097,
+      "grad_norm": 0.18831832706928253,
+      "learning_rate": 0.001,
+      "loss": 1.2021,
+      "step": 10470
+    },
+    {
+      "epoch": 0.3341943301763449,
+      "grad_norm": 0.18308256566524506,
+      "learning_rate": 0.001,
+      "loss": 1.1916,
+      "step": 10480
+    },
+    {
+      "epoch": 0.3345132178959788,
+      "grad_norm": 0.17930716276168823,
+      "learning_rate": 0.001,
+      "loss": 1.1848,
+      "step": 10490
+    },
+    {
+      "epoch": 0.3348321056156127,
+      "grad_norm": 0.19042780995368958,
+      "learning_rate": 0.001,
+      "loss": 1.1918,
+      "step": 10500
+    },
+    {
+      "epoch": 0.33515099333524667,
+      "grad_norm": 0.18928125500679016,
+      "learning_rate": 0.001,
+      "loss": 1.2009,
+      "step": 10510
+    },
+    {
+      "epoch": 0.33546988105488057,
+      "grad_norm": 0.18939156830310822,
+      "learning_rate": 0.001,
+      "loss": 1.2123,
+      "step": 10520
+    },
+    {
+      "epoch": 0.33578876877451447,
+      "grad_norm": 0.18722973763942719,
+      "learning_rate": 0.001,
+      "loss": 1.1999,
+      "step": 10530
+    },
+    {
+      "epoch": 0.3361076564941484,
+      "grad_norm": 0.18385377526283264,
+      "learning_rate": 0.001,
+      "loss": 1.1748,
+      "step": 10540
+    },
+    {
+      "epoch": 0.3364265442137823,
+      "grad_norm": 0.18305833637714386,
+      "learning_rate": 0.001,
+      "loss": 1.1826,
+      "step": 10550
+    },
+    {
+      "epoch": 0.3367454319334162,
+      "grad_norm": 0.18611466884613037,
+      "learning_rate": 0.001,
+      "loss": 1.1928,
+      "step": 10560
+    },
+    {
+      "epoch": 0.33706431965305017,
+      "grad_norm": 0.19067902863025665,
+      "learning_rate": 0.001,
+      "loss": 1.1842,
+      "step": 10570
+    },
+    {
+      "epoch": 0.33738320737268407,
+      "grad_norm": 0.1923217922449112,
+      "learning_rate": 0.001,
+      "loss": 1.1901,
+      "step": 10580
+    },
+    {
+      "epoch": 0.337702095092318,
+      "grad_norm": 0.18926678597927094,
+      "learning_rate": 0.001,
+      "loss": 1.1637,
+      "step": 10590
+    },
+    {
+      "epoch": 0.3380209828119519,
+      "grad_norm": 0.19210688769817352,
+      "learning_rate": 0.001,
+      "loss": 1.1846,
+      "step": 10600
+    },
+    {
+      "epoch": 0.3383398705315858,
+      "grad_norm": 0.18888874351978302,
+      "learning_rate": 0.001,
+      "loss": 1.1892,
+      "step": 10610
+    },
+    {
+      "epoch": 0.33865875825121977,
+      "grad_norm": 0.1817292869091034,
+      "learning_rate": 0.001,
+      "loss": 1.2155,
+      "step": 10620
+    },
+    {
+      "epoch": 0.33897764597085367,
+      "grad_norm": 0.1859046071767807,
+      "learning_rate": 0.001,
+      "loss": 1.1924,
+      "step": 10630
+    },
+    {
+      "epoch": 0.33929653369048757,
+      "grad_norm": 0.19133812189102173,
+      "learning_rate": 0.001,
+      "loss": 1.1985,
+      "step": 10640
+    },
+    {
+      "epoch": 0.3396154214101215,
+      "grad_norm": 0.18890315294265747,
+      "learning_rate": 0.001,
+      "loss": 1.1963,
+      "step": 10650
+    },
+    {
+      "epoch": 0.3399343091297554,
+      "grad_norm": 0.1901807337999344,
+      "learning_rate": 0.001,
+      "loss": 1.1926,
+      "step": 10660
+    },
+    {
+      "epoch": 0.3402531968493893,
+      "grad_norm": 0.17766164243221283,
+      "learning_rate": 0.001,
+      "loss": 1.2001,
+      "step": 10670
+    },
+    {
+      "epoch": 0.34057208456902327,
+      "grad_norm": 0.18734650313854218,
+      "learning_rate": 0.001,
+      "loss": 1.2038,
+      "step": 10680
+    },
+    {
+      "epoch": 0.34089097228865717,
+      "grad_norm": 0.1862167865037918,
+      "learning_rate": 0.001,
+      "loss": 1.1942,
+      "step": 10690
+    },
+    {
+      "epoch": 0.34120986000829107,
+      "grad_norm": 0.18553335964679718,
+      "learning_rate": 0.001,
+      "loss": 1.1926,
+      "step": 10700
+    },
+    {
+      "epoch": 0.341528747727925,
+      "grad_norm": 0.18530946969985962,
+      "learning_rate": 0.001,
+      "loss": 1.1901,
+      "step": 10710
+    },
+    {
+      "epoch": 0.3418476354475589,
+      "grad_norm": 0.18436305224895477,
+      "learning_rate": 0.001,
+      "loss": 1.1963,
+      "step": 10720
+    },
+    {
+      "epoch": 0.3421665231671928,
+      "grad_norm": 0.1863793581724167,
+      "learning_rate": 0.001,
+      "loss": 1.1721,
+      "step": 10730
+    },
+    {
+      "epoch": 0.34248541088682677,
+      "grad_norm": 0.19209203124046326,
+      "learning_rate": 0.001,
+      "loss": 1.1999,
+      "step": 10740
+    },
+    {
+      "epoch": 0.34280429860646067,
+      "grad_norm": 0.182088702917099,
+      "learning_rate": 0.001,
+      "loss": 1.1859,
+      "step": 10750
+    },
+    {
+      "epoch": 0.34312318632609456,
+      "grad_norm": 0.18213383853435516,
+      "learning_rate": 0.001,
+      "loss": 1.1793,
+      "step": 10760
+    },
+    {
+      "epoch": 0.3434420740457285,
+      "grad_norm": 0.18192435801029205,
+      "learning_rate": 0.001,
+      "loss": 1.1703,
+      "step": 10770
+    },
+    {
+      "epoch": 0.3437609617653624,
+      "grad_norm": 0.1949847936630249,
+      "learning_rate": 0.001,
+      "loss": 1.1938,
+      "step": 10780
+    },
+    {
+      "epoch": 0.3440798494849963,
+      "grad_norm": 0.18721050024032593,
+      "learning_rate": 0.001,
+      "loss": 1.1833,
+      "step": 10790
+    },
+    {
+      "epoch": 0.34439873720463027,
+      "grad_norm": 0.18720798194408417,
+      "learning_rate": 0.001,
+      "loss": 1.194,
+      "step": 10800
+    },
+    {
+      "epoch": 0.34471762492426417,
+      "grad_norm": 0.18595951795578003,
+      "learning_rate": 0.001,
+      "loss": 1.2108,
+      "step": 10810
+    },
+    {
+      "epoch": 0.34503651264389806,
+      "grad_norm": 0.1828535795211792,
+      "learning_rate": 0.001,
+      "loss": 1.1993,
+      "step": 10820
+    },
+    {
+      "epoch": 0.345355400363532,
+      "grad_norm": 0.181146040558815,
+      "learning_rate": 0.001,
+      "loss": 1.1879,
+      "step": 10830
+    },
+    {
+      "epoch": 0.3456742880831659,
+      "grad_norm": 0.18184798955917358,
+      "learning_rate": 0.001,
+      "loss": 1.1761,
+      "step": 10840
+    },
+    {
+      "epoch": 0.3459931758027998,
+      "grad_norm": 0.18391500413417816,
+      "learning_rate": 0.001,
+      "loss": 1.1768,
+      "step": 10850
+    },
+    {
+      "epoch": 0.34631206352243377,
+      "grad_norm": 0.18120771646499634,
+      "learning_rate": 0.001,
+      "loss": 1.1782,
+      "step": 10860
+    },
+    {
+      "epoch": 0.34663095124206766,
+      "grad_norm": 0.18713876605033875,
+      "learning_rate": 0.001,
+      "loss": 1.194,
+      "step": 10870
+    },
+    {
+      "epoch": 0.34694983896170156,
+      "grad_norm": 0.18283255398273468,
+      "learning_rate": 0.001,
+      "loss": 1.2069,
+      "step": 10880
+    },
+    {
+      "epoch": 0.3472687266813355,
+      "grad_norm": 0.1826285868883133,
+      "learning_rate": 0.001,
+      "loss": 1.1933,
+      "step": 10890
+    },
+    {
+      "epoch": 0.3475876144009694,
+      "grad_norm": 0.19150082767009735,
+      "learning_rate": 0.001,
+      "loss": 1.199,
+      "step": 10900
+    },
+    {
+      "epoch": 0.3479065021206033,
+      "grad_norm": 0.18925228714942932,
+      "learning_rate": 0.001,
+      "loss": 1.1748,
+      "step": 10910
+    },
+    {
+      "epoch": 0.34822538984023726,
+      "grad_norm": 0.18681123852729797,
+      "learning_rate": 0.001,
+      "loss": 1.1938,
+      "step": 10920
+    },
+    {
+      "epoch": 0.34854427755987116,
+      "grad_norm": 0.19334517419338226,
+      "learning_rate": 0.001,
+      "loss": 1.1997,
+      "step": 10930
+    },
+    {
+      "epoch": 0.34886316527950506,
+      "grad_norm": 0.19396010041236877,
+      "learning_rate": 0.001,
+      "loss": 1.1772,
+      "step": 10940
+    },
+    {
+      "epoch": 0.349182052999139,
+      "grad_norm": 0.18436571955680847,
+      "learning_rate": 0.001,
+      "loss": 1.2072,
+      "step": 10950
+    },
+    {
+      "epoch": 0.3495009407187729,
+      "grad_norm": 0.18500304222106934,
+      "learning_rate": 0.001,
+      "loss": 1.1882,
+      "step": 10960
+    },
+    {
+      "epoch": 0.3498198284384068,
+      "grad_norm": 0.18106575310230255,
+      "learning_rate": 0.001,
+      "loss": 1.1967,
+      "step": 10970
+    },
+    {
+      "epoch": 0.35013871615804076,
+      "grad_norm": 0.18210440874099731,
+      "learning_rate": 0.001,
+      "loss": 1.1908,
+      "step": 10980
+    },
+    {
+      "epoch": 0.35045760387767466,
+      "grad_norm": 0.1854126900434494,
+      "learning_rate": 0.001,
+      "loss": 1.1698,
+      "step": 10990
+    },
+    {
+      "epoch": 0.3507764915973086,
+      "grad_norm": 0.18545810878276825,
+      "learning_rate": 0.001,
+      "loss": 1.185,
+      "step": 11000
+    },
+    {
+      "epoch": 0.3510953793169425,
+      "grad_norm": 0.18830567598342896,
+      "learning_rate": 0.001,
+      "loss": 1.2,
+      "step": 11010
+    },
+    {
+      "epoch": 0.3514142670365764,
+      "grad_norm": 0.18633617460727692,
+      "learning_rate": 0.001,
+      "loss": 1.1788,
+      "step": 11020
+    },
+    {
+      "epoch": 0.35173315475621036,
+      "grad_norm": 0.1790749579668045,
+      "learning_rate": 0.001,
+      "loss": 1.1621,
+      "step": 11030
+    },
+    {
+      "epoch": 0.35205204247584426,
+      "grad_norm": 0.18545784056186676,
+      "learning_rate": 0.001,
+      "loss": 1.1776,
+      "step": 11040
+    },
+    {
+      "epoch": 0.35237093019547816,
+      "grad_norm": 0.18230289220809937,
+      "learning_rate": 0.001,
+      "loss": 1.1611,
+      "step": 11050
+    },
+    {
+      "epoch": 0.3526898179151121,
+      "grad_norm": 0.1841791570186615,
+      "learning_rate": 0.001,
+      "loss": 1.1863,
+      "step": 11060
+    },
+    {
+      "epoch": 0.353008705634746,
+      "grad_norm": 0.18114890158176422,
+      "learning_rate": 0.001,
+      "loss": 1.1791,
+      "step": 11070
+    },
+    {
+      "epoch": 0.3533275933543799,
+      "grad_norm": 0.18278636038303375,
+      "learning_rate": 0.001,
+      "loss": 1.1774,
+      "step": 11080
+    },
+    {
+      "epoch": 0.35364648107401386,
+      "grad_norm": 0.18547603487968445,
+      "learning_rate": 0.001,
+      "loss": 1.1663,
+      "step": 11090
+    },
+    {
+      "epoch": 0.35396536879364776,
+      "grad_norm": 0.18347465991973877,
+      "learning_rate": 0.001,
+      "loss": 1.1656,
+      "step": 11100
+    },
+    {
+      "epoch": 0.35428425651328166,
+      "grad_norm": 0.1870725154876709,
+      "learning_rate": 0.001,
+      "loss": 1.1898,
+      "step": 11110
+    },
+    {
+      "epoch": 0.3546031442329156,
+      "grad_norm": 0.18504945933818817,
+      "learning_rate": 0.001,
+      "loss": 1.193,
+      "step": 11120
+    },
+    {
+      "epoch": 0.3549220319525495,
+      "grad_norm": 0.185967355966568,
+      "learning_rate": 0.001,
+      "loss": 1.191,
+      "step": 11130
+    },
+    {
+      "epoch": 0.3552409196721834,
+      "grad_norm": 0.18489550054073334,
+      "learning_rate": 0.001,
+      "loss": 1.1814,
+      "step": 11140
+    },
+    {
+      "epoch": 0.35555980739181736,
+      "grad_norm": 0.18694624304771423,
+      "learning_rate": 0.001,
+      "loss": 1.1817,
+      "step": 11150
+    },
+    {
+      "epoch": 0.35587869511145126,
+      "grad_norm": 0.1808748096227646,
+      "learning_rate": 0.001,
+      "loss": 1.1878,
+      "step": 11160
+    },
+    {
+      "epoch": 0.35619758283108516,
+      "grad_norm": 0.17861339449882507,
+      "learning_rate": 0.001,
+      "loss": 1.1849,
+      "step": 11170
+    },
+    {
+      "epoch": 0.3565164705507191,
+      "grad_norm": 0.19131644070148468,
+      "learning_rate": 0.001,
+      "loss": 1.1769,
+      "step": 11180
+    },
+    {
+      "epoch": 0.356835358270353,
+      "grad_norm": 0.18798422813415527,
+      "learning_rate": 0.001,
+      "loss": 1.1763,
+      "step": 11190
+    },
+    {
+      "epoch": 0.3571542459899869,
+      "grad_norm": 0.1764545887708664,
+      "learning_rate": 0.001,
+      "loss": 1.1528,
+      "step": 11200
+    },
+    {
+      "epoch": 0.35747313370962086,
+      "grad_norm": 0.18498189747333527,
+      "learning_rate": 0.001,
+      "loss": 1.1927,
+      "step": 11210
+    },
+    {
+      "epoch": 0.35779202142925476,
+      "grad_norm": 0.17672859132289886,
+      "learning_rate": 0.001,
+      "loss": 1.1782,
+      "step": 11220
+    },
+    {
+      "epoch": 0.35811090914888866,
+      "grad_norm": 0.18820945918560028,
+      "learning_rate": 0.001,
+      "loss": 1.2013,
+      "step": 11230
+    },
+    {
+      "epoch": 0.3584297968685226,
+      "grad_norm": 0.18688707053661346,
+      "learning_rate": 0.001,
+      "loss": 1.1976,
+      "step": 11240
+    },
+    {
+      "epoch": 0.3587486845881565,
+      "grad_norm": 0.18575413525104523,
+      "learning_rate": 0.001,
+      "loss": 1.176,
+      "step": 11250
+    },
+    {
+      "epoch": 0.3590675723077904,
+      "grad_norm": 0.18457651138305664,
+      "learning_rate": 0.001,
+      "loss": 1.1669,
+      "step": 11260
+    },
+    {
+      "epoch": 0.35938646002742436,
+      "grad_norm": 0.1832694411277771,
+      "learning_rate": 0.001,
+      "loss": 1.192,
+      "step": 11270
+    },
+    {
+      "epoch": 0.35970534774705826,
+      "grad_norm": 0.1787375658750534,
+      "learning_rate": 0.001,
+      "loss": 1.16,
+      "step": 11280
+    },
+    {
+      "epoch": 0.36002423546669216,
+      "grad_norm": 0.19141420722007751,
+      "learning_rate": 0.001,
+      "loss": 1.171,
+      "step": 11290
+    },
+    {
+      "epoch": 0.3603431231863261,
+      "grad_norm": 0.18656058609485626,
+      "learning_rate": 0.001,
+      "loss": 1.1834,
+      "step": 11300
+    },
+    {
+      "epoch": 0.36066201090596,
+      "grad_norm": 0.17995838820934296,
+      "learning_rate": 0.001,
+      "loss": 1.1693,
+      "step": 11310
+    },
+    {
+      "epoch": 0.3609808986255939,
+      "grad_norm": 0.1849241405725479,
+      "learning_rate": 0.001,
+      "loss": 1.1671,
+      "step": 11320
+    },
+    {
+      "epoch": 0.36129978634522786,
+      "grad_norm": 0.18031972646713257,
+      "learning_rate": 0.001,
+      "loss": 1.1673,
+      "step": 11330
+    },
+    {
+      "epoch": 0.36161867406486176,
+      "grad_norm": 0.17960664629936218,
+      "learning_rate": 0.001,
+      "loss": 1.1894,
+      "step": 11340
+    },
+    {
+      "epoch": 0.36193756178449565,
+      "grad_norm": 0.18871818482875824,
+      "learning_rate": 0.001,
+      "loss": 1.1815,
+      "step": 11350
+    },
+    {
+      "epoch": 0.3622564495041296,
+      "grad_norm": 0.1836932748556137,
+      "learning_rate": 0.001,
+      "loss": 1.1872,
+      "step": 11360
+    },
+    {
+      "epoch": 0.3625753372237635,
+      "grad_norm": 0.18293382227420807,
+      "learning_rate": 0.001,
+      "loss": 1.1752,
+      "step": 11370
+    },
+    {
+      "epoch": 0.3628942249433974,
+      "grad_norm": 0.1888611763715744,
+      "learning_rate": 0.001,
+      "loss": 1.1745,
+      "step": 11380
+    },
+    {
+      "epoch": 0.36321311266303136,
+      "grad_norm": 0.18882402777671814,
+      "learning_rate": 0.001,
+      "loss": 1.1753,
+      "step": 11390
+    },
+    {
+      "epoch": 0.36353200038266525,
+      "grad_norm": 0.18266329169273376,
+      "learning_rate": 0.001,
+      "loss": 1.1863,
+      "step": 11400
+    },
+    {
+      "epoch": 0.3638508881022992,
+      "grad_norm": 0.18598082661628723,
+      "learning_rate": 0.001,
+      "loss": 1.1694,
+      "step": 11410
+    },
+    {
+      "epoch": 0.3641697758219331,
+      "grad_norm": 0.18703070282936096,
+      "learning_rate": 0.001,
+      "loss": 1.1781,
+      "step": 11420
+    },
+    {
+      "epoch": 0.364488663541567,
+      "grad_norm": 0.18162168562412262,
+      "learning_rate": 0.001,
+      "loss": 1.1567,
+      "step": 11430
+    },
+    {
+      "epoch": 0.36480755126120096,
+      "grad_norm": 0.1802840381860733,
+      "learning_rate": 0.001,
+      "loss": 1.1776,
+      "step": 11440
+    },
+    {
+      "epoch": 0.36512643898083486,
+      "grad_norm": 0.18248935043811798,
+      "learning_rate": 0.001,
+      "loss": 1.1772,
+      "step": 11450
+    },
+    {
+      "epoch": 0.36544532670046875,
+      "grad_norm": 0.17799849808216095,
+      "learning_rate": 0.001,
+      "loss": 1.1591,
+      "step": 11460
+    },
+    {
+      "epoch": 0.3657642144201027,
+      "grad_norm": 0.18336179852485657,
+      "learning_rate": 0.001,
+      "loss": 1.1777,
+      "step": 11470
+    },
+    {
+      "epoch": 0.3660831021397366,
+      "grad_norm": 0.18439556658267975,
+      "learning_rate": 0.001,
+      "loss": 1.1685,
+      "step": 11480
+    },
+    {
+      "epoch": 0.3664019898593705,
+      "grad_norm": 0.1861710548400879,
+      "learning_rate": 0.001,
+      "loss": 1.1785,
+      "step": 11490
+    },
+    {
+      "epoch": 0.36672087757900446,
+      "grad_norm": 0.1906169354915619,
+      "learning_rate": 0.001,
+      "loss": 1.1917,
+      "step": 11500
+    },
+    {
+      "epoch": 0.36703976529863835,
+      "grad_norm": 0.17864377796649933,
+      "learning_rate": 0.001,
+      "loss": 1.1622,
+      "step": 11510
+    },
+    {
+      "epoch": 0.36735865301827225,
+      "grad_norm": 0.18871931731700897,
+      "learning_rate": 0.001,
+      "loss": 1.189,
+      "step": 11520
+    },
+    {
+      "epoch": 0.3676775407379062,
+      "grad_norm": 0.1843557208776474,
+      "learning_rate": 0.001,
+      "loss": 1.1632,
+      "step": 11530
+    },
+    {
+      "epoch": 0.3679964284575401,
+      "grad_norm": 0.1822815090417862,
+      "learning_rate": 0.001,
+      "loss": 1.1522,
+      "step": 11540
+    },
+    {
+      "epoch": 0.368315316177174,
+      "grad_norm": 0.18551307916641235,
+      "learning_rate": 0.001,
+      "loss": 1.1796,
+      "step": 11550
+    },
+    {
+      "epoch": 0.36863420389680795,
+      "grad_norm": 0.1817987561225891,
+      "learning_rate": 0.001,
+      "loss": 1.1682,
+      "step": 11560
+    },
+    {
+      "epoch": 0.36895309161644185,
+      "grad_norm": 0.1884773075580597,
+      "learning_rate": 0.001,
+      "loss": 1.1818,
+      "step": 11570
+    },
+    {
+      "epoch": 0.36927197933607575,
+      "grad_norm": 0.18173812329769135,
+      "learning_rate": 0.001,
+      "loss": 1.1507,
+      "step": 11580
+    },
+    {
+      "epoch": 0.3695908670557097,
+      "grad_norm": 0.1824948489665985,
+      "learning_rate": 0.001,
+      "loss": 1.1586,
+      "step": 11590
+    },
+    {
+      "epoch": 0.3699097547753436,
+      "grad_norm": 0.1829361766576767,
+      "learning_rate": 0.001,
+      "loss": 1.1953,
+      "step": 11600
+    },
+    {
+      "epoch": 0.3702286424949775,
+      "grad_norm": 0.18169468641281128,
+      "learning_rate": 0.001,
+      "loss": 1.1786,
+      "step": 11610
+    },
+    {
+      "epoch": 0.37054753021461145,
+      "grad_norm": 0.18068274855613708,
+      "learning_rate": 0.001,
+      "loss": 1.1862,
+      "step": 11620
+    },
+    {
+      "epoch": 0.37086641793424535,
+      "grad_norm": 0.1841064840555191,
+      "learning_rate": 0.001,
+      "loss": 1.1768,
+      "step": 11630
+    },
+    {
+      "epoch": 0.37118530565387925,
+      "grad_norm": 0.18492387235164642,
+      "learning_rate": 0.001,
+      "loss": 1.198,
+      "step": 11640
+    },
+    {
+      "epoch": 0.3715041933735132,
+      "grad_norm": 0.17407332360744476,
+      "learning_rate": 0.001,
+      "loss": 1.1688,
+      "step": 11650
+    },
+    {
+      "epoch": 0.3718230810931471,
+      "grad_norm": 0.18511362373828888,
+      "learning_rate": 0.001,
+      "loss": 1.1905,
+      "step": 11660
+    },
+    {
+      "epoch": 0.372141968812781,
+      "grad_norm": 0.18631397187709808,
+      "learning_rate": 0.001,
+      "loss": 1.1934,
+      "step": 11670
+    },
+    {
+      "epoch": 0.37246085653241495,
+      "grad_norm": 0.18116828799247742,
+      "learning_rate": 0.001,
+      "loss": 1.1727,
+      "step": 11680
+    },
+    {
+      "epoch": 0.37277974425204885,
+      "grad_norm": 0.18609020113945007,
+      "learning_rate": 0.001,
+      "loss": 1.1718,
+      "step": 11690
+    },
+    {
+      "epoch": 0.37309863197168275,
+      "grad_norm": 0.1801086664199829,
+      "learning_rate": 0.001,
+      "loss": 1.1578,
+      "step": 11700
+    },
+    {
+      "epoch": 0.3734175196913167,
+      "grad_norm": 0.17963343858718872,
+      "learning_rate": 0.001,
+      "loss": 1.1743,
+      "step": 11710
+    },
+    {
+      "epoch": 0.3737364074109506,
+      "grad_norm": 0.17758981883525848,
+      "learning_rate": 0.001,
+      "loss": 1.1653,
+      "step": 11720
+    },
+    {
+      "epoch": 0.3740552951305845,
+      "grad_norm": 0.18511025607585907,
+      "learning_rate": 0.001,
+      "loss": 1.1803,
+      "step": 11730
+    },
+    {
+      "epoch": 0.37437418285021845,
+      "grad_norm": 0.18526363372802734,
+      "learning_rate": 0.001,
+      "loss": 1.1813,
+      "step": 11740
+    },
+    {
+      "epoch": 0.37469307056985235,
+      "grad_norm": 0.17796054482460022,
+      "learning_rate": 0.001,
+      "loss": 1.1909,
+      "step": 11750
+    },
+    {
+      "epoch": 0.37501195828948625,
+      "grad_norm": 0.18078628182411194,
+      "learning_rate": 0.001,
+      "loss": 1.1823,
+      "step": 11760
+    },
+    {
+      "epoch": 0.3753308460091202,
+      "grad_norm": 0.18209408223628998,
+      "learning_rate": 0.001,
+      "loss": 1.1798,
+      "step": 11770
+    },
+    {
+      "epoch": 0.3756497337287541,
+      "grad_norm": 0.18052178621292114,
+      "learning_rate": 0.001,
+      "loss": 1.1704,
+      "step": 11780
+    },
+    {
+      "epoch": 0.375968621448388,
+      "grad_norm": 0.18084560334682465,
+      "learning_rate": 0.001,
+      "loss": 1.1969,
+      "step": 11790
+    },
+    {
+      "epoch": 0.37628750916802195,
+      "grad_norm": 0.17806032299995422,
+      "learning_rate": 0.001,
+      "loss": 1.1739,
+      "step": 11800
+    },
+    {
+      "epoch": 0.37660639688765585,
+      "grad_norm": 0.1794821321964264,
+      "learning_rate": 0.001,
+      "loss": 1.1806,
+      "step": 11810
+    },
+    {
+      "epoch": 0.37692528460728975,
+      "grad_norm": 0.1821807473897934,
+      "learning_rate": 0.001,
+      "loss": 1.1794,
+      "step": 11820
+    },
+    {
+      "epoch": 0.3772441723269237,
+      "grad_norm": 0.18536563217639923,
+      "learning_rate": 0.001,
+      "loss": 1.1574,
+      "step": 11830
+    },
+    {
+      "epoch": 0.3775630600465576,
+      "grad_norm": 0.17974741756916046,
+      "learning_rate": 0.001,
+      "loss": 1.1795,
+      "step": 11840
+    },
+    {
+      "epoch": 0.37788194776619155,
+      "grad_norm": 0.1832747459411621,
+      "learning_rate": 0.001,
+      "loss": 1.175,
+      "step": 11850
+    },
+    {
+      "epoch": 0.37820083548582545,
+      "grad_norm": 0.17949554324150085,
+      "learning_rate": 0.001,
+      "loss": 1.1824,
+      "step": 11860
+    },
+    {
+      "epoch": 0.37851972320545935,
+      "grad_norm": 0.17655141651630402,
+      "learning_rate": 0.001,
+      "loss": 1.178,
+      "step": 11870
+    },
+    {
+      "epoch": 0.3788386109250933,
+      "grad_norm": 0.17882749438285828,
+      "learning_rate": 0.001,
+      "loss": 1.1756,
+      "step": 11880
+    },
+    {
+      "epoch": 0.3791574986447272,
+      "grad_norm": 0.17046712338924408,
+      "learning_rate": 0.001,
+      "loss": 1.1677,
+      "step": 11890
+    },
+    {
+      "epoch": 0.3794763863643611,
+      "grad_norm": 0.18203288316726685,
+      "learning_rate": 0.001,
+      "loss": 1.1608,
+      "step": 11900
+    },
+    {
+      "epoch": 0.37979527408399505,
+      "grad_norm": 0.18050289154052734,
+      "learning_rate": 0.001,
+      "loss": 1.1797,
+      "step": 11910
+    },
+    {
+      "epoch": 0.38011416180362895,
+      "grad_norm": 0.17975153028964996,
+      "learning_rate": 0.001,
+      "loss": 1.1846,
+      "step": 11920
+    },
+    {
+      "epoch": 0.38043304952326285,
+      "grad_norm": 0.17785479128360748,
+      "learning_rate": 0.001,
+      "loss": 1.1494,
+      "step": 11930
+    },
+    {
+      "epoch": 0.3807519372428968,
+      "grad_norm": 0.18440751731395721,
+      "learning_rate": 0.001,
+      "loss": 1.1433,
+      "step": 11940
+    },
+    {
+      "epoch": 0.3810708249625307,
+      "grad_norm": 0.17766568064689636,
+      "learning_rate": 0.001,
+      "loss": 1.1682,
+      "step": 11950
+    },
+    {
+      "epoch": 0.3813897126821646,
+      "grad_norm": 0.17816171050071716,
+      "learning_rate": 0.001,
+      "loss": 1.1793,
+      "step": 11960
+    },
+    {
+      "epoch": 0.38170860040179855,
+      "grad_norm": 0.1773807853460312,
+      "learning_rate": 0.001,
+      "loss": 1.1751,
+      "step": 11970
+    },
+    {
+      "epoch": 0.38202748812143245,
+      "grad_norm": 0.17774201929569244,
+      "learning_rate": 0.001,
+      "loss": 1.1522,
+      "step": 11980
+    },
+    {
+      "epoch": 0.38234637584106634,
+      "grad_norm": 0.184428870677948,
+      "learning_rate": 0.001,
+      "loss": 1.1644,
+      "step": 11990
+    },
+    {
+      "epoch": 0.3826652635607003,
+      "grad_norm": 0.18758635222911835,
+      "learning_rate": 0.001,
+      "loss": 1.1649,
+      "step": 12000
+    },
+    {
+      "epoch": 0.3829841512803342,
+      "grad_norm": 0.18008406460285187,
+      "learning_rate": 0.001,
+      "loss": 1.1769,
+      "step": 12010
+    },
+    {
+      "epoch": 0.3833030389999681,
+      "grad_norm": 0.18150904774665833,
+      "learning_rate": 0.001,
+      "loss": 1.1926,
+      "step": 12020
+    },
+    {
+      "epoch": 0.38362192671960205,
+      "grad_norm": 0.18088509142398834,
+      "learning_rate": 0.001,
+      "loss": 1.1672,
+      "step": 12030
+    },
+    {
+      "epoch": 0.38394081443923594,
+      "grad_norm": 0.17663919925689697,
+      "learning_rate": 0.001,
+      "loss": 1.1621,
+      "step": 12040
+    },
+    {
+      "epoch": 0.38425970215886984,
+      "grad_norm": 0.17578722536563873,
+      "learning_rate": 0.001,
+      "loss": 1.1701,
+      "step": 12050
+    },
+    {
+      "epoch": 0.3845785898785038,
+      "grad_norm": 0.17906726896762848,
+      "learning_rate": 0.001,
+      "loss": 1.1585,
+      "step": 12060
+    },
+    {
+      "epoch": 0.3848974775981377,
+      "grad_norm": 0.17861928045749664,
+      "learning_rate": 0.001,
+      "loss": 1.1749,
+      "step": 12070
+    },
+    {
+      "epoch": 0.3852163653177716,
+      "grad_norm": 0.18296191096305847,
+      "learning_rate": 0.001,
+      "loss": 1.1684,
+      "step": 12080
+    },
+    {
+      "epoch": 0.38553525303740555,
+      "grad_norm": 0.18413016200065613,
+      "learning_rate": 0.001,
+      "loss": 1.1622,
+      "step": 12090
+    },
+    {
+      "epoch": 0.38585414075703944,
+      "grad_norm": 0.18453866243362427,
+      "learning_rate": 0.001,
+      "loss": 1.169,
+      "step": 12100
+    },
+    {
+      "epoch": 0.38617302847667334,
+      "grad_norm": 0.1800256073474884,
+      "learning_rate": 0.001,
+      "loss": 1.1711,
+      "step": 12110
+    },
+    {
+      "epoch": 0.3864919161963073,
+      "grad_norm": 0.18271325528621674,
+      "learning_rate": 0.001,
+      "loss": 1.1779,
+      "step": 12120
+    },
+    {
+      "epoch": 0.3868108039159412,
+      "grad_norm": 0.18513017892837524,
+      "learning_rate": 0.001,
+      "loss": 1.1789,
+      "step": 12130
+    },
+    {
+      "epoch": 0.3871296916355751,
+      "grad_norm": 0.17271831631660461,
+      "learning_rate": 0.001,
+      "loss": 1.1441,
+      "step": 12140
+    },
+    {
+      "epoch": 0.38744857935520904,
+      "grad_norm": 0.17758046090602875,
+      "learning_rate": 0.001,
+      "loss": 1.1834,
+      "step": 12150
+    },
+    {
+      "epoch": 0.38776746707484294,
+      "grad_norm": 0.17706969380378723,
+      "learning_rate": 0.001,
+      "loss": 1.1809,
+      "step": 12160
+    },
+    {
+      "epoch": 0.38808635479447684,
+      "grad_norm": 0.18004131317138672,
+      "learning_rate": 0.001,
+      "loss": 1.1454,
+      "step": 12170
+    },
+    {
+      "epoch": 0.3884052425141108,
+      "grad_norm": 0.17989245057106018,
+      "learning_rate": 0.001,
+      "loss": 1.1868,
+      "step": 12180
+    },
+    {
+      "epoch": 0.3887241302337447,
+      "grad_norm": 0.18094083666801453,
+      "learning_rate": 0.001,
+      "loss": 1.1606,
+      "step": 12190
+    },
+    {
+      "epoch": 0.3890430179533786,
+      "grad_norm": 0.1783522516489029,
+      "learning_rate": 0.001,
+      "loss": 1.1639,
+      "step": 12200
+    },
+    {
+      "epoch": 0.38936190567301254,
+      "grad_norm": 0.18109023571014404,
+      "learning_rate": 0.001,
+      "loss": 1.168,
+      "step": 12210
+    },
+    {
+      "epoch": 0.38968079339264644,
+      "grad_norm": 0.17977750301361084,
+      "learning_rate": 0.001,
+      "loss": 1.1695,
+      "step": 12220
+    },
+    {
+      "epoch": 0.38999968111228034,
+      "grad_norm": 0.17573317885398865,
+      "learning_rate": 0.001,
+      "loss": 1.1569,
+      "step": 12230
+    },
+    {
+      "epoch": 0.3903185688319143,
+      "grad_norm": 0.17866761982440948,
+      "learning_rate": 0.001,
+      "loss": 1.1573,
+      "step": 12240
+    },
+    {
+      "epoch": 0.3906374565515482,
+      "grad_norm": 0.1950611174106598,
+      "learning_rate": 0.001,
+      "loss": 1.1459,
+      "step": 12250
+    },
+    {
+      "epoch": 0.39095634427118214,
+      "grad_norm": 0.18050596117973328,
+      "learning_rate": 0.001,
+      "loss": 1.1614,
+      "step": 12260
+    },
+    {
+      "epoch": 0.39127523199081604,
+      "grad_norm": 0.17666298151016235,
+      "learning_rate": 0.001,
+      "loss": 1.1762,
+      "step": 12270
+    },
+    {
+      "epoch": 0.39159411971044994,
+      "grad_norm": 0.18661339581012726,
+      "learning_rate": 0.001,
+      "loss": 1.1631,
+      "step": 12280
+    },
+    {
+      "epoch": 0.3919130074300839,
+      "grad_norm": 0.18043366074562073,
+      "learning_rate": 0.001,
+      "loss": 1.1614,
+      "step": 12290
+    },
+    {
+      "epoch": 0.3922318951497178,
+      "grad_norm": 0.1795511692762375,
+      "learning_rate": 0.001,
+      "loss": 1.18,
+      "step": 12300
+    },
+    {
+      "epoch": 0.3925507828693517,
+      "grad_norm": 0.18567945063114166,
+      "learning_rate": 0.001,
+      "loss": 1.1677,
+      "step": 12310
+    },
+    {
+      "epoch": 0.39286967058898564,
+      "grad_norm": 0.18031778931617737,
+      "learning_rate": 0.001,
+      "loss": 1.1733,
+      "step": 12320
+    },
+    {
+      "epoch": 0.39318855830861954,
+      "grad_norm": 0.1776386946439743,
+      "learning_rate": 0.001,
+      "loss": 1.1709,
+      "step": 12330
+    },
+    {
+      "epoch": 0.39350744602825344,
+      "grad_norm": 0.18277223408222198,
+      "learning_rate": 0.001,
+      "loss": 1.1522,
+      "step": 12340
+    },
+    {
+      "epoch": 0.3938263337478874,
+      "grad_norm": 0.1731247454881668,
+      "learning_rate": 0.001,
+      "loss": 1.1509,
+      "step": 12350
+    },
+    {
+      "epoch": 0.3941452214675213,
+      "grad_norm": 0.18343017995357513,
+      "learning_rate": 0.001,
+      "loss": 1.1909,
+      "step": 12360
+    },
+    {
+      "epoch": 0.3944641091871552,
+      "grad_norm": 0.1750156581401825,
+      "learning_rate": 0.001,
+      "loss": 1.1576,
+      "step": 12370
+    },
+    {
+      "epoch": 0.39478299690678914,
+      "grad_norm": 0.18162328004837036,
+      "learning_rate": 0.001,
+      "loss": 1.1551,
+      "step": 12380
+    },
+    {
+      "epoch": 0.39510188462642304,
+      "grad_norm": 0.18877561390399933,
+      "learning_rate": 0.001,
+      "loss": 1.1558,
+      "step": 12390
+    },
+    {
+      "epoch": 0.39542077234605694,
+      "grad_norm": 0.18337975442409515,
+      "learning_rate": 0.001,
+      "loss": 1.1701,
+      "step": 12400
+    },
+    {
+      "epoch": 0.3957396600656909,
+      "grad_norm": 0.18324777483940125,
+      "learning_rate": 0.001,
+      "loss": 1.155,
+      "step": 12410
+    },
+    {
+      "epoch": 0.3960585477853248,
+      "grad_norm": 0.18144458532333374,
+      "learning_rate": 0.001,
+      "loss": 1.1681,
+      "step": 12420
+    },
+    {
+      "epoch": 0.3963774355049587,
+      "grad_norm": 0.18112270534038544,
+      "learning_rate": 0.001,
+      "loss": 1.171,
+      "step": 12430
+    },
+    {
+      "epoch": 0.39669632322459264,
+      "grad_norm": 0.18132545053958893,
+      "learning_rate": 0.001,
+      "loss": 1.1454,
+      "step": 12440
+    },
+    {
+      "epoch": 0.39701521094422654,
+      "grad_norm": 0.18609032034873962,
+      "learning_rate": 0.001,
+      "loss": 1.169,
+      "step": 12450
+    },
+    {
+      "epoch": 0.39733409866386044,
+      "grad_norm": 0.17792826890945435,
+      "learning_rate": 0.001,
+      "loss": 1.1535,
+      "step": 12460
+    },
+    {
+      "epoch": 0.3976529863834944,
+      "grad_norm": 0.17896825075149536,
+      "learning_rate": 0.001,
+      "loss": 1.1752,
+      "step": 12470
+    },
+    {
+      "epoch": 0.3979718741031283,
+      "grad_norm": 0.18214313685894012,
+      "learning_rate": 0.001,
+      "loss": 1.1752,
+      "step": 12480
+    },
+    {
+      "epoch": 0.3982907618227622,
+      "grad_norm": 0.17771407961845398,
+      "learning_rate": 0.001,
+      "loss": 1.1647,
+      "step": 12490
+    },
+    {
+      "epoch": 0.39860964954239614,
+      "grad_norm": 0.17843423783779144,
+      "learning_rate": 0.001,
+      "loss": 1.1795,
+      "step": 12500
+    },
+    {
+      "epoch": 0.39892853726203004,
+      "grad_norm": 0.17646920680999756,
+      "learning_rate": 0.001,
+      "loss": 1.1538,
+      "step": 12510
+    },
+    {
+      "epoch": 0.39924742498166393,
+      "grad_norm": 0.1750120222568512,
+      "learning_rate": 0.001,
+      "loss": 1.157,
+      "step": 12520
+    },
+    {
+      "epoch": 0.3995663127012979,
+      "grad_norm": 0.17426247894763947,
+      "learning_rate": 0.001,
+      "loss": 1.1755,
+      "step": 12530
+    },
+    {
+      "epoch": 0.3998852004209318,
+      "grad_norm": 0.17977719008922577,
+      "learning_rate": 0.001,
+      "loss": 1.162,
+      "step": 12540
+    },
+    {
+      "epoch": 0.4002040881405657,
+      "grad_norm": 0.1777566522359848,
+      "learning_rate": 0.001,
+      "loss": 1.1392,
+      "step": 12550
+    },
+    {
+      "epoch": 0.40052297586019964,
+      "grad_norm": 0.17475956678390503,
+      "learning_rate": 0.001,
+      "loss": 1.1584,
+      "step": 12560
+    },
+    {
+      "epoch": 0.40084186357983353,
+      "grad_norm": 0.1797400861978531,
+      "learning_rate": 0.001,
+      "loss": 1.1841,
+      "step": 12570
+    },
+    {
+      "epoch": 0.40116075129946743,
+      "grad_norm": 0.18256771564483643,
+      "learning_rate": 0.001,
+      "loss": 1.1384,
+      "step": 12580
+    },
+    {
+      "epoch": 0.4014796390191014,
+      "grad_norm": 0.1778700202703476,
+      "learning_rate": 0.001,
+      "loss": 1.1823,
+      "step": 12590
+    },
+    {
+      "epoch": 0.4017985267387353,
+      "grad_norm": 0.17812106013298035,
+      "learning_rate": 0.001,
+      "loss": 1.1546,
+      "step": 12600
+    },
+    {
+      "epoch": 0.4021174144583692,
+      "grad_norm": 0.17732104659080505,
+      "learning_rate": 0.001,
+      "loss": 1.1618,
+      "step": 12610
+    },
+    {
+      "epoch": 0.40243630217800314,
+      "grad_norm": 0.1758512556552887,
+      "learning_rate": 0.001,
+      "loss": 1.1683,
+      "step": 12620
+    },
+    {
+      "epoch": 0.40275518989763703,
+      "grad_norm": 0.17862358689308167,
+      "learning_rate": 0.001,
+      "loss": 1.1444,
+      "step": 12630
+    },
+    {
+      "epoch": 0.40307407761727093,
+      "grad_norm": 0.17952404916286469,
+      "learning_rate": 0.001,
+      "loss": 1.1415,
+      "step": 12640
+    },
+    {
+      "epoch": 0.4033929653369049,
+      "grad_norm": 0.17920245230197906,
+      "learning_rate": 0.001,
+      "loss": 1.1672,
+      "step": 12650
+    },
+    {
+      "epoch": 0.4037118530565388,
+      "grad_norm": 0.18261930346488953,
+      "learning_rate": 0.001,
+      "loss": 1.1579,
+      "step": 12660
+    },
+    {
+      "epoch": 0.40403074077617274,
+      "grad_norm": 0.17810742557048798,
+      "learning_rate": 0.001,
+      "loss": 1.1572,
+      "step": 12670
+    },
+    {
+      "epoch": 0.40434962849580663,
+      "grad_norm": 0.18469861149787903,
+      "learning_rate": 0.001,
+      "loss": 1.1627,
+      "step": 12680
+    },
+    {
+      "epoch": 0.40466851621544053,
+      "grad_norm": 0.1746484488248825,
+      "learning_rate": 0.001,
+      "loss": 1.15,
+      "step": 12690
+    },
+    {
+      "epoch": 0.4049874039350745,
+      "grad_norm": 0.1788916438817978,
+      "learning_rate": 0.001,
+      "loss": 1.1631,
+      "step": 12700
+    },
+    {
+      "epoch": 0.4053062916547084,
+      "grad_norm": 0.17759430408477783,
+      "learning_rate": 0.001,
+      "loss": 1.148,
+      "step": 12710
+    },
+    {
+      "epoch": 0.4056251793743423,
+      "grad_norm": 0.18186287581920624,
+      "learning_rate": 0.001,
+      "loss": 1.1633,
+      "step": 12720
+    },
+    {
+      "epoch": 0.40594406709397624,
+      "grad_norm": 0.18196360766887665,
+      "learning_rate": 0.001,
+      "loss": 1.1485,
+      "step": 12730
+    },
+    {
+      "epoch": 0.40626295481361013,
+      "grad_norm": 0.1794370859861374,
+      "learning_rate": 0.001,
+      "loss": 1.1638,
+      "step": 12740
+    },
+    {
+      "epoch": 0.40658184253324403,
+      "grad_norm": 0.17594540119171143,
+      "learning_rate": 0.001,
+      "loss": 1.1521,
+      "step": 12750
+    },
+    {
+      "epoch": 0.406900730252878,
+      "grad_norm": 0.1787782609462738,
+      "learning_rate": 0.001,
+      "loss": 1.1568,
+      "step": 12760
+    },
+    {
+      "epoch": 0.4072196179725119,
+      "grad_norm": 0.1763012856245041,
+      "learning_rate": 0.001,
+      "loss": 1.1754,
+      "step": 12770
+    },
+    {
+      "epoch": 0.4075385056921458,
+      "grad_norm": 0.18189765512943268,
+      "learning_rate": 0.001,
+      "loss": 1.1529,
+      "step": 12780
+    },
+    {
+      "epoch": 0.40785739341177973,
+      "grad_norm": 0.18166020512580872,
+      "learning_rate": 0.001,
+      "loss": 1.1541,
+      "step": 12790
+    },
+    {
+      "epoch": 0.40817628113141363,
+      "grad_norm": 0.1810879111289978,
+      "learning_rate": 0.001,
+      "loss": 1.1717,
+      "step": 12800
+    },
+    {
+      "epoch": 0.40849516885104753,
+      "grad_norm": 0.17724236845970154,
+      "learning_rate": 0.001,
+      "loss": 1.1492,
+      "step": 12810
+    },
+    {
+      "epoch": 0.4088140565706815,
+      "grad_norm": 0.18092194199562073,
+      "learning_rate": 0.001,
+      "loss": 1.1522,
+      "step": 12820
+    },
+    {
+      "epoch": 0.4091329442903154,
+      "grad_norm": 0.17586788535118103,
+      "learning_rate": 0.001,
+      "loss": 1.1487,
+      "step": 12830
+    },
+    {
+      "epoch": 0.4094518320099493,
+      "grad_norm": 0.17758262157440186,
+      "learning_rate": 0.001,
+      "loss": 1.1283,
+      "step": 12840
+    },
+    {
+      "epoch": 0.40977071972958323,
+      "grad_norm": 0.1868557333946228,
+      "learning_rate": 0.001,
+      "loss": 1.1606,
+      "step": 12850
+    },
+    {
+      "epoch": 0.41008960744921713,
+      "grad_norm": 0.1756649762392044,
+      "learning_rate": 0.001,
+      "loss": 1.1413,
+      "step": 12860
+    },
+    {
+      "epoch": 0.41040849516885103,
+      "grad_norm": 0.1724313199520111,
+      "learning_rate": 0.001,
+      "loss": 1.1465,
+      "step": 12870
+    },
+    {
+      "epoch": 0.410727382888485,
+      "grad_norm": 0.17643587291240692,
+      "learning_rate": 0.001,
+      "loss": 1.1592,
+      "step": 12880
+    },
+    {
+      "epoch": 0.4110462706081189,
+      "grad_norm": 0.18537591397762299,
+      "learning_rate": 0.001,
+      "loss": 1.1638,
+      "step": 12890
+    },
+    {
+      "epoch": 0.4113651583277528,
+      "grad_norm": 0.1723945140838623,
+      "learning_rate": 0.001,
+      "loss": 1.1469,
+      "step": 12900
+    },
+    {
+      "epoch": 0.41168404604738673,
+      "grad_norm": 0.1771482527256012,
+      "learning_rate": 0.001,
+      "loss": 1.1432,
+      "step": 12910
+    },
+    {
+      "epoch": 0.41200293376702063,
+      "grad_norm": 0.1836037039756775,
+      "learning_rate": 0.001,
+      "loss": 1.1522,
+      "step": 12920
+    },
+    {
+      "epoch": 0.4123218214866545,
+      "grad_norm": 0.1788443922996521,
+      "learning_rate": 0.001,
+      "loss": 1.1467,
+      "step": 12930
+    },
+    {
+      "epoch": 0.4126407092062885,
+      "grad_norm": 0.18017710745334625,
+      "learning_rate": 0.001,
+      "loss": 1.1655,
+      "step": 12940
+    },
+    {
+      "epoch": 0.4129595969259224,
+      "grad_norm": 0.18141914904117584,
+      "learning_rate": 0.001,
+      "loss": 1.1614,
+      "step": 12950
+    },
+    {
+      "epoch": 0.4132784846455563,
+      "grad_norm": 0.1714852750301361,
+      "learning_rate": 0.001,
+      "loss": 1.1592,
+      "step": 12960
+    },
+    {
+      "epoch": 0.41359737236519023,
+      "grad_norm": 0.18428003787994385,
+      "learning_rate": 0.001,
+      "loss": 1.1645,
+      "step": 12970
+    },
+    {
+      "epoch": 0.41391626008482413,
+      "grad_norm": 0.18629293143749237,
+      "learning_rate": 0.001,
+      "loss": 1.1676,
+      "step": 12980
+    },
+    {
+      "epoch": 0.414235147804458,
+      "grad_norm": 0.18700791895389557,
+      "learning_rate": 0.001,
+      "loss": 1.1527,
+      "step": 12990
+    },
+    {
+      "epoch": 0.414554035524092,
+      "grad_norm": 0.17667566239833832,
+      "learning_rate": 0.001,
+      "loss": 1.154,
+      "step": 13000
+    },
+    {
+      "epoch": 0.4148729232437259,
+      "grad_norm": 0.17916272580623627,
+      "learning_rate": 0.001,
+      "loss": 1.1339,
+      "step": 13010
+    },
+    {
+      "epoch": 0.4151918109633598,
+      "grad_norm": 0.18255506455898285,
+      "learning_rate": 0.001,
+      "loss": 1.171,
+      "step": 13020
+    },
+    {
+      "epoch": 0.41551069868299373,
+      "grad_norm": 0.17822222411632538,
+      "learning_rate": 0.001,
+      "loss": 1.1656,
+      "step": 13030
+    },
+    {
+      "epoch": 0.4158295864026276,
+      "grad_norm": 0.1775846928358078,
+      "learning_rate": 0.001,
+      "loss": 1.1466,
+      "step": 13040
+    },
+    {
+      "epoch": 0.4161484741222615,
+      "grad_norm": 0.17505867779254913,
+      "learning_rate": 0.001,
+      "loss": 1.1528,
+      "step": 13050
+    },
+    {
+      "epoch": 0.4164673618418955,
+      "grad_norm": 0.18390125036239624,
+      "learning_rate": 0.001,
+      "loss": 1.1696,
+      "step": 13060
+    },
+    {
+      "epoch": 0.4167862495615294,
+      "grad_norm": 0.17525775730609894,
+      "learning_rate": 0.001,
+      "loss": 1.153,
+      "step": 13070
+    },
+    {
+      "epoch": 0.41710513728116333,
+      "grad_norm": 0.17498883605003357,
+      "learning_rate": 0.001,
+      "loss": 1.1583,
+      "step": 13080
+    },
+    {
+      "epoch": 0.4174240250007972,
+      "grad_norm": 0.18329474329948425,
+      "learning_rate": 0.001,
+      "loss": 1.1583,
+      "step": 13090
+    },
+    {
+      "epoch": 0.4177429127204311,
+      "grad_norm": 0.17313571274280548,
+      "learning_rate": 0.001,
+      "loss": 1.1477,
+      "step": 13100
+    },
+    {
+      "epoch": 0.4180618004400651,
+      "grad_norm": 0.1746223270893097,
+      "learning_rate": 0.001,
+      "loss": 1.1566,
+      "step": 13110
+    },
+    {
+      "epoch": 0.418380688159699,
+      "grad_norm": 0.18266509473323822,
+      "learning_rate": 0.001,
+      "loss": 1.1692,
+      "step": 13120
+    },
+    {
+      "epoch": 0.4186995758793329,
+      "grad_norm": 0.17773644626140594,
+      "learning_rate": 0.001,
+      "loss": 1.1571,
+      "step": 13130
+    },
+    {
+      "epoch": 0.41901846359896683,
+      "grad_norm": 0.1756133735179901,
+      "learning_rate": 0.001,
+      "loss": 1.1497,
+      "step": 13140
+    },
+    {
+      "epoch": 0.4193373513186007,
+      "grad_norm": 0.17332470417022705,
+      "learning_rate": 0.001,
+      "loss": 1.163,
+      "step": 13150
+    },
+    {
+      "epoch": 0.4196562390382346,
+      "grad_norm": 0.17600354552268982,
+      "learning_rate": 0.001,
+      "loss": 1.1471,
+      "step": 13160
+    },
+    {
+      "epoch": 0.4199751267578686,
+      "grad_norm": 0.17582380771636963,
+      "learning_rate": 0.001,
+      "loss": 1.1549,
+      "step": 13170
+    },
+    {
+      "epoch": 0.4202940144775025,
+      "grad_norm": 0.1780974119901657,
+      "learning_rate": 0.001,
+      "loss": 1.1389,
+      "step": 13180
+    },
+    {
+      "epoch": 0.4206129021971364,
+      "grad_norm": 0.18114261329174042,
+      "learning_rate": 0.001,
+      "loss": 1.1524,
+      "step": 13190
+    },
+    {
+      "epoch": 0.4209317899167703,
+      "grad_norm": 0.1800422966480255,
+      "learning_rate": 0.001,
+      "loss": 1.1694,
+      "step": 13200
+    },
+    {
+      "epoch": 0.4212506776364042,
+      "grad_norm": 0.17944826185703278,
+      "learning_rate": 0.001,
+      "loss": 1.1596,
+      "step": 13210
+    },
+    {
+      "epoch": 0.4215695653560381,
+      "grad_norm": 0.17792369425296783,
+      "learning_rate": 0.001,
+      "loss": 1.1504,
+      "step": 13220
+    },
+    {
+      "epoch": 0.4218884530756721,
+      "grad_norm": 0.17227941751480103,
+      "learning_rate": 0.001,
+      "loss": 1.1581,
+      "step": 13230
+    },
+    {
+      "epoch": 0.422207340795306,
+      "grad_norm": 0.18069201707839966,
+      "learning_rate": 0.001,
+      "loss": 1.1401,
+      "step": 13240
+    },
+    {
+      "epoch": 0.42252622851493987,
+      "grad_norm": 0.17398959398269653,
+      "learning_rate": 0.001,
+      "loss": 1.1329,
+      "step": 13250
+    },
+    {
+      "epoch": 0.4228451162345738,
+      "grad_norm": 0.17893588542938232,
+      "learning_rate": 0.001,
+      "loss": 1.1442,
+      "step": 13260
+    },
+    {
+      "epoch": 0.4231640039542077,
+      "grad_norm": 0.18032392859458923,
+      "learning_rate": 0.001,
+      "loss": 1.1578,
+      "step": 13270
+    },
+    {
+      "epoch": 0.4234828916738416,
+      "grad_norm": 0.17575973272323608,
+      "learning_rate": 0.001,
+      "loss": 1.1347,
+      "step": 13280
+    },
+    {
+      "epoch": 0.4238017793934756,
+      "grad_norm": 0.1713789403438568,
+      "learning_rate": 0.001,
+      "loss": 1.1405,
+      "step": 13290
+    },
+    {
+      "epoch": 0.4241206671131095,
+      "grad_norm": 0.17849451303482056,
+      "learning_rate": 0.001,
+      "loss": 1.1668,
+      "step": 13300
+    },
+    {
+      "epoch": 0.42443955483274337,
+      "grad_norm": 0.17444877326488495,
+      "learning_rate": 0.001,
+      "loss": 1.1665,
+      "step": 13310
+    },
+    {
+      "epoch": 0.4247584425523773,
+      "grad_norm": 0.1770418882369995,
+      "learning_rate": 0.001,
+      "loss": 1.1431,
+      "step": 13320
+    },
+    {
+      "epoch": 0.4250773302720112,
+      "grad_norm": 0.17722870409488678,
+      "learning_rate": 0.001,
+      "loss": 1.152,
+      "step": 13330
+    },
+    {
+      "epoch": 0.4253962179916451,
+      "grad_norm": 0.17850705981254578,
+      "learning_rate": 0.001,
+      "loss": 1.1465,
+      "step": 13340
+    },
+    {
+      "epoch": 0.4257151057112791,
+      "grad_norm": 0.18088005483150482,
+      "learning_rate": 0.001,
+      "loss": 1.1511,
+      "step": 13350
+    },
+    {
+      "epoch": 0.42603399343091297,
+      "grad_norm": 0.18208032846450806,
+      "learning_rate": 0.001,
+      "loss": 1.1367,
+      "step": 13360
+    },
+    {
+      "epoch": 0.42635288115054687,
+      "grad_norm": 0.18329893052577972,
+      "learning_rate": 0.001,
+      "loss": 1.1551,
+      "step": 13370
+    },
+    {
+      "epoch": 0.4266717688701808,
+      "grad_norm": 0.17308764159679413,
+      "learning_rate": 0.001,
+      "loss": 1.1335,
+      "step": 13380
+    },
+    {
+      "epoch": 0.4269906565898147,
+      "grad_norm": 0.17596715688705444,
+      "learning_rate": 0.001,
+      "loss": 1.1395,
+      "step": 13390
+    },
+    {
+      "epoch": 0.4273095443094486,
+      "grad_norm": 0.17568935453891754,
+      "learning_rate": 0.001,
+      "loss": 1.1381,
+      "step": 13400
+    },
+    {
+      "epoch": 0.4276284320290826,
+      "grad_norm": 0.1743965744972229,
+      "learning_rate": 0.001,
+      "loss": 1.1283,
+      "step": 13410
+    },
+    {
+      "epoch": 0.42794731974871647,
+      "grad_norm": 0.18175481259822845,
+      "learning_rate": 0.001,
+      "loss": 1.1598,
+      "step": 13420
+    },
+    {
+      "epoch": 0.42826620746835037,
+      "grad_norm": 0.17291270196437836,
+      "learning_rate": 0.001,
+      "loss": 1.1515,
+      "step": 13430
+    },
+    {
+      "epoch": 0.4285850951879843,
+      "grad_norm": 0.18205703794956207,
+      "learning_rate": 0.001,
+      "loss": 1.1602,
+      "step": 13440
+    },
+    {
+      "epoch": 0.4289039829076182,
+      "grad_norm": 0.1734030842781067,
+      "learning_rate": 0.001,
+      "loss": 1.1451,
+      "step": 13450
+    },
+    {
+      "epoch": 0.4292228706272521,
+      "grad_norm": 0.17287306487560272,
+      "learning_rate": 0.001,
+      "loss": 1.148,
+      "step": 13460
+    },
+    {
+      "epoch": 0.42954175834688607,
+      "grad_norm": 0.1736667901277542,
+      "learning_rate": 0.001,
+      "loss": 1.15,
+      "step": 13470
+    },
+    {
+      "epoch": 0.42986064606651997,
+      "grad_norm": 0.17618460953235626,
+      "learning_rate": 0.001,
+      "loss": 1.165,
+      "step": 13480
+    },
+    {
+      "epoch": 0.4301795337861539,
+      "grad_norm": 0.16635005176067352,
+      "learning_rate": 0.001,
+      "loss": 1.1414,
+      "step": 13490
+    },
+    {
+      "epoch": 0.4304984215057878,
+      "grad_norm": 0.17537757754325867,
+      "learning_rate": 0.001,
+      "loss": 1.1448,
+      "step": 13500
+    },
+    {
+      "epoch": 0.4308173092254217,
+      "grad_norm": 0.1771899312734604,
+      "learning_rate": 0.001,
+      "loss": 1.1423,
+      "step": 13510
+    },
+    {
+      "epoch": 0.43113619694505567,
+      "grad_norm": 0.17241476476192474,
+      "learning_rate": 0.001,
+      "loss": 1.1694,
+      "step": 13520
+    },
+    {
+      "epoch": 0.43145508466468957,
+      "grad_norm": 0.17704904079437256,
+      "learning_rate": 0.001,
+      "loss": 1.1405,
+      "step": 13530
+    },
+    {
+      "epoch": 0.43177397238432347,
+      "grad_norm": 0.17959144711494446,
+      "learning_rate": 0.001,
+      "loss": 1.15,
+      "step": 13540
+    },
+    {
+      "epoch": 0.4320928601039574,
+      "grad_norm": 0.17902454733848572,
+      "learning_rate": 0.001,
+      "loss": 1.1579,
+      "step": 13550
+    },
+    {
+      "epoch": 0.4324117478235913,
+      "grad_norm": 0.1775130033493042,
+      "learning_rate": 0.001,
+      "loss": 1.1479,
+      "step": 13560
+    },
+    {
+      "epoch": 0.4327306355432252,
+      "grad_norm": 0.17502056062221527,
+      "learning_rate": 0.001,
+      "loss": 1.1513,
+      "step": 13570
+    },
+    {
+      "epoch": 0.43304952326285917,
+      "grad_norm": 0.18220333755016327,
+      "learning_rate": 0.001,
+      "loss": 1.1607,
+      "step": 13580
+    },
+    {
+      "epoch": 0.43336841098249307,
+      "grad_norm": 0.1754351109266281,
+      "learning_rate": 0.001,
+      "loss": 1.1423,
+      "step": 13590
+    },
+    {
+      "epoch": 0.43368729870212697,
+      "grad_norm": 0.17384397983551025,
+      "learning_rate": 0.001,
+      "loss": 1.1395,
+      "step": 13600
+    },
+    {
+      "epoch": 0.4340061864217609,
+      "grad_norm": 0.17870260775089264,
+      "learning_rate": 0.001,
+      "loss": 1.1468,
+      "step": 13610
+    },
+    {
+      "epoch": 0.4343250741413948,
+      "grad_norm": 0.1795383095741272,
+      "learning_rate": 0.001,
+      "loss": 1.1653,
+      "step": 13620
+    },
+    {
+      "epoch": 0.4346439618610287,
+      "grad_norm": 0.1721075177192688,
+      "learning_rate": 0.001,
+      "loss": 1.1291,
+      "step": 13630
+    },
+    {
+      "epoch": 0.43496284958066267,
+      "grad_norm": 0.17633718252182007,
+      "learning_rate": 0.001,
+      "loss": 1.154,
+      "step": 13640
+    },
+    {
+      "epoch": 0.43528173730029657,
+      "grad_norm": 0.17714627087116241,
+      "learning_rate": 0.001,
+      "loss": 1.147,
+      "step": 13650
+    },
+    {
+      "epoch": 0.43560062501993047,
+      "grad_norm": 0.17243249714374542,
+      "learning_rate": 0.001,
+      "loss": 1.1351,
+      "step": 13660
+    },
+    {
+      "epoch": 0.4359195127395644,
+      "grad_norm": 0.17836323380470276,
+      "learning_rate": 0.001,
+      "loss": 1.1531,
+      "step": 13670
+    },
+    {
+      "epoch": 0.4362384004591983,
+      "grad_norm": 0.1793064922094345,
+      "learning_rate": 0.001,
+      "loss": 1.1409,
+      "step": 13680
+    },
+    {
+      "epoch": 0.4365572881788322,
+      "grad_norm": 0.17111380398273468,
+      "learning_rate": 0.001,
+      "loss": 1.1363,
+      "step": 13690
+    },
+    {
+      "epoch": 0.43687617589846617,
+      "grad_norm": 0.18050673604011536,
+      "learning_rate": 0.001,
+      "loss": 1.1549,
+      "step": 13700
+    },
+    {
+      "epoch": 0.43719506361810007,
+      "grad_norm": 0.17576226592063904,
+      "learning_rate": 0.001,
+      "loss": 1.1389,
+      "step": 13710
+    },
+    {
+      "epoch": 0.43751395133773396,
+      "grad_norm": 0.16893474757671356,
+      "learning_rate": 0.001,
+      "loss": 1.1222,
+      "step": 13720
+    },
+    {
+      "epoch": 0.4378328390573679,
+      "grad_norm": 0.17337141931056976,
+      "learning_rate": 0.001,
+      "loss": 1.1638,
+      "step": 13730
+    },
+    {
+      "epoch": 0.4381517267770018,
+      "grad_norm": 0.17256872355937958,
+      "learning_rate": 0.001,
+      "loss": 1.1492,
+      "step": 13740
+    },
+    {
+      "epoch": 0.4384706144966357,
+      "grad_norm": 0.17770156264305115,
+      "learning_rate": 0.001,
+      "loss": 1.1438,
+      "step": 13750
+    },
+    {
+      "epoch": 0.43878950221626967,
+      "grad_norm": 0.17507579922676086,
+      "learning_rate": 0.001,
+      "loss": 1.1391,
+      "step": 13760
+    },
+    {
+      "epoch": 0.43910838993590356,
+      "grad_norm": 0.18140773475170135,
+      "learning_rate": 0.001,
+      "loss": 1.1493,
+      "step": 13770
+    },
+    {
+      "epoch": 0.43942727765553746,
+      "grad_norm": 0.18639299273490906,
+      "learning_rate": 0.001,
+      "loss": 1.1545,
+      "step": 13780
+    },
+    {
+      "epoch": 0.4397461653751714,
+      "grad_norm": 0.176975280046463,
+      "learning_rate": 0.001,
+      "loss": 1.1519,
+      "step": 13790
+    },
+    {
+      "epoch": 0.4400650530948053,
+      "grad_norm": 0.17814040184020996,
+      "learning_rate": 0.001,
+      "loss": 1.1446,
+      "step": 13800
+    },
+    {
+      "epoch": 0.4403839408144392,
+      "grad_norm": 0.1774367243051529,
+      "learning_rate": 0.001,
+      "loss": 1.1355,
+      "step": 13810
+    },
+    {
+      "epoch": 0.44070282853407317,
+      "grad_norm": 0.18180528283119202,
+      "learning_rate": 0.001,
+      "loss": 1.1372,
+      "step": 13820
+    },
+    {
+      "epoch": 0.44102171625370706,
+      "grad_norm": 0.1772371381521225,
+      "learning_rate": 0.001,
+      "loss": 1.1409,
+      "step": 13830
+    },
+    {
+      "epoch": 0.44134060397334096,
+      "grad_norm": 0.18037815392017365,
+      "learning_rate": 0.001,
+      "loss": 1.1297,
+      "step": 13840
+    },
+    {
+      "epoch": 0.4416594916929749,
+      "grad_norm": 0.17408838868141174,
+      "learning_rate": 0.001,
+      "loss": 1.1401,
+      "step": 13850
+    },
+    {
+      "epoch": 0.4419783794126088,
+      "grad_norm": 0.1728648841381073,
+      "learning_rate": 0.001,
+      "loss": 1.1397,
+      "step": 13860
+    },
+    {
+      "epoch": 0.4422972671322427,
+      "grad_norm": 0.17422670125961304,
+      "learning_rate": 0.001,
+      "loss": 1.1305,
+      "step": 13870
+    },
+    {
+      "epoch": 0.44261615485187666,
+      "grad_norm": 0.17485372722148895,
+      "learning_rate": 0.001,
+      "loss": 1.1217,
+      "step": 13880
+    },
+    {
+      "epoch": 0.44293504257151056,
+      "grad_norm": 0.17043618857860565,
+      "learning_rate": 0.001,
+      "loss": 1.1286,
+      "step": 13890
+    },
+    {
+      "epoch": 0.4432539302911445,
+      "grad_norm": 0.17486435174942017,
+      "learning_rate": 0.001,
+      "loss": 1.1489,
+      "step": 13900
+    },
+    {
+      "epoch": 0.4435728180107784,
+      "grad_norm": 0.1703348606824875,
+      "learning_rate": 0.001,
+      "loss": 1.1316,
+      "step": 13910
+    },
+    {
+      "epoch": 0.4438917057304123,
+      "grad_norm": 0.17141640186309814,
+      "learning_rate": 0.001,
+      "loss": 1.1498,
+      "step": 13920
+    },
+    {
+      "epoch": 0.44421059345004626,
+      "grad_norm": 0.17244957387447357,
+      "learning_rate": 0.001,
+      "loss": 1.1553,
+      "step": 13930
+    },
+    {
+      "epoch": 0.44452948116968016,
+      "grad_norm": 0.17857715487480164,
+      "learning_rate": 0.001,
+      "loss": 1.1266,
+      "step": 13940
+    },
+    {
+      "epoch": 0.44484836888931406,
+      "grad_norm": 0.17840982973575592,
+      "learning_rate": 0.001,
+      "loss": 1.1417,
+      "step": 13950
+    },
+    {
+      "epoch": 0.445167256608948,
+      "grad_norm": 0.17459362745285034,
+      "learning_rate": 0.001,
+      "loss": 1.1621,
+      "step": 13960
+    },
+    {
+      "epoch": 0.4454861443285819,
+      "grad_norm": 0.17860184609889984,
+      "learning_rate": 0.001,
+      "loss": 1.1337,
+      "step": 13970
+    },
+    {
+      "epoch": 0.4458050320482158,
+      "grad_norm": 0.1751028597354889,
+      "learning_rate": 0.001,
+      "loss": 1.143,
+      "step": 13980
+    },
+    {
+      "epoch": 0.44612391976784976,
+      "grad_norm": 0.1695946753025055,
+      "learning_rate": 0.001,
+      "loss": 1.1387,
+      "step": 13990
+    },
+    {
+      "epoch": 0.44644280748748366,
+      "grad_norm": 0.18408192694187164,
+      "learning_rate": 0.001,
+      "loss": 1.1318,
+      "step": 14000
+    },
+    {
+      "epoch": 0.44676169520711756,
+      "grad_norm": 0.17113225162029266,
+      "learning_rate": 0.001,
+      "loss": 1.1492,
+      "step": 14010
+    },
+    {
+      "epoch": 0.4470805829267515,
+      "grad_norm": 0.17978601157665253,
+      "learning_rate": 0.001,
+      "loss": 1.1497,
+      "step": 14020
+    },
+    {
+      "epoch": 0.4473994706463854,
+      "grad_norm": 0.18019011616706848,
+      "learning_rate": 0.001,
+      "loss": 1.147,
+      "step": 14030
+    },
+    {
+      "epoch": 0.4477183583660193,
+      "grad_norm": 0.17102354764938354,
+      "learning_rate": 0.001,
+      "loss": 1.1339,
+      "step": 14040
+    },
+    {
+      "epoch": 0.44803724608565326,
+      "grad_norm": 0.17504438757896423,
+      "learning_rate": 0.001,
+      "loss": 1.1222,
+      "step": 14050
+    },
+    {
+      "epoch": 0.44835613380528716,
+      "grad_norm": 0.18110573291778564,
+      "learning_rate": 0.001,
+      "loss": 1.1286,
+      "step": 14060
+    },
+    {
+      "epoch": 0.44867502152492106,
+      "grad_norm": 0.17824122309684753,
+      "learning_rate": 0.001,
+      "loss": 1.1247,
+      "step": 14070
+    },
+    {
+      "epoch": 0.448993909244555,
+      "grad_norm": 0.17802254855632782,
+      "learning_rate": 0.001,
+      "loss": 1.1375,
+      "step": 14080
+    },
+    {
+      "epoch": 0.4493127969641889,
+      "grad_norm": 0.1781211942434311,
+      "learning_rate": 0.001,
+      "loss": 1.1347,
+      "step": 14090
+    },
+    {
+      "epoch": 0.4496316846838228,
+      "grad_norm": 0.17185848951339722,
+      "learning_rate": 0.001,
+      "loss": 1.1378,
+      "step": 14100
+    },
+    {
+      "epoch": 0.44995057240345676,
+      "grad_norm": 0.1789436936378479,
+      "learning_rate": 0.001,
+      "loss": 1.1426,
+      "step": 14110
+    },
+    {
+      "epoch": 0.45026946012309066,
+      "grad_norm": 0.17003895342350006,
+      "learning_rate": 0.001,
+      "loss": 1.1282,
+      "step": 14120
+    },
+    {
+      "epoch": 0.45058834784272456,
+      "grad_norm": 0.17219972610473633,
+      "learning_rate": 0.001,
+      "loss": 1.1324,
+      "step": 14130
+    },
+    {
+      "epoch": 0.4509072355623585,
+      "grad_norm": 0.1720394790172577,
+      "learning_rate": 0.001,
+      "loss": 1.1298,
+      "step": 14140
+    },
+    {
+      "epoch": 0.4512261232819924,
+      "grad_norm": 0.17621329426765442,
+      "learning_rate": 0.001,
+      "loss": 1.1602,
+      "step": 14150
+    },
+    {
+      "epoch": 0.4515450110016263,
+      "grad_norm": 0.17632067203521729,
+      "learning_rate": 0.001,
+      "loss": 1.1334,
+      "step": 14160
+    },
+    {
+      "epoch": 0.45186389872126026,
+      "grad_norm": 0.16688014566898346,
+      "learning_rate": 0.001,
+      "loss": 1.1238,
+      "step": 14170
+    },
+    {
+      "epoch": 0.45218278644089416,
+      "grad_norm": 0.17131859064102173,
+      "learning_rate": 0.001,
+      "loss": 1.1355,
+      "step": 14180
+    },
+    {
+      "epoch": 0.45250167416052806,
+      "grad_norm": 0.16830040514469147,
+      "learning_rate": 0.001,
+      "loss": 1.1219,
+      "step": 14190
+    },
+    {
+      "epoch": 0.452820561880162,
+      "grad_norm": 0.17178581655025482,
+      "learning_rate": 0.001,
+      "loss": 1.1489,
+      "step": 14200
+    },
+    {
+      "epoch": 0.4531394495997959,
+      "grad_norm": 0.17017489671707153,
+      "learning_rate": 0.001,
+      "loss": 1.1487,
+      "step": 14210
+    },
+    {
+      "epoch": 0.4534583373194298,
+      "grad_norm": 0.17112135887145996,
+      "learning_rate": 0.001,
+      "loss": 1.1372,
+      "step": 14220
+    },
+    {
+      "epoch": 0.45377722503906376,
+      "grad_norm": 0.1778554469347,
+      "learning_rate": 0.001,
+      "loss": 1.1248,
+      "step": 14230
+    },
+    {
+      "epoch": 0.45409611275869766,
+      "grad_norm": 0.17707771062850952,
+      "learning_rate": 0.001,
+      "loss": 1.1243,
+      "step": 14240
+    },
+    {
+      "epoch": 0.45441500047833155,
+      "grad_norm": 0.17464545369148254,
+      "learning_rate": 0.001,
+      "loss": 1.1392,
+      "step": 14250
+    },
+    {
+      "epoch": 0.4547338881979655,
+      "grad_norm": 0.17195813357830048,
+      "learning_rate": 0.001,
+      "loss": 1.1511,
+      "step": 14260
+    },
+    {
+      "epoch": 0.4550527759175994,
+      "grad_norm": 0.1783856600522995,
+      "learning_rate": 0.001,
+      "loss": 1.1345,
+      "step": 14270
+    },
+    {
+      "epoch": 0.4553716636372333,
+      "grad_norm": 0.1746705174446106,
+      "learning_rate": 0.001,
+      "loss": 1.1424,
+      "step": 14280
+    },
+    {
+      "epoch": 0.45569055135686726,
+      "grad_norm": 0.18295544385910034,
+      "learning_rate": 0.001,
+      "loss": 1.1479,
+      "step": 14290
+    },
+    {
+      "epoch": 0.45600943907650116,
+      "grad_norm": 0.16954278945922852,
+      "learning_rate": 0.001,
+      "loss": 1.1473,
+      "step": 14300
+    },
+    {
+      "epoch": 0.45632832679613505,
+      "grad_norm": 0.17534016072750092,
+      "learning_rate": 0.001,
+      "loss": 1.148,
+      "step": 14310
+    },
+    {
+      "epoch": 0.456647214515769,
+      "grad_norm": 0.17889931797981262,
+      "learning_rate": 0.001,
+      "loss": 1.1374,
+      "step": 14320
+    },
+    {
+      "epoch": 0.4569661022354029,
+      "grad_norm": 0.17737379670143127,
+      "learning_rate": 0.001,
+      "loss": 1.1311,
+      "step": 14330
+    },
+    {
+      "epoch": 0.45728498995503686,
+      "grad_norm": 0.17604157328605652,
+      "learning_rate": 0.001,
+      "loss": 1.1262,
+      "step": 14340
+    },
+    {
+      "epoch": 0.45760387767467076,
+      "grad_norm": 0.17472808063030243,
+      "learning_rate": 0.001,
+      "loss": 1.1366,
+      "step": 14350
+    },
+    {
+      "epoch": 0.45792276539430465,
+      "grad_norm": 0.1687580943107605,
+      "learning_rate": 0.001,
+      "loss": 1.1317,
+      "step": 14360
+    },
+    {
+      "epoch": 0.4582416531139386,
+      "grad_norm": 0.1738176941871643,
+      "learning_rate": 0.001,
+      "loss": 1.1597,
+      "step": 14370
+    },
+    {
+      "epoch": 0.4585605408335725,
+      "grad_norm": 0.17493782937526703,
+      "learning_rate": 0.001,
+      "loss": 1.1475,
+      "step": 14380
+    },
+    {
+      "epoch": 0.4588794285532064,
+      "grad_norm": 0.17473384737968445,
+      "learning_rate": 0.001,
+      "loss": 1.1285,
+      "step": 14390
+    },
+    {
+      "epoch": 0.45919831627284036,
+      "grad_norm": 0.17252366244792938,
+      "learning_rate": 0.001,
+      "loss": 1.1527,
+      "step": 14400
+    },
+    {
+      "epoch": 0.45951720399247425,
+      "grad_norm": 0.17336592078208923,
+      "learning_rate": 0.001,
+      "loss": 1.1074,
+      "step": 14410
+    },
+    {
+      "epoch": 0.45983609171210815,
+      "grad_norm": 0.1781364381313324,
+      "learning_rate": 0.001,
+      "loss": 1.1315,
+      "step": 14420
+    },
+    {
+      "epoch": 0.4601549794317421,
+      "grad_norm": 0.1762470006942749,
+      "learning_rate": 0.001,
+      "loss": 1.1317,
+      "step": 14430
+    },
+    {
+      "epoch": 0.460473867151376,
+      "grad_norm": 0.17392881214618683,
+      "learning_rate": 0.001,
+      "loss": 1.1483,
+      "step": 14440
+    },
+    {
+      "epoch": 0.4607927548710099,
+      "grad_norm": 0.17682306468486786,
+      "learning_rate": 0.001,
+      "loss": 1.1258,
+      "step": 14450
+    },
+    {
+      "epoch": 0.46111164259064386,
+      "grad_norm": 0.1736285388469696,
+      "learning_rate": 0.001,
+      "loss": 1.1405,
+      "step": 14460
+    },
+    {
+      "epoch": 0.46143053031027775,
+      "grad_norm": 0.1780506670475006,
+      "learning_rate": 0.001,
+      "loss": 1.1301,
+      "step": 14470
+    },
+    {
+      "epoch": 0.46174941802991165,
+      "grad_norm": 0.17654019594192505,
+      "learning_rate": 0.001,
+      "loss": 1.119,
+      "step": 14480
+    },
+    {
+      "epoch": 0.4620683057495456,
+      "grad_norm": 0.17297044396400452,
+      "learning_rate": 0.001,
+      "loss": 1.131,
+      "step": 14490
+    },
+    {
+      "epoch": 0.4623871934691795,
+      "grad_norm": 0.17622877657413483,
+      "learning_rate": 0.001,
+      "loss": 1.1345,
+      "step": 14500
+    },
+    {
+      "epoch": 0.4627060811888134,
+      "grad_norm": 0.18307894468307495,
+      "learning_rate": 0.001,
+      "loss": 1.1295,
+      "step": 14510
+    },
+    {
+      "epoch": 0.46302496890844735,
+      "grad_norm": 0.17374959588050842,
+      "learning_rate": 0.001,
+      "loss": 1.1271,
+      "step": 14520
+    },
+    {
+      "epoch": 0.46334385662808125,
+      "grad_norm": 0.18093709647655487,
+      "learning_rate": 0.001,
+      "loss": 1.156,
+      "step": 14530
+    },
+    {
+      "epoch": 0.46366274434771515,
+      "grad_norm": 0.16554683446884155,
+      "learning_rate": 0.001,
+      "loss": 1.1208,
+      "step": 14540
+    },
+    {
+      "epoch": 0.4639816320673491,
+      "grad_norm": 0.1772506684064865,
+      "learning_rate": 0.001,
+      "loss": 1.1486,
+      "step": 14550
+    },
+    {
+      "epoch": 0.464300519786983,
+      "grad_norm": 0.1779564768075943,
+      "learning_rate": 0.001,
+      "loss": 1.135,
+      "step": 14560
+    },
+    {
+      "epoch": 0.4646194075066169,
+      "grad_norm": 0.17387722432613373,
+      "learning_rate": 0.001,
+      "loss": 1.1349,
+      "step": 14570
+    },
+    {
+      "epoch": 0.46493829522625085,
+      "grad_norm": 0.1813112050294876,
+      "learning_rate": 0.001,
+      "loss": 1.1294,
+      "step": 14580
+    },
+    {
+      "epoch": 0.46525718294588475,
+      "grad_norm": 0.17628876864910126,
+      "learning_rate": 0.001,
+      "loss": 1.1273,
+      "step": 14590
+    },
+    {
+      "epoch": 0.46557607066551865,
+      "grad_norm": 0.17620135843753815,
+      "learning_rate": 0.001,
+      "loss": 1.1346,
+      "step": 14600
+    },
+    {
+      "epoch": 0.4658949583851526,
+      "grad_norm": 0.17300722002983093,
+      "learning_rate": 0.001,
+      "loss": 1.1397,
+      "step": 14610
+    },
+    {
+      "epoch": 0.4662138461047865,
+      "grad_norm": 0.1759226769208908,
+      "learning_rate": 0.001,
+      "loss": 1.1327,
+      "step": 14620
+    },
+    {
+      "epoch": 0.4665327338244204,
+      "grad_norm": 0.17027615010738373,
+      "learning_rate": 0.001,
+      "loss": 1.1353,
+      "step": 14630
+    },
+    {
+      "epoch": 0.46685162154405435,
+      "grad_norm": 0.17422300577163696,
+      "learning_rate": 0.001,
+      "loss": 1.1139,
+      "step": 14640
+    },
+    {
+      "epoch": 0.46717050926368825,
+      "grad_norm": 0.1712965965270996,
+      "learning_rate": 0.001,
+      "loss": 1.1373,
+      "step": 14650
+    },
+    {
+      "epoch": 0.46748939698332215,
+      "grad_norm": 0.17639024555683136,
+      "learning_rate": 0.001,
+      "loss": 1.1278,
+      "step": 14660
+    },
+    {
+      "epoch": 0.4678082847029561,
+      "grad_norm": 0.17248843610286713,
+      "learning_rate": 0.001,
+      "loss": 1.1298,
+      "step": 14670
+    },
+    {
+      "epoch": 0.46812717242259,
+      "grad_norm": 0.1724853813648224,
+      "learning_rate": 0.001,
+      "loss": 1.1185,
+      "step": 14680
+    },
+    {
+      "epoch": 0.4684460601422239,
+      "grad_norm": 0.1779608428478241,
+      "learning_rate": 0.001,
+      "loss": 1.1258,
+      "step": 14690
+    },
+    {
+      "epoch": 0.46876494786185785,
+      "grad_norm": 0.17226381599903107,
+      "learning_rate": 0.001,
+      "loss": 1.1276,
+      "step": 14700
+    },
+    {
+      "epoch": 0.46908383558149175,
+      "grad_norm": 0.1835961490869522,
+      "learning_rate": 0.001,
+      "loss": 1.1435,
+      "step": 14710
+    },
+    {
+      "epoch": 0.46940272330112565,
+      "grad_norm": 0.17311832308769226,
+      "learning_rate": 0.001,
+      "loss": 1.1353,
+      "step": 14720
+    },
+    {
+      "epoch": 0.4697216110207596,
+      "grad_norm": 0.1746002435684204,
+      "learning_rate": 0.001,
+      "loss": 1.1284,
+      "step": 14730
+    },
+    {
+      "epoch": 0.4700404987403935,
+      "grad_norm": 0.1797454059123993,
+      "learning_rate": 0.001,
+      "loss": 1.1441,
+      "step": 14740
+    },
+    {
+      "epoch": 0.47035938646002745,
+      "grad_norm": 0.1763915866613388,
+      "learning_rate": 0.001,
+      "loss": 1.1451,
+      "step": 14750
+    },
+    {
+      "epoch": 0.47067827417966135,
+      "grad_norm": 0.16806401312351227,
+      "learning_rate": 0.001,
+      "loss": 1.1159,
+      "step": 14760
+    },
+    {
+      "epoch": 0.47099716189929525,
+      "grad_norm": 0.17495940625667572,
+      "learning_rate": 0.001,
+      "loss": 1.1063,
+      "step": 14770
+    },
+    {
+      "epoch": 0.4713160496189292,
+      "grad_norm": 0.16822218894958496,
+      "learning_rate": 0.001,
+      "loss": 1.1362,
+      "step": 14780
+    },
+    {
+      "epoch": 0.4716349373385631,
+      "grad_norm": 0.18633721768856049,
+      "learning_rate": 0.001,
+      "loss": 1.1432,
+      "step": 14790
+    },
+    {
+      "epoch": 0.471953825058197,
+      "grad_norm": 0.17266841232776642,
+      "learning_rate": 0.001,
+      "loss": 1.1402,
+      "step": 14800
+    },
+    {
+      "epoch": 0.47227271277783095,
+      "grad_norm": 0.1750502586364746,
+      "learning_rate": 0.001,
+      "loss": 1.1287,
+      "step": 14810
+    },
+    {
+      "epoch": 0.47259160049746485,
+      "grad_norm": 0.17796385288238525,
+      "learning_rate": 0.001,
+      "loss": 1.1362,
+      "step": 14820
+    },
+    {
+      "epoch": 0.47291048821709875,
+      "grad_norm": 0.17094792425632477,
+      "learning_rate": 0.001,
+      "loss": 1.1424,
+      "step": 14830
+    },
+    {
+      "epoch": 0.4732293759367327,
+      "grad_norm": 0.1744205206632614,
+      "learning_rate": 0.001,
+      "loss": 1.143,
+      "step": 14840
+    },
+    {
+      "epoch": 0.4735482636563666,
+      "grad_norm": 0.16905353963375092,
+      "learning_rate": 0.001,
+      "loss": 1.1341,
+      "step": 14850
+    },
+    {
+      "epoch": 0.4738671513760005,
+      "grad_norm": 0.17296987771987915,
+      "learning_rate": 0.001,
+      "loss": 1.121,
+      "step": 14860
+    },
+    {
+      "epoch": 0.47418603909563445,
+      "grad_norm": 0.17449218034744263,
+      "learning_rate": 0.001,
+      "loss": 1.1205,
+      "step": 14870
+    },
+    {
+      "epoch": 0.47450492681526835,
+      "grad_norm": 0.17594942450523376,
+      "learning_rate": 0.001,
+      "loss": 1.1289,
+      "step": 14880
+    },
+    {
+      "epoch": 0.47482381453490224,
+      "grad_norm": 0.17545631527900696,
+      "learning_rate": 0.001,
+      "loss": 1.1348,
+      "step": 14890
+    },
+    {
+      "epoch": 0.4751427022545362,
+      "grad_norm": 0.17231062054634094,
+      "learning_rate": 0.001,
+      "loss": 1.1314,
+      "step": 14900
+    },
+    {
+      "epoch": 0.4754615899741701,
+      "grad_norm": 0.17671747505664825,
+      "learning_rate": 0.001,
+      "loss": 1.1464,
+      "step": 14910
+    },
+    {
+      "epoch": 0.475780477693804,
+      "grad_norm": 0.17096923291683197,
+      "learning_rate": 0.001,
+      "loss": 1.1296,
+      "step": 14920
+    },
+    {
+      "epoch": 0.47609936541343795,
+      "grad_norm": 0.1766541302204132,
+      "learning_rate": 0.001,
+      "loss": 1.138,
+      "step": 14930
+    },
+    {
+      "epoch": 0.47641825313307185,
+      "grad_norm": 0.172738716006279,
+      "learning_rate": 0.001,
+      "loss": 1.1477,
+      "step": 14940
+    },
+    {
+      "epoch": 0.47673714085270574,
+      "grad_norm": 0.17469772696495056,
+      "learning_rate": 0.001,
+      "loss": 1.1195,
+      "step": 14950
+    },
+    {
+      "epoch": 0.4770560285723397,
+      "grad_norm": 0.1674947440624237,
+      "learning_rate": 0.001,
+      "loss": 1.1367,
+      "step": 14960
+    },
+    {
+      "epoch": 0.4773749162919736,
+      "grad_norm": 0.1782265305519104,
+      "learning_rate": 0.001,
+      "loss": 1.1374,
+      "step": 14970
+    },
+    {
+      "epoch": 0.4776938040116075,
+      "grad_norm": 0.17530570924282074,
+      "learning_rate": 0.001,
+      "loss": 1.122,
+      "step": 14980
+    },
+    {
+      "epoch": 0.47801269173124145,
+      "grad_norm": 0.16821470856666565,
+      "learning_rate": 0.001,
+      "loss": 1.116,
+      "step": 14990
+    },
+    {
+      "epoch": 0.47833157945087534,
+      "grad_norm": 0.17344245314598083,
+      "learning_rate": 0.001,
+      "loss": 1.1444,
+      "step": 15000
+    },
+    {
+      "epoch": 0.47865046717050924,
+      "grad_norm": 0.17519459128379822,
+      "learning_rate": 0.001,
+      "loss": 1.1424,
+      "step": 15010
+    },
+    {
+      "epoch": 0.4789693548901432,
+      "grad_norm": 0.17834389209747314,
+      "learning_rate": 0.001,
+      "loss": 1.1181,
+      "step": 15020
+    },
+    {
+      "epoch": 0.4792882426097771,
+      "grad_norm": 0.1712467074394226,
+      "learning_rate": 0.001,
+      "loss": 1.1416,
+      "step": 15030
+    },
+    {
+      "epoch": 0.479607130329411,
+      "grad_norm": 0.16900737583637238,
+      "learning_rate": 0.001,
+      "loss": 1.1191,
+      "step": 15040
+    },
+    {
+      "epoch": 0.47992601804904494,
+      "grad_norm": 0.17559660971164703,
+      "learning_rate": 0.001,
+      "loss": 1.1333,
+      "step": 15050
+    },
+    {
+      "epoch": 0.48024490576867884,
+      "grad_norm": 0.17230333387851715,
+      "learning_rate": 0.001,
+      "loss": 1.1261,
+      "step": 15060
+    },
+    {
+      "epoch": 0.48056379348831274,
+      "grad_norm": 0.1761026233434677,
+      "learning_rate": 0.001,
+      "loss": 1.1384,
+      "step": 15070
+    },
+    {
+      "epoch": 0.4808826812079467,
+      "grad_norm": 0.17489396035671234,
+      "learning_rate": 0.001,
+      "loss": 1.1341,
+      "step": 15080
+    },
+    {
+      "epoch": 0.4812015689275806,
+      "grad_norm": 0.1864793449640274,
+      "learning_rate": 0.001,
+      "loss": 1.1252,
+      "step": 15090
+    },
+    {
+      "epoch": 0.4815204566472145,
+      "grad_norm": 0.1766740083694458,
+      "learning_rate": 0.001,
+      "loss": 1.1406,
+      "step": 15100
+    },
+    {
+      "epoch": 0.48183934436684844,
+      "grad_norm": 0.17469963431358337,
+      "learning_rate": 0.001,
+      "loss": 1.1129,
+      "step": 15110
+    },
+    {
+      "epoch": 0.48215823208648234,
+      "grad_norm": 0.17255161702632904,
+      "learning_rate": 0.001,
+      "loss": 1.1078,
+      "step": 15120
+    },
+    {
+      "epoch": 0.48247711980611624,
+      "grad_norm": 0.17505139112472534,
+      "learning_rate": 0.001,
+      "loss": 1.1173,
+      "step": 15130
+    },
+    {
+      "epoch": 0.4827960075257502,
+      "grad_norm": 0.17324578762054443,
+      "learning_rate": 0.001,
+      "loss": 1.126,
+      "step": 15140
+    },
+    {
+      "epoch": 0.4831148952453841,
+      "grad_norm": 0.1717669516801834,
+      "learning_rate": 0.001,
+      "loss": 1.1094,
+      "step": 15150
+    },
+    {
+      "epoch": 0.48343378296501804,
+      "grad_norm": 0.1693733185529709,
+      "learning_rate": 0.001,
+      "loss": 1.1206,
+      "step": 15160
+    },
+    {
+      "epoch": 0.48375267068465194,
+      "grad_norm": 0.1819078028202057,
+      "learning_rate": 0.001,
+      "loss": 1.1259,
+      "step": 15170
+    },
+    {
+      "epoch": 0.48407155840428584,
+      "grad_norm": 0.18116608262062073,
+      "learning_rate": 0.001,
+      "loss": 1.1226,
+      "step": 15180
+    },
+    {
+      "epoch": 0.4843904461239198,
+      "grad_norm": 0.1702233999967575,
+      "learning_rate": 0.001,
+      "loss": 1.1141,
+      "step": 15190
+    },
+    {
+      "epoch": 0.4847093338435537,
+      "grad_norm": 0.16998600959777832,
+      "learning_rate": 0.001,
+      "loss": 1.1519,
+      "step": 15200
+    },
+    {
+      "epoch": 0.4850282215631876,
+      "grad_norm": 0.17265725135803223,
+      "learning_rate": 0.001,
+      "loss": 1.1135,
+      "step": 15210
+    },
+    {
+      "epoch": 0.48534710928282154,
+      "grad_norm": 0.1677086502313614,
+      "learning_rate": 0.001,
+      "loss": 1.1275,
+      "step": 15220
+    },
+    {
+      "epoch": 0.48566599700245544,
+      "grad_norm": 0.1745789349079132,
+      "learning_rate": 0.001,
+      "loss": 1.1331,
+      "step": 15230
+    },
+    {
+      "epoch": 0.48598488472208934,
+      "grad_norm": 0.1719202697277069,
+      "learning_rate": 0.001,
+      "loss": 1.1365,
+      "step": 15240
+    },
+    {
+      "epoch": 0.4863037724417233,
+      "grad_norm": 0.17276684939861298,
+      "learning_rate": 0.001,
+      "loss": 1.1176,
+      "step": 15250
+    },
+    {
+      "epoch": 0.4866226601613572,
+      "grad_norm": 0.17195747792720795,
+      "learning_rate": 0.001,
+      "loss": 1.1238,
+      "step": 15260
+    },
+    {
+      "epoch": 0.4869415478809911,
+      "grad_norm": 0.16910310089588165,
+      "learning_rate": 0.001,
+      "loss": 1.1206,
+      "step": 15270
+    },
+    {
+      "epoch": 0.48726043560062504,
+      "grad_norm": 0.1696513593196869,
+      "learning_rate": 0.001,
+      "loss": 1.1058,
+      "step": 15280
+    },
+    {
+      "epoch": 0.48757932332025894,
+      "grad_norm": 0.170016348361969,
+      "learning_rate": 0.001,
+      "loss": 1.1272,
+      "step": 15290
+    },
+    {
+      "epoch": 0.48789821103989284,
+      "grad_norm": 0.16925400495529175,
+      "learning_rate": 0.001,
+      "loss": 1.1151,
+      "step": 15300
+    },
+    {
+      "epoch": 0.4882170987595268,
+      "grad_norm": 0.17601697146892548,
+      "learning_rate": 0.001,
+      "loss": 1.1072,
+      "step": 15310
+    },
+    {
+      "epoch": 0.4885359864791607,
+      "grad_norm": 0.17617474496364594,
+      "learning_rate": 0.001,
+      "loss": 1.1257,
+      "step": 15320
+    },
+    {
+      "epoch": 0.4888548741987946,
+      "grad_norm": 0.17635928094387054,
+      "learning_rate": 0.001,
+      "loss": 1.126,
+      "step": 15330
+    },
+    {
+      "epoch": 0.48917376191842854,
+      "grad_norm": 0.17766179144382477,
+      "learning_rate": 0.001,
+      "loss": 1.1111,
+      "step": 15340
+    },
+    {
+      "epoch": 0.48949264963806244,
+      "grad_norm": 0.16769689321517944,
+      "learning_rate": 0.001,
+      "loss": 1.1188,
+      "step": 15350
+    },
+    {
+      "epoch": 0.48981153735769634,
+      "grad_norm": 0.17344802618026733,
+      "learning_rate": 0.001,
+      "loss": 1.1076,
+      "step": 15360
+    },
+    {
+      "epoch": 0.4901304250773303,
+      "grad_norm": 0.17091096937656403,
+      "learning_rate": 0.001,
+      "loss": 1.1342,
+      "step": 15370
+    },
+    {
+      "epoch": 0.4904493127969642,
+      "grad_norm": 0.1739739030599594,
+      "learning_rate": 0.001,
+      "loss": 1.1275,
+      "step": 15380
+    },
+    {
+      "epoch": 0.4907682005165981,
+      "grad_norm": 0.17534139752388,
+      "learning_rate": 0.001,
+      "loss": 1.1352,
+      "step": 15390
+    },
+    {
+      "epoch": 0.49108708823623204,
+      "grad_norm": 0.17085367441177368,
+      "learning_rate": 0.001,
+      "loss": 1.1117,
+      "step": 15400
+    },
+    {
+      "epoch": 0.49140597595586594,
+      "grad_norm": 0.17168210446834564,
+      "learning_rate": 0.001,
+      "loss": 1.1283,
+      "step": 15410
+    },
+    {
+      "epoch": 0.49172486367549983,
+      "grad_norm": 0.1765969693660736,
+      "learning_rate": 0.001,
+      "loss": 1.1319,
+      "step": 15420
+    },
+    {
+      "epoch": 0.4920437513951338,
+      "grad_norm": 0.1726970672607422,
+      "learning_rate": 0.001,
+      "loss": 1.1264,
+      "step": 15430
+    },
+    {
+      "epoch": 0.4923626391147677,
+      "grad_norm": 0.16879424452781677,
+      "learning_rate": 0.001,
+      "loss": 1.1417,
+      "step": 15440
+    },
+    {
+      "epoch": 0.4926815268344016,
+      "grad_norm": 0.17404453456401825,
+      "learning_rate": 0.001,
+      "loss": 1.1529,
+      "step": 15450
+    },
+    {
+      "epoch": 0.49300041455403554,
+      "grad_norm": 0.16755877435207367,
+      "learning_rate": 0.001,
+      "loss": 1.14,
+      "step": 15460
+    },
+    {
+      "epoch": 0.49331930227366944,
+      "grad_norm": 0.17902107536792755,
+      "learning_rate": 0.001,
+      "loss": 1.1458,
+      "step": 15470
+    },
+    {
+      "epoch": 0.49363818999330333,
+      "grad_norm": 0.16922469437122345,
+      "learning_rate": 0.001,
+      "loss": 1.114,
+      "step": 15480
+    },
+    {
+      "epoch": 0.4939570777129373,
+      "grad_norm": 0.1739080250263214,
+      "learning_rate": 0.001,
+      "loss": 1.1414,
+      "step": 15490
+    },
+    {
+      "epoch": 0.4942759654325712,
+      "grad_norm": 0.17348088324069977,
+      "learning_rate": 0.001,
+      "loss": 1.1178,
+      "step": 15500
+    },
+    {
+      "epoch": 0.4945948531522051,
+      "grad_norm": 0.1783573478460312,
+      "learning_rate": 0.001,
+      "loss": 1.1251,
+      "step": 15510
+    },
+    {
+      "epoch": 0.49491374087183904,
+      "grad_norm": 0.17203998565673828,
+      "learning_rate": 0.001,
+      "loss": 1.1375,
+      "step": 15520
+    },
+    {
+      "epoch": 0.49523262859147293,
+      "grad_norm": 0.17022830247879028,
+      "learning_rate": 0.001,
+      "loss": 1.1246,
+      "step": 15530
+    },
+    {
+      "epoch": 0.49555151631110683,
+      "grad_norm": 0.16698627173900604,
+      "learning_rate": 0.001,
+      "loss": 1.1226,
+      "step": 15540
+    },
+    {
+      "epoch": 0.4958704040307408,
+      "grad_norm": 0.17425251007080078,
+      "learning_rate": 0.001,
+      "loss": 1.1304,
+      "step": 15550
+    },
+    {
+      "epoch": 0.4961892917503747,
+      "grad_norm": 0.17078517377376556,
+      "learning_rate": 0.001,
+      "loss": 1.1178,
+      "step": 15560
+    },
+    {
+      "epoch": 0.49650817947000864,
+      "grad_norm": 0.16847193241119385,
+      "learning_rate": 0.001,
+      "loss": 1.1301,
+      "step": 15570
+    },
+    {
+      "epoch": 0.49682706718964254,
+      "grad_norm": 0.17440183460712433,
+      "learning_rate": 0.001,
+      "loss": 1.133,
+      "step": 15580
+    },
+    {
+      "epoch": 0.49714595490927643,
+      "grad_norm": 0.17285966873168945,
+      "learning_rate": 0.001,
+      "loss": 1.1362,
+      "step": 15590
+    },
+    {
+      "epoch": 0.4974648426289104,
+      "grad_norm": 0.1749267727136612,
+      "learning_rate": 0.001,
+      "loss": 1.1195,
+      "step": 15600
+    },
+    {
+      "epoch": 0.4977837303485443,
+      "grad_norm": 0.17195169627666473,
+      "learning_rate": 0.001,
+      "loss": 1.1165,
+      "step": 15610
+    },
+    {
+      "epoch": 0.4981026180681782,
+      "grad_norm": 0.1727495789527893,
+      "learning_rate": 0.001,
+      "loss": 1.1264,
+      "step": 15620
+    },
+    {
+      "epoch": 0.49842150578781214,
+      "grad_norm": 0.17486001551151276,
+      "learning_rate": 0.001,
+      "loss": 1.1146,
+      "step": 15630
+    },
+    {
+      "epoch": 0.49874039350744603,
+      "grad_norm": 0.17329901456832886,
+      "learning_rate": 0.001,
+      "loss": 1.1258,
+      "step": 15640
+    },
+    {
+      "epoch": 0.49905928122707993,
+      "grad_norm": 0.17345182597637177,
+      "learning_rate": 0.001,
+      "loss": 1.1229,
+      "step": 15650
+    },
+    {
+      "epoch": 0.4993781689467139,
+      "grad_norm": 0.17703242599964142,
+      "learning_rate": 0.001,
+      "loss": 1.1452,
+      "step": 15660
+    },
+    {
+      "epoch": 0.4996970566663478,
+      "grad_norm": 0.17290346324443817,
+      "learning_rate": 0.001,
+      "loss": 1.1236,
+      "step": 15670
+    },
+    {
+      "epoch": 0.5000159443859817,
+      "grad_norm": 0.17458419501781464,
+      "learning_rate": 0.001,
+      "loss": 1.1279,
+      "step": 15680
+    },
+    {
+      "epoch": 0.5003348321056156,
+      "grad_norm": 0.17450138926506042,
+      "learning_rate": 0.001,
+      "loss": 1.1207,
+      "step": 15690
+    },
+    {
+      "epoch": 0.5006537198252495,
+      "grad_norm": 0.17006513476371765,
+      "learning_rate": 0.001,
+      "loss": 1.1216,
+      "step": 15700
+    },
+    {
+      "epoch": 0.5009726075448835,
+      "grad_norm": 0.1693337857723236,
+      "learning_rate": 0.001,
+      "loss": 1.1245,
+      "step": 15710
+    },
+    {
+      "epoch": 0.5012914952645173,
+      "grad_norm": 0.16920679807662964,
+      "learning_rate": 0.001,
+      "loss": 1.1386,
+      "step": 15720
+    },
+    {
+      "epoch": 0.5016103829841513,
+      "grad_norm": 0.17201140522956848,
+      "learning_rate": 0.001,
+      "loss": 1.1135,
+      "step": 15730
+    },
+    {
+      "epoch": 0.5019292707037852,
+      "grad_norm": 0.1762687861919403,
+      "learning_rate": 0.001,
+      "loss": 1.1462,
+      "step": 15740
+    },
+    {
+      "epoch": 0.5022481584234191,
+      "grad_norm": 0.17598646879196167,
+      "learning_rate": 0.001,
+      "loss": 1.1259,
+      "step": 15750
+    },
+    {
+      "epoch": 0.502567046143053,
+      "grad_norm": 0.17038299143314362,
+      "learning_rate": 0.001,
+      "loss": 1.109,
+      "step": 15760
+    },
+    {
+      "epoch": 0.502885933862687,
+      "grad_norm": 0.18112069368362427,
+      "learning_rate": 0.001,
+      "loss": 1.1322,
+      "step": 15770
+    },
+    {
+      "epoch": 0.5032048215823208,
+      "grad_norm": 0.1772948056459427,
+      "learning_rate": 0.001,
+      "loss": 1.1253,
+      "step": 15780
+    },
+    {
+      "epoch": 0.5035237093019548,
+      "grad_norm": 0.16839788854122162,
+      "learning_rate": 0.001,
+      "loss": 1.1199,
+      "step": 15790
+    },
+    {
+      "epoch": 0.5038425970215887,
+      "grad_norm": 0.17266593873500824,
+      "learning_rate": 0.001,
+      "loss": 1.1166,
+      "step": 15800
+    },
+    {
+      "epoch": 0.5041614847412226,
+      "grad_norm": 0.1770016849040985,
+      "learning_rate": 0.001,
+      "loss": 1.1319,
+      "step": 15810
+    },
+    {
+      "epoch": 0.5044803724608565,
+      "grad_norm": 0.16987302899360657,
+      "learning_rate": 0.001,
+      "loss": 1.1176,
+      "step": 15820
+    },
+    {
+      "epoch": 0.5047992601804905,
+      "grad_norm": 0.16873012483119965,
+      "learning_rate": 0.001,
+      "loss": 1.1114,
+      "step": 15830
+    },
+    {
+      "epoch": 0.5051181479001243,
+      "grad_norm": 0.17353899776935577,
+      "learning_rate": 0.001,
+      "loss": 1.1298,
+      "step": 15840
+    },
+    {
+      "epoch": 0.5054370356197583,
+      "grad_norm": 0.17694109678268433,
+      "learning_rate": 0.001,
+      "loss": 1.1314,
+      "step": 15850
+    },
+    {
+      "epoch": 0.5057559233393922,
+      "grad_norm": 0.17422708868980408,
+      "learning_rate": 0.001,
+      "loss": 1.1295,
+      "step": 15860
+    },
+    {
+      "epoch": 0.5060748110590261,
+      "grad_norm": 0.17685523629188538,
+      "learning_rate": 0.001,
+      "loss": 1.1136,
+      "step": 15870
+    },
+    {
+      "epoch": 0.50639369877866,
+      "grad_norm": 0.16311325132846832,
+      "learning_rate": 0.001,
+      "loss": 1.1278,
+      "step": 15880
+    },
+    {
+      "epoch": 0.506712586498294,
+      "grad_norm": 0.1693374216556549,
+      "learning_rate": 0.001,
+      "loss": 1.1117,
+      "step": 15890
+    },
+    {
+      "epoch": 0.5070314742179278,
+      "grad_norm": 0.1808386892080307,
+      "learning_rate": 0.001,
+      "loss": 1.1157,
+      "step": 15900
+    },
+    {
+      "epoch": 0.5073503619375618,
+      "grad_norm": 0.17110781371593475,
+      "learning_rate": 0.001,
+      "loss": 1.1191,
+      "step": 15910
+    },
+    {
+      "epoch": 0.5076692496571957,
+      "grad_norm": 0.1752341389656067,
+      "learning_rate": 0.001,
+      "loss": 1.1173,
+      "step": 15920
+    },
+    {
+      "epoch": 0.5079881373768296,
+      "grad_norm": 0.17590519785881042,
+      "learning_rate": 0.001,
+      "loss": 1.123,
+      "step": 15930
+    },
+    {
+      "epoch": 0.5083070250964635,
+      "grad_norm": 0.16483011841773987,
+      "learning_rate": 0.001,
+      "loss": 1.114,
+      "step": 15940
+    },
+    {
+      "epoch": 0.5086259128160975,
+      "grad_norm": 0.1718084067106247,
+      "learning_rate": 0.001,
+      "loss": 1.1205,
+      "step": 15950
+    },
+    {
+      "epoch": 0.5089448005357313,
+      "grad_norm": 0.17069004476070404,
+      "learning_rate": 0.001,
+      "loss": 1.1278,
+      "step": 15960
+    },
+    {
+      "epoch": 0.5092636882553653,
+      "grad_norm": 0.1754763126373291,
+      "learning_rate": 0.001,
+      "loss": 1.1312,
+      "step": 15970
+    },
+    {
+      "epoch": 0.5095825759749992,
+      "grad_norm": 0.1660955846309662,
+      "learning_rate": 0.001,
+      "loss": 1.1153,
+      "step": 15980
+    },
+    {
+      "epoch": 0.5099014636946331,
+      "grad_norm": 0.16823166608810425,
+      "learning_rate": 0.001,
+      "loss": 1.1049,
+      "step": 15990
+    },
+    {
+      "epoch": 0.510220351414267,
+      "grad_norm": 0.17043384909629822,
+      "learning_rate": 0.001,
+      "loss": 1.1104,
+      "step": 16000
+    },
+    {
+      "epoch": 0.510539239133901,
+      "grad_norm": 0.17292456328868866,
+      "learning_rate": 0.001,
+      "loss": 1.119,
+      "step": 16010
+    },
+    {
+      "epoch": 0.5108581268535348,
+      "grad_norm": 0.17111191153526306,
+      "learning_rate": 0.001,
+      "loss": 1.1254,
+      "step": 16020
+    },
+    {
+      "epoch": 0.5111770145731688,
+      "grad_norm": 0.16945838928222656,
+      "learning_rate": 0.001,
+      "loss": 1.099,
+      "step": 16030
+    },
+    {
+      "epoch": 0.5114959022928027,
+      "grad_norm": 0.17108163237571716,
+      "learning_rate": 0.001,
+      "loss": 1.1369,
+      "step": 16040
+    },
+    {
+      "epoch": 0.5118147900124366,
+      "grad_norm": 0.17080144584178925,
+      "learning_rate": 0.001,
+      "loss": 1.1138,
+      "step": 16050
+    },
+    {
+      "epoch": 0.5121336777320705,
+      "grad_norm": 0.17457976937294006,
+      "learning_rate": 0.001,
+      "loss": 1.1342,
+      "step": 16060
+    },
+    {
+      "epoch": 0.5124525654517045,
+      "grad_norm": 0.17557039856910706,
+      "learning_rate": 0.001,
+      "loss": 1.0928,
+      "step": 16070
+    },
+    {
+      "epoch": 0.5127714531713383,
+      "grad_norm": 0.17394962906837463,
+      "learning_rate": 0.001,
+      "loss": 1.113,
+      "step": 16080
+    },
+    {
+      "epoch": 0.5130903408909723,
+      "grad_norm": 0.17363347113132477,
+      "learning_rate": 0.001,
+      "loss": 1.1287,
+      "step": 16090
+    },
+    {
+      "epoch": 0.5134092286106062,
+      "grad_norm": 0.17071598768234253,
+      "learning_rate": 0.001,
+      "loss": 1.1229,
+      "step": 16100
+    },
+    {
+      "epoch": 0.5137281163302401,
+      "grad_norm": 0.17206592857837677,
+      "learning_rate": 0.001,
+      "loss": 1.1182,
+      "step": 16110
+    },
+    {
+      "epoch": 0.514047004049874,
+      "grad_norm": 0.16682256758213043,
+      "learning_rate": 0.001,
+      "loss": 1.1171,
+      "step": 16120
+    },
+    {
+      "epoch": 0.514365891769508,
+      "grad_norm": 0.16921396553516388,
+      "learning_rate": 0.001,
+      "loss": 1.1136,
+      "step": 16130
+    },
+    {
+      "epoch": 0.5146847794891418,
+      "grad_norm": 0.17058788239955902,
+      "learning_rate": 0.001,
+      "loss": 1.1172,
+      "step": 16140
+    },
+    {
+      "epoch": 0.5150036672087758,
+      "grad_norm": 0.17264387011528015,
+      "learning_rate": 0.001,
+      "loss": 1.1265,
+      "step": 16150
+    },
+    {
+      "epoch": 0.5153225549284097,
+      "grad_norm": 0.1710854172706604,
+      "learning_rate": 0.001,
+      "loss": 1.1195,
+      "step": 16160
+    },
+    {
+      "epoch": 0.5156414426480436,
+      "grad_norm": 0.18194888532161713,
+      "learning_rate": 0.001,
+      "loss": 1.1026,
+      "step": 16170
+    },
+    {
+      "epoch": 0.5159603303676775,
+      "grad_norm": 0.17689311504364014,
+      "learning_rate": 0.001,
+      "loss": 1.106,
+      "step": 16180
+    },
+    {
+      "epoch": 0.5162792180873115,
+      "grad_norm": 0.16910506784915924,
+      "learning_rate": 0.001,
+      "loss": 1.1127,
+      "step": 16190
+    },
+    {
+      "epoch": 0.5165981058069454,
+      "grad_norm": 0.16860614717006683,
+      "learning_rate": 0.001,
+      "loss": 1.1228,
+      "step": 16200
+    },
+    {
+      "epoch": 0.5169169935265793,
+      "grad_norm": 0.17780789732933044,
+      "learning_rate": 0.001,
+      "loss": 1.1163,
+      "step": 16210
+    },
+    {
+      "epoch": 0.5172358812462132,
+      "grad_norm": 0.1708582043647766,
+      "learning_rate": 0.001,
+      "loss": 1.1167,
+      "step": 16220
+    },
+    {
+      "epoch": 0.5175547689658472,
+      "grad_norm": 0.16993635892868042,
+      "learning_rate": 0.001,
+      "loss": 1.1147,
+      "step": 16230
+    },
+    {
+      "epoch": 0.517873656685481,
+      "grad_norm": 0.17231173813343048,
+      "learning_rate": 0.001,
+      "loss": 1.1149,
+      "step": 16240
+    },
+    {
+      "epoch": 0.518192544405115,
+      "grad_norm": 0.17344434559345245,
+      "learning_rate": 0.001,
+      "loss": 1.1283,
+      "step": 16250
+    },
+    {
+      "epoch": 0.5185114321247489,
+      "grad_norm": 0.17418132722377777,
+      "learning_rate": 0.001,
+      "loss": 1.1177,
+      "step": 16260
+    },
+    {
+      "epoch": 0.5188303198443828,
+      "grad_norm": 0.1698276549577713,
+      "learning_rate": 0.001,
+      "loss": 1.1043,
+      "step": 16270
+    },
+    {
+      "epoch": 0.5191492075640167,
+      "grad_norm": 0.16663305461406708,
+      "learning_rate": 0.001,
+      "loss": 1.1038,
+      "step": 16280
+    },
+    {
+      "epoch": 0.5194680952836507,
+      "grad_norm": 0.1778172254562378,
+      "learning_rate": 0.001,
+      "loss": 1.1117,
+      "step": 16290
+    },
+    {
+      "epoch": 0.5197869830032845,
+      "grad_norm": 0.1671564280986786,
+      "learning_rate": 0.001,
+      "loss": 1.1325,
+      "step": 16300
+    },
+    {
+      "epoch": 0.5201058707229185,
+      "grad_norm": 0.17214436829090118,
+      "learning_rate": 0.001,
+      "loss": 1.1304,
+      "step": 16310
+    },
+    {
+      "epoch": 0.5204247584425524,
+      "grad_norm": 0.1681654304265976,
+      "learning_rate": 0.001,
+      "loss": 1.113,
+      "step": 16320
+    },
+    {
+      "epoch": 0.5207436461621863,
+      "grad_norm": 0.1728064864873886,
+      "learning_rate": 0.001,
+      "loss": 1.1161,
+      "step": 16330
+    },
+    {
+      "epoch": 0.5210625338818202,
+      "grad_norm": 0.17555077373981476,
+      "learning_rate": 0.001,
+      "loss": 1.1304,
+      "step": 16340
+    },
+    {
+      "epoch": 0.5213814216014542,
+      "grad_norm": 0.17947399616241455,
+      "learning_rate": 0.001,
+      "loss": 1.1088,
+      "step": 16350
+    },
+    {
+      "epoch": 0.521700309321088,
+      "grad_norm": 0.16473524272441864,
+      "learning_rate": 0.001,
+      "loss": 1.0931,
+      "step": 16360
+    },
+    {
+      "epoch": 0.522019197040722,
+      "grad_norm": 0.17586977779865265,
+      "learning_rate": 0.001,
+      "loss": 1.1067,
+      "step": 16370
+    },
+    {
+      "epoch": 0.5223380847603559,
+      "grad_norm": 0.16663765907287598,
+      "learning_rate": 0.001,
+      "loss": 1.1028,
+      "step": 16380
+    },
+    {
+      "epoch": 0.5226569724799898,
+      "grad_norm": 0.17395268380641937,
+      "learning_rate": 0.001,
+      "loss": 1.1454,
+      "step": 16390
+    },
+    {
+      "epoch": 0.5229758601996237,
+      "grad_norm": 0.1764364093542099,
+      "learning_rate": 0.001,
+      "loss": 1.1085,
+      "step": 16400
+    },
+    {
+      "epoch": 0.5232947479192577,
+      "grad_norm": 0.17778083682060242,
+      "learning_rate": 0.001,
+      "loss": 1.1096,
+      "step": 16410
+    },
+    {
+      "epoch": 0.5236136356388915,
+      "grad_norm": 0.1722065955400467,
+      "learning_rate": 0.001,
+      "loss": 1.1066,
+      "step": 16420
+    },
+    {
+      "epoch": 0.5239325233585255,
+      "grad_norm": 0.17221935093402863,
+      "learning_rate": 0.001,
+      "loss": 1.1341,
+      "step": 16430
+    },
+    {
+      "epoch": 0.5242514110781594,
+      "grad_norm": 0.1714521199464798,
+      "learning_rate": 0.001,
+      "loss": 1.1257,
+      "step": 16440
+    },
+    {
+      "epoch": 0.5245702987977933,
+      "grad_norm": 0.17682762444019318,
+      "learning_rate": 0.001,
+      "loss": 1.1281,
+      "step": 16450
+    },
+    {
+      "epoch": 0.5248891865174272,
+      "grad_norm": 0.17583733797073364,
+      "learning_rate": 0.001,
+      "loss": 1.118,
+      "step": 16460
+    },
+    {
+      "epoch": 0.5252080742370612,
+      "grad_norm": 0.172967791557312,
+      "learning_rate": 0.001,
+      "loss": 1.1237,
+      "step": 16470
+    },
+    {
+      "epoch": 0.525526961956695,
+      "grad_norm": 0.16899310052394867,
+      "learning_rate": 0.001,
+      "loss": 1.1354,
+      "step": 16480
+    },
+    {
+      "epoch": 0.525845849676329,
+      "grad_norm": 0.1808301955461502,
+      "learning_rate": 0.001,
+      "loss": 1.1272,
+      "step": 16490
+    },
+    {
+      "epoch": 0.5261647373959629,
+      "grad_norm": 0.17048123478889465,
+      "learning_rate": 0.001,
+      "loss": 1.0992,
+      "step": 16500
+    },
+    {
+      "epoch": 0.5264836251155968,
+      "grad_norm": 0.16553248465061188,
+      "learning_rate": 0.001,
+      "loss": 1.104,
+      "step": 16510
+    },
+    {
+      "epoch": 0.5268025128352307,
+      "grad_norm": 0.1753700077533722,
+      "learning_rate": 0.001,
+      "loss": 1.119,
+      "step": 16520
+    },
+    {
+      "epoch": 0.5271214005548647,
+      "grad_norm": 0.16744454205036163,
+      "learning_rate": 0.001,
+      "loss": 1.1107,
+      "step": 16530
+    },
+    {
+      "epoch": 0.5274402882744985,
+      "grad_norm": 0.17314577102661133,
+      "learning_rate": 0.001,
+      "loss": 1.1442,
+      "step": 16540
+    },
+    {
+      "epoch": 0.5277591759941325,
+      "grad_norm": 0.16656264662742615,
+      "learning_rate": 0.001,
+      "loss": 1.0986,
+      "step": 16550
+    },
+    {
+      "epoch": 0.5280780637137664,
+      "grad_norm": 0.1720547080039978,
+      "learning_rate": 0.001,
+      "loss": 1.1202,
+      "step": 16560
+    },
+    {
+      "epoch": 0.5283969514334003,
+      "grad_norm": 0.17319488525390625,
+      "learning_rate": 0.001,
+      "loss": 1.1054,
+      "step": 16570
+    },
+    {
+      "epoch": 0.5287158391530342,
+      "grad_norm": 0.1704757660627365,
+      "learning_rate": 0.001,
+      "loss": 1.0962,
+      "step": 16580
+    },
+    {
+      "epoch": 0.5290347268726682,
+      "grad_norm": 0.17540287971496582,
+      "learning_rate": 0.001,
+      "loss": 1.1066,
+      "step": 16590
+    },
+    {
+      "epoch": 0.529353614592302,
+      "grad_norm": 0.16792285442352295,
+      "learning_rate": 0.001,
+      "loss": 1.1111,
+      "step": 16600
+    },
+    {
+      "epoch": 0.529672502311936,
+      "grad_norm": 0.1729079931974411,
+      "learning_rate": 0.001,
+      "loss": 1.1015,
+      "step": 16610
+    },
+    {
+      "epoch": 0.5299913900315699,
+      "grad_norm": 0.16947157680988312,
+      "learning_rate": 0.001,
+      "loss": 1.1383,
+      "step": 16620
+    },
+    {
+      "epoch": 0.5303102777512038,
+      "grad_norm": 0.17262917757034302,
+      "learning_rate": 0.001,
+      "loss": 1.1178,
+      "step": 16630
+    },
+    {
+      "epoch": 0.5306291654708377,
+      "grad_norm": 0.1714227795600891,
+      "learning_rate": 0.001,
+      "loss": 1.1154,
+      "step": 16640
+    },
+    {
+      "epoch": 0.5309480531904717,
+      "grad_norm": 0.16897515952587128,
+      "learning_rate": 0.001,
+      "loss": 1.1197,
+      "step": 16650
+    },
+    {
+      "epoch": 0.5312669409101055,
+      "grad_norm": 0.167450949549675,
+      "learning_rate": 0.001,
+      "loss": 1.118,
+      "step": 16660
+    },
+    {
+      "epoch": 0.5315858286297395,
+      "grad_norm": 0.16560305655002594,
+      "learning_rate": 0.001,
+      "loss": 1.0995,
+      "step": 16670
+    },
+    {
+      "epoch": 0.5319047163493734,
+      "grad_norm": 0.16930148005485535,
+      "learning_rate": 0.001,
+      "loss": 1.1458,
+      "step": 16680
+    },
+    {
+      "epoch": 0.5322236040690073,
+      "grad_norm": 0.16743934154510498,
+      "learning_rate": 0.001,
+      "loss": 1.1063,
+      "step": 16690
+    },
+    {
+      "epoch": 0.5325424917886412,
+      "grad_norm": 0.17878761887550354,
+      "learning_rate": 0.001,
+      "loss": 1.1078,
+      "step": 16700
+    },
+    {
+      "epoch": 0.5328613795082752,
+      "grad_norm": 0.166554257273674,
+      "learning_rate": 0.001,
+      "loss": 1.1031,
+      "step": 16710
+    },
+    {
+      "epoch": 0.533180267227909,
+      "grad_norm": 0.17427927255630493,
+      "learning_rate": 0.001,
+      "loss": 1.1299,
+      "step": 16720
+    },
+    {
+      "epoch": 0.533499154947543,
+      "grad_norm": 0.16897770762443542,
+      "learning_rate": 0.001,
+      "loss": 1.1172,
+      "step": 16730
+    },
+    {
+      "epoch": 0.5338180426671769,
+      "grad_norm": 0.16430383920669556,
+      "learning_rate": 0.001,
+      "loss": 1.1102,
+      "step": 16740
+    },
+    {
+      "epoch": 0.5341369303868108,
+      "grad_norm": 0.16883912682533264,
+      "learning_rate": 0.001,
+      "loss": 1.1037,
+      "step": 16750
+    },
+    {
+      "epoch": 0.5344558181064447,
+      "grad_norm": 0.1662365049123764,
+      "learning_rate": 0.001,
+      "loss": 1.1239,
+      "step": 16760
+    },
+    {
+      "epoch": 0.5347747058260787,
+      "grad_norm": 0.16690349578857422,
+      "learning_rate": 0.001,
+      "loss": 1.1167,
+      "step": 16770
+    },
+    {
+      "epoch": 0.5350935935457125,
+      "grad_norm": 0.17271411418914795,
+      "learning_rate": 0.001,
+      "loss": 1.1211,
+      "step": 16780
+    },
+    {
+      "epoch": 0.5354124812653465,
+      "grad_norm": 0.17028991878032684,
+      "learning_rate": 0.001,
+      "loss": 1.1196,
+      "step": 16790
+    },
+    {
+      "epoch": 0.5357313689849804,
+      "grad_norm": 0.17466682195663452,
+      "learning_rate": 0.001,
+      "loss": 1.1097,
+      "step": 16800
+    },
+    {
+      "epoch": 0.5360502567046143,
+      "grad_norm": 0.17412136495113373,
+      "learning_rate": 0.001,
+      "loss": 1.1349,
+      "step": 16810
+    },
+    {
+      "epoch": 0.5363691444242482,
+      "grad_norm": 0.17085425555706024,
+      "learning_rate": 0.001,
+      "loss": 1.1139,
+      "step": 16820
+    },
+    {
+      "epoch": 0.5366880321438822,
+      "grad_norm": 0.1674584150314331,
+      "learning_rate": 0.001,
+      "loss": 1.111,
+      "step": 16830
+    },
+    {
+      "epoch": 0.537006919863516,
+      "grad_norm": 0.17785713076591492,
+      "learning_rate": 0.001,
+      "loss": 1.1393,
+      "step": 16840
+    },
+    {
+      "epoch": 0.53732580758315,
+      "grad_norm": 0.17034348845481873,
+      "learning_rate": 0.001,
+      "loss": 1.1224,
+      "step": 16850
+    },
+    {
+      "epoch": 0.5376446953027839,
+      "grad_norm": 0.1680379956960678,
+      "learning_rate": 0.001,
+      "loss": 1.1181,
+      "step": 16860
+    },
+    {
+      "epoch": 0.5379635830224178,
+      "grad_norm": 0.17434479296207428,
+      "learning_rate": 0.001,
+      "loss": 1.108,
+      "step": 16870
+    },
+    {
+      "epoch": 0.5382824707420517,
+      "grad_norm": 0.1710536777973175,
+      "learning_rate": 0.001,
+      "loss": 1.0958,
+      "step": 16880
+    },
+    {
+      "epoch": 0.5386013584616857,
+      "grad_norm": 0.1662193238735199,
+      "learning_rate": 0.001,
+      "loss": 1.1238,
+      "step": 16890
+    },
+    {
+      "epoch": 0.5389202461813195,
+      "grad_norm": 0.16588041186332703,
+      "learning_rate": 0.001,
+      "loss": 1.1196,
+      "step": 16900
+    },
+    {
+      "epoch": 0.5392391339009535,
+      "grad_norm": 0.16906344890594482,
+      "learning_rate": 0.001,
+      "loss": 1.1005,
+      "step": 16910
+    },
+    {
+      "epoch": 0.5395580216205874,
+      "grad_norm": 0.16948343813419342,
+      "learning_rate": 0.001,
+      "loss": 1.1265,
+      "step": 16920
+    },
+    {
+      "epoch": 0.5398769093402213,
+      "grad_norm": 0.17260994017124176,
+      "learning_rate": 0.001,
+      "loss": 1.1098,
+      "step": 16930
+    },
+    {
+      "epoch": 0.5401957970598552,
+      "grad_norm": 0.16991999745368958,
+      "learning_rate": 0.001,
+      "loss": 1.1175,
+      "step": 16940
+    },
+    {
+      "epoch": 0.5405146847794892,
+      "grad_norm": 0.17070095241069794,
+      "learning_rate": 0.001,
+      "loss": 1.1061,
+      "step": 16950
+    },
+    {
+      "epoch": 0.540833572499123,
+      "grad_norm": 0.17698924243450165,
+      "learning_rate": 0.001,
+      "loss": 1.1124,
+      "step": 16960
+    },
+    {
+      "epoch": 0.541152460218757,
+      "grad_norm": 0.1651303917169571,
+      "learning_rate": 0.001,
+      "loss": 1.1122,
+      "step": 16970
+    },
+    {
+      "epoch": 0.5414713479383909,
+      "grad_norm": 0.17084306478500366,
+      "learning_rate": 0.001,
+      "loss": 1.1069,
+      "step": 16980
+    },
+    {
+      "epoch": 0.5417902356580248,
+      "grad_norm": 0.1684514433145523,
+      "learning_rate": 0.001,
+      "loss": 1.1282,
+      "step": 16990
+    },
+    {
+      "epoch": 0.5421091233776587,
+      "grad_norm": 0.17040887475013733,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 17000
+    },
+    {
+      "epoch": 0.5424280110972927,
+      "grad_norm": 0.16266708076000214,
+      "learning_rate": 0.001,
+      "loss": 1.0929,
+      "step": 17010
+    },
+    {
+      "epoch": 0.5427468988169266,
+      "grad_norm": 0.1671130657196045,
+      "learning_rate": 0.001,
+      "loss": 1.0842,
+      "step": 17020
+    },
+    {
+      "epoch": 0.5430657865365605,
+      "grad_norm": 0.17035824060440063,
+      "learning_rate": 0.001,
+      "loss": 1.1192,
+      "step": 17030
+    },
+    {
+      "epoch": 0.5433846742561944,
+      "grad_norm": 0.16444888710975647,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 17040
+    },
+    {
+      "epoch": 0.5437035619758284,
+      "grad_norm": 0.16725726425647736,
+      "learning_rate": 0.001,
+      "loss": 1.0976,
+      "step": 17050
+    },
+    {
+      "epoch": 0.5440224496954622,
+      "grad_norm": 0.16788360476493835,
+      "learning_rate": 0.001,
+      "loss": 1.1231,
+      "step": 17060
+    },
+    {
+      "epoch": 0.5443413374150962,
+      "grad_norm": 0.1691126525402069,
+      "learning_rate": 0.001,
+      "loss": 1.1243,
+      "step": 17070
+    },
+    {
+      "epoch": 0.5446602251347301,
+      "grad_norm": 0.17056216299533844,
+      "learning_rate": 0.001,
+      "loss": 1.1253,
+      "step": 17080
+    },
+    {
+      "epoch": 0.544979112854364,
+      "grad_norm": 0.17879615724086761,
+      "learning_rate": 0.001,
+      "loss": 1.1165,
+      "step": 17090
+    },
+    {
+      "epoch": 0.5452980005739979,
+      "grad_norm": 0.17294946312904358,
+      "learning_rate": 0.001,
+      "loss": 1.1075,
+      "step": 17100
+    },
+    {
+      "epoch": 0.5456168882936319,
+      "grad_norm": 0.16905255615711212,
+      "learning_rate": 0.001,
+      "loss": 1.1197,
+      "step": 17110
+    },
+    {
+      "epoch": 0.5459357760132657,
+      "grad_norm": 0.17166034877300262,
+      "learning_rate": 0.001,
+      "loss": 1.1225,
+      "step": 17120
+    },
+    {
+      "epoch": 0.5462546637328997,
+      "grad_norm": 0.1762578934431076,
+      "learning_rate": 0.001,
+      "loss": 1.1156,
+      "step": 17130
+    },
+    {
+      "epoch": 0.5465735514525336,
+      "grad_norm": 0.17545616626739502,
+      "learning_rate": 0.001,
+      "loss": 1.1052,
+      "step": 17140
+    },
+    {
+      "epoch": 0.5468924391721675,
+      "grad_norm": 0.17133742570877075,
+      "learning_rate": 0.001,
+      "loss": 1.1096,
+      "step": 17150
+    },
+    {
+      "epoch": 0.5472113268918014,
+      "grad_norm": 0.16983424127101898,
+      "learning_rate": 0.001,
+      "loss": 1.1073,
+      "step": 17160
+    },
+    {
+      "epoch": 0.5475302146114354,
+      "grad_norm": 0.16490906476974487,
+      "learning_rate": 0.001,
+      "loss": 1.1016,
+      "step": 17170
+    },
+    {
+      "epoch": 0.5478491023310692,
+      "grad_norm": 0.17924512922763824,
+      "learning_rate": 0.001,
+      "loss": 1.1254,
+      "step": 17180
+    },
+    {
+      "epoch": 0.5481679900507032,
+      "grad_norm": 0.17216329276561737,
+      "learning_rate": 0.001,
+      "loss": 1.1011,
+      "step": 17190
+    },
+    {
+      "epoch": 0.5484868777703371,
+      "grad_norm": 0.1712387204170227,
+      "learning_rate": 0.001,
+      "loss": 1.1206,
+      "step": 17200
+    },
+    {
+      "epoch": 0.548805765489971,
+      "grad_norm": 0.1688283532857895,
+      "learning_rate": 0.001,
+      "loss": 1.1118,
+      "step": 17210
+    },
+    {
+      "epoch": 0.5491246532096049,
+      "grad_norm": 0.17269356548786163,
+      "learning_rate": 0.001,
+      "loss": 1.1031,
+      "step": 17220
+    },
+    {
+      "epoch": 0.5494435409292389,
+      "grad_norm": 0.167024627327919,
+      "learning_rate": 0.001,
+      "loss": 1.0996,
+      "step": 17230
+    },
+    {
+      "epoch": 0.5497624286488727,
+      "grad_norm": 0.17057278752326965,
+      "learning_rate": 0.001,
+      "loss": 1.1167,
+      "step": 17240
+    },
+    {
+      "epoch": 0.5500813163685067,
+      "grad_norm": 0.1737520545721054,
+      "learning_rate": 0.001,
+      "loss": 1.1163,
+      "step": 17250
+    },
+    {
+      "epoch": 0.5504002040881406,
+      "grad_norm": 0.1737147867679596,
+      "learning_rate": 0.001,
+      "loss": 1.1085,
+      "step": 17260
+    },
+    {
+      "epoch": 0.5507190918077745,
+      "grad_norm": 0.17490079998970032,
+      "learning_rate": 0.001,
+      "loss": 1.1343,
+      "step": 17270
+    },
+    {
+      "epoch": 0.5510379795274084,
+      "grad_norm": 0.16712409257888794,
+      "learning_rate": 0.001,
+      "loss": 1.0932,
+      "step": 17280
+    },
+    {
+      "epoch": 0.5513568672470424,
+      "grad_norm": 0.1729220300912857,
+      "learning_rate": 0.001,
+      "loss": 1.1298,
+      "step": 17290
+    },
+    {
+      "epoch": 0.5516757549666762,
+      "grad_norm": 0.17015966773033142,
+      "learning_rate": 0.001,
+      "loss": 1.1075,
+      "step": 17300
+    },
+    {
+      "epoch": 0.5519946426863102,
+      "grad_norm": 0.1695127785205841,
+      "learning_rate": 0.001,
+      "loss": 1.1123,
+      "step": 17310
+    },
+    {
+      "epoch": 0.5523135304059441,
+      "grad_norm": 0.16665105521678925,
+      "learning_rate": 0.001,
+      "loss": 1.1105,
+      "step": 17320
+    },
+    {
+      "epoch": 0.552632418125578,
+      "grad_norm": 0.16556957364082336,
+      "learning_rate": 0.001,
+      "loss": 1.0983,
+      "step": 17330
+    },
+    {
+      "epoch": 0.5529513058452119,
+      "grad_norm": 0.1715242713689804,
+      "learning_rate": 0.001,
+      "loss": 1.1118,
+      "step": 17340
+    },
+    {
+      "epoch": 0.5532701935648459,
+      "grad_norm": 0.166187584400177,
+      "learning_rate": 0.001,
+      "loss": 1.1107,
+      "step": 17350
+    },
+    {
+      "epoch": 0.5535890812844797,
+      "grad_norm": 0.1719885617494583,
+      "learning_rate": 0.001,
+      "loss": 1.0866,
+      "step": 17360
+    },
+    {
+      "epoch": 0.5539079690041137,
+      "grad_norm": 0.16900859773159027,
+      "learning_rate": 0.001,
+      "loss": 1.1024,
+      "step": 17370
+    },
+    {
+      "epoch": 0.5542268567237476,
+      "grad_norm": 0.17110826075077057,
+      "learning_rate": 0.001,
+      "loss": 1.1181,
+      "step": 17380
+    },
+    {
+      "epoch": 0.5545457444433814,
+      "grad_norm": 0.1805061548948288,
+      "learning_rate": 0.001,
+      "loss": 1.1095,
+      "step": 17390
+    },
+    {
+      "epoch": 0.5548646321630154,
+      "grad_norm": 0.16934873163700104,
+      "learning_rate": 0.001,
+      "loss": 1.1336,
+      "step": 17400
+    },
+    {
+      "epoch": 0.5551835198826494,
+      "grad_norm": 0.17308500409126282,
+      "learning_rate": 0.001,
+      "loss": 1.1165,
+      "step": 17410
+    },
+    {
+      "epoch": 0.5555024076022832,
+      "grad_norm": 0.17419050633907318,
+      "learning_rate": 0.001,
+      "loss": 1.0959,
+      "step": 17420
+    },
+    {
+      "epoch": 0.5558212953219172,
+      "grad_norm": 0.17350651323795319,
+      "learning_rate": 0.001,
+      "loss": 1.1082,
+      "step": 17430
+    },
+    {
+      "epoch": 0.5561401830415511,
+      "grad_norm": 0.1707797795534134,
+      "learning_rate": 0.001,
+      "loss": 1.1107,
+      "step": 17440
+    },
+    {
+      "epoch": 0.556459070761185,
+      "grad_norm": 0.16700954735279083,
+      "learning_rate": 0.001,
+      "loss": 1.1141,
+      "step": 17450
+    },
+    {
+      "epoch": 0.5567779584808189,
+      "grad_norm": 0.1669778972864151,
+      "learning_rate": 0.001,
+      "loss": 1.113,
+      "step": 17460
+    },
+    {
+      "epoch": 0.5570968462004529,
+      "grad_norm": 0.16680364310741425,
+      "learning_rate": 0.001,
+      "loss": 1.1146,
+      "step": 17470
+    },
+    {
+      "epoch": 0.5574157339200867,
+      "grad_norm": 0.16998562216758728,
+      "learning_rate": 0.001,
+      "loss": 1.0902,
+      "step": 17480
+    },
+    {
+      "epoch": 0.5577346216397207,
+      "grad_norm": 0.16862469911575317,
+      "learning_rate": 0.001,
+      "loss": 1.1176,
+      "step": 17490
+    },
+    {
+      "epoch": 0.5580535093593546,
+      "grad_norm": 0.16966623067855835,
+      "learning_rate": 0.001,
+      "loss": 1.1113,
+      "step": 17500
+    },
+    {
+      "epoch": 0.5583723970789884,
+      "grad_norm": 0.1724562793970108,
+      "learning_rate": 0.001,
+      "loss": 1.1101,
+      "step": 17510
+    },
+    {
+      "epoch": 0.5586912847986224,
+      "grad_norm": 0.17634367942810059,
+      "learning_rate": 0.001,
+      "loss": 1.1112,
+      "step": 17520
+    },
+    {
+      "epoch": 0.5590101725182564,
+      "grad_norm": 0.17674317955970764,
+      "learning_rate": 0.001,
+      "loss": 1.0956,
+      "step": 17530
+    },
+    {
+      "epoch": 0.5593290602378902,
+      "grad_norm": 0.16610750555992126,
+      "learning_rate": 0.001,
+      "loss": 1.1033,
+      "step": 17540
+    },
+    {
+      "epoch": 0.5596479479575242,
+      "grad_norm": 0.17032527923583984,
+      "learning_rate": 0.001,
+      "loss": 1.1341,
+      "step": 17550
+    },
+    {
+      "epoch": 0.5599668356771581,
+      "grad_norm": 0.17055188119411469,
+      "learning_rate": 0.001,
+      "loss": 1.0943,
+      "step": 17560
+    },
+    {
+      "epoch": 0.560285723396792,
+      "grad_norm": 0.18244749307632446,
+      "learning_rate": 0.001,
+      "loss": 1.1105,
+      "step": 17570
+    },
+    {
+      "epoch": 0.5606046111164259,
+      "grad_norm": 0.17542274296283722,
+      "learning_rate": 0.001,
+      "loss": 1.1116,
+      "step": 17580
+    },
+    {
+      "epoch": 0.5609234988360599,
+      "grad_norm": 0.168791264295578,
+      "learning_rate": 0.001,
+      "loss": 1.1051,
+      "step": 17590
+    },
+    {
+      "epoch": 0.5612423865556937,
+      "grad_norm": 0.16718518733978271,
+      "learning_rate": 0.001,
+      "loss": 1.1019,
+      "step": 17600
+    },
+    {
+      "epoch": 0.5615612742753276,
+      "grad_norm": 0.16974017024040222,
+      "learning_rate": 0.001,
+      "loss": 1.1123,
+      "step": 17610
+    },
+    {
+      "epoch": 0.5618801619949616,
+      "grad_norm": 0.16427423059940338,
+      "learning_rate": 0.001,
+      "loss": 1.1102,
+      "step": 17620
+    },
+    {
+      "epoch": 0.5621990497145954,
+      "grad_norm": 0.1736641824245453,
+      "learning_rate": 0.001,
+      "loss": 1.1146,
+      "step": 17630
+    },
+    {
+      "epoch": 0.5625179374342294,
+      "grad_norm": 0.1725444495677948,
+      "learning_rate": 0.001,
+      "loss": 1.0983,
+      "step": 17640
+    },
+    {
+      "epoch": 0.5628368251538634,
+      "grad_norm": 0.16529996693134308,
+      "learning_rate": 0.001,
+      "loss": 1.0841,
+      "step": 17650
+    },
+    {
+      "epoch": 0.5631557128734972,
+      "grad_norm": 0.17500054836273193,
+      "learning_rate": 0.001,
+      "loss": 1.1078,
+      "step": 17660
+    },
+    {
+      "epoch": 0.5634746005931311,
+      "grad_norm": 0.17414040863513947,
+      "learning_rate": 0.001,
+      "loss": 1.106,
+      "step": 17670
+    },
+    {
+      "epoch": 0.5637934883127651,
+      "grad_norm": 0.17169852554798126,
+      "learning_rate": 0.001,
+      "loss": 1.1011,
+      "step": 17680
+    },
+    {
+      "epoch": 0.5641123760323989,
+      "grad_norm": 0.1719132363796234,
+      "learning_rate": 0.001,
+      "loss": 1.1165,
+      "step": 17690
+    },
+    {
+      "epoch": 0.5644312637520329,
+      "grad_norm": 0.17163269221782684,
+      "learning_rate": 0.001,
+      "loss": 1.1082,
+      "step": 17700
+    },
+    {
+      "epoch": 0.5647501514716669,
+      "grad_norm": 0.16895899176597595,
+      "learning_rate": 0.001,
+      "loss": 1.1132,
+      "step": 17710
+    },
+    {
+      "epoch": 0.5650690391913007,
+      "grad_norm": 0.1689912974834442,
+      "learning_rate": 0.001,
+      "loss": 1.1078,
+      "step": 17720
+    },
+    {
+      "epoch": 0.5653879269109346,
+      "grad_norm": 0.17176558077335358,
+      "learning_rate": 0.001,
+      "loss": 1.0839,
+      "step": 17730
+    },
+    {
+      "epoch": 0.5657068146305686,
+      "grad_norm": 0.16738194227218628,
+      "learning_rate": 0.001,
+      "loss": 1.102,
+      "step": 17740
+    },
+    {
+      "epoch": 0.5660257023502024,
+      "grad_norm": 0.17044048011302948,
+      "learning_rate": 0.001,
+      "loss": 1.0937,
+      "step": 17750
+    },
+    {
+      "epoch": 0.5663445900698364,
+      "grad_norm": 0.16983206570148468,
+      "learning_rate": 0.001,
+      "loss": 1.1066,
+      "step": 17760
+    },
+    {
+      "epoch": 0.5666634777894703,
+      "grad_norm": 0.17275819182395935,
+      "learning_rate": 0.001,
+      "loss": 1.0966,
+      "step": 17770
+    },
+    {
+      "epoch": 0.5669823655091042,
+      "grad_norm": 0.1728706955909729,
+      "learning_rate": 0.001,
+      "loss": 1.1057,
+      "step": 17780
+    },
+    {
+      "epoch": 0.5673012532287381,
+      "grad_norm": 0.17265573143959045,
+      "learning_rate": 0.001,
+      "loss": 1.1164,
+      "step": 17790
+    },
+    {
+      "epoch": 0.5676201409483721,
+      "grad_norm": 0.16493971645832062,
+      "learning_rate": 0.001,
+      "loss": 1.0871,
+      "step": 17800
+    },
+    {
+      "epoch": 0.5679390286680059,
+      "grad_norm": 0.17371971905231476,
+      "learning_rate": 0.001,
+      "loss": 1.1067,
+      "step": 17810
+    },
+    {
+      "epoch": 0.5682579163876399,
+      "grad_norm": 0.17413072288036346,
+      "learning_rate": 0.001,
+      "loss": 1.1055,
+      "step": 17820
+    },
+    {
+      "epoch": 0.5685768041072738,
+      "grad_norm": 0.16485251486301422,
+      "learning_rate": 0.001,
+      "loss": 1.097,
+      "step": 17830
+    },
+    {
+      "epoch": 0.5688956918269078,
+      "grad_norm": 0.17777450382709503,
+      "learning_rate": 0.001,
+      "loss": 1.095,
+      "step": 17840
+    },
+    {
+      "epoch": 0.5692145795465416,
+      "grad_norm": 0.16648739576339722,
+      "learning_rate": 0.001,
+      "loss": 1.1036,
+      "step": 17850
+    },
+    {
+      "epoch": 0.5695334672661756,
+      "grad_norm": 0.16905073821544647,
+      "learning_rate": 0.001,
+      "loss": 1.0986,
+      "step": 17860
+    },
+    {
+      "epoch": 0.5698523549858096,
+      "grad_norm": 0.16770754754543304,
+      "learning_rate": 0.001,
+      "loss": 1.0964,
+      "step": 17870
+    },
+    {
+      "epoch": 0.5701712427054434,
+      "grad_norm": 0.17330941557884216,
+      "learning_rate": 0.001,
+      "loss": 1.1084,
+      "step": 17880
+    },
+    {
+      "epoch": 0.5704901304250773,
+      "grad_norm": 0.16497695446014404,
+      "learning_rate": 0.001,
+      "loss": 1.0997,
+      "step": 17890
+    },
+    {
+      "epoch": 0.5708090181447113,
+      "grad_norm": 0.1721220165491104,
+      "learning_rate": 0.001,
+      "loss": 1.1014,
+      "step": 17900
+    },
+    {
+      "epoch": 0.5711279058643451,
+      "grad_norm": 0.16339078545570374,
+      "learning_rate": 0.001,
+      "loss": 1.1042,
+      "step": 17910
+    },
+    {
+      "epoch": 0.5714467935839791,
+      "grad_norm": 0.16810251772403717,
+      "learning_rate": 0.001,
+      "loss": 1.1007,
+      "step": 17920
+    },
+    {
+      "epoch": 0.571765681303613,
+      "grad_norm": 0.1646348237991333,
+      "learning_rate": 0.001,
+      "loss": 1.1022,
+      "step": 17930
+    },
+    {
+      "epoch": 0.5720845690232469,
+      "grad_norm": 0.16776010394096375,
+      "learning_rate": 0.001,
+      "loss": 1.118,
+      "step": 17940
+    },
+    {
+      "epoch": 0.5724034567428808,
+      "grad_norm": 0.1724771410226822,
+      "learning_rate": 0.001,
+      "loss": 1.0982,
+      "step": 17950
+    },
+    {
+      "epoch": 0.5727223444625148,
+      "grad_norm": 0.1659904420375824,
+      "learning_rate": 0.001,
+      "loss": 1.0859,
+      "step": 17960
+    },
+    {
+      "epoch": 0.5730412321821486,
+      "grad_norm": 0.16321036219596863,
+      "learning_rate": 0.001,
+      "loss": 1.0966,
+      "step": 17970
+    },
+    {
+      "epoch": 0.5733601199017826,
+      "grad_norm": 0.17762115597724915,
+      "learning_rate": 0.001,
+      "loss": 1.1058,
+      "step": 17980
+    },
+    {
+      "epoch": 0.5736790076214165,
+      "grad_norm": 0.16604764759540558,
+      "learning_rate": 0.001,
+      "loss": 1.1161,
+      "step": 17990
+    },
+    {
+      "epoch": 0.5739978953410504,
+      "grad_norm": 0.17151857912540436,
+      "learning_rate": 0.001,
+      "loss": 1.0966,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5743167830606843,
+      "grad_norm": 0.17127293348312378,
+      "learning_rate": 0.001,
+      "loss": 1.0915,
+      "step": 18010
+    },
+    {
+      "epoch": 0.5746356707803183,
+      "grad_norm": 0.17190389335155487,
+      "learning_rate": 0.001,
+      "loss": 1.1058,
+      "step": 18020
+    },
+    {
+      "epoch": 0.5749545584999521,
+      "grad_norm": 0.16845406591892242,
+      "learning_rate": 0.001,
+      "loss": 1.1093,
+      "step": 18030
+    },
+    {
+      "epoch": 0.5752734462195861,
+      "grad_norm": 0.1646239012479782,
+      "learning_rate": 0.001,
+      "loss": 1.091,
+      "step": 18040
+    },
+    {
+      "epoch": 0.57559233393922,
+      "grad_norm": 0.1710473597049713,
+      "learning_rate": 0.001,
+      "loss": 1.0982,
+      "step": 18050
+    },
+    {
+      "epoch": 0.5759112216588539,
+      "grad_norm": 0.17428575456142426,
+      "learning_rate": 0.001,
+      "loss": 1.096,
+      "step": 18060
+    },
+    {
+      "epoch": 0.5762301093784878,
+      "grad_norm": 0.16330930590629578,
+      "learning_rate": 0.001,
+      "loss": 1.0996,
+      "step": 18070
+    },
+    {
+      "epoch": 0.5765489970981218,
+      "grad_norm": 0.16600897908210754,
+      "learning_rate": 0.001,
+      "loss": 1.1,
+      "step": 18080
+    },
+    {
+      "epoch": 0.5768678848177556,
+      "grad_norm": 0.17178833484649658,
+      "learning_rate": 0.001,
+      "loss": 1.1241,
+      "step": 18090
+    },
+    {
+      "epoch": 0.5771867725373896,
+      "grad_norm": 0.16852764785289764,
+      "learning_rate": 0.001,
+      "loss": 1.1157,
+      "step": 18100
+    },
+    {
+      "epoch": 0.5775056602570235,
+      "grad_norm": 0.16611634194850922,
+      "learning_rate": 0.001,
+      "loss": 1.0974,
+      "step": 18110
+    },
+    {
+      "epoch": 0.5778245479766574,
+      "grad_norm": 0.17456525564193726,
+      "learning_rate": 0.001,
+      "loss": 1.1091,
+      "step": 18120
+    },
+    {
+      "epoch": 0.5781434356962913,
+      "grad_norm": 0.16914406418800354,
+      "learning_rate": 0.001,
+      "loss": 1.0874,
+      "step": 18130
+    },
+    {
+      "epoch": 0.5784623234159253,
+      "grad_norm": 0.17011883854866028,
+      "learning_rate": 0.001,
+      "loss": 1.0998,
+      "step": 18140
+    },
+    {
+      "epoch": 0.5787812111355591,
+      "grad_norm": 0.16776974499225616,
+      "learning_rate": 0.001,
+      "loss": 1.0735,
+      "step": 18150
+    },
+    {
+      "epoch": 0.5791000988551931,
+      "grad_norm": 0.16967648267745972,
+      "learning_rate": 0.001,
+      "loss": 1.0962,
+      "step": 18160
+    },
+    {
+      "epoch": 0.579418986574827,
+      "grad_norm": 0.16465380787849426,
+      "learning_rate": 0.001,
+      "loss": 1.0846,
+      "step": 18170
+    },
+    {
+      "epoch": 0.5797378742944609,
+      "grad_norm": 0.16524037718772888,
+      "learning_rate": 0.001,
+      "loss": 1.1098,
+      "step": 18180
+    },
+    {
+      "epoch": 0.5800567620140948,
+      "grad_norm": 0.1675986647605896,
+      "learning_rate": 0.001,
+      "loss": 1.1059,
+      "step": 18190
+    },
+    {
+      "epoch": 0.5803756497337288,
+      "grad_norm": 0.17379865050315857,
+      "learning_rate": 0.001,
+      "loss": 1.1106,
+      "step": 18200
+    },
+    {
+      "epoch": 0.5806945374533626,
+      "grad_norm": 0.1684434860944748,
+      "learning_rate": 0.001,
+      "loss": 1.0985,
+      "step": 18210
+    },
+    {
+      "epoch": 0.5810134251729966,
+      "grad_norm": 0.16824005544185638,
+      "learning_rate": 0.001,
+      "loss": 1.1036,
+      "step": 18220
+    },
+    {
+      "epoch": 0.5813323128926305,
+      "grad_norm": 0.17007438838481903,
+      "learning_rate": 0.001,
+      "loss": 1.1254,
+      "step": 18230
+    },
+    {
+      "epoch": 0.5816512006122644,
+      "grad_norm": 0.16796430945396423,
+      "learning_rate": 0.001,
+      "loss": 1.0957,
+      "step": 18240
+    },
+    {
+      "epoch": 0.5819700883318983,
+      "grad_norm": 0.16562458872795105,
+      "learning_rate": 0.001,
+      "loss": 1.1029,
+      "step": 18250
+    },
+    {
+      "epoch": 0.5822889760515323,
+      "grad_norm": 0.17029453814029694,
+      "learning_rate": 0.001,
+      "loss": 1.0963,
+      "step": 18260
+    },
+    {
+      "epoch": 0.5826078637711661,
+      "grad_norm": 0.17549550533294678,
+      "learning_rate": 0.001,
+      "loss": 1.1081,
+      "step": 18270
+    },
+    {
+      "epoch": 0.5829267514908001,
+      "grad_norm": 0.17263686656951904,
+      "learning_rate": 0.001,
+      "loss": 1.0888,
+      "step": 18280
+    },
+    {
+      "epoch": 0.583245639210434,
+      "grad_norm": 0.16540630161762238,
+      "learning_rate": 0.001,
+      "loss": 1.1016,
+      "step": 18290
+    },
+    {
+      "epoch": 0.5835645269300679,
+      "grad_norm": 0.17288976907730103,
+      "learning_rate": 0.001,
+      "loss": 1.0994,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5838834146497018,
+      "grad_norm": 0.16975054144859314,
+      "learning_rate": 0.001,
+      "loss": 1.0882,
+      "step": 18310
+    },
+    {
+      "epoch": 0.5842023023693358,
+      "grad_norm": 0.16850727796554565,
+      "learning_rate": 0.001,
+      "loss": 1.102,
+      "step": 18320
+    },
+    {
+      "epoch": 0.5845211900889696,
+      "grad_norm": 0.17005418241024017,
+      "learning_rate": 0.001,
+      "loss": 1.1047,
+      "step": 18330
+    },
+    {
+      "epoch": 0.5848400778086036,
+      "grad_norm": 0.17504164576530457,
+      "learning_rate": 0.001,
+      "loss": 1.0929,
+      "step": 18340
+    },
+    {
+      "epoch": 0.5851589655282375,
+      "grad_norm": 0.16485849022865295,
+      "learning_rate": 0.001,
+      "loss": 1.1031,
+      "step": 18350
+    },
+    {
+      "epoch": 0.5854778532478714,
+      "grad_norm": 0.16495245695114136,
+      "learning_rate": 0.001,
+      "loss": 1.1203,
+      "step": 18360
+    },
+    {
+      "epoch": 0.5857967409675053,
+      "grad_norm": 0.1715477854013443,
+      "learning_rate": 0.001,
+      "loss": 1.116,
+      "step": 18370
+    },
+    {
+      "epoch": 0.5861156286871393,
+      "grad_norm": 0.18126294016838074,
+      "learning_rate": 0.001,
+      "loss": 1.1229,
+      "step": 18380
+    },
+    {
+      "epoch": 0.5864345164067731,
+      "grad_norm": 0.16917794942855835,
+      "learning_rate": 0.001,
+      "loss": 1.0919,
+      "step": 18390
+    },
+    {
+      "epoch": 0.5867534041264071,
+      "grad_norm": 0.17057420313358307,
+      "learning_rate": 0.001,
+      "loss": 1.0914,
+      "step": 18400
+    },
+    {
+      "epoch": 0.587072291846041,
+      "grad_norm": 0.16627562046051025,
+      "learning_rate": 0.001,
+      "loss": 1.0981,
+      "step": 18410
+    },
+    {
+      "epoch": 0.5873911795656749,
+      "grad_norm": 0.17235489189624786,
+      "learning_rate": 0.001,
+      "loss": 1.1172,
+      "step": 18420
+    },
+    {
+      "epoch": 0.5877100672853088,
+      "grad_norm": 0.16684970259666443,
+      "learning_rate": 0.001,
+      "loss": 1.1117,
+      "step": 18430
+    },
+    {
+      "epoch": 0.5880289550049428,
+      "grad_norm": 0.1660860925912857,
+      "learning_rate": 0.001,
+      "loss": 1.1038,
+      "step": 18440
+    },
+    {
+      "epoch": 0.5883478427245766,
+      "grad_norm": 0.16902343928813934,
+      "learning_rate": 0.001,
+      "loss": 1.1023,
+      "step": 18450
+    },
+    {
+      "epoch": 0.5886667304442106,
+      "grad_norm": 0.16275769472122192,
+      "learning_rate": 0.001,
+      "loss": 1.0823,
+      "step": 18460
+    },
+    {
+      "epoch": 0.5889856181638445,
+      "grad_norm": 0.16918061673641205,
+      "learning_rate": 0.001,
+      "loss": 1.0965,
+      "step": 18470
+    },
+    {
+      "epoch": 0.5893045058834784,
+      "grad_norm": 0.16743110120296478,
+      "learning_rate": 0.001,
+      "loss": 1.1065,
+      "step": 18480
+    },
+    {
+      "epoch": 0.5896233936031123,
+      "grad_norm": 0.16440755128860474,
+      "learning_rate": 0.001,
+      "loss": 1.0934,
+      "step": 18490
+    },
+    {
+      "epoch": 0.5899422813227463,
+      "grad_norm": 0.16429467499256134,
+      "learning_rate": 0.001,
+      "loss": 1.1044,
+      "step": 18500
+    },
+    {
+      "epoch": 0.5902611690423801,
+      "grad_norm": 0.17269568145275116,
+      "learning_rate": 0.001,
+      "loss": 1.1021,
+      "step": 18510
+    },
+    {
+      "epoch": 0.5905800567620141,
+      "grad_norm": 0.17350298166275024,
+      "learning_rate": 0.001,
+      "loss": 1.1106,
+      "step": 18520
+    },
+    {
+      "epoch": 0.590898944481648,
+      "grad_norm": 0.17228271067142487,
+      "learning_rate": 0.001,
+      "loss": 1.0906,
+      "step": 18530
+    },
+    {
+      "epoch": 0.5912178322012819,
+      "grad_norm": 0.17125701904296875,
+      "learning_rate": 0.001,
+      "loss": 1.1031,
+      "step": 18540
+    },
+    {
+      "epoch": 0.5915367199209158,
+      "grad_norm": 0.17287059128284454,
+      "learning_rate": 0.001,
+      "loss": 1.1035,
+      "step": 18550
+    },
+    {
+      "epoch": 0.5918556076405498,
+      "grad_norm": 0.17267583310604095,
+      "learning_rate": 0.001,
+      "loss": 1.122,
+      "step": 18560
+    },
+    {
+      "epoch": 0.5921744953601836,
+      "grad_norm": 0.1665111482143402,
+      "learning_rate": 0.001,
+      "loss": 1.0852,
+      "step": 18570
+    },
+    {
+      "epoch": 0.5924933830798176,
+      "grad_norm": 0.18036475777626038,
+      "learning_rate": 0.001,
+      "loss": 1.0897,
+      "step": 18580
+    },
+    {
+      "epoch": 0.5928122707994515,
+      "grad_norm": 0.16732090711593628,
+      "learning_rate": 0.001,
+      "loss": 1.1005,
+      "step": 18590
+    },
+    {
+      "epoch": 0.5931311585190854,
+      "grad_norm": 0.16416814923286438,
+      "learning_rate": 0.001,
+      "loss": 1.0994,
+      "step": 18600
+    },
+    {
+      "epoch": 0.5934500462387193,
+      "grad_norm": 0.1659676730632782,
+      "learning_rate": 0.001,
+      "loss": 1.1058,
+      "step": 18610
+    },
+    {
+      "epoch": 0.5937689339583533,
+      "grad_norm": 0.17046456038951874,
+      "learning_rate": 0.001,
+      "loss": 1.1051,
+      "step": 18620
+    },
+    {
+      "epoch": 0.5940878216779871,
+      "grad_norm": 0.17397566139698029,
+      "learning_rate": 0.001,
+      "loss": 1.1013,
+      "step": 18630
+    },
+    {
+      "epoch": 0.5944067093976211,
+      "grad_norm": 0.1660611480474472,
+      "learning_rate": 0.001,
+      "loss": 1.1004,
+      "step": 18640
+    },
+    {
+      "epoch": 0.594725597117255,
+      "grad_norm": 0.1655200570821762,
+      "learning_rate": 0.001,
+      "loss": 1.1037,
+      "step": 18650
+    },
+    {
+      "epoch": 0.5950444848368889,
+      "grad_norm": 0.16909074783325195,
+      "learning_rate": 0.001,
+      "loss": 1.0759,
+      "step": 18660
+    },
+    {
+      "epoch": 0.5953633725565228,
+      "grad_norm": 0.17524762451648712,
+      "learning_rate": 0.001,
+      "loss": 1.092,
+      "step": 18670
+    },
+    {
+      "epoch": 0.5956822602761568,
+      "grad_norm": 0.1684812307357788,
+      "learning_rate": 0.001,
+      "loss": 1.1036,
+      "step": 18680
+    },
+    {
+      "epoch": 0.5960011479957907,
+      "grad_norm": 0.1623716950416565,
+      "learning_rate": 0.001,
+      "loss": 1.0907,
+      "step": 18690
+    },
+    {
+      "epoch": 0.5963200357154246,
+      "grad_norm": 0.16788646578788757,
+      "learning_rate": 0.001,
+      "loss": 1.0807,
+      "step": 18700
+    },
+    {
+      "epoch": 0.5966389234350585,
+      "grad_norm": 0.1718616485595703,
+      "learning_rate": 0.001,
+      "loss": 1.1057,
+      "step": 18710
+    },
+    {
+      "epoch": 0.5969578111546925,
+      "grad_norm": 0.17129045724868774,
+      "learning_rate": 0.001,
+      "loss": 1.0989,
+      "step": 18720
+    },
+    {
+      "epoch": 0.5972766988743263,
+      "grad_norm": 0.16963504254817963,
+      "learning_rate": 0.001,
+      "loss": 1.0987,
+      "step": 18730
+    },
+    {
+      "epoch": 0.5975955865939603,
+      "grad_norm": 0.16359004378318787,
+      "learning_rate": 0.001,
+      "loss": 1.1128,
+      "step": 18740
+    },
+    {
+      "epoch": 0.5979144743135942,
+      "grad_norm": 0.16450457274913788,
+      "learning_rate": 0.001,
+      "loss": 1.084,
+      "step": 18750
+    },
+    {
+      "epoch": 0.5982333620332281,
+      "grad_norm": 0.16531401872634888,
+      "learning_rate": 0.001,
+      "loss": 1.105,
+      "step": 18760
+    },
+    {
+      "epoch": 0.598552249752862,
+      "grad_norm": 0.16693365573883057,
+      "learning_rate": 0.001,
+      "loss": 1.0991,
+      "step": 18770
+    },
+    {
+      "epoch": 0.598871137472496,
+      "grad_norm": 0.1713932305574417,
+      "learning_rate": 0.001,
+      "loss": 1.0912,
+      "step": 18780
+    },
+    {
+      "epoch": 0.5991900251921298,
+      "grad_norm": 0.16845330595970154,
+      "learning_rate": 0.001,
+      "loss": 1.1299,
+      "step": 18790
+    },
+    {
+      "epoch": 0.5995089129117638,
+      "grad_norm": 0.1709652543067932,
+      "learning_rate": 0.001,
+      "loss": 1.0666,
+      "step": 18800
+    },
+    {
+      "epoch": 0.5998278006313977,
+      "grad_norm": 0.16776952147483826,
+      "learning_rate": 0.001,
+      "loss": 1.0857,
+      "step": 18810
+    },
+    {
+      "epoch": 0.6001466883510316,
+      "grad_norm": 0.17375540733337402,
+      "learning_rate": 0.001,
+      "loss": 1.1064,
+      "step": 18820
+    },
+    {
+      "epoch": 0.6004655760706655,
+      "grad_norm": 0.16971836984157562,
+      "learning_rate": 0.001,
+      "loss": 1.0992,
+      "step": 18830
+    },
+    {
+      "epoch": 0.6007844637902995,
+      "grad_norm": 0.17225322127342224,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 18840
+    },
+    {
+      "epoch": 0.6011033515099333,
+      "grad_norm": 0.172017440199852,
+      "learning_rate": 0.001,
+      "loss": 1.1328,
+      "step": 18850
+    },
+    {
+      "epoch": 0.6014222392295673,
+      "grad_norm": 0.16906633973121643,
+      "learning_rate": 0.001,
+      "loss": 1.0944,
+      "step": 18860
+    },
+    {
+      "epoch": 0.6017411269492012,
+      "grad_norm": 0.17072793841362,
+      "learning_rate": 0.001,
+      "loss": 1.1012,
+      "step": 18870
+    },
+    {
+      "epoch": 0.6020600146688351,
+      "grad_norm": 0.16955476999282837,
+      "learning_rate": 0.001,
+      "loss": 1.0825,
+      "step": 18880
+    },
+    {
+      "epoch": 0.602378902388469,
+      "grad_norm": 0.16245858371257782,
+      "learning_rate": 0.001,
+      "loss": 1.0948,
+      "step": 18890
+    },
+    {
+      "epoch": 0.602697790108103,
+      "grad_norm": 0.1670025736093521,
+      "learning_rate": 0.001,
+      "loss": 1.0995,
+      "step": 18900
+    },
+    {
+      "epoch": 0.6030166778277368,
+      "grad_norm": 0.1683637946844101,
+      "learning_rate": 0.001,
+      "loss": 1.1024,
+      "step": 18910
+    },
+    {
+      "epoch": 0.6033355655473708,
+      "grad_norm": 0.17176498472690582,
+      "learning_rate": 0.001,
+      "loss": 1.1075,
+      "step": 18920
+    },
+    {
+      "epoch": 0.6036544532670047,
+      "grad_norm": 0.16669496893882751,
+      "learning_rate": 0.001,
+      "loss": 1.0736,
+      "step": 18930
+    },
+    {
+      "epoch": 0.6039733409866386,
+      "grad_norm": 0.17387279868125916,
+      "learning_rate": 0.001,
+      "loss": 1.0768,
+      "step": 18940
+    },
+    {
+      "epoch": 0.6042922287062725,
+      "grad_norm": 0.170260950922966,
+      "learning_rate": 0.001,
+      "loss": 1.1012,
+      "step": 18950
+    },
+    {
+      "epoch": 0.6046111164259065,
+      "grad_norm": 0.17083097994327545,
+      "learning_rate": 0.001,
+      "loss": 1.1049,
+      "step": 18960
+    },
+    {
+      "epoch": 0.6049300041455403,
+      "grad_norm": 0.17165276408195496,
+      "learning_rate": 0.001,
+      "loss": 1.101,
+      "step": 18970
+    },
+    {
+      "epoch": 0.6052488918651743,
+      "grad_norm": 0.17530037462711334,
+      "learning_rate": 0.001,
+      "loss": 1.1065,
+      "step": 18980
+    },
+    {
+      "epoch": 0.6055677795848082,
+      "grad_norm": 0.16949541866779327,
+      "learning_rate": 0.001,
+      "loss": 1.1038,
+      "step": 18990
+    },
+    {
+      "epoch": 0.6058866673044421,
+      "grad_norm": 0.17218689620494843,
+      "learning_rate": 0.001,
+      "loss": 1.098,
+      "step": 19000
+    },
+    {
+      "epoch": 0.606205555024076,
+      "grad_norm": 0.18472091853618622,
+      "learning_rate": 0.001,
+      "loss": 1.0901,
+      "step": 19010
+    },
+    {
+      "epoch": 0.60652444274371,
+      "grad_norm": 0.1716701090335846,
+      "learning_rate": 0.001,
+      "loss": 1.1091,
+      "step": 19020
+    },
+    {
+      "epoch": 0.6068433304633438,
+      "grad_norm": 0.1648435741662979,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 19030
+    },
+    {
+      "epoch": 0.6071622181829778,
+      "grad_norm": 0.16991214454174042,
+      "learning_rate": 0.001,
+      "loss": 1.0827,
+      "step": 19040
+    },
+    {
+      "epoch": 0.6074811059026117,
+      "grad_norm": 0.17173060774803162,
+      "learning_rate": 0.001,
+      "loss": 1.0895,
+      "step": 19050
+    },
+    {
+      "epoch": 0.6077999936222456,
+      "grad_norm": 0.16547681391239166,
+      "learning_rate": 0.001,
+      "loss": 1.0994,
+      "step": 19060
+    },
+    {
+      "epoch": 0.6081188813418795,
+      "grad_norm": 0.16734836995601654,
+      "learning_rate": 0.001,
+      "loss": 1.0989,
+      "step": 19070
+    },
+    {
+      "epoch": 0.6084377690615135,
+      "grad_norm": 0.17032268643379211,
+      "learning_rate": 0.001,
+      "loss": 1.0928,
+      "step": 19080
+    },
+    {
+      "epoch": 0.6087566567811473,
+      "grad_norm": 0.17411461472511292,
+      "learning_rate": 0.001,
+      "loss": 1.0737,
+      "step": 19090
+    },
+    {
+      "epoch": 0.6090755445007813,
+      "grad_norm": 0.1729476898908615,
+      "learning_rate": 0.001,
+      "loss": 1.0995,
+      "step": 19100
+    },
+    {
+      "epoch": 0.6093944322204152,
+      "grad_norm": 0.17751654982566833,
+      "learning_rate": 0.001,
+      "loss": 1.0767,
+      "step": 19110
+    },
+    {
+      "epoch": 0.6097133199400491,
+      "grad_norm": 0.1721947342157364,
+      "learning_rate": 0.001,
+      "loss": 1.0965,
+      "step": 19120
+    },
+    {
+      "epoch": 0.610032207659683,
+      "grad_norm": 0.1693287491798401,
+      "learning_rate": 0.001,
+      "loss": 1.1138,
+      "step": 19130
+    },
+    {
+      "epoch": 0.610351095379317,
+      "grad_norm": 0.16716571152210236,
+      "learning_rate": 0.001,
+      "loss": 1.1116,
+      "step": 19140
+    },
+    {
+      "epoch": 0.6106699830989508,
+      "grad_norm": 0.16647756099700928,
+      "learning_rate": 0.001,
+      "loss": 1.1066,
+      "step": 19150
+    },
+    {
+      "epoch": 0.6109888708185848,
+      "grad_norm": 0.1653558611869812,
+      "learning_rate": 0.001,
+      "loss": 1.0802,
+      "step": 19160
+    },
+    {
+      "epoch": 0.6113077585382187,
+      "grad_norm": 0.16676802933216095,
+      "learning_rate": 0.001,
+      "loss": 1.1088,
+      "step": 19170
+    },
+    {
+      "epoch": 0.6116266462578526,
+      "grad_norm": 0.1702679842710495,
+      "learning_rate": 0.001,
+      "loss": 1.1055,
+      "step": 19180
+    },
+    {
+      "epoch": 0.6119455339774865,
+      "grad_norm": 0.16800037026405334,
+      "learning_rate": 0.001,
+      "loss": 1.104,
+      "step": 19190
+    },
+    {
+      "epoch": 0.6122644216971205,
+      "grad_norm": 0.17141634225845337,
+      "learning_rate": 0.001,
+      "loss": 1.0885,
+      "step": 19200
+    },
+    {
+      "epoch": 0.6125833094167543,
+      "grad_norm": 0.16200610995292664,
+      "learning_rate": 0.001,
+      "loss": 1.0737,
+      "step": 19210
+    },
+    {
+      "epoch": 0.6129021971363883,
+      "grad_norm": 0.1708095520734787,
+      "learning_rate": 0.001,
+      "loss": 1.1009,
+      "step": 19220
+    },
+    {
+      "epoch": 0.6132210848560222,
+      "grad_norm": 0.16531571745872498,
+      "learning_rate": 0.001,
+      "loss": 1.0825,
+      "step": 19230
+    },
+    {
+      "epoch": 0.6135399725756561,
+      "grad_norm": 0.1678553968667984,
+      "learning_rate": 0.001,
+      "loss": 1.0928,
+      "step": 19240
+    },
+    {
+      "epoch": 0.61385886029529,
+      "grad_norm": 0.16929244995117188,
+      "learning_rate": 0.001,
+      "loss": 1.0762,
+      "step": 19250
+    },
+    {
+      "epoch": 0.614177748014924,
+      "grad_norm": 0.1687399297952652,
+      "learning_rate": 0.001,
+      "loss": 1.1068,
+      "step": 19260
+    },
+    {
+      "epoch": 0.6144966357345578,
+      "grad_norm": 0.1622915267944336,
+      "learning_rate": 0.001,
+      "loss": 1.1062,
+      "step": 19270
+    },
+    {
+      "epoch": 0.6148155234541918,
+      "grad_norm": 0.16613072156906128,
+      "learning_rate": 0.001,
+      "loss": 1.0977,
+      "step": 19280
+    },
+    {
+      "epoch": 0.6151344111738257,
+      "grad_norm": 0.16244791448116302,
+      "learning_rate": 0.001,
+      "loss": 1.0794,
+      "step": 19290
+    },
+    {
+      "epoch": 0.6154532988934596,
+      "grad_norm": 0.17199251055717468,
+      "learning_rate": 0.001,
+      "loss": 1.1048,
+      "step": 19300
+    },
+    {
+      "epoch": 0.6157721866130935,
+      "grad_norm": 0.1752566248178482,
+      "learning_rate": 0.001,
+      "loss": 1.1059,
+      "step": 19310
+    },
+    {
+      "epoch": 0.6160910743327275,
+      "grad_norm": 0.17897622287273407,
+      "learning_rate": 0.001,
+      "loss": 1.0919,
+      "step": 19320
+    },
+    {
+      "epoch": 0.6164099620523613,
+      "grad_norm": 0.1731637865304947,
+      "learning_rate": 0.001,
+      "loss": 1.1154,
+      "step": 19330
+    },
+    {
+      "epoch": 0.6167288497719953,
+      "grad_norm": 0.16791284084320068,
+      "learning_rate": 0.001,
+      "loss": 1.0889,
+      "step": 19340
+    },
+    {
+      "epoch": 0.6170477374916292,
+      "grad_norm": 0.16614709794521332,
+      "learning_rate": 0.001,
+      "loss": 1.0938,
+      "step": 19350
+    },
+    {
+      "epoch": 0.6173666252112631,
+      "grad_norm": 0.1709696352481842,
+      "learning_rate": 0.001,
+      "loss": 1.0956,
+      "step": 19360
+    },
+    {
+      "epoch": 0.617685512930897,
+      "grad_norm": 0.16510511934757233,
+      "learning_rate": 0.001,
+      "loss": 1.0989,
+      "step": 19370
+    },
+    {
+      "epoch": 0.618004400650531,
+      "grad_norm": 0.16931302845478058,
+      "learning_rate": 0.001,
+      "loss": 1.1062,
+      "step": 19380
+    },
+    {
+      "epoch": 0.6183232883701648,
+      "grad_norm": 0.17815431952476501,
+      "learning_rate": 0.001,
+      "loss": 1.1119,
+      "step": 19390
+    },
+    {
+      "epoch": 0.6186421760897988,
+      "grad_norm": 0.1690226048231125,
+      "learning_rate": 0.001,
+      "loss": 1.0876,
+      "step": 19400
+    },
+    {
+      "epoch": 0.6189610638094327,
+      "grad_norm": 0.16781018674373627,
+      "learning_rate": 0.001,
+      "loss": 1.0896,
+      "step": 19410
+    },
+    {
+      "epoch": 0.6192799515290666,
+      "grad_norm": 0.16898411512374878,
+      "learning_rate": 0.001,
+      "loss": 1.1129,
+      "step": 19420
+    },
+    {
+      "epoch": 0.6195988392487005,
+      "grad_norm": 0.17247016727924347,
+      "learning_rate": 0.001,
+      "loss": 1.0907,
+      "step": 19430
+    },
+    {
+      "epoch": 0.6199177269683345,
+      "grad_norm": 0.1620088517665863,
+      "learning_rate": 0.001,
+      "loss": 1.0984,
+      "step": 19440
+    },
+    {
+      "epoch": 0.6202366146879683,
+      "grad_norm": 0.16331516206264496,
+      "learning_rate": 0.001,
+      "loss": 1.0794,
+      "step": 19450
+    },
+    {
+      "epoch": 0.6205555024076023,
+      "grad_norm": 0.1672041267156601,
+      "learning_rate": 0.001,
+      "loss": 1.114,
+      "step": 19460
+    },
+    {
+      "epoch": 0.6208743901272362,
+      "grad_norm": 0.16793550550937653,
+      "learning_rate": 0.001,
+      "loss": 1.0828,
+      "step": 19470
+    },
+    {
+      "epoch": 0.6211932778468701,
+      "grad_norm": 0.16448058187961578,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 19480
+    },
+    {
+      "epoch": 0.621512165566504,
+      "grad_norm": 0.1713237166404724,
+      "learning_rate": 0.001,
+      "loss": 1.098,
+      "step": 19490
+    },
+    {
+      "epoch": 0.621831053286138,
+      "grad_norm": 0.17198623716831207,
+      "learning_rate": 0.001,
+      "loss": 1.1004,
+      "step": 19500
+    },
+    {
+      "epoch": 0.6221499410057719,
+      "grad_norm": 0.16575346887111664,
+      "learning_rate": 0.001,
+      "loss": 1.1014,
+      "step": 19510
+    },
+    {
+      "epoch": 0.6224688287254058,
+      "grad_norm": 0.1719786524772644,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 19520
+    },
+    {
+      "epoch": 0.6227877164450397,
+      "grad_norm": 0.16646794974803925,
+      "learning_rate": 0.001,
+      "loss": 1.1097,
+      "step": 19530
+    },
+    {
+      "epoch": 0.6231066041646737,
+      "grad_norm": 0.16815558075904846,
+      "learning_rate": 0.001,
+      "loss": 1.0888,
+      "step": 19540
+    },
+    {
+      "epoch": 0.6234254918843075,
+      "grad_norm": 0.1675843745470047,
+      "learning_rate": 0.001,
+      "loss": 1.0896,
+      "step": 19550
+    },
+    {
+      "epoch": 0.6237443796039415,
+      "grad_norm": 0.1759590357542038,
+      "learning_rate": 0.001,
+      "loss": 1.1131,
+      "step": 19560
+    },
+    {
+      "epoch": 0.6240632673235754,
+      "grad_norm": 0.17111027240753174,
+      "learning_rate": 0.001,
+      "loss": 1.1067,
+      "step": 19570
+    },
+    {
+      "epoch": 0.6243821550432093,
+      "grad_norm": 0.17272379994392395,
+      "learning_rate": 0.001,
+      "loss": 1.101,
+      "step": 19580
+    },
+    {
+      "epoch": 0.6247010427628432,
+      "grad_norm": 0.16584420204162598,
+      "learning_rate": 0.001,
+      "loss": 1.0948,
+      "step": 19590
+    },
+    {
+      "epoch": 0.6250199304824772,
+      "grad_norm": 0.16466067731380463,
+      "learning_rate": 0.001,
+      "loss": 1.0887,
+      "step": 19600
+    },
+    {
+      "epoch": 0.625338818202111,
+      "grad_norm": 0.1696598380804062,
+      "learning_rate": 0.001,
+      "loss": 1.0999,
+      "step": 19610
+    },
+    {
+      "epoch": 0.625657705921745,
+      "grad_norm": 0.1659633368253708,
+      "learning_rate": 0.001,
+      "loss": 1.1112,
+      "step": 19620
+    },
+    {
+      "epoch": 0.6259765936413789,
+      "grad_norm": 0.17163695394992828,
+      "learning_rate": 0.001,
+      "loss": 1.086,
+      "step": 19630
+    },
+    {
+      "epoch": 0.6262954813610128,
+      "grad_norm": 0.17734724283218384,
+      "learning_rate": 0.001,
+      "loss": 1.099,
+      "step": 19640
+    },
+    {
+      "epoch": 0.6266143690806467,
+      "grad_norm": 0.16526159644126892,
+      "learning_rate": 0.001,
+      "loss": 1.0978,
+      "step": 19650
+    },
+    {
+      "epoch": 0.6269332568002807,
+      "grad_norm": 0.1740102618932724,
+      "learning_rate": 0.001,
+      "loss": 1.0921,
+      "step": 19660
+    },
+    {
+      "epoch": 0.6272521445199145,
+      "grad_norm": 0.17085231840610504,
+      "learning_rate": 0.001,
+      "loss": 1.091,
+      "step": 19670
+    },
+    {
+      "epoch": 0.6275710322395485,
+      "grad_norm": 0.17608655989170074,
+      "learning_rate": 0.001,
+      "loss": 1.1068,
+      "step": 19680
+    },
+    {
+      "epoch": 0.6278899199591824,
+      "grad_norm": 0.16677728295326233,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 19690
+    },
+    {
+      "epoch": 0.6282088076788163,
+      "grad_norm": 0.17247577011585236,
+      "learning_rate": 0.001,
+      "loss": 1.0922,
+      "step": 19700
+    },
+    {
+      "epoch": 0.6285276953984502,
+      "grad_norm": 0.16981416940689087,
+      "learning_rate": 0.001,
+      "loss": 1.0772,
+      "step": 19710
+    },
+    {
+      "epoch": 0.6288465831180842,
+      "grad_norm": 0.16509251296520233,
+      "learning_rate": 0.001,
+      "loss": 1.0939,
+      "step": 19720
+    },
+    {
+      "epoch": 0.629165470837718,
+      "grad_norm": 0.16891643404960632,
+      "learning_rate": 0.001,
+      "loss": 1.0857,
+      "step": 19730
+    },
+    {
+      "epoch": 0.629484358557352,
+      "grad_norm": 0.16418832540512085,
+      "learning_rate": 0.001,
+      "loss": 1.0903,
+      "step": 19740
+    },
+    {
+      "epoch": 0.6298032462769859,
+      "grad_norm": 0.16804587841033936,
+      "learning_rate": 0.001,
+      "loss": 1.0838,
+      "step": 19750
+    },
+    {
+      "epoch": 0.6301221339966198,
+      "grad_norm": 0.1670849472284317,
+      "learning_rate": 0.001,
+      "loss": 1.0839,
+      "step": 19760
+    },
+    {
+      "epoch": 0.6304410217162537,
+      "grad_norm": 0.17061035335063934,
+      "learning_rate": 0.001,
+      "loss": 1.0819,
+      "step": 19770
+    },
+    {
+      "epoch": 0.6307599094358877,
+      "grad_norm": 0.16851983964443207,
+      "learning_rate": 0.001,
+      "loss": 1.085,
+      "step": 19780
+    },
+    {
+      "epoch": 0.6310787971555215,
+      "grad_norm": 0.16248421370983124,
+      "learning_rate": 0.001,
+      "loss": 1.0959,
+      "step": 19790
+    },
+    {
+      "epoch": 0.6313976848751555,
+      "grad_norm": 0.17037761211395264,
+      "learning_rate": 0.001,
+      "loss": 1.0885,
+      "step": 19800
+    },
+    {
+      "epoch": 0.6317165725947894,
+      "grad_norm": 0.1663527637720108,
+      "learning_rate": 0.001,
+      "loss": 1.0947,
+      "step": 19810
+    },
+    {
+      "epoch": 0.6320354603144233,
+      "grad_norm": 0.17086707055568695,
+      "learning_rate": 0.001,
+      "loss": 1.0895,
+      "step": 19820
+    },
+    {
+      "epoch": 0.6323543480340572,
+      "grad_norm": 0.17373676598072052,
+      "learning_rate": 0.001,
+      "loss": 1.0898,
+      "step": 19830
+    },
+    {
+      "epoch": 0.6326732357536912,
+      "grad_norm": 0.1695992350578308,
+      "learning_rate": 0.001,
+      "loss": 1.105,
+      "step": 19840
+    },
+    {
+      "epoch": 0.632992123473325,
+      "grad_norm": 0.1636921763420105,
+      "learning_rate": 0.001,
+      "loss": 1.0857,
+      "step": 19850
+    },
+    {
+      "epoch": 0.633311011192959,
+      "grad_norm": 0.17723794281482697,
+      "learning_rate": 0.001,
+      "loss": 1.0957,
+      "step": 19860
+    },
+    {
+      "epoch": 0.6336298989125929,
+      "grad_norm": 0.16286857426166534,
+      "learning_rate": 0.001,
+      "loss": 1.0859,
+      "step": 19870
+    },
+    {
+      "epoch": 0.6339487866322268,
+      "grad_norm": 0.17155247926712036,
+      "learning_rate": 0.001,
+      "loss": 1.0921,
+      "step": 19880
+    },
+    {
+      "epoch": 0.6342676743518607,
+      "grad_norm": 0.16349942982196808,
+      "learning_rate": 0.001,
+      "loss": 1.0939,
+      "step": 19890
+    },
+    {
+      "epoch": 0.6345865620714947,
+      "grad_norm": 0.1695716679096222,
+      "learning_rate": 0.001,
+      "loss": 1.095,
+      "step": 19900
+    },
+    {
+      "epoch": 0.6349054497911285,
+      "grad_norm": 0.17381371557712555,
+      "learning_rate": 0.001,
+      "loss": 1.1043,
+      "step": 19910
+    },
+    {
+      "epoch": 0.6352243375107625,
+      "grad_norm": 0.1696106344461441,
+      "learning_rate": 0.001,
+      "loss": 1.099,
+      "step": 19920
+    },
+    {
+      "epoch": 0.6355432252303964,
+      "grad_norm": 0.17153136432170868,
+      "learning_rate": 0.001,
+      "loss": 1.0776,
+      "step": 19930
+    },
+    {
+      "epoch": 0.6358621129500303,
+      "grad_norm": 0.17578768730163574,
+      "learning_rate": 0.001,
+      "loss": 1.0836,
+      "step": 19940
+    },
+    {
+      "epoch": 0.6361810006696642,
+      "grad_norm": 0.16676296293735504,
+      "learning_rate": 0.001,
+      "loss": 1.0752,
+      "step": 19950
+    },
+    {
+      "epoch": 0.6364998883892982,
+      "grad_norm": 0.17143648862838745,
+      "learning_rate": 0.001,
+      "loss": 1.0936,
+      "step": 19960
+    },
+    {
+      "epoch": 0.636818776108932,
+      "grad_norm": 0.172052264213562,
+      "learning_rate": 0.001,
+      "loss": 1.1046,
+      "step": 19970
+    },
+    {
+      "epoch": 0.637137663828566,
+      "grad_norm": 0.16690969467163086,
+      "learning_rate": 0.001,
+      "loss": 1.1054,
+      "step": 19980
+    },
+    {
+      "epoch": 0.6374565515481999,
+      "grad_norm": 0.16814520955085754,
+      "learning_rate": 0.001,
+      "loss": 1.0981,
+      "step": 19990
+    },
+    {
+      "epoch": 0.6377754392678338,
+      "grad_norm": 0.16802173852920532,
+      "learning_rate": 0.001,
+      "loss": 1.0911,
+      "step": 20000
+    },
+    {
+      "epoch": 0.6380943269874677,
+      "grad_norm": 0.16470474004745483,
+      "learning_rate": 0.001,
+      "loss": 1.1015,
+      "step": 20010
+    },
+    {
+      "epoch": 0.6384132147071017,
+      "grad_norm": 0.1644125133752823,
+      "learning_rate": 0.001,
+      "loss": 1.1007,
+      "step": 20020
+    },
+    {
+      "epoch": 0.6387321024267355,
+      "grad_norm": 0.17056648433208466,
+      "learning_rate": 0.001,
+      "loss": 1.1168,
+      "step": 20030
+    },
+    {
+      "epoch": 0.6390509901463695,
+      "grad_norm": 0.1640109121799469,
+      "learning_rate": 0.001,
+      "loss": 1.0898,
+      "step": 20040
+    },
+    {
+      "epoch": 0.6393698778660034,
+      "grad_norm": 0.16083675622940063,
+      "learning_rate": 0.001,
+      "loss": 1.0826,
+      "step": 20050
+    },
+    {
+      "epoch": 0.6396887655856373,
+      "grad_norm": 0.1644384115934372,
+      "learning_rate": 0.001,
+      "loss": 1.0919,
+      "step": 20060
+    },
+    {
+      "epoch": 0.6400076533052712,
+      "grad_norm": 0.1658620685338974,
+      "learning_rate": 0.001,
+      "loss": 1.0918,
+      "step": 20070
+    },
+    {
+      "epoch": 0.6403265410249052,
+      "grad_norm": 0.17355003952980042,
+      "learning_rate": 0.001,
+      "loss": 1.087,
+      "step": 20080
+    },
+    {
+      "epoch": 0.640645428744539,
+      "grad_norm": 0.15890207886695862,
+      "learning_rate": 0.001,
+      "loss": 1.0992,
+      "step": 20090
+    },
+    {
+      "epoch": 0.640964316464173,
+      "grad_norm": 0.17783460021018982,
+      "learning_rate": 0.001,
+      "loss": 1.1057,
+      "step": 20100
+    },
+    {
+      "epoch": 0.6412832041838069,
+      "grad_norm": 0.16919948160648346,
+      "learning_rate": 0.001,
+      "loss": 1.079,
+      "step": 20110
+    },
+    {
+      "epoch": 0.6416020919034408,
+      "grad_norm": 0.16504615545272827,
+      "learning_rate": 0.001,
+      "loss": 1.1018,
+      "step": 20120
+    },
+    {
+      "epoch": 0.6419209796230747,
+      "grad_norm": 0.1690061092376709,
+      "learning_rate": 0.001,
+      "loss": 1.0985,
+      "step": 20130
+    },
+    {
+      "epoch": 0.6422398673427087,
+      "grad_norm": 0.16797204315662384,
+      "learning_rate": 0.001,
+      "loss": 1.0923,
+      "step": 20140
+    },
+    {
+      "epoch": 0.6425587550623425,
+      "grad_norm": 0.1663273572921753,
+      "learning_rate": 0.001,
+      "loss": 1.0958,
+      "step": 20150
+    },
+    {
+      "epoch": 0.6428776427819765,
+      "grad_norm": 0.16364677250385284,
+      "learning_rate": 0.001,
+      "loss": 1.0996,
+      "step": 20160
+    },
+    {
+      "epoch": 0.6431965305016104,
+      "grad_norm": 0.17317451536655426,
+      "learning_rate": 0.001,
+      "loss": 1.0985,
+      "step": 20170
+    },
+    {
+      "epoch": 0.6435154182212443,
+      "grad_norm": 0.1685129553079605,
+      "learning_rate": 0.001,
+      "loss": 1.0944,
+      "step": 20180
+    },
+    {
+      "epoch": 0.6438343059408782,
+      "grad_norm": 0.16640588641166687,
+      "learning_rate": 0.001,
+      "loss": 1.064,
+      "step": 20190
+    },
+    {
+      "epoch": 0.6441531936605122,
+      "grad_norm": 0.16327440738677979,
+      "learning_rate": 0.001,
+      "loss": 1.0887,
+      "step": 20200
+    },
+    {
+      "epoch": 0.644472081380146,
+      "grad_norm": 0.16816017031669617,
+      "learning_rate": 0.001,
+      "loss": 1.1073,
+      "step": 20210
+    },
+    {
+      "epoch": 0.64479096909978,
+      "grad_norm": 0.16961325705051422,
+      "learning_rate": 0.001,
+      "loss": 1.0847,
+      "step": 20220
+    },
+    {
+      "epoch": 0.6451098568194139,
+      "grad_norm": 0.17201103270053864,
+      "learning_rate": 0.001,
+      "loss": 1.0926,
+      "step": 20230
+    },
+    {
+      "epoch": 0.6454287445390477,
+      "grad_norm": 0.16639626026153564,
+      "learning_rate": 0.001,
+      "loss": 1.0798,
+      "step": 20240
+    },
+    {
+      "epoch": 0.6457476322586817,
+      "grad_norm": 0.16544729471206665,
+      "learning_rate": 0.001,
+      "loss": 1.0707,
+      "step": 20250
+    },
+    {
+      "epoch": 0.6460665199783157,
+      "grad_norm": 0.1765151172876358,
+      "learning_rate": 0.001,
+      "loss": 1.0868,
+      "step": 20260
+    },
+    {
+      "epoch": 0.6463854076979495,
+      "grad_norm": 0.1677905023097992,
+      "learning_rate": 0.001,
+      "loss": 1.0962,
+      "step": 20270
+    },
+    {
+      "epoch": 0.6467042954175835,
+      "grad_norm": 0.16907256841659546,
+      "learning_rate": 0.001,
+      "loss": 1.1083,
+      "step": 20280
+    },
+    {
+      "epoch": 0.6470231831372174,
+      "grad_norm": 0.16835346817970276,
+      "learning_rate": 0.001,
+      "loss": 1.0787,
+      "step": 20290
+    },
+    {
+      "epoch": 0.6473420708568512,
+      "grad_norm": 0.17407572269439697,
+      "learning_rate": 0.001,
+      "loss": 1.0952,
+      "step": 20300
+    },
+    {
+      "epoch": 0.6476609585764852,
+      "grad_norm": 0.17516972124576569,
+      "learning_rate": 0.001,
+      "loss": 1.0838,
+      "step": 20310
+    },
+    {
+      "epoch": 0.6479798462961192,
+      "grad_norm": 0.16547511518001556,
+      "learning_rate": 0.001,
+      "loss": 1.0819,
+      "step": 20320
+    },
+    {
+      "epoch": 0.6482987340157531,
+      "grad_norm": 0.16972166299819946,
+      "learning_rate": 0.001,
+      "loss": 1.0617,
+      "step": 20330
+    },
+    {
+      "epoch": 0.648617621735387,
+      "grad_norm": 0.17067791521549225,
+      "learning_rate": 0.001,
+      "loss": 1.0962,
+      "step": 20340
+    },
+    {
+      "epoch": 0.6489365094550209,
+      "grad_norm": 0.16942836344242096,
+      "learning_rate": 0.001,
+      "loss": 1.1061,
+      "step": 20350
+    },
+    {
+      "epoch": 0.6492553971746549,
+      "grad_norm": 0.16373704373836517,
+      "learning_rate": 0.001,
+      "loss": 1.094,
+      "step": 20360
+    },
+    {
+      "epoch": 0.6495742848942887,
+      "grad_norm": 0.1656084656715393,
+      "learning_rate": 0.001,
+      "loss": 1.0858,
+      "step": 20370
+    },
+    {
+      "epoch": 0.6498931726139227,
+      "grad_norm": 0.1641538143157959,
+      "learning_rate": 0.001,
+      "loss": 1.1047,
+      "step": 20380
+    },
+    {
+      "epoch": 0.6502120603335566,
+      "grad_norm": 0.16529102623462677,
+      "learning_rate": 0.0009972136778050526,
+      "loss": 1.0647,
+      "step": 20390
+    },
+    {
+      "epoch": 0.6505309480531904,
+      "grad_norm": 0.17615461349487305,
+      "learning_rate": 0.0009930383279382303,
+      "loss": 1.0964,
+      "step": 20400
+    },
+    {
+      "epoch": 0.6508498357728244,
+      "grad_norm": 0.17260606586933136,
+      "learning_rate": 0.000988880460329121,
+      "loss": 1.0724,
+      "step": 20410
+    },
+    {
+      "epoch": 0.6511687234924584,
+      "grad_norm": 0.16743755340576172,
+      "learning_rate": 0.000984740001779228,
+      "loss": 1.0932,
+      "step": 20420
+    },
+    {
+      "epoch": 0.6514876112120922,
+      "grad_norm": 0.1746867150068283,
+      "learning_rate": 0.000980616879396537,
+      "loss": 1.0988,
+      "step": 20430
+    },
+    {
+      "epoch": 0.6518064989317262,
+      "grad_norm": 0.1718806028366089,
+      "learning_rate": 0.0009765110205942349,
+      "loss": 1.1176,
+      "step": 20440
+    },
+    {
+      "epoch": 0.6521253866513601,
+      "grad_norm": 0.1670590043067932,
+      "learning_rate": 0.0009724223530894298,
+      "loss": 1.0917,
+      "step": 20450
+    },
+    {
+      "epoch": 0.652444274370994,
+      "grad_norm": 0.1681852489709854,
+      "learning_rate": 0.00096835080490188,
+      "loss": 1.0839,
+      "step": 20460
+    },
+    {
+      "epoch": 0.6527631620906279,
+      "grad_norm": 0.17223966121673584,
+      "learning_rate": 0.0009642963043527262,
+      "loss": 1.0833,
+      "step": 20470
+    },
+    {
+      "epoch": 0.6530820498102619,
+      "grad_norm": 0.16178889572620392,
+      "learning_rate": 0.0009602587800632295,
+      "loss": 1.0536,
+      "step": 20480
+    },
+    {
+      "epoch": 0.6534009375298957,
+      "grad_norm": 0.1683199554681778,
+      "learning_rate": 0.0009562381609535146,
+      "loss": 1.0837,
+      "step": 20490
+    },
+    {
+      "epoch": 0.6537198252495297,
+      "grad_norm": 0.17109110951423645,
+      "learning_rate": 0.0009522343762413196,
+      "loss": 1.0786,
+      "step": 20500
+    },
+    {
+      "epoch": 0.6540387129691636,
+      "grad_norm": 0.1645098775625229,
+      "learning_rate": 0.0009482473554407485,
+      "loss": 1.0779,
+      "step": 20510
+    },
+    {
+      "epoch": 0.6543576006887974,
+      "grad_norm": 0.1678561270236969,
+      "learning_rate": 0.000944277028361031,
+      "loss": 1.0829,
+      "step": 20520
+    },
+    {
+      "epoch": 0.6546764884084314,
+      "grad_norm": 0.16512475907802582,
+      "learning_rate": 0.0009403233251052866,
+      "loss": 1.068,
+      "step": 20530
+    },
+    {
+      "epoch": 0.6549953761280654,
+      "grad_norm": 0.1729077249765396,
+      "learning_rate": 0.0009363861760692945,
+      "loss": 1.0914,
+      "step": 20540
+    },
+    {
+      "epoch": 0.6553142638476992,
+      "grad_norm": 0.16553378105163574,
+      "learning_rate": 0.0009324655119402678,
+      "loss": 1.0798,
+      "step": 20550
+    },
+    {
+      "epoch": 0.6556331515673332,
+      "grad_norm": 0.16073313355445862,
+      "learning_rate": 0.0009285612636956329,
+      "loss": 1.0699,
+      "step": 20560
+    },
+    {
+      "epoch": 0.6559520392869671,
+      "grad_norm": 0.16968417167663574,
+      "learning_rate": 0.0009246733626018155,
+      "loss": 1.0918,
+      "step": 20570
+    },
+    {
+      "epoch": 0.656270927006601,
+      "grad_norm": 0.1629115641117096,
+      "learning_rate": 0.0009208017402130296,
+      "loss": 1.0878,
+      "step": 20580
+    },
+    {
+      "epoch": 0.6565898147262349,
+      "grad_norm": 0.16741321980953217,
+      "learning_rate": 0.0009169463283700727,
+      "loss": 1.0725,
+      "step": 20590
+    },
+    {
+      "epoch": 0.6569087024458689,
+      "grad_norm": 0.16493017971515656,
+      "learning_rate": 0.0009131070591991262,
+      "loss": 1.0786,
+      "step": 20600
+    },
+    {
+      "epoch": 0.6572275901655027,
+      "grad_norm": 0.1635984629392624,
+      "learning_rate": 0.00090928386511056,
+      "loss": 1.0826,
+      "step": 20610
+    },
+    {
+      "epoch": 0.6575464778851366,
+      "grad_norm": 0.16892851889133453,
+      "learning_rate": 0.0009054766787977433,
+      "loss": 1.0711,
+      "step": 20620
+    },
+    {
+      "epoch": 0.6578653656047706,
+      "grad_norm": 0.1606256067752838,
+      "learning_rate": 0.0009016854332358588,
+      "loss": 1.056,
+      "step": 20630
+    },
+    {
+      "epoch": 0.6581842533244044,
+      "grad_norm": 0.16923518478870392,
+      "learning_rate": 0.0008979100616807236,
+      "loss": 1.1132,
+      "step": 20640
+    },
+    {
+      "epoch": 0.6585031410440384,
+      "grad_norm": 0.17301279306411743,
+      "learning_rate": 0.0008941504976676136,
+      "loss": 1.0764,
+      "step": 20650
+    },
+    {
+      "epoch": 0.6588220287636724,
+      "grad_norm": 0.1645253598690033,
+      "learning_rate": 0.0008904066750100934,
+      "loss": 1.0599,
+      "step": 20660
+    },
+    {
+      "epoch": 0.6591409164833062,
+      "grad_norm": 0.16737417876720428,
+      "learning_rate": 0.0008866785277988516,
+      "loss": 1.0728,
+      "step": 20670
+    },
+    {
+      "epoch": 0.6594598042029401,
+      "grad_norm": 0.15950274467468262,
+      "learning_rate": 0.00088296599040054,
+      "loss": 1.0716,
+      "step": 20680
+    },
+    {
+      "epoch": 0.6597786919225741,
+      "grad_norm": 0.1676434576511383,
+      "learning_rate": 0.0008792689974566183,
+      "loss": 1.0655,
+      "step": 20690
+    },
+    {
+      "epoch": 0.6600975796422079,
+      "grad_norm": 0.16000451147556305,
+      "learning_rate": 0.0008755874838822034,
+      "loss": 1.0722,
+      "step": 20700
+    },
+    {
+      "epoch": 0.6604164673618419,
+      "grad_norm": 0.1731007993221283,
+      "learning_rate": 0.0008719213848649239,
+      "loss": 1.0753,
+      "step": 20710
+    },
+    {
+      "epoch": 0.6607353550814759,
+      "grad_norm": 0.16616320610046387,
+      "learning_rate": 0.0008682706358637785,
+      "loss": 1.0395,
+      "step": 20720
+    },
+    {
+      "epoch": 0.6610542428011097,
+      "grad_norm": 0.1657589226961136,
+      "learning_rate": 0.0008646351726080005,
+      "loss": 1.0694,
+      "step": 20730
+    },
+    {
+      "epoch": 0.6613731305207436,
+      "grad_norm": 0.1682083159685135,
+      "learning_rate": 0.0008610149310959256,
+      "loss": 1.0667,
+      "step": 20740
+    },
+    {
+      "epoch": 0.6616920182403776,
+      "grad_norm": 0.16832314431667328,
+      "learning_rate": 0.0008574098475938659,
+      "loss": 1.0754,
+      "step": 20750
+    },
+    {
+      "epoch": 0.6620109059600114,
+      "grad_norm": 0.1608540415763855,
+      "learning_rate": 0.000853819858634987,
+      "loss": 1.0568,
+      "step": 20760
+    },
+    {
+      "epoch": 0.6623297936796454,
+      "grad_norm": 0.1716543585062027,
+      "learning_rate": 0.0008502449010181915,
+      "loss": 1.0788,
+      "step": 20770
+    },
+    {
+      "epoch": 0.6626486813992793,
+      "grad_norm": 0.16817323863506317,
+      "learning_rate": 0.0008466849118070059,
+      "loss": 1.0665,
+      "step": 20780
+    },
+    {
+      "epoch": 0.6629675691189132,
+      "grad_norm": 0.16492323577404022,
+      "learning_rate": 0.0008431398283284729,
+      "loss": 1.0679,
+      "step": 20790
+    },
+    {
+      "epoch": 0.6632864568385471,
+      "grad_norm": 0.16486430168151855,
+      "learning_rate": 0.0008396095881720477,
+      "loss": 1.0807,
+      "step": 20800
+    },
+    {
+      "epoch": 0.6636053445581811,
+      "grad_norm": 0.16902168095111847,
+      "learning_rate": 0.0008360941291884995,
+      "loss": 1.0478,
+      "step": 20810
+    },
+    {
+      "epoch": 0.6639242322778149,
+      "grad_norm": 0.1692483276128769,
+      "learning_rate": 0.0008325933894888175,
+      "loss": 1.0753,
+      "step": 20820
+    },
+    {
+      "epoch": 0.6642431199974489,
+      "grad_norm": 0.16759838163852692,
+      "learning_rate": 0.0008291073074431209,
+      "loss": 1.0563,
+      "step": 20830
+    },
+    {
+      "epoch": 0.6645620077170828,
+      "grad_norm": 0.16476279497146606,
+      "learning_rate": 0.0008256358216795744,
+      "loss": 1.0563,
+      "step": 20840
+    },
+    {
+      "epoch": 0.6648808954367167,
+      "grad_norm": 0.16580504179000854,
+      "learning_rate": 0.0008221788710833077,
+      "loss": 1.0632,
+      "step": 20850
+    },
+    {
+      "epoch": 0.6651997831563506,
+      "grad_norm": 0.16249163448810577,
+      "learning_rate": 0.0008187363947953391,
+      "loss": 1.0545,
+      "step": 20860
+    },
+    {
+      "epoch": 0.6655186708759846,
+      "grad_norm": 0.16070739924907684,
+      "learning_rate": 0.0008153083322115049,
+      "loss": 1.0685,
+      "step": 20870
+    },
+    {
+      "epoch": 0.6658375585956184,
+      "grad_norm": 0.1714041382074356,
+      "learning_rate": 0.0008118946229813915,
+      "loss": 1.0571,
+      "step": 20880
+    },
+    {
+      "epoch": 0.6661564463152524,
+      "grad_norm": 0.1640019714832306,
+      "learning_rate": 0.0008084952070072738,
+      "loss": 1.0558,
+      "step": 20890
+    },
+    {
+      "epoch": 0.6664753340348863,
+      "grad_norm": 0.16367119550704956,
+      "learning_rate": 0.0008051100244430569,
+      "loss": 1.0645,
+      "step": 20900
+    },
+    {
+      "epoch": 0.6667942217545202,
+      "grad_norm": 0.16876378655433655,
+      "learning_rate": 0.0008017390156932223,
+      "loss": 1.0564,
+      "step": 20910
+    },
+    {
+      "epoch": 0.6671131094741541,
+      "grad_norm": 0.1563488394021988,
+      "learning_rate": 0.0007983821214117789,
+      "loss": 1.0362,
+      "step": 20920
+    },
+    {
+      "epoch": 0.6674319971937881,
+      "grad_norm": 0.1577063649892807,
+      "learning_rate": 0.0007950392825012183,
+      "loss": 1.0588,
+      "step": 20930
+    },
+    {
+      "epoch": 0.6677508849134219,
+      "grad_norm": 0.16953368484973907,
+      "learning_rate": 0.0007917104401114743,
+      "loss": 1.0446,
+      "step": 20940
+    },
+    {
+      "epoch": 0.6680697726330559,
+      "grad_norm": 0.17253628373146057,
+      "learning_rate": 0.000788395535638887,
+      "loss": 1.0621,
+      "step": 20950
+    },
+    {
+      "epoch": 0.6683886603526898,
+      "grad_norm": 0.1640058308839798,
+      "learning_rate": 0.0007850945107251707,
+      "loss": 1.0624,
+      "step": 20960
+    },
+    {
+      "epoch": 0.6687075480723237,
+      "grad_norm": 0.16152934730052948,
+      "learning_rate": 0.000781807307256387,
+      "loss": 1.0424,
+      "step": 20970
+    },
+    {
+      "epoch": 0.6690264357919576,
+      "grad_norm": 0.15980294346809387,
+      "learning_rate": 0.0007785338673619216,
+      "loss": 1.0482,
+      "step": 20980
+    },
+    {
+      "epoch": 0.6693453235115916,
+      "grad_norm": 0.16612927615642548,
+      "learning_rate": 0.0007752741334134652,
+      "loss": 1.0475,
+      "step": 20990
+    },
+    {
+      "epoch": 0.6696642112312254,
+      "grad_norm": 0.16465890407562256,
+      "learning_rate": 0.0007720280480239992,
+      "loss": 1.0668,
+      "step": 21000
+    },
+    {
+      "epoch": 0.6699830989508594,
+      "grad_norm": 0.1595165729522705,
+      "learning_rate": 0.0007687955540467853,
+      "loss": 1.0507,
+      "step": 21010
+    },
+    {
+      "epoch": 0.6703019866704933,
+      "grad_norm": 0.16686223447322845,
+      "learning_rate": 0.0007655765945743598,
+      "loss": 1.0484,
+      "step": 21020
+    },
+    {
+      "epoch": 0.6706208743901272,
+      "grad_norm": 0.1746356338262558,
+      "learning_rate": 0.0007623711129375314,
+      "loss": 1.0527,
+      "step": 21030
+    },
+    {
+      "epoch": 0.6709397621097611,
+      "grad_norm": 0.16428135335445404,
+      "learning_rate": 0.0007591790527043832,
+      "loss": 1.0506,
+      "step": 21040
+    },
+    {
+      "epoch": 0.6712586498293951,
+      "grad_norm": 0.15915681421756744,
+      "learning_rate": 0.0007560003576792802,
+      "loss": 1.0527,
+      "step": 21050
+    },
+    {
+      "epoch": 0.6715775375490289,
+      "grad_norm": 0.1672055721282959,
+      "learning_rate": 0.0007528349719018794,
+      "loss": 1.0528,
+      "step": 21060
+    },
+    {
+      "epoch": 0.6718964252686629,
+      "grad_norm": 0.1689584106206894,
+      "learning_rate": 0.0007496828396461442,
+      "loss": 1.0495,
+      "step": 21070
+    },
+    {
+      "epoch": 0.6722153129882968,
+      "grad_norm": 0.16179822385311127,
+      "learning_rate": 0.0007465439054193641,
+      "loss": 1.0713,
+      "step": 21080
+    },
+    {
+      "epoch": 0.6725342007079307,
+      "grad_norm": 0.1632169634103775,
+      "learning_rate": 0.0007434181139611777,
+      "loss": 1.0444,
+      "step": 21090
+    },
+    {
+      "epoch": 0.6728530884275646,
+      "grad_norm": 0.16350291669368744,
+      "learning_rate": 0.0007403054102425991,
+      "loss": 1.0191,
+      "step": 21100
+    },
+    {
+      "epoch": 0.6731719761471986,
+      "grad_norm": 0.16606953740119934,
+      "learning_rate": 0.0007372057394650503,
+      "loss": 1.0697,
+      "step": 21110
+    },
+    {
+      "epoch": 0.6734908638668324,
+      "grad_norm": 0.16665737330913544,
+      "learning_rate": 0.0007341190470593954,
+      "loss": 1.045,
+      "step": 21120
+    },
+    {
+      "epoch": 0.6738097515864664,
+      "grad_norm": 0.1595230996608734,
+      "learning_rate": 0.0007310452786849806,
+      "loss": 1.0286,
+      "step": 21130
+    },
+    {
+      "epoch": 0.6741286393061003,
+      "grad_norm": 0.1614699512720108,
+      "learning_rate": 0.0007279843802286769,
+      "loss": 1.045,
+      "step": 21140
+    },
+    {
+      "epoch": 0.6744475270257342,
+      "grad_norm": 0.16494181752204895,
+      "learning_rate": 0.0007249362978039282,
+      "loss": 1.0395,
+      "step": 21150
+    },
+    {
+      "epoch": 0.6747664147453681,
+      "grad_norm": 0.164317786693573,
+      "learning_rate": 0.0007219009777498024,
+      "loss": 1.0667,
+      "step": 21160
+    },
+    {
+      "epoch": 0.6750853024650021,
+      "grad_norm": 0.16172905266284943,
+      "learning_rate": 0.0007188783666300463,
+      "loss": 1.027,
+      "step": 21170
+    },
+    {
+      "epoch": 0.675404190184636,
+      "grad_norm": 0.16476504504680634,
+      "learning_rate": 0.000715868411232145,
+      "loss": 1.0615,
+      "step": 21180
+    },
+    {
+      "epoch": 0.6757230779042699,
+      "grad_norm": 0.1657302975654602,
+      "learning_rate": 0.0007128710585663859,
+      "loss": 1.0207,
+      "step": 21190
+    },
+    {
+      "epoch": 0.6760419656239038,
+      "grad_norm": 0.1605929583311081,
+      "learning_rate": 0.0007098862558649246,
+      "loss": 1.0271,
+      "step": 21200
+    },
+    {
+      "epoch": 0.6763608533435378,
+      "grad_norm": 0.16407832503318787,
+      "learning_rate": 0.000706913950580857,
+      "loss": 1.0121,
+      "step": 21210
+    },
+    {
+      "epoch": 0.6766797410631716,
+      "grad_norm": 0.15979787707328796,
+      "learning_rate": 0.000703954090387293,
+      "loss": 1.034,
+      "step": 21220
+    },
+    {
+      "epoch": 0.6769986287828056,
+      "grad_norm": 0.1594686061143875,
+      "learning_rate": 0.0007010066231764369,
+      "loss": 1.0199,
+      "step": 21230
+    },
+    {
+      "epoch": 0.6773175165024395,
+      "grad_norm": 0.16522930562496185,
+      "learning_rate": 0.0006980714970586688,
+      "loss": 1.0329,
+      "step": 21240
+    },
+    {
+      "epoch": 0.6776364042220734,
+      "grad_norm": 0.16075100004673004,
+      "learning_rate": 0.0006951486603616313,
+      "loss": 1.0463,
+      "step": 21250
+    },
+    {
+      "epoch": 0.6779552919417073,
+      "grad_norm": 0.1620459407567978,
+      "learning_rate": 0.0006922380616293202,
+      "loss": 1.0279,
+      "step": 21260
+    },
+    {
+      "epoch": 0.6782741796613413,
+      "grad_norm": 0.17030924558639526,
+      "learning_rate": 0.0006893396496211784,
+      "loss": 1.0406,
+      "step": 21270
+    },
+    {
+      "epoch": 0.6785930673809751,
+      "grad_norm": 0.16657640039920807,
+      "learning_rate": 0.0006864533733111942,
+      "loss": 1.0515,
+      "step": 21280
+    },
+    {
+      "epoch": 0.6789119551006091,
+      "grad_norm": 0.16453786194324493,
+      "learning_rate": 0.0006835791818870018,
+      "loss": 1.0203,
+      "step": 21290
+    },
+    {
+      "epoch": 0.679230842820243,
+      "grad_norm": 0.1666298508644104,
+      "learning_rate": 0.0006807170247489883,
+      "loss": 1.0369,
+      "step": 21300
+    },
+    {
+      "epoch": 0.6795497305398769,
+      "grad_norm": 0.16200481355190277,
+      "learning_rate": 0.0006778668515094021,
+      "loss": 1.0376,
+      "step": 21310
+    },
+    {
+      "epoch": 0.6798686182595108,
+      "grad_norm": 0.16168266534805298,
+      "learning_rate": 0.0006750286119914656,
+      "loss": 1.0258,
+      "step": 21320
+    },
+    {
+      "epoch": 0.6801875059791448,
+      "grad_norm": 0.16025297343730927,
+      "learning_rate": 0.0006722022562284926,
+      "loss": 1.0484,
+      "step": 21330
+    },
+    {
+      "epoch": 0.6805063936987786,
+      "grad_norm": 0.16829195618629456,
+      "learning_rate": 0.000669387734463008,
+      "loss": 1.0187,
+      "step": 21340
+    },
+    {
+      "epoch": 0.6808252814184126,
+      "grad_norm": 0.16115964949131012,
+      "learning_rate": 0.0006665849971458721,
+      "loss": 1.046,
+      "step": 21350
+    },
+    {
+      "epoch": 0.6811441691380465,
+      "grad_norm": 0.1624351292848587,
+      "learning_rate": 0.0006637939949354081,
+      "loss": 1.0218,
+      "step": 21360
+    },
+    {
+      "epoch": 0.6814630568576804,
+      "grad_norm": 0.1657162606716156,
+      "learning_rate": 0.000661014678696534,
+      "loss": 1.0321,
+      "step": 21370
+    },
+    {
+      "epoch": 0.6817819445773143,
+      "grad_norm": 0.16228361427783966,
+      "learning_rate": 0.0006582469994998967,
+      "loss": 1.0553,
+      "step": 21380
+    },
+    {
+      "epoch": 0.6821008322969483,
+      "grad_norm": 0.16864708065986633,
+      "learning_rate": 0.0006554909086210115,
+      "loss": 1.0299,
+      "step": 21390
+    },
+    {
+      "epoch": 0.6824197200165821,
+      "grad_norm": 0.16412338614463806,
+      "learning_rate": 0.0006527463575394037,
+      "loss": 1.0311,
+      "step": 21400
+    },
+    {
+      "epoch": 0.6827386077362161,
+      "grad_norm": 0.15431377291679382,
+      "learning_rate": 0.0006500132979377546,
+      "loss": 1.0314,
+      "step": 21410
+    },
+    {
+      "epoch": 0.68305749545585,
+      "grad_norm": 0.16200804710388184,
+      "learning_rate": 0.0006472916817010511,
+      "loss": 1.0223,
+      "step": 21420
+    },
+    {
+      "epoch": 0.6833763831754839,
+      "grad_norm": 0.16413038969039917,
+      "learning_rate": 0.0006445814609157381,
+      "loss": 1.0238,
+      "step": 21430
+    },
+    {
+      "epoch": 0.6836952708951178,
+      "grad_norm": 0.1650371551513672,
+      "learning_rate": 0.0006418825878688756,
+      "loss": 1.0315,
+      "step": 21440
+    },
+    {
+      "epoch": 0.6840141586147518,
+      "grad_norm": 0.16407425701618195,
+      "learning_rate": 0.0006391950150472985,
+      "loss": 1.0199,
+      "step": 21450
+    },
+    {
+      "epoch": 0.6843330463343856,
+      "grad_norm": 0.1609756350517273,
+      "learning_rate": 0.0006365186951367798,
+      "loss": 1.0235,
+      "step": 21460
+    },
+    {
+      "epoch": 0.6846519340540196,
+      "grad_norm": 0.16075292229652405,
+      "learning_rate": 0.0006338535810211983,
+      "loss": 1.0267,
+      "step": 21470
+    },
+    {
+      "epoch": 0.6849708217736535,
+      "grad_norm": 0.15738439559936523,
+      "learning_rate": 0.0006311996257817083,
+      "loss": 1.0226,
+      "step": 21480
+    },
+    {
+      "epoch": 0.6852897094932874,
+      "grad_norm": 0.16439034044742584,
+      "learning_rate": 0.0006285567826959148,
+      "loss": 1.0295,
+      "step": 21490
+    },
+    {
+      "epoch": 0.6856085972129213,
+      "grad_norm": 0.16584716737270355,
+      "learning_rate": 0.0006259250052370493,
+      "loss": 1.0455,
+      "step": 21500
+    },
+    {
+      "epoch": 0.6859274849325553,
+      "grad_norm": 0.15936855971813202,
+      "learning_rate": 0.0006233042470731524,
+      "loss": 1.0172,
+      "step": 21510
+    },
+    {
+      "epoch": 0.6862463726521891,
+      "grad_norm": 0.15984030067920685,
+      "learning_rate": 0.0006206944620662569,
+      "loss": 1.0297,
+      "step": 21520
+    },
+    {
+      "epoch": 0.6865652603718231,
+      "grad_norm": 0.15906420350074768,
+      "learning_rate": 0.0006180956042715764,
+      "loss": 1.016,
+      "step": 21530
+    },
+    {
+      "epoch": 0.686884148091457,
+      "grad_norm": 0.16585637629032135,
+      "learning_rate": 0.0006155076279366958,
+      "loss": 1.0237,
+      "step": 21540
+    },
+    {
+      "epoch": 0.6872030358110909,
+      "grad_norm": 0.1607387810945511,
+      "learning_rate": 0.0006129304875007661,
+      "loss": 1.0156,
+      "step": 21550
+    },
+    {
+      "epoch": 0.6875219235307248,
+      "grad_norm": 0.16103488206863403,
+      "learning_rate": 0.0006103641375937023,
+      "loss": 1.0243,
+      "step": 21560
+    },
+    {
+      "epoch": 0.6878408112503588,
+      "grad_norm": 0.16695347428321838,
+      "learning_rate": 0.0006078085330353851,
+      "loss": 1.0086,
+      "step": 21570
+    },
+    {
+      "epoch": 0.6881596989699926,
+      "grad_norm": 0.16508355736732483,
+      "learning_rate": 0.0006052636288348644,
+      "loss": 0.9974,
+      "step": 21580
+    },
+    {
+      "epoch": 0.6884785866896266,
+      "grad_norm": 0.16593703627586365,
+      "learning_rate": 0.0006027293801895685,
+      "loss": 1.0211,
+      "step": 21590
+    },
+    {
+      "epoch": 0.6887974744092605,
+      "grad_norm": 0.16203592717647552,
+      "learning_rate": 0.0006002057424845144,
+      "loss": 1.0126,
+      "step": 21600
+    },
+    {
+      "epoch": 0.6891163621288944,
+      "grad_norm": 0.16561388969421387,
+      "learning_rate": 0.0005976926712915233,
+      "loss": 1.0207,
+      "step": 21610
+    },
+    {
+      "epoch": 0.6894352498485283,
+      "grad_norm": 0.1665153205394745,
+      "learning_rate": 0.0005951901223684372,
+      "loss": 1.0224,
+      "step": 21620
+    },
+    {
+      "epoch": 0.6897541375681623,
+      "grad_norm": 0.1631123572587967,
+      "learning_rate": 0.0005926980516583412,
+      "loss": 1.007,
+      "step": 21630
+    },
+    {
+      "epoch": 0.6900730252877961,
+      "grad_norm": 0.16275134682655334,
+      "learning_rate": 0.0005902164152887875,
+      "loss": 1.0144,
+      "step": 21640
+    },
+    {
+      "epoch": 0.6903919130074301,
+      "grad_norm": 0.16036659479141235,
+      "learning_rate": 0.0005877451695710226,
+      "loss": 1.0222,
+      "step": 21650
+    },
+    {
+      "epoch": 0.690710800727064,
+      "grad_norm": 0.15921048820018768,
+      "learning_rate": 0.0005852842709992187,
+      "loss": 1.0201,
+      "step": 21660
+    },
+    {
+      "epoch": 0.6910296884466979,
+      "grad_norm": 0.1643412709236145,
+      "learning_rate": 0.0005828336762497074,
+      "loss": 1.0071,
+      "step": 21670
+    },
+    {
+      "epoch": 0.6913485761663318,
+      "grad_norm": 0.16244398057460785,
+      "learning_rate": 0.0005803933421802178,
+      "loss": 1.0255,
+      "step": 21680
+    },
+    {
+      "epoch": 0.6916674638859658,
+      "grad_norm": 0.16339954733848572,
+      "learning_rate": 0.0005779632258291156,
+      "loss": 1.0173,
+      "step": 21690
+    },
+    {
+      "epoch": 0.6919863516055996,
+      "grad_norm": 0.16126830875873566,
+      "learning_rate": 0.0005755432844146483,
+      "loss": 1.018,
+      "step": 21700
+    },
+    {
+      "epoch": 0.6923052393252336,
+      "grad_norm": 0.1622725874185562,
+      "learning_rate": 0.0005731334753341907,
+      "loss": 1.0112,
+      "step": 21710
+    },
+    {
+      "epoch": 0.6926241270448675,
+      "grad_norm": 0.1686084270477295,
+      "learning_rate": 0.0005707337561634957,
+      "loss": 1.0075,
+      "step": 21720
+    },
+    {
+      "epoch": 0.6929430147645014,
+      "grad_norm": 0.16552889347076416,
+      "learning_rate": 0.0005683440846559473,
+      "loss": 1.0229,
+      "step": 21730
+    },
+    {
+      "epoch": 0.6932619024841353,
+      "grad_norm": 0.1642754077911377,
+      "learning_rate": 0.0005659644187418168,
+      "loss": 1.0056,
+      "step": 21740
+    },
+    {
+      "epoch": 0.6935807902037693,
+      "grad_norm": 0.1628817319869995,
+      "learning_rate": 0.0005635947165275219,
+      "loss": 1.0098,
+      "step": 21750
+    },
+    {
+      "epoch": 0.6938996779234031,
+      "grad_norm": 0.16525153815746307,
+      "learning_rate": 0.0005612349362948896,
+      "loss": 1.0185,
+      "step": 21760
+    },
+    {
+      "epoch": 0.6942185656430371,
+      "grad_norm": 0.1632060408592224,
+      "learning_rate": 0.0005588850365004215,
+      "loss": 1.0148,
+      "step": 21770
+    },
+    {
+      "epoch": 0.694537453362671,
+      "grad_norm": 0.1638498157262802,
+      "learning_rate": 0.0005565449757745625,
+      "loss": 1.0189,
+      "step": 21780
+    },
+    {
+      "epoch": 0.6948563410823049,
+      "grad_norm": 0.15921951830387115,
+      "learning_rate": 0.0005542147129209725,
+      "loss": 1.012,
+      "step": 21790
+    },
+    {
+      "epoch": 0.6951752288019388,
+      "grad_norm": 0.1655687838792801,
+      "learning_rate": 0.0005518942069158012,
+      "loss": 1.0176,
+      "step": 21800
+    },
+    {
+      "epoch": 0.6954941165215728,
+      "grad_norm": 0.16503483057022095,
+      "learning_rate": 0.0005495834169069658,
+      "loss": 1.0059,
+      "step": 21810
+    },
+    {
+      "epoch": 0.6958130042412066,
+      "grad_norm": 0.1610029637813568,
+      "learning_rate": 0.0005472823022134319,
+      "loss": 1.0158,
+      "step": 21820
+    },
+    {
+      "epoch": 0.6961318919608406,
+      "grad_norm": 0.1622898131608963,
+      "learning_rate": 0.000544990822324497,
+      "loss": 0.997,
+      "step": 21830
+    },
+    {
+      "epoch": 0.6964507796804745,
+      "grad_norm": 0.16136029362678528,
+      "learning_rate": 0.0005427089368990779,
+      "loss": 1.0004,
+      "step": 21840
+    },
+    {
+      "epoch": 0.6967696674001084,
+      "grad_norm": 0.16808313131332397,
+      "learning_rate": 0.0005404366057649998,
+      "loss": 1.0151,
+      "step": 21850
+    },
+    {
+      "epoch": 0.6970885551197423,
+      "grad_norm": 0.16658589243888855,
+      "learning_rate": 0.00053817378891829,
+      "loss": 1.0045,
+      "step": 21860
+    },
+    {
+      "epoch": 0.6974074428393763,
+      "grad_norm": 0.16027787327766418,
+      "learning_rate": 0.0005359204465224725,
+      "loss": 1.0188,
+      "step": 21870
+    },
+    {
+      "epoch": 0.6977263305590101,
+      "grad_norm": 0.1679868847131729,
+      "learning_rate": 0.0005336765389078676,
+      "loss": 1.0053,
+      "step": 21880
+    },
+    {
+      "epoch": 0.6980452182786441,
+      "grad_norm": 0.16373640298843384,
+      "learning_rate": 0.000531442026570893,
+      "loss": 1.0139,
+      "step": 21890
+    },
+    {
+      "epoch": 0.698364105998278,
+      "grad_norm": 0.16460253298282623,
+      "learning_rate": 0.0005292168701733688,
+      "loss": 1.0089,
+      "step": 21900
+    },
+    {
+      "epoch": 0.6986829937179119,
+      "grad_norm": 0.16029703617095947,
+      "learning_rate": 0.0005270010305418245,
+      "loss": 0.9898,
+      "step": 21910
+    },
+    {
+      "epoch": 0.6990018814375458,
+      "grad_norm": 0.16033555567264557,
+      "learning_rate": 0.0005247944686668097,
+      "loss": 0.9906,
+      "step": 21920
+    },
+    {
+      "epoch": 0.6993207691571798,
+      "grad_norm": 0.16450348496437073,
+      "learning_rate": 0.0005225971457022069,
+      "loss": 1.0056,
+      "step": 21930
+    },
+    {
+      "epoch": 0.6996396568768136,
+      "grad_norm": 0.16469576954841614,
+      "learning_rate": 0.0005204090229645483,
+      "loss": 0.9883,
+      "step": 21940
+    },
+    {
+      "epoch": 0.6999585445964476,
+      "grad_norm": 0.15795905888080597,
+      "learning_rate": 0.0005182300619323341,
+      "loss": 0.9956,
+      "step": 21950
+    },
+    {
+      "epoch": 0.7002774323160815,
+      "grad_norm": 0.1648419350385666,
+      "learning_rate": 0.0005160602242453551,
+      "loss": 1.0061,
+      "step": 21960
+    },
+    {
+      "epoch": 0.7005963200357154,
+      "grad_norm": 0.16443631052970886,
+      "learning_rate": 0.0005138994717040161,
+      "loss": 0.9819,
+      "step": 21970
+    },
+    {
+      "epoch": 0.7009152077553493,
+      "grad_norm": 0.16426272690296173,
+      "learning_rate": 0.0005117477662686652,
+      "loss": 1.0071,
+      "step": 21980
+    },
+    {
+      "epoch": 0.7012340954749833,
+      "grad_norm": 0.1597410887479782,
+      "learning_rate": 0.0005096050700589225,
+      "loss": 1.0085,
+      "step": 21990
+    },
+    {
+      "epoch": 0.7015529831946172,
+      "grad_norm": 0.16440628468990326,
+      "learning_rate": 0.0005074713453530141,
+      "loss": 0.9989,
+      "step": 22000
+    },
+    {
+      "epoch": 0.7018718709142511,
+      "grad_norm": 0.15793122351169586,
+      "learning_rate": 0.0005053465545871075,
+      "loss": 1.0088,
+      "step": 22010
+    },
+    {
+      "epoch": 0.702190758633885,
+      "grad_norm": 0.17422626912593842,
+      "learning_rate": 0.0005032306603546511,
+      "loss": 0.9918,
+      "step": 22020
+    },
+    {
+      "epoch": 0.702509646353519,
+      "grad_norm": 0.1663317233324051,
+      "learning_rate": 0.0005011236254057146,
+      "loss": 1.0069,
+      "step": 22030
+    },
+    {
+      "epoch": 0.7028285340731528,
+      "grad_norm": 0.1605159342288971,
+      "learning_rate": 0.0004990254126463343,
+      "loss": 0.9984,
+      "step": 22040
+    },
+    {
+      "epoch": 0.7031474217927868,
+      "grad_norm": 0.166510671377182,
+      "learning_rate": 0.0004969359851378588,
+      "loss": 1.0014,
+      "step": 22050
+    },
+    {
+      "epoch": 0.7034663095124207,
+      "grad_norm": 0.16222979128360748,
+      "learning_rate": 0.0004948553060963001,
+      "loss": 0.9773,
+      "step": 22060
+    },
+    {
+      "epoch": 0.7037851972320546,
+      "grad_norm": 0.17086096107959747,
+      "learning_rate": 0.0004927833388916852,
+      "loss": 1.0113,
+      "step": 22070
+    },
+    {
+      "epoch": 0.7041040849516885,
+      "grad_norm": 0.16001038253307343,
+      "learning_rate": 0.0004907200470474113,
+      "loss": 0.9999,
+      "step": 22080
+    },
+    {
+      "epoch": 0.7044229726713225,
+      "grad_norm": 0.1599177122116089,
+      "learning_rate": 0.0004886653942396035,
+      "loss": 0.9879,
+      "step": 22090
+    },
+    {
+      "epoch": 0.7047418603909563,
+      "grad_norm": 0.15955601632595062,
+      "learning_rate": 0.00048661934429647597,
+      "loss": 1.0056,
+      "step": 22100
+    },
+    {
+      "epoch": 0.7050607481105903,
+      "grad_norm": 0.168484166264534,
+      "learning_rate": 0.0004845818611976946,
+      "loss": 0.9948,
+      "step": 22110
+    },
+    {
+      "epoch": 0.7053796358302242,
+      "grad_norm": 0.1610783189535141,
+      "learning_rate": 0.0004825529090737429,
+      "loss": 0.9842,
+      "step": 22120
+    },
+    {
+      "epoch": 0.7056985235498581,
+      "grad_norm": 0.1566549837589264,
+      "learning_rate": 0.0004805324522052906,
+      "loss": 1.0026,
+      "step": 22130
+    },
+    {
+      "epoch": 0.706017411269492,
+      "grad_norm": 0.15728874504566193,
+      "learning_rate": 0.000478520455022565,
+      "loss": 0.9897,
+      "step": 22140
+    },
+    {
+      "epoch": 0.706336298989126,
+      "grad_norm": 0.1586431860923767,
+      "learning_rate": 0.0004765168821047247,
+      "loss": 0.9911,
+      "step": 22150
+    },
+    {
+      "epoch": 0.7066551867087598,
+      "grad_norm": 0.1617947518825531,
+      "learning_rate": 0.0004745216981792355,
+      "loss": 1.0057,
+      "step": 22160
+    },
+    {
+      "epoch": 0.7069740744283938,
+      "grad_norm": 0.15989385545253754,
+      "learning_rate": 0.00047253486812125044,
+      "loss": 0.9716,
+      "step": 22170
+    },
+    {
+      "epoch": 0.7072929621480277,
+      "grad_norm": 0.16351810097694397,
+      "learning_rate": 0.0004705563569529904,
+      "loss": 0.9917,
+      "step": 22180
+    },
+    {
+      "epoch": 0.7076118498676616,
+      "grad_norm": 0.1647985577583313,
+      "learning_rate": 0.00046858612984312904,
+      "loss": 1.0022,
+      "step": 22190
+    },
+    {
+      "epoch": 0.7079307375872955,
+      "grad_norm": 0.16196519136428833,
+      "learning_rate": 0.00046662415210617933,
+      "loss": 0.9898,
+      "step": 22200
+    },
+    {
+      "epoch": 0.7082496253069295,
+      "grad_norm": 0.15809550881385803,
+      "learning_rate": 0.00046467038920188283,
+      "loss": 0.991,
+      "step": 22210
+    },
+    {
+      "epoch": 0.7085685130265633,
+      "grad_norm": 0.1606694906949997,
+      "learning_rate": 0.0004627248067346017,
+      "loss": 0.9884,
+      "step": 22220
+    },
+    {
+      "epoch": 0.7088874007461973,
+      "grad_norm": 0.15880940854549408,
+      "learning_rate": 0.0004607873704527135,
+      "loss": 1.0084,
+      "step": 22230
+    },
+    {
+      "epoch": 0.7092062884658312,
+      "grad_norm": 0.1667298674583435,
+      "learning_rate": 0.00045885804624800757,
+      "loss": 1.0052,
+      "step": 22240
+    },
+    {
+      "epoch": 0.7095251761854651,
+      "grad_norm": 0.15849944949150085,
+      "learning_rate": 0.000456936800155085,
+      "loss": 0.9801,
+      "step": 22250
+    },
+    {
+      "epoch": 0.709844063905099,
+      "grad_norm": 0.16091540455818176,
+      "learning_rate": 0.00045502359835076047,
+      "loss": 0.9847,
+      "step": 22260
+    },
+    {
+      "epoch": 0.710162951624733,
+      "grad_norm": 0.165780708193779,
+      "learning_rate": 0.00045311840715346694,
+      "loss": 0.9868,
+      "step": 22270
+    },
+    {
+      "epoch": 0.7104818393443668,
+      "grad_norm": 0.16307060420513153,
+      "learning_rate": 0.0004512211930226627,
+      "loss": 0.9873,
+      "step": 22280
+    },
+    {
+      "epoch": 0.7108007270640008,
+      "grad_norm": 0.16260948777198792,
+      "learning_rate": 0.0004493319225582409,
+      "loss": 0.9976,
+      "step": 22290
+    },
+    {
+      "epoch": 0.7111196147836347,
+      "grad_norm": 0.16536939144134521,
+      "learning_rate": 0.00044745056249994127,
+      "loss": 0.9958,
+      "step": 22300
+    },
+    {
+      "epoch": 0.7114385025032686,
+      "grad_norm": 0.16163595020771027,
+      "learning_rate": 0.00044557707972676475,
+      "loss": 0.9889,
+      "step": 22310
+    },
+    {
+      "epoch": 0.7117573902229025,
+      "grad_norm": 0.1603289395570755,
+      "learning_rate": 0.0004437114412563908,
+      "loss": 0.9947,
+      "step": 22320
+    },
+    {
+      "epoch": 0.7120762779425365,
+      "grad_norm": 0.16332769393920898,
+      "learning_rate": 0.0004418536142445961,
+      "loss": 0.9886,
+      "step": 22330
+    },
+    {
+      "epoch": 0.7123951656621703,
+      "grad_norm": 0.16538774967193604,
+      "learning_rate": 0.0004400035659846766,
+      "loss": 0.9816,
+      "step": 22340
+    },
+    {
+      "epoch": 0.7127140533818043,
+      "grad_norm": 0.16004317998886108,
+      "learning_rate": 0.00043816126390687195,
+      "loss": 0.9704,
+      "step": 22350
+    },
+    {
+      "epoch": 0.7130329411014382,
+      "grad_norm": 0.1620021015405655,
+      "learning_rate": 0.0004363266755777918,
+      "loss": 0.9873,
+      "step": 22360
+    },
+    {
+      "epoch": 0.7133518288210721,
+      "grad_norm": 0.1575515866279602,
+      "learning_rate": 0.00043449976869984496,
+      "loss": 0.9837,
+      "step": 22370
+    },
+    {
+      "epoch": 0.713670716540706,
+      "grad_norm": 0.15856903791427612,
+      "learning_rate": 0.00043268051111067067,
+      "loss": 0.9711,
+      "step": 22380
+    },
+    {
+      "epoch": 0.71398960426034,
+      "grad_norm": 0.161849707365036,
+      "learning_rate": 0.00043086887078257267,
+      "loss": 0.9754,
+      "step": 22390
+    },
+    {
+      "epoch": 0.7143084919799738,
+      "grad_norm": 0.16092804074287415,
+      "learning_rate": 0.00042906481582195513,
+      "loss": 0.9665,
+      "step": 22400
+    },
+    {
+      "epoch": 0.7146273796996078,
+      "grad_norm": 0.16372103989124298,
+      "learning_rate": 0.0004272683144687611,
+      "loss": 0.997,
+      "step": 22410
+    },
+    {
+      "epoch": 0.7149462674192417,
+      "grad_norm": 0.1614946871995926,
+      "learning_rate": 0.0004254793350959138,
+      "loss": 0.9884,
+      "step": 22420
+    },
+    {
+      "epoch": 0.7152651551388756,
+      "grad_norm": 0.15983077883720398,
+      "learning_rate": 0.00042369784620875905,
+      "loss": 0.9708,
+      "step": 22430
+    },
+    {
+      "epoch": 0.7155840428585095,
+      "grad_norm": 0.16505417227745056,
+      "learning_rate": 0.00042192381644451176,
+      "loss": 0.9783,
+      "step": 22440
+    },
+    {
+      "epoch": 0.7159029305781435,
+      "grad_norm": 0.1651798039674759,
+      "learning_rate": 0.0004201572145717032,
+      "loss": 0.9792,
+      "step": 22450
+    },
+    {
+      "epoch": 0.7162218182977773,
+      "grad_norm": 0.16053563356399536,
+      "learning_rate": 0.0004183980094896312,
+      "loss": 0.9881,
+      "step": 22460
+    },
+    {
+      "epoch": 0.7165407060174113,
+      "grad_norm": 0.1612672060728073,
+      "learning_rate": 0.0004166461702278128,
+      "loss": 0.9749,
+      "step": 22470
+    },
+    {
+      "epoch": 0.7168595937370452,
+      "grad_norm": 0.1621503382921219,
+      "learning_rate": 0.0004149016659454389,
+      "loss": 0.9765,
+      "step": 22480
+    },
+    {
+      "epoch": 0.7171784814566791,
+      "grad_norm": 0.16153399646282196,
+      "learning_rate": 0.00041316446593083145,
+      "loss": 0.9799,
+      "step": 22490
+    },
+    {
+      "epoch": 0.717497369176313,
+      "grad_norm": 0.16426432132720947,
+      "learning_rate": 0.00041143453960090277,
+      "loss": 0.9826,
+      "step": 22500
+    },
+    {
+      "epoch": 0.717816256895947,
+      "grad_norm": 0.16212931275367737,
+      "learning_rate": 0.0004097118565006169,
+      "loss": 0.9799,
+      "step": 22510
+    },
+    {
+      "epoch": 0.7181351446155808,
+      "grad_norm": 0.16201704740524292,
+      "learning_rate": 0.00040799638630245356,
+      "loss": 0.9711,
+      "step": 22520
+    },
+    {
+      "epoch": 0.7184540323352148,
+      "grad_norm": 0.1674143671989441,
+      "learning_rate": 0.0004062880988058746,
+      "loss": 0.9731,
+      "step": 22530
+    },
+    {
+      "epoch": 0.7187729200548487,
+      "grad_norm": 0.1614939272403717,
+      "learning_rate": 0.0004045869639367917,
+      "loss": 0.9775,
+      "step": 22540
+    },
+    {
+      "epoch": 0.7190918077744826,
+      "grad_norm": 0.1606423556804657,
+      "learning_rate": 0.0004028929517470373,
+      "loss": 0.9737,
+      "step": 22550
+    },
+    {
+      "epoch": 0.7194106954941165,
+      "grad_norm": 0.15913470089435577,
+      "learning_rate": 0.00040120603241383754,
+      "loss": 0.9726,
+      "step": 22560
+    },
+    {
+      "epoch": 0.7197295832137505,
+      "grad_norm": 0.16026009619235992,
+      "learning_rate": 0.0003995261762392866,
+      "loss": 0.9765,
+      "step": 22570
+    },
+    {
+      "epoch": 0.7200484709333843,
+      "grad_norm": 0.1610075980424881,
+      "learning_rate": 0.0003978533536498247,
+      "loss": 0.9736,
+      "step": 22580
+    },
+    {
+      "epoch": 0.7203673586530183,
+      "grad_norm": 0.1646539866924286,
+      "learning_rate": 0.0003961875351957167,
+      "loss": 0.97,
+      "step": 22590
+    },
+    {
+      "epoch": 0.7206862463726522,
+      "grad_norm": 0.16013289988040924,
+      "learning_rate": 0.000394528691550534,
+      "loss": 0.9765,
+      "step": 22600
+    },
+    {
+      "epoch": 0.7210051340922861,
+      "grad_norm": 0.15997812151908875,
+      "learning_rate": 0.0003928767935106386,
+      "loss": 0.9622,
+      "step": 22610
+    },
+    {
+      "epoch": 0.72132402181192,
+      "grad_norm": 0.15702947974205017,
+      "learning_rate": 0.0003912318119946682,
+      "loss": 0.9789,
+      "step": 22620
+    },
+    {
+      "epoch": 0.721642909531554,
+      "grad_norm": 0.16323722898960114,
+      "learning_rate": 0.0003895937180430247,
+      "loss": 0.9719,
+      "step": 22630
+    },
+    {
+      "epoch": 0.7219617972511878,
+      "grad_norm": 0.15763328969478607,
+      "learning_rate": 0.0003879624828173645,
+      "loss": 0.9613,
+      "step": 22640
+    },
+    {
+      "epoch": 0.7222806849708218,
+      "grad_norm": 0.16371284425258636,
+      "learning_rate": 0.0003863380776000905,
+      "loss": 0.974,
+      "step": 22650
+    },
+    {
+      "epoch": 0.7225995726904557,
+      "grad_norm": 0.168775275349617,
+      "learning_rate": 0.0003847204737938466,
+      "loss": 0.9674,
+      "step": 22660
+    },
+    {
+      "epoch": 0.7229184604100896,
+      "grad_norm": 0.16056294739246368,
+      "learning_rate": 0.0003831096429210144,
+      "loss": 0.9665,
+      "step": 22670
+    },
+    {
+      "epoch": 0.7232373481297235,
+      "grad_norm": 0.16246573626995087,
+      "learning_rate": 0.00038150555662321163,
+      "loss": 0.9709,
+      "step": 22680
+    },
+    {
+      "epoch": 0.7235562358493575,
+      "grad_norm": 0.16221703588962555,
+      "learning_rate": 0.0003799081866607931,
+      "loss": 0.9827,
+      "step": 22690
+    },
+    {
+      "epoch": 0.7238751235689913,
+      "grad_norm": 0.15954597294330597,
+      "learning_rate": 0.00037831750491235344,
+      "loss": 0.9786,
+      "step": 22700
+    },
+    {
+      "epoch": 0.7241940112886253,
+      "grad_norm": 0.16421189904212952,
+      "learning_rate": 0.0003767334833742322,
+      "loss": 0.9726,
+      "step": 22710
+    },
+    {
+      "epoch": 0.7245128990082592,
+      "grad_norm": 0.1675567775964737,
+      "learning_rate": 0.0003751560941600205,
+      "loss": 0.9591,
+      "step": 22720
+    },
+    {
+      "epoch": 0.7248317867278931,
+      "grad_norm": 0.16072584688663483,
+      "learning_rate": 0.0003735853095000704,
+      "loss": 0.9649,
+      "step": 22730
+    },
+    {
+      "epoch": 0.725150674447527,
+      "grad_norm": 0.16682232916355133,
+      "learning_rate": 0.0003720211017410059,
+      "loss": 0.9567,
+      "step": 22740
+    },
+    {
+      "epoch": 0.725469562167161,
+      "grad_norm": 0.16807231307029724,
+      "learning_rate": 0.0003704634433452362,
+      "loss": 0.9624,
+      "step": 22750
+    },
+    {
+      "epoch": 0.7257884498867948,
+      "grad_norm": 0.1716778427362442,
+      "learning_rate": 0.0003689123068904708,
+      "loss": 0.969,
+      "step": 22760
+    },
+    {
+      "epoch": 0.7261073376064288,
+      "grad_norm": 0.17154444754123688,
+      "learning_rate": 0.00036736766506923683,
+      "loss": 0.9727,
+      "step": 22770
+    },
+    {
+      "epoch": 0.7264262253260627,
+      "grad_norm": 0.16088175773620605,
+      "learning_rate": 0.00036582949068839814,
+      "loss": 0.9604,
+      "step": 22780
+    },
+    {
+      "epoch": 0.7267451130456966,
+      "grad_norm": 0.16620905697345734,
+      "learning_rate": 0.0003642977566686768,
+      "loss": 0.9853,
+      "step": 22790
+    },
+    {
+      "epoch": 0.7270640007653305,
+      "grad_norm": 0.15978989005088806,
+      "learning_rate": 0.00036277243604417616,
+      "loss": 0.9629,
+      "step": 22800
+    },
+    {
+      "epoch": 0.7273828884849645,
+      "grad_norm": 0.16269829869270325,
+      "learning_rate": 0.00036125350196190614,
+      "loss": 0.975,
+      "step": 22810
+    },
+    {
+      "epoch": 0.7277017762045984,
+      "grad_norm": 0.16348029673099518,
+      "learning_rate": 0.0003597409276813109,
+      "loss": 0.9833,
+      "step": 22820
+    },
+    {
+      "epoch": 0.7280206639242323,
+      "grad_norm": 0.16305918991565704,
+      "learning_rate": 0.0003582346865737974,
+      "loss": 0.9713,
+      "step": 22830
+    },
+    {
+      "epoch": 0.7283395516438662,
+      "grad_norm": 0.16581085324287415,
+      "learning_rate": 0.0003567347521222671,
+      "loss": 0.983,
+      "step": 22840
+    },
+    {
+      "epoch": 0.7286584393635002,
+      "grad_norm": 0.1627841591835022,
+      "learning_rate": 0.00035524109792064903,
+      "loss": 0.9846,
+      "step": 22850
+    },
+    {
+      "epoch": 0.728977327083134,
+      "grad_norm": 0.16312730312347412,
+      "learning_rate": 0.00035375369767343465,
+      "loss": 0.9612,
+      "step": 22860
+    },
+    {
+      "epoch": 0.729296214802768,
+      "grad_norm": 0.16436518728733063,
+      "learning_rate": 0.0003522725251952154,
+      "loss": 0.9677,
+      "step": 22870
+    },
+    {
+      "epoch": 0.7296151025224019,
+      "grad_norm": 0.1663922518491745,
+      "learning_rate": 0.0003507975544102212,
+      "loss": 0.9509,
+      "step": 22880
+    },
+    {
+      "epoch": 0.7299339902420358,
+      "grad_norm": 0.16606448590755463,
+      "learning_rate": 0.000349328759351862,
+      "loss": 0.9618,
+      "step": 22890
+    },
+    {
+      "epoch": 0.7302528779616697,
+      "grad_norm": 0.1651606410741806,
+      "learning_rate": 0.00034786611416226987,
+      "loss": 0.9774,
+      "step": 22900
+    },
+    {
+      "epoch": 0.7305717656813037,
+      "grad_norm": 0.15751025080680847,
+      "learning_rate": 0.0003464095930918445,
+      "loss": 0.9614,
+      "step": 22910
+    },
+    {
+      "epoch": 0.7308906534009375,
+      "grad_norm": 0.16441288590431213,
+      "learning_rate": 0.0003449591704987995,
+      "loss": 0.9568,
+      "step": 22920
+    },
+    {
+      "epoch": 0.7312095411205715,
+      "grad_norm": 0.16570603847503662,
+      "learning_rate": 0.0003435148208487109,
+      "loss": 0.9729,
+      "step": 22930
+    },
+    {
+      "epoch": 0.7315284288402054,
+      "grad_norm": 0.16226355731487274,
+      "learning_rate": 0.0003420765187140679,
+      "loss": 0.9594,
+      "step": 22940
+    },
+    {
+      "epoch": 0.7318473165598393,
+      "grad_norm": 0.16443803906440735,
+      "learning_rate": 0.00034064423877382523,
+      "loss": 0.9567,
+      "step": 22950
+    },
+    {
+      "epoch": 0.7321662042794732,
+      "grad_norm": 0.16529656946659088,
+      "learning_rate": 0.000339217955812957,
+      "loss": 0.9604,
+      "step": 22960
+    },
+    {
+      "epoch": 0.7324850919991072,
+      "grad_norm": 0.160089910030365,
+      "learning_rate": 0.0003377976447220132,
+      "loss": 0.949,
+      "step": 22970
+    },
+    {
+      "epoch": 0.732803979718741,
+      "grad_norm": 0.1677403301000595,
+      "learning_rate": 0.0003363832804966775,
+      "loss": 0.958,
+      "step": 22980
+    },
+    {
+      "epoch": 0.733122867438375,
+      "grad_norm": 0.16066356003284454,
+      "learning_rate": 0.00033497483823732686,
+      "loss": 0.948,
+      "step": 22990
+    },
+    {
+      "epoch": 0.7334417551580089,
+      "grad_norm": 0.1701483428478241,
+      "learning_rate": 0.0003335722931485937,
+      "loss": 0.97,
+      "step": 23000
+    },
+    {
+      "epoch": 0.7337606428776428,
+      "grad_norm": 0.16166579723358154,
+      "learning_rate": 0.00033217562053892876,
+      "loss": 0.9516,
+      "step": 23010
+    },
+    {
+      "epoch": 0.7340795305972767,
+      "grad_norm": 0.16624568402767181,
+      "learning_rate": 0.00033078479582016675,
+      "loss": 0.9675,
+      "step": 23020
+    },
+    {
+      "epoch": 0.7343984183169107,
+      "grad_norm": 0.1629163771867752,
+      "learning_rate": 0.0003293997945070935,
+      "loss": 0.9766,
+      "step": 23030
+    },
+    {
+      "epoch": 0.7347173060365445,
+      "grad_norm": 0.16117964684963226,
+      "learning_rate": 0.00032802059221701484,
+      "loss": 0.949,
+      "step": 23040
+    },
+    {
+      "epoch": 0.7350361937561785,
+      "grad_norm": 0.1631103754043579,
+      "learning_rate": 0.00032664716466932733,
+      "loss": 0.9792,
+      "step": 23050
+    },
+    {
+      "epoch": 0.7353550814758124,
+      "grad_norm": 0.1656711995601654,
+      "learning_rate": 0.0003252794876850907,
+      "loss": 0.9605,
+      "step": 23060
+    },
+    {
+      "epoch": 0.7356739691954463,
+      "grad_norm": 0.16462354362010956,
+      "learning_rate": 0.00032391753718660234,
+      "loss": 0.955,
+      "step": 23070
+    },
+    {
+      "epoch": 0.7359928569150802,
+      "grad_norm": 0.15977884829044342,
+      "learning_rate": 0.00032256128919697346,
+      "loss": 0.9651,
+      "step": 23080
+    },
+    {
+      "epoch": 0.7363117446347142,
+      "grad_norm": 0.16425421833992004,
+      "learning_rate": 0.00032121071983970694,
+      "loss": 0.9403,
+      "step": 23090
+    },
+    {
+      "epoch": 0.736630632354348,
+      "grad_norm": 0.16412289440631866,
+      "learning_rate": 0.0003198658053382767,
+      "loss": 0.9581,
+      "step": 23100
+    },
+    {
+      "epoch": 0.736949520073982,
+      "grad_norm": 0.16336853802204132,
+      "learning_rate": 0.0003185265220157095,
+      "loss": 0.9439,
+      "step": 23110
+    },
+    {
+      "epoch": 0.7372684077936159,
+      "grad_norm": 0.1620175540447235,
+      "learning_rate": 0.00031719284629416804,
+      "loss": 0.9593,
+      "step": 23120
+    },
+    {
+      "epoch": 0.7375872955132498,
+      "grad_norm": 0.16276094317436218,
+      "learning_rate": 0.0003158647546945357,
+      "loss": 0.9502,
+      "step": 23130
+    },
+    {
+      "epoch": 0.7379061832328837,
+      "grad_norm": 0.17003417015075684,
+      "learning_rate": 0.0003145422238360034,
+      "loss": 0.9791,
+      "step": 23140
+    },
+    {
+      "epoch": 0.7382250709525177,
+      "grad_norm": 0.16041214764118195,
+      "learning_rate": 0.0003132252304356578,
+      "loss": 0.9757,
+      "step": 23150
+    },
+    {
+      "epoch": 0.7385439586721515,
+      "grad_norm": 0.16528736054897308,
+      "learning_rate": 0.00031191375130807145,
+      "loss": 0.9597,
+      "step": 23160
+    },
+    {
+      "epoch": 0.7388628463917855,
+      "grad_norm": 0.16135884821414948,
+      "learning_rate": 0.0003106077633648948,
+      "loss": 0.9646,
+      "step": 23170
+    },
+    {
+      "epoch": 0.7391817341114194,
+      "grad_norm": 0.1672159880399704,
+      "learning_rate": 0.0003093072436144496,
+      "loss": 0.9544,
+      "step": 23180
+    },
+    {
+      "epoch": 0.7395006218310533,
+      "grad_norm": 0.1631564199924469,
+      "learning_rate": 0.00030801216916132403,
+      "loss": 0.9518,
+      "step": 23190
+    },
+    {
+      "epoch": 0.7398195095506872,
+      "grad_norm": 0.15956902503967285,
+      "learning_rate": 0.00030672251720596967,
+      "loss": 0.9487,
+      "step": 23200
+    },
+    {
+      "epoch": 0.7401383972703212,
+      "grad_norm": 0.16405710577964783,
+      "learning_rate": 0.0003054382650443004,
+      "loss": 0.962,
+      "step": 23210
+    },
+    {
+      "epoch": 0.740457284989955,
+      "grad_norm": 0.16590505838394165,
+      "learning_rate": 0.0003041593900672922,
+      "loss": 0.9454,
+      "step": 23220
+    },
+    {
+      "epoch": 0.740776172709589,
+      "grad_norm": 0.16362358629703522,
+      "learning_rate": 0.00030288586976058574,
+      "loss": 0.9398,
+      "step": 23230
+    },
+    {
+      "epoch": 0.7410950604292229,
+      "grad_norm": 0.17358466982841492,
+      "learning_rate": 0.00030161768170408935,
+      "loss": 0.9452,
+      "step": 23240
+    },
+    {
+      "epoch": 0.7414139481488567,
+      "grad_norm": 0.1667748987674713,
+      "learning_rate": 0.0003003548035715848,
+      "loss": 0.9467,
+      "step": 23250
+    },
+    {
+      "epoch": 0.7417328358684907,
+      "grad_norm": 0.16409194469451904,
+      "learning_rate": 0.0002990972131303341,
+      "loss": 0.9482,
+      "step": 23260
+    },
+    {
+      "epoch": 0.7420517235881247,
+      "grad_norm": 0.16616195440292358,
+      "learning_rate": 0.0002978448882406881,
+      "loss": 0.9478,
+      "step": 23270
+    },
+    {
+      "epoch": 0.7423706113077585,
+      "grad_norm": 0.16837824881076813,
+      "learning_rate": 0.00029659780685569674,
+      "loss": 0.9413,
+      "step": 23280
+    },
+    {
+      "epoch": 0.7426894990273925,
+      "grad_norm": 0.16764934360980988,
+      "learning_rate": 0.00029535594702072087,
+      "loss": 0.942,
+      "step": 23290
+    },
+    {
+      "epoch": 0.7430083867470264,
+      "grad_norm": 0.1687283217906952,
+      "learning_rate": 0.0002941192868730457,
+      "loss": 0.9677,
+      "step": 23300
+    },
+    {
+      "epoch": 0.7433272744666602,
+      "grad_norm": 0.16668742895126343,
+      "learning_rate": 0.00029288780464149593,
+      "loss": 0.9544,
+      "step": 23310
+    },
+    {
+      "epoch": 0.7436461621862942,
+      "grad_norm": 0.16436542570590973,
+      "learning_rate": 0.00029166147864605294,
+      "loss": 0.9604,
+      "step": 23320
+    },
+    {
+      "epoch": 0.7439650499059282,
+      "grad_norm": 0.16353577375411987,
+      "learning_rate": 0.0002904402872974721,
+      "loss": 0.9582,
+      "step": 23330
+    },
+    {
+      "epoch": 0.744283937625562,
+      "grad_norm": 0.16399306058883667,
+      "learning_rate": 0.00028922420909690367,
+      "loss": 0.9537,
+      "step": 23340
+    },
+    {
+      "epoch": 0.744602825345196,
+      "grad_norm": 0.16380132734775543,
+      "learning_rate": 0.00028801322263551397,
+      "loss": 0.9527,
+      "step": 23350
+    },
+    {
+      "epoch": 0.7449217130648299,
+      "grad_norm": 0.16486787796020508,
+      "learning_rate": 0.0002868073065941083,
+      "loss": 0.9344,
+      "step": 23360
+    },
+    {
+      "epoch": 0.7452406007844637,
+      "grad_norm": 0.1594451665878296,
+      "learning_rate": 0.00028560643974275587,
+      "loss": 0.9459,
+      "step": 23370
+    },
+    {
+      "epoch": 0.7455594885040977,
+      "grad_norm": 0.16545240581035614,
+      "learning_rate": 0.00028441060094041583,
+      "loss": 0.9464,
+      "step": 23380
+    },
+    {
+      "epoch": 0.7458783762237317,
+      "grad_norm": 0.170954629778862,
+      "learning_rate": 0.0002832197691345653,
+      "loss": 0.9508,
+      "step": 23390
+    },
+    {
+      "epoch": 0.7461972639433655,
+      "grad_norm": 0.1691250205039978,
+      "learning_rate": 0.0002820339233608287,
+      "loss": 0.9528,
+      "step": 23400
+    },
+    {
+      "epoch": 0.7465161516629994,
+      "grad_norm": 0.1623115986585617,
+      "learning_rate": 0.00028085304274260857,
+      "loss": 0.9598,
+      "step": 23410
+    },
+    {
+      "epoch": 0.7468350393826334,
+      "grad_norm": 0.1692652702331543,
+      "learning_rate": 0.0002796771064907181,
+      "loss": 0.952,
+      "step": 23420
+    },
+    {
+      "epoch": 0.7471539271022672,
+      "grad_norm": 0.16102375090122223,
+      "learning_rate": 0.0002785060939030151,
+      "loss": 0.9432,
+      "step": 23430
+    },
+    {
+      "epoch": 0.7474728148219012,
+      "grad_norm": 0.1656680405139923,
+      "learning_rate": 0.0002773399843640378,
+      "loss": 0.9331,
+      "step": 23440
+    },
+    {
+      "epoch": 0.7477917025415352,
+      "grad_norm": 0.16000087559223175,
+      "learning_rate": 0.00027617875734464145,
+      "loss": 0.9358,
+      "step": 23450
+    },
+    {
+      "epoch": 0.748110590261169,
+      "grad_norm": 0.17037495970726013,
+      "learning_rate": 0.00027502239240163715,
+      "loss": 0.9616,
+      "step": 23460
+    },
+    {
+      "epoch": 0.748429477980803,
+      "grad_norm": 0.16514143347740173,
+      "learning_rate": 0.00027387086917743224,
+      "loss": 0.9491,
+      "step": 23470
+    },
+    {
+      "epoch": 0.7487483657004369,
+      "grad_norm": 0.1671529859304428,
+      "learning_rate": 0.0002727241673996714,
+      "loss": 0.9391,
+      "step": 23480
+    },
+    {
+      "epoch": 0.7490672534200707,
+      "grad_norm": 0.16395649313926697,
+      "learning_rate": 0.00027158226688088006,
+      "loss": 0.9431,
+      "step": 23490
+    },
+    {
+      "epoch": 0.7493861411397047,
+      "grad_norm": 0.17046767473220825,
+      "learning_rate": 0.0002704451475181089,
+      "loss": 0.9499,
+      "step": 23500
+    },
+    {
+      "epoch": 0.7497050288593387,
+      "grad_norm": 0.15811336040496826,
+      "learning_rate": 0.00026931278929257993,
+      "loss": 0.9359,
+      "step": 23510
+    },
+    {
+      "epoch": 0.7500239165789725,
+      "grad_norm": 0.16410349309444427,
+      "learning_rate": 0.00026818517226933437,
+      "loss": 0.9351,
+      "step": 23520
+    },
+    {
+      "epoch": 0.7503428042986064,
+      "grad_norm": 0.16768516600131989,
+      "learning_rate": 0.00026706227659688107,
+      "loss": 0.9462,
+      "step": 23530
+    },
+    {
+      "epoch": 0.7506616920182404,
+      "grad_norm": 0.16026252508163452,
+      "learning_rate": 0.00026594408250684776,
+      "loss": 0.9439,
+      "step": 23540
+    },
+    {
+      "epoch": 0.7509805797378742,
+      "grad_norm": 0.16729193925857544,
+      "learning_rate": 0.00026483057031363234,
+      "loss": 0.9493,
+      "step": 23550
+    },
+    {
+      "epoch": 0.7512994674575082,
+      "grad_norm": 0.16243338584899902,
+      "learning_rate": 0.00026372172041405677,
+      "loss": 0.9404,
+      "step": 23560
+    },
+    {
+      "epoch": 0.7516183551771422,
+      "grad_norm": 0.169617161154747,
+      "learning_rate": 0.0002626175132870219,
+      "loss": 0.9504,
+      "step": 23570
+    },
+    {
+      "epoch": 0.751937242896776,
+      "grad_norm": 0.1605086624622345,
+      "learning_rate": 0.0002615179294931637,
+      "loss": 0.9399,
+      "step": 23580
+    },
+    {
+      "epoch": 0.75225613061641,
+      "grad_norm": 0.16321012377738953,
+      "learning_rate": 0.000260422949674511,
+      "loss": 0.9362,
+      "step": 23590
+    },
+    {
+      "epoch": 0.7525750183360439,
+      "grad_norm": 0.16131918132305145,
+      "learning_rate": 0.00025933255455414493,
+      "loss": 0.9342,
+      "step": 23600
+    },
+    {
+      "epoch": 0.7528939060556777,
+      "grad_norm": 0.16513609886169434,
+      "learning_rate": 0.0002582467249358593,
+      "loss": 0.95,
+      "step": 23610
+    },
+    {
+      "epoch": 0.7532127937753117,
+      "grad_norm": 0.1651507318019867,
+      "learning_rate": 0.0002571654417038226,
+      "loss": 0.9507,
+      "step": 23620
+    },
+    {
+      "epoch": 0.7535316814949456,
+      "grad_norm": 0.1660146713256836,
+      "learning_rate": 0.0002560886858222419,
+      "loss": 0.9526,
+      "step": 23630
+    },
+    {
+      "epoch": 0.7538505692145795,
+      "grad_norm": 0.15882445871829987,
+      "learning_rate": 0.0002550164383350272,
+      "loss": 0.9243,
+      "step": 23640
+    },
+    {
+      "epoch": 0.7541694569342134,
+      "grad_norm": 0.17136244475841522,
+      "learning_rate": 0.0002539486803654581,
+      "loss": 0.9448,
+      "step": 23650
+    },
+    {
+      "epoch": 0.7544883446538474,
+      "grad_norm": 0.1678006798028946,
+      "learning_rate": 0.0002528853931158513,
+      "loss": 0.9429,
+      "step": 23660
+    },
+    {
+      "epoch": 0.7548072323734814,
+      "grad_norm": 0.16127173602581024,
+      "learning_rate": 0.0002518265578672297,
+      "loss": 0.9366,
+      "step": 23670
+    },
+    {
+      "epoch": 0.7551261200931152,
+      "grad_norm": 0.16299548745155334,
+      "learning_rate": 0.00025077215597899276,
+      "loss": 0.9378,
+      "step": 23680
+    },
+    {
+      "epoch": 0.7554450078127491,
+      "grad_norm": 0.1708274483680725,
+      "learning_rate": 0.0002497221688885888,
+      "loss": 0.9601,
+      "step": 23690
+    },
+    {
+      "epoch": 0.7557638955323831,
+      "grad_norm": 0.16434860229492188,
+      "learning_rate": 0.0002486765781111873,
+      "loss": 0.9401,
+      "step": 23700
+    },
+    {
+      "epoch": 0.7560827832520169,
+      "grad_norm": 0.1635703295469284,
+      "learning_rate": 0.0002476353652393545,
+      "loss": 0.9476,
+      "step": 23710
+    },
+    {
+      "epoch": 0.7564016709716509,
+      "grad_norm": 0.17095215618610382,
+      "learning_rate": 0.0002465985119427286,
+      "loss": 0.9291,
+      "step": 23720
+    },
+    {
+      "epoch": 0.7567205586912849,
+      "grad_norm": 0.1732560396194458,
+      "learning_rate": 0.0002455659999676974,
+      "loss": 0.9374,
+      "step": 23730
+    },
+    {
+      "epoch": 0.7570394464109187,
+      "grad_norm": 0.16333815455436707,
+      "learning_rate": 0.0002445378111370768,
+      "loss": 0.9406,
+      "step": 23740
+    },
+    {
+      "epoch": 0.7573583341305526,
+      "grad_norm": 0.16199623048305511,
+      "learning_rate": 0.00024351392734979106,
+      "loss": 0.934,
+      "step": 23750
+    },
+    {
+      "epoch": 0.7576772218501866,
+      "grad_norm": 0.16906283795833588,
+      "learning_rate": 0.00024249433058055368,
+      "loss": 0.9252,
+      "step": 23760
+    },
+    {
+      "epoch": 0.7579961095698204,
+      "grad_norm": 0.17197982966899872,
+      "learning_rate": 0.00024147900287955056,
+      "loss": 0.9489,
+      "step": 23770
+    },
+    {
+      "epoch": 0.7583149972894544,
+      "grad_norm": 0.16192105412483215,
+      "learning_rate": 0.0002404679263721236,
+      "loss": 0.9261,
+      "step": 23780
+    },
+    {
+      "epoch": 0.7586338850090883,
+      "grad_norm": 0.16491258144378662,
+      "learning_rate": 0.00023946108325845628,
+      "loss": 0.9431,
+      "step": 23790
+    },
+    {
+      "epoch": 0.7589527727287222,
+      "grad_norm": 0.16750048100948334,
+      "learning_rate": 0.00023845845581326017,
+      "loss": 0.9591,
+      "step": 23800
+    },
+    {
+      "epoch": 0.7592716604483561,
+      "grad_norm": 0.17201946675777435,
+      "learning_rate": 0.00023746002638546287,
+      "loss": 0.926,
+      "step": 23810
+    },
+    {
+      "epoch": 0.7595905481679901,
+      "grad_norm": 0.16682586073875427,
+      "learning_rate": 0.00023646577739789735,
+      "loss": 0.9395,
+      "step": 23820
+    },
+    {
+      "epoch": 0.7599094358876239,
+      "grad_norm": 0.16354051232337952,
+      "learning_rate": 0.00023547569134699248,
+      "loss": 0.9419,
+      "step": 23830
+    },
+    {
+      "epoch": 0.7602283236072579,
+      "grad_norm": 0.16591399908065796,
+      "learning_rate": 0.00023448975080246477,
+      "loss": 0.938,
+      "step": 23840
+    },
+    {
+      "epoch": 0.7605472113268918,
+      "grad_norm": 0.16553214192390442,
+      "learning_rate": 0.00023350793840701175,
+      "loss": 0.9359,
+      "step": 23850
+    },
+    {
+      "epoch": 0.7608660990465257,
+      "grad_norm": 0.16791701316833496,
+      "learning_rate": 0.00023253023687600625,
+      "loss": 0.928,
+      "step": 23860
+    },
+    {
+      "epoch": 0.7611849867661596,
+      "grad_norm": 0.16159778833389282,
+      "learning_rate": 0.00023155662899719193,
+      "loss": 0.9554,
+      "step": 23870
+    },
+    {
+      "epoch": 0.7615038744857936,
+      "grad_norm": 0.1624593585729599,
+      "learning_rate": 0.00023058709763038065,
+      "loss": 0.9385,
+      "step": 23880
+    },
+    {
+      "epoch": 0.7618227622054274,
+      "grad_norm": 0.16371050477027893,
+      "learning_rate": 0.00022962162570715043,
+      "loss": 0.943,
+      "step": 23890
+    },
+    {
+      "epoch": 0.7621416499250614,
+      "grad_norm": 0.17248646914958954,
+      "learning_rate": 0.00022866019623054502,
+      "loss": 0.9253,
+      "step": 23900
+    },
+    {
+      "epoch": 0.7624605376446953,
+      "grad_norm": 0.16547873616218567,
+      "learning_rate": 0.00022770279227477477,
+      "loss": 0.9403,
+      "step": 23910
+    },
+    {
+      "epoch": 0.7627794253643292,
+      "grad_norm": 0.16172504425048828,
+      "learning_rate": 0.00022674939698491854,
+      "loss": 0.9285,
+      "step": 23920
+    },
+    {
+      "epoch": 0.7630983130839631,
+      "grad_norm": 0.16042064130306244,
+      "learning_rate": 0.000225799993576627,
+      "loss": 0.9295,
+      "step": 23930
+    },
+    {
+      "epoch": 0.7634172008035971,
+      "grad_norm": 0.17115207016468048,
+      "learning_rate": 0.0002248545653358273,
+      "loss": 0.943,
+      "step": 23940
+    },
+    {
+      "epoch": 0.7637360885232309,
+      "grad_norm": 0.1645117849111557,
+      "learning_rate": 0.00022391309561842852,
+      "loss": 0.9374,
+      "step": 23950
+    },
+    {
+      "epoch": 0.7640549762428649,
+      "grad_norm": 0.16595900058746338,
+      "learning_rate": 0.00022297556785002887,
+      "loss": 0.936,
+      "step": 23960
+    },
+    {
+      "epoch": 0.7643738639624988,
+      "grad_norm": 0.16360650956630707,
+      "learning_rate": 0.0002220419655256239,
+      "loss": 0.9239,
+      "step": 23970
+    },
+    {
+      "epoch": 0.7646927516821327,
+      "grad_norm": 0.1633913815021515,
+      "learning_rate": 0.00022111227220931578,
+      "loss": 0.9372,
+      "step": 23980
+    },
+    {
+      "epoch": 0.7650116394017666,
+      "grad_norm": 0.16824324429035187,
+      "learning_rate": 0.00022018647153402418,
+      "loss": 0.9355,
+      "step": 23990
+    },
+    {
+      "epoch": 0.7653305271214006,
+      "grad_norm": 0.16444505751132965,
+      "learning_rate": 0.00021926454720119792,
+      "loss": 0.9482,
+      "step": 24000
+    },
+    {
+      "epoch": 0.7656494148410344,
+      "grad_norm": 0.16452354192733765,
+      "learning_rate": 0.00021834648298052806,
+      "loss": 0.9165,
+      "step": 24010
+    },
+    {
+      "epoch": 0.7659683025606684,
+      "grad_norm": 0.16021281480789185,
+      "learning_rate": 0.00021743226270966244,
+      "loss": 0.9326,
+      "step": 24020
+    },
+    {
+      "epoch": 0.7662871902803023,
+      "grad_norm": 0.17205902934074402,
+      "learning_rate": 0.00021652187029392068,
+      "loss": 0.9492,
+      "step": 24030
+    },
+    {
+      "epoch": 0.7666060779999362,
+      "grad_norm": 0.16032502055168152,
+      "learning_rate": 0.00021561528970601124,
+      "loss": 0.939,
+      "step": 24040
+    },
+    {
+      "epoch": 0.7669249657195701,
+      "grad_norm": 0.1687769591808319,
+      "learning_rate": 0.00021471250498574909,
+      "loss": 0.936,
+      "step": 24050
+    },
+    {
+      "epoch": 0.7672438534392041,
+      "grad_norm": 0.163095161318779,
+      "learning_rate": 0.0002138135002397747,
+      "loss": 0.9395,
+      "step": 24060
+    },
+    {
+      "epoch": 0.7675627411588379,
+      "grad_norm": 0.16366122663021088,
+      "learning_rate": 0.0002129182596412743,
+      "loss": 0.9528,
+      "step": 24070
+    },
+    {
+      "epoch": 0.7678816288784719,
+      "grad_norm": 0.16610878705978394,
+      "learning_rate": 0.00021202676742970136,
+      "loss": 0.9416,
+      "step": 24080
+    },
+    {
+      "epoch": 0.7682005165981058,
+      "grad_norm": 0.16417580842971802,
+      "learning_rate": 0.0002111390079104989,
+      "loss": 0.9317,
+      "step": 24090
+    },
+    {
+      "epoch": 0.7685194043177397,
+      "grad_norm": 0.16696830093860626,
+      "learning_rate": 0.00021025496545482323,
+      "loss": 0.9359,
+      "step": 24100
+    },
+    {
+      "epoch": 0.7688382920373736,
+      "grad_norm": 0.1716535985469818,
+      "learning_rate": 0.0002093746244992691,
+      "loss": 0.9396,
+      "step": 24110
+    },
+    {
+      "epoch": 0.7691571797570076,
+      "grad_norm": 0.1632964015007019,
+      "learning_rate": 0.0002084979695455954,
+      "loss": 0.9292,
+      "step": 24120
+    },
+    {
+      "epoch": 0.7694760674766414,
+      "grad_norm": 0.16584451496601105,
+      "learning_rate": 0.00020762498516045232,
+      "loss": 0.9326,
+      "step": 24130
+    },
+    {
+      "epoch": 0.7697949551962754,
+      "grad_norm": 0.16385923326015472,
+      "learning_rate": 0.00020675565597510982,
+      "loss": 0.9329,
+      "step": 24140
+    },
+    {
+      "epoch": 0.7701138429159093,
+      "grad_norm": 0.1649266630411148,
+      "learning_rate": 0.00020588996668518696,
+      "loss": 0.9371,
+      "step": 24150
+    },
+    {
+      "epoch": 0.7704327306355432,
+      "grad_norm": 0.16053041815757751,
+      "learning_rate": 0.00020502790205038248,
+      "loss": 0.92,
+      "step": 24160
+    },
+    {
+      "epoch": 0.7707516183551771,
+      "grad_norm": 0.16492949426174164,
+      "learning_rate": 0.00020416944689420654,
+      "loss": 0.9235,
+      "step": 24170
+    },
+    {
+      "epoch": 0.7710705060748111,
+      "grad_norm": 0.16709771752357483,
+      "learning_rate": 0.00020331458610371345,
+      "loss": 0.9385,
+      "step": 24180
+    },
+    {
+      "epoch": 0.7713893937944449,
+      "grad_norm": 0.16922813653945923,
+      "learning_rate": 0.00020246330462923582,
+      "loss": 0.9301,
+      "step": 24190
+    },
+    {
+      "epoch": 0.7717082815140789,
+      "grad_norm": 0.1640113890171051,
+      "learning_rate": 0.00020161558748411925,
+      "loss": 0.9211,
+      "step": 24200
+    },
+    {
+      "epoch": 0.7720271692337128,
+      "grad_norm": 0.16404199600219727,
+      "learning_rate": 0.00020077141974445887,
+      "loss": 0.9278,
+      "step": 24210
+    },
+    {
+      "epoch": 0.7723460569533467,
+      "grad_norm": 0.16532927751541138,
+      "learning_rate": 0.00019993078654883636,
+      "loss": 0.9219,
+      "step": 24220
+    },
+    {
+      "epoch": 0.7726649446729806,
+      "grad_norm": 0.1662260890007019,
+      "learning_rate": 0.00019909367309805842,
+      "loss": 0.9054,
+      "step": 24230
+    },
+    {
+      "epoch": 0.7729838323926146,
+      "grad_norm": 0.16583295166492462,
+      "learning_rate": 0.0001982600646548962,
+      "loss": 0.9262,
+      "step": 24240
+    },
+    {
+      "epoch": 0.7733027201122484,
+      "grad_norm": 0.16515842080116272,
+      "learning_rate": 0.0001974299465438259,
+      "loss": 0.9179,
+      "step": 24250
+    },
+    {
+      "epoch": 0.7736216078318824,
+      "grad_norm": 0.17167966067790985,
+      "learning_rate": 0.00019660330415077035,
+      "loss": 0.9172,
+      "step": 24260
+    },
+    {
+      "epoch": 0.7739404955515163,
+      "grad_norm": 0.1663963943719864,
+      "learning_rate": 0.00019578012292284172,
+      "loss": 0.933,
+      "step": 24270
+    },
+    {
+      "epoch": 0.7742593832711502,
+      "grad_norm": 0.16700772941112518,
+      "learning_rate": 0.0001949603883680855,
+      "loss": 0.9381,
+      "step": 24280
+    },
+    {
+      "epoch": 0.7745782709907841,
+      "grad_norm": 0.1680431216955185,
+      "learning_rate": 0.0001941440860552251,
+      "loss": 0.9357,
+      "step": 24290
+    },
+    {
+      "epoch": 0.7748971587104181,
+      "grad_norm": 0.16372446715831757,
+      "learning_rate": 0.00019333120161340792,
+      "loss": 0.9299,
+      "step": 24300
+    },
+    {
+      "epoch": 0.7752160464300519,
+      "grad_norm": 0.16472508013248444,
+      "learning_rate": 0.00019252172073195239,
+      "loss": 0.931,
+      "step": 24310
+    },
+    {
+      "epoch": 0.7755349341496859,
+      "grad_norm": 0.17051909863948822,
+      "learning_rate": 0.00019171562916009603,
+      "loss": 0.9378,
+      "step": 24320
+    },
+    {
+      "epoch": 0.7758538218693198,
+      "grad_norm": 0.16289319097995758,
+      "learning_rate": 0.00019091291270674447,
+      "loss": 0.9212,
+      "step": 24330
+    },
+    {
+      "epoch": 0.7761727095889537,
+      "grad_norm": 0.1658150851726532,
+      "learning_rate": 0.00019011355724022166,
+      "loss": 0.9321,
+      "step": 24340
+    },
+    {
+      "epoch": 0.7764915973085876,
+      "grad_norm": 0.16301952302455902,
+      "learning_rate": 0.0001893175486880212,
+      "loss": 0.9341,
+      "step": 24350
+    },
+    {
+      "epoch": 0.7768104850282216,
+      "grad_norm": 0.1663392335176468,
+      "learning_rate": 0.0001885248730365585,
+      "loss": 0.9267,
+      "step": 24360
+    },
+    {
+      "epoch": 0.7771293727478554,
+      "grad_norm": 0.16546833515167236,
+      "learning_rate": 0.00018773551633092397,
+      "loss": 0.9157,
+      "step": 24370
+    },
+    {
+      "epoch": 0.7774482604674894,
+      "grad_norm": 0.16583992540836334,
+      "learning_rate": 0.00018694946467463756,
+      "loss": 0.9084,
+      "step": 24380
+    },
+    {
+      "epoch": 0.7777671481871233,
+      "grad_norm": 0.1641332507133484,
+      "learning_rate": 0.00018616670422940394,
+      "loss": 0.9277,
+      "step": 24390
+    },
+    {
+      "epoch": 0.7780860359067572,
+      "grad_norm": 0.171635240316391,
+      "learning_rate": 0.00018538722121486895,
+      "loss": 0.929,
+      "step": 24400
+    },
+    {
+      "epoch": 0.7784049236263911,
+      "grad_norm": 0.16188062727451324,
+      "learning_rate": 0.00018461100190837707,
+      "loss": 0.9308,
+      "step": 24410
+    },
+    {
+      "epoch": 0.7787238113460251,
+      "grad_norm": 0.16226285696029663,
+      "learning_rate": 0.0001838380326447297,
+      "loss": 0.9362,
+      "step": 24420
+    },
+    {
+      "epoch": 0.7790426990656589,
+      "grad_norm": 0.16537430882453918,
+      "learning_rate": 0.00018306829981594458,
+      "loss": 0.9271,
+      "step": 24430
+    },
+    {
+      "epoch": 0.7793615867852929,
+      "grad_norm": 0.163283571600914,
+      "learning_rate": 0.0001823017898710165,
+      "loss": 0.9274,
+      "step": 24440
+    },
+    {
+      "epoch": 0.7796804745049268,
+      "grad_norm": 0.161969393491745,
+      "learning_rate": 0.00018153848931567836,
+      "loss": 0.9175,
+      "step": 24450
+    },
+    {
+      "epoch": 0.7799993622245607,
+      "grad_norm": 0.16395604610443115,
+      "learning_rate": 0.00018077838471216377,
+      "loss": 0.928,
+      "step": 24460
+    },
+    {
+      "epoch": 0.7803182499441946,
+      "grad_norm": 0.1620090901851654,
+      "learning_rate": 0.00018002146267897054,
+      "loss": 0.92,
+      "step": 24470
+    },
+    {
+      "epoch": 0.7806371376638286,
+      "grad_norm": 0.16308417916297913,
+      "learning_rate": 0.00017926770989062511,
+      "loss": 0.9098,
+      "step": 24480
+    },
+    {
+      "epoch": 0.7809560253834625,
+      "grad_norm": 0.1651722639799118,
+      "learning_rate": 0.0001785171130774477,
+      "loss": 0.9244,
+      "step": 24490
+    },
+    {
+      "epoch": 0.7812749131030964,
+      "grad_norm": 0.1706753969192505,
+      "learning_rate": 0.00017776965902531916,
+      "loss": 0.9302,
+      "step": 24500
+    },
+    {
+      "epoch": 0.7815938008227303,
+      "grad_norm": 0.16880400478839874,
+      "learning_rate": 0.00017702533457544784,
+      "loss": 0.9094,
+      "step": 24510
+    },
+    {
+      "epoch": 0.7819126885423643,
+      "grad_norm": 0.16843248903751373,
+      "learning_rate": 0.00017628412662413823,
+      "loss": 0.9308,
+      "step": 24520
+    },
+    {
+      "epoch": 0.7822315762619981,
+      "grad_norm": 0.16641728579998016,
+      "learning_rate": 0.0001755460221225603,
+      "loss": 0.9351,
+      "step": 24530
+    },
+    {
+      "epoch": 0.7825504639816321,
+      "grad_norm": 0.16158147156238556,
+      "learning_rate": 0.00017481100807651963,
+      "loss": 0.9155,
+      "step": 24540
+    },
+    {
+      "epoch": 0.782869351701266,
+      "grad_norm": 0.16641920804977417,
+      "learning_rate": 0.00017407907154622863,
+      "loss": 0.9249,
+      "step": 24550
+    },
+    {
+      "epoch": 0.7831882394208999,
+      "grad_norm": 0.16956974565982819,
+      "learning_rate": 0.00017335019964607887,
+      "loss": 0.9187,
+      "step": 24560
+    },
+    {
+      "epoch": 0.7835071271405338,
+      "grad_norm": 0.16971714794635773,
+      "learning_rate": 0.00017262437954441417,
+      "loss": 0.9259,
+      "step": 24570
+    },
+    {
+      "epoch": 0.7838260148601678,
+      "grad_norm": 0.16633319854736328,
+      "learning_rate": 0.00017190159846330476,
+      "loss": 0.9217,
+      "step": 24580
+    },
+    {
+      "epoch": 0.7841449025798016,
+      "grad_norm": 0.16065815091133118,
+      "learning_rate": 0.00017118184367832215,
+      "loss": 0.927,
+      "step": 24590
+    },
+    {
+      "epoch": 0.7844637902994356,
+      "grad_norm": 0.1626022458076477,
+      "learning_rate": 0.00017046510251831525,
+      "loss": 0.9235,
+      "step": 24600
+    },
+    {
+      "epoch": 0.7847826780190695,
+      "grad_norm": 0.1678556501865387,
+      "learning_rate": 0.0001697513623651875,
+      "loss": 0.9168,
+      "step": 24610
+    },
+    {
+      "epoch": 0.7851015657387034,
+      "grad_norm": 0.16368116438388824,
+      "learning_rate": 0.00016904061065367424,
+      "loss": 0.907,
+      "step": 24620
+    },
+    {
+      "epoch": 0.7854204534583373,
+      "grad_norm": 0.16854064166545868,
+      "learning_rate": 0.00016833283487112187,
+      "loss": 0.9207,
+      "step": 24630
+    },
+    {
+      "epoch": 0.7857393411779713,
+      "grad_norm": 0.16941261291503906,
+      "learning_rate": 0.00016762802255726757,
+      "loss": 0.9172,
+      "step": 24640
+    },
+    {
+      "epoch": 0.7860582288976051,
+      "grad_norm": 0.1684810221195221,
+      "learning_rate": 0.0001669261613040197,
+      "loss": 0.9164,
+      "step": 24650
+    },
+    {
+      "epoch": 0.7863771166172391,
+      "grad_norm": 0.16645093262195587,
+      "learning_rate": 0.00016622723875523958,
+      "loss": 0.9237,
+      "step": 24660
+    },
+    {
+      "epoch": 0.786696004336873,
+      "grad_norm": 0.16230951249599457,
+      "learning_rate": 0.0001655312426065239,
+      "loss": 0.9133,
+      "step": 24670
+    },
+    {
+      "epoch": 0.7870148920565069,
+      "grad_norm": 0.1660119891166687,
+      "learning_rate": 0.00016483816060498802,
+      "loss": 0.9321,
+      "step": 24680
+    },
+    {
+      "epoch": 0.7873337797761408,
+      "grad_norm": 0.16883163154125214,
+      "learning_rate": 0.00016414798054905036,
+      "loss": 0.9268,
+      "step": 24690
+    },
+    {
+      "epoch": 0.7876526674957748,
+      "grad_norm": 0.16361361742019653,
+      "learning_rate": 0.0001634606902882176,
+      "loss": 0.9082,
+      "step": 24700
+    },
+    {
+      "epoch": 0.7879715552154086,
+      "grad_norm": 0.16677947342395782,
+      "learning_rate": 0.00016277627772287072,
+      "loss": 0.926,
+      "step": 24710
+    },
+    {
+      "epoch": 0.7882904429350426,
+      "grad_norm": 0.16485854983329773,
+      "learning_rate": 0.00016209473080405187,
+      "loss": 0.9066,
+      "step": 24720
+    },
+    {
+      "epoch": 0.7886093306546765,
+      "grad_norm": 0.16697336733341217,
+      "learning_rate": 0.0001614160375332526,
+      "loss": 0.9246,
+      "step": 24730
+    },
+    {
+      "epoch": 0.7889282183743104,
+      "grad_norm": 0.17086774110794067,
+      "learning_rate": 0.00016074018596220224,
+      "loss": 0.9235,
+      "step": 24740
+    },
+    {
+      "epoch": 0.7892471060939443,
+      "grad_norm": 0.1602909117937088,
+      "learning_rate": 0.00016006716419265783,
+      "loss": 0.9371,
+      "step": 24750
+    },
+    {
+      "epoch": 0.7895659938135783,
+      "grad_norm": 0.16252830624580383,
+      "learning_rate": 0.00015939696037619444,
+      "loss": 0.9139,
+      "step": 24760
+    },
+    {
+      "epoch": 0.7898848815332121,
+      "grad_norm": 0.1664387583732605,
+      "learning_rate": 0.00015872956271399674,
+      "loss": 0.9048,
+      "step": 24770
+    },
+    {
+      "epoch": 0.7902037692528461,
+      "grad_norm": 0.162310391664505,
+      "learning_rate": 0.00015806495945665133,
+      "loss": 0.9271,
+      "step": 24780
+    },
+    {
+      "epoch": 0.79052265697248,
+      "grad_norm": 0.16639670729637146,
+      "learning_rate": 0.00015740313890393964,
+      "loss": 0.9401,
+      "step": 24790
+    },
+    {
+      "epoch": 0.7908415446921139,
+      "grad_norm": 0.16429823637008667,
+      "learning_rate": 0.00015674408940463216,
+      "loss": 0.9052,
+      "step": 24800
+    },
+    {
+      "epoch": 0.7911604324117478,
+      "grad_norm": 0.16192969679832458,
+      "learning_rate": 0.00015608779935628333,
+      "loss": 0.9241,
+      "step": 24810
+    },
+    {
+      "epoch": 0.7914793201313818,
+      "grad_norm": 0.17283402383327484,
+      "learning_rate": 0.0001554342572050271,
+      "loss": 0.9164,
+      "step": 24820
+    },
+    {
+      "epoch": 0.7917982078510156,
+      "grad_norm": 0.17254789173603058,
+      "learning_rate": 0.00015478345144537376,
+      "loss": 0.9205,
+      "step": 24830
+    },
+    {
+      "epoch": 0.7921170955706496,
+      "grad_norm": 0.16515520215034485,
+      "learning_rate": 0.0001541353706200072,
+      "loss": 0.9109,
+      "step": 24840
+    },
+    {
+      "epoch": 0.7924359832902835,
+      "grad_norm": 0.1679113805294037,
+      "learning_rate": 0.0001534900033195833,
+      "loss": 0.9232,
+      "step": 24850
+    },
+    {
+      "epoch": 0.7927548710099174,
+      "grad_norm": 0.17153725028038025,
+      "learning_rate": 0.00015284733818252897,
+      "loss": 0.9125,
+      "step": 24860
+    },
+    {
+      "epoch": 0.7930737587295513,
+      "grad_norm": 0.16140176355838776,
+      "learning_rate": 0.00015220736389484244,
+      "loss": 0.9065,
+      "step": 24870
+    },
+    {
+      "epoch": 0.7933926464491853,
+      "grad_norm": 0.16564328968524933,
+      "learning_rate": 0.00015157006918989363,
+      "loss": 0.9144,
+      "step": 24880
+    },
+    {
+      "epoch": 0.7937115341688191,
+      "grad_norm": 0.16771960258483887,
+      "learning_rate": 0.00015093544284822607,
+      "loss": 0.909,
+      "step": 24890
+    },
+    {
+      "epoch": 0.7940304218884531,
+      "grad_norm": 0.1626899540424347,
+      "learning_rate": 0.0001503034736973594,
+      "loss": 0.9096,
+      "step": 24900
+    },
+    {
+      "epoch": 0.794349309608087,
+      "grad_norm": 0.1682693064212799,
+      "learning_rate": 0.00014967415061159254,
+      "loss": 0.9203,
+      "step": 24910
+    },
+    {
+      "epoch": 0.7946681973277209,
+      "grad_norm": 0.16387136280536652,
+      "learning_rate": 0.00014904746251180796,
+      "loss": 0.9158,
+      "step": 24920
+    },
+    {
+      "epoch": 0.7949870850473548,
+      "grad_norm": 0.16813243925571442,
+      "learning_rate": 0.00014842339836527651,
+      "loss": 0.907,
+      "step": 24930
+    },
+    {
+      "epoch": 0.7953059727669888,
+      "grad_norm": 0.16818828880786896,
+      "learning_rate": 0.00014780194718546334,
+      "loss": 0.9139,
+      "step": 24940
+    },
+    {
+      "epoch": 0.7956248604866226,
+      "grad_norm": 0.17012420296669006,
+      "learning_rate": 0.00014718309803183436,
+      "loss": 0.9245,
+      "step": 24950
+    },
+    {
+      "epoch": 0.7959437482062566,
+      "grad_norm": 0.16938473284244537,
+      "learning_rate": 0.00014656684000966363,
+      "loss": 0.911,
+      "step": 24960
+    },
+    {
+      "epoch": 0.7962626359258905,
+      "grad_norm": 0.16450780630111694,
+      "learning_rate": 0.00014595316226984173,
+      "loss": 0.9091,
+      "step": 24970
+    },
+    {
+      "epoch": 0.7965815236455244,
+      "grad_norm": 0.16333532333374023,
+      "learning_rate": 0.00014534205400868448,
+      "loss": 0.9147,
+      "step": 24980
+    },
+    {
+      "epoch": 0.7969004113651583,
+      "grad_norm": 0.16380712389945984,
+      "learning_rate": 0.000144733504467743,
+      "loss": 0.8932,
+      "step": 24990
+    },
+    {
+      "epoch": 0.7972192990847923,
+      "grad_norm": 0.166630357503891,
+      "learning_rate": 0.00014412750293361419,
+      "loss": 0.9106,
+      "step": 25000
+    },
+    {
+      "epoch": 0.7975381868044261,
+      "grad_norm": 0.16908857226371765,
+      "learning_rate": 0.00014352403873775206,
+      "loss": 0.9279,
+      "step": 25010
+    },
+    {
+      "epoch": 0.7978570745240601,
+      "grad_norm": 0.17311139404773712,
+      "learning_rate": 0.0001429231012562802,
+      "loss": 0.9077,
+      "step": 25020
+    },
+    {
+      "epoch": 0.798175962243694,
+      "grad_norm": 0.17072586715221405,
+      "learning_rate": 0.00014232467990980434,
+      "loss": 0.904,
+      "step": 25030
+    },
+    {
+      "epoch": 0.7984948499633279,
+      "grad_norm": 0.17837849259376526,
+      "learning_rate": 0.0001417287641632264,
+      "loss": 0.9207,
+      "step": 25040
+    },
+    {
+      "epoch": 0.7988137376829618,
+      "grad_norm": 0.16402915120124817,
+      "learning_rate": 0.00014113534352555893,
+      "loss": 0.9072,
+      "step": 25050
+    },
+    {
+      "epoch": 0.7991326254025958,
+      "grad_norm": 0.16501149535179138,
+      "learning_rate": 0.00014054440754974036,
+      "loss": 0.9053,
+      "step": 25060
+    },
+    {
+      "epoch": 0.7994515131222296,
+      "grad_norm": 0.16770212352275848,
+      "learning_rate": 0.00013995594583245116,
+      "loss": 0.9241,
+      "step": 25070
+    },
+    {
+      "epoch": 0.7997704008418636,
+      "grad_norm": 0.17260852456092834,
+      "learning_rate": 0.0001393699480139307,
+      "loss": 0.9161,
+      "step": 25080
+    },
+    {
+      "epoch": 0.8000892885614975,
+      "grad_norm": 0.16932493448257446,
+      "learning_rate": 0.00013878640377779487,
+      "loss": 0.9239,
+      "step": 25090
+    },
+    {
+      "epoch": 0.8004081762811314,
+      "grad_norm": 0.16582724452018738,
+      "learning_rate": 0.00013820530285085425,
+      "loss": 0.9012,
+      "step": 25100
+    },
+    {
+      "epoch": 0.8007270640007653,
+      "grad_norm": 0.16309213638305664,
+      "learning_rate": 0.0001376266350029336,
+      "loss": 0.9143,
+      "step": 25110
+    },
+    {
+      "epoch": 0.8010459517203993,
+      "grad_norm": 0.16737112402915955,
+      "learning_rate": 0.0001370503900466916,
+      "loss": 0.9116,
+      "step": 25120
+    },
+    {
+      "epoch": 0.8013648394400331,
+      "grad_norm": 0.16866783797740936,
+      "learning_rate": 0.00013647655783744143,
+      "loss": 0.9052,
+      "step": 25130
+    },
+    {
+      "epoch": 0.8016837271596671,
+      "grad_norm": 0.16859377920627594,
+      "learning_rate": 0.00013590512827297215,
+      "loss": 0.9124,
+      "step": 25140
+    },
+    {
+      "epoch": 0.802002614879301,
+      "grad_norm": 0.1713414490222931,
+      "learning_rate": 0.00013533609129337112,
+      "loss": 0.9035,
+      "step": 25150
+    },
+    {
+      "epoch": 0.8023215025989349,
+      "grad_norm": 0.16575460135936737,
+      "learning_rate": 0.00013476943688084665,
+      "loss": 0.926,
+      "step": 25160
+    },
+    {
+      "epoch": 0.8026403903185688,
+      "grad_norm": 0.17065243422985077,
+      "learning_rate": 0.00013420515505955158,
+      "loss": 0.8947,
+      "step": 25170
+    },
+    {
+      "epoch": 0.8029592780382028,
+      "grad_norm": 0.16338184475898743,
+      "learning_rate": 0.000133643235895408,
+      "loss": 0.9037,
+      "step": 25180
+    },
+    {
+      "epoch": 0.8032781657578366,
+      "grad_norm": 0.17414377629756927,
+      "learning_rate": 0.00013308366949593199,
+      "loss": 0.9007,
+      "step": 25190
+    },
+    {
+      "epoch": 0.8035970534774706,
+      "grad_norm": 0.1691446453332901,
+      "learning_rate": 0.0001325264460100597,
+      "loss": 0.91,
+      "step": 25200
+    },
+    {
+      "epoch": 0.8039159411971045,
+      "grad_norm": 0.17124170064926147,
+      "learning_rate": 0.0001319715556279738,
+      "loss": 0.9199,
+      "step": 25210
+    },
+    {
+      "epoch": 0.8042348289167384,
+      "grad_norm": 0.16459155082702637,
+      "learning_rate": 0.00013141898858093086,
+      "loss": 0.9072,
+      "step": 25220
+    },
+    {
+      "epoch": 0.8045537166363723,
+      "grad_norm": 0.1644795536994934,
+      "learning_rate": 0.00013086873514108925,
+      "loss": 0.8996,
+      "step": 25230
+    },
+    {
+      "epoch": 0.8048726043560063,
+      "grad_norm": 0.16637784242630005,
+      "learning_rate": 0.00013032078562133812,
+      "loss": 0.9232,
+      "step": 25240
+    },
+    {
+      "epoch": 0.8051914920756401,
+      "grad_norm": 0.165706068277359,
+      "learning_rate": 0.0001297751303751266,
+      "loss": 0.9191,
+      "step": 25250
+    },
+    {
+      "epoch": 0.8055103797952741,
+      "grad_norm": 0.1714121401309967,
+      "learning_rate": 0.00012923175979629407,
+      "loss": 0.891,
+      "step": 25260
+    },
+    {
+      "epoch": 0.805829267514908,
+      "grad_norm": 0.1578255593776703,
+      "learning_rate": 0.00012869066431890117,
+      "loss": 0.8875,
+      "step": 25270
+    },
+    {
+      "epoch": 0.8061481552345419,
+      "grad_norm": 0.16436845064163208,
+      "learning_rate": 0.00012815183441706112,
+      "loss": 0.9124,
+      "step": 25280
+    },
+    {
+      "epoch": 0.8064670429541758,
+      "grad_norm": 0.165314182639122,
+      "learning_rate": 0.0001276152606047724,
+      "loss": 0.9131,
+      "step": 25290
+    },
+    {
+      "epoch": 0.8067859306738098,
+      "grad_norm": 0.1633354127407074,
+      "learning_rate": 0.0001270809334357514,
+      "loss": 0.9082,
+      "step": 25300
+    },
+    {
+      "epoch": 0.8071048183934437,
+      "grad_norm": 0.1653420776128769,
+      "learning_rate": 0.00012654884350326619,
+      "loss": 0.8937,
+      "step": 25310
+    },
+    {
+      "epoch": 0.8074237061130776,
+      "grad_norm": 0.16683243215084076,
+      "learning_rate": 0.00012601898143997108,
+      "loss": 0.9267,
+      "step": 25320
+    },
+    {
+      "epoch": 0.8077425938327115,
+      "grad_norm": 0.16653725504875183,
+      "learning_rate": 0.00012549133791774162,
+      "loss": 0.9094,
+      "step": 25330
+    },
+    {
+      "epoch": 0.8080614815523455,
+      "grad_norm": 0.16630671918392181,
+      "learning_rate": 0.0001249659036475103,
+      "loss": 0.9072,
+      "step": 25340
+    },
+    {
+      "epoch": 0.8083803692719793,
+      "grad_norm": 0.16600552201271057,
+      "learning_rate": 0.00012444266937910312,
+      "loss": 0.9107,
+      "step": 25350
+    },
+    {
+      "epoch": 0.8086992569916133,
+      "grad_norm": 0.17081421613693237,
+      "learning_rate": 0.0001239216259010767,
+      "loss": 0.9019,
+      "step": 25360
+    },
+    {
+      "epoch": 0.8090181447112472,
+      "grad_norm": 0.16658015549182892,
+      "learning_rate": 0.00012340276404055616,
+      "loss": 0.8995,
+      "step": 25370
+    },
+    {
+      "epoch": 0.8093370324308811,
+      "grad_norm": 0.16746315360069275,
+      "learning_rate": 0.00012288607466307355,
+      "loss": 0.915,
+      "step": 25380
+    },
+    {
+      "epoch": 0.809655920150515,
+      "grad_norm": 0.1671326905488968,
+      "learning_rate": 0.0001223715486724071,
+      "loss": 0.9013,
+      "step": 25390
+    },
+    {
+      "epoch": 0.809974807870149,
+      "grad_norm": 0.1631750613451004,
+      "learning_rate": 0.00012185917701042106,
+      "loss": 0.9014,
+      "step": 25400
+    },
+    {
+      "epoch": 0.8102936955897828,
+      "grad_norm": 0.16608569025993347,
+      "learning_rate": 0.0001213489506569063,
+      "loss": 0.9044,
+      "step": 25410
+    },
+    {
+      "epoch": 0.8106125833094168,
+      "grad_norm": 0.17144687473773956,
+      "learning_rate": 0.00012084086062942134,
+      "loss": 0.9071,
+      "step": 25420
+    },
+    {
+      "epoch": 0.8109314710290507,
+      "grad_norm": 0.17095890641212463,
+      "learning_rate": 0.00012033489798313444,
+      "loss": 0.9074,
+      "step": 25430
+    },
+    {
+      "epoch": 0.8112503587486846,
+      "grad_norm": 0.16820640861988068,
+      "learning_rate": 0.00011983105381066592,
+      "loss": 0.9162,
+      "step": 25440
+    },
+    {
+      "epoch": 0.8115692464683185,
+      "grad_norm": 0.16968011856079102,
+      "learning_rate": 0.00011932931924193155,
+      "loss": 0.9058,
+      "step": 25450
+    },
+    {
+      "epoch": 0.8118881341879525,
+      "grad_norm": 0.16316497325897217,
+      "learning_rate": 0.0001188296854439862,
+      "loss": 0.8939,
+      "step": 25460
+    },
+    {
+      "epoch": 0.8122070219075863,
+      "grad_norm": 0.17123737931251526,
+      "learning_rate": 0.00011833214362086844,
+      "loss": 0.9167,
+      "step": 25470
+    },
+    {
+      "epoch": 0.8125259096272203,
+      "grad_norm": 0.16038616001605988,
+      "learning_rate": 0.00011783668501344572,
+      "loss": 0.9195,
+      "step": 25480
+    },
+    {
+      "epoch": 0.8128447973468542,
+      "grad_norm": 0.17204549908638,
+      "learning_rate": 0.00011734330089926006,
+      "loss": 0.9142,
+      "step": 25490
+    },
+    {
+      "epoch": 0.8131636850664881,
+      "grad_norm": 0.17099328339099884,
+      "learning_rate": 0.0001168519825923746,
+      "loss": 0.9152,
+      "step": 25500
+    },
+    {
+      "epoch": 0.813482572786122,
+      "grad_norm": 0.16785970330238342,
+      "learning_rate": 0.00011636272144322059,
+      "loss": 0.9089,
+      "step": 25510
+    },
+    {
+      "epoch": 0.813801460505756,
+      "grad_norm": 0.1722290962934494,
+      "learning_rate": 0.00011587550883844523,
+      "loss": 0.9049,
+      "step": 25520
+    },
+    {
+      "epoch": 0.8141203482253898,
+      "grad_norm": 0.17191378772258759,
+      "learning_rate": 0.00011539033620075986,
+      "loss": 0.9104,
+      "step": 25530
+    },
+    {
+      "epoch": 0.8144392359450238,
+      "grad_norm": 0.16544301807880402,
+      "learning_rate": 0.00011490719498878924,
+      "loss": 0.9116,
+      "step": 25540
+    },
+    {
+      "epoch": 0.8147581236646577,
+      "grad_norm": 0.1681593805551529,
+      "learning_rate": 0.00011442607669692085,
+      "loss": 0.9086,
+      "step": 25550
+    },
+    {
+      "epoch": 0.8150770113842916,
+      "grad_norm": 0.16554562747478485,
+      "learning_rate": 0.00011394697285515537,
+      "loss": 0.9004,
+      "step": 25560
+    },
+    {
+      "epoch": 0.8153958991039255,
+      "grad_norm": 0.16348989307880402,
+      "learning_rate": 0.0001134698750289575,
+      "loss": 0.9104,
+      "step": 25570
+    },
+    {
+      "epoch": 0.8157147868235595,
+      "grad_norm": 0.16586485505104065,
+      "learning_rate": 0.00011299477481910747,
+      "loss": 0.9067,
+      "step": 25580
+    },
+    {
+      "epoch": 0.8160336745431933,
+      "grad_norm": 0.17197948694229126,
+      "learning_rate": 0.00011252166386155319,
+      "loss": 0.9176,
+      "step": 25590
+    },
+    {
+      "epoch": 0.8163525622628273,
+      "grad_norm": 0.16917942464351654,
+      "learning_rate": 0.00011205053382726299,
+      "loss": 0.9088,
+      "step": 25600
+    },
+    {
+      "epoch": 0.8166714499824612,
+      "grad_norm": 0.16606974601745605,
+      "learning_rate": 0.00011158137642207893,
+      "loss": 0.9034,
+      "step": 25610
+    },
+    {
+      "epoch": 0.8169903377020951,
+      "grad_norm": 0.16527631878852844,
+      "learning_rate": 0.00011111418338657102,
+      "loss": 0.908,
+      "step": 25620
+    },
+    {
+      "epoch": 0.817309225421729,
+      "grad_norm": 0.16847580671310425,
+      "learning_rate": 0.00011064894649589143,
+      "loss": 0.9171,
+      "step": 25630
+    },
+    {
+      "epoch": 0.817628113141363,
+      "grad_norm": 0.16356876492500305,
+      "learning_rate": 0.00011018565755962999,
+      "loss": 0.912,
+      "step": 25640
+    },
+    {
+      "epoch": 0.8179470008609968,
+      "grad_norm": 0.16655093431472778,
+      "learning_rate": 0.00010972430842166995,
+      "loss": 0.8906,
+      "step": 25650
+    },
+    {
+      "epoch": 0.8182658885806308,
+      "grad_norm": 0.1708131730556488,
+      "learning_rate": 0.0001092648909600443,
+      "loss": 0.9081,
+      "step": 25660
+    },
+    {
+      "epoch": 0.8185847763002647,
+      "grad_norm": 0.16948804259300232,
+      "learning_rate": 0.00010880739708679283,
+      "loss": 0.9052,
+      "step": 25670
+    },
+    {
+      "epoch": 0.8189036640198986,
+      "grad_norm": 0.17297212779521942,
+      "learning_rate": 0.00010835181874781979,
+      "loss": 0.9129,
+      "step": 25680
+    },
+    {
+      "epoch": 0.8192225517395325,
+      "grad_norm": 0.17286381125450134,
+      "learning_rate": 0.00010789814792275205,
+      "loss": 0.9137,
+      "step": 25690
+    },
+    {
+      "epoch": 0.8195414394591665,
+      "grad_norm": 0.17813879251480103,
+      "learning_rate": 0.00010744637662479786,
+      "loss": 0.9146,
+      "step": 25700
+    },
+    {
+      "epoch": 0.8198603271788003,
+      "grad_norm": 0.16722270846366882,
+      "learning_rate": 0.00010699649690060641,
+      "loss": 0.9048,
+      "step": 25710
+    },
+    {
+      "epoch": 0.8201792148984343,
+      "grad_norm": 0.1656230092048645,
+      "learning_rate": 0.00010654850083012758,
+      "loss": 0.889,
+      "step": 25720
+    },
+    {
+      "epoch": 0.8204981026180682,
+      "grad_norm": 0.16876403987407684,
+      "learning_rate": 0.00010610238052647271,
+      "loss": 0.9064,
+      "step": 25730
+    },
+    {
+      "epoch": 0.8208169903377021,
+      "grad_norm": 0.17110751569271088,
+      "learning_rate": 0.00010565812813577561,
+      "loss": 0.9029,
+      "step": 25740
+    },
+    {
+      "epoch": 0.821135878057336,
+      "grad_norm": 0.16737671196460724,
+      "learning_rate": 0.00010521573583705442,
+      "loss": 0.9041,
+      "step": 25750
+    },
+    {
+      "epoch": 0.82145476577697,
+      "grad_norm": 0.16565321385860443,
+      "learning_rate": 0.00010477519584207379,
+      "loss": 0.8823,
+      "step": 25760
+    },
+    {
+      "epoch": 0.8217736534966038,
+      "grad_norm": 0.17009054124355316,
+      "learning_rate": 0.00010433650039520788,
+      "loss": 0.9021,
+      "step": 25770
+    },
+    {
+      "epoch": 0.8220925412162378,
+      "grad_norm": 0.1713910698890686,
+      "learning_rate": 0.00010389964177330376,
+      "loss": 0.9013,
+      "step": 25780
+    },
+    {
+      "epoch": 0.8224114289358717,
+      "grad_norm": 0.171029731631279,
+      "learning_rate": 0.00010346461228554552,
+      "loss": 0.9077,
+      "step": 25790
+    },
+    {
+      "epoch": 0.8227303166555056,
+      "grad_norm": 0.1669735312461853,
+      "learning_rate": 0.00010303140427331876,
+      "loss": 0.8842,
+      "step": 25800
+    },
+    {
+      "epoch": 0.8230492043751395,
+      "grad_norm": 0.16688182950019836,
+      "learning_rate": 0.00010260001011007584,
+      "loss": 0.9053,
+      "step": 25810
+    },
+    {
+      "epoch": 0.8233680920947735,
+      "grad_norm": 0.16837717592716217,
+      "learning_rate": 0.00010217042220120158,
+      "loss": 0.9078,
+      "step": 25820
+    },
+    {
+      "epoch": 0.8236869798144073,
+      "grad_norm": 0.1698017567396164,
+      "learning_rate": 0.00010174263298387963,
+      "loss": 0.8976,
+      "step": 25830
+    },
+    {
+      "epoch": 0.8240058675340413,
+      "grad_norm": 0.16503947973251343,
+      "learning_rate": 0.00010131663492695925,
+      "loss": 0.8984,
+      "step": 25840
+    },
+    {
+      "epoch": 0.8243247552536752,
+      "grad_norm": 0.1713607758283615,
+      "learning_rate": 0.00010089242053082271,
+      "loss": 0.9078,
+      "step": 25850
+    },
+    {
+      "epoch": 0.824643642973309,
+      "grad_norm": 0.1704990118741989,
+      "learning_rate": 0.00010046998232725337,
+      "loss": 0.895,
+      "step": 25860
+    },
+    {
+      "epoch": 0.824962530692943,
+      "grad_norm": 0.16618500649929047,
+      "learning_rate": 0.00010004931287930405,
+      "loss": 0.8866,
+      "step": 25870
+    },
+    {
+      "epoch": 0.825281418412577,
+      "grad_norm": 0.16616980731487274,
+      "learning_rate": 9.96304047811663e-05,
+      "loss": 0.8938,
+      "step": 25880
+    },
+    {
+      "epoch": 0.8256003061322108,
+      "grad_norm": 0.1703374981880188,
+      "learning_rate": 9.921325065803983e-05,
+      "loss": 0.9068,
+      "step": 25890
+    },
+    {
+      "epoch": 0.8259191938518448,
+      "grad_norm": 0.17246967554092407,
+      "learning_rate": 9.879784316600278e-05,
+      "loss": 0.8894,
+      "step": 25900
+    },
+    {
+      "epoch": 0.8262380815714787,
+      "grad_norm": 0.17710429430007935,
+      "learning_rate": 9.838417499188239e-05,
+      "loss": 0.9046,
+      "step": 25910
+    },
+    {
+      "epoch": 0.8265569692911126,
+      "grad_norm": 0.1626577228307724,
+      "learning_rate": 9.79722388531263e-05,
+      "loss": 0.9032,
+      "step": 25920
+    },
+    {
+      "epoch": 0.8268758570107465,
+      "grad_norm": 0.16949373483657837,
+      "learning_rate": 9.756202749767429e-05,
+      "loss": 0.896,
+      "step": 25930
+    },
+    {
+      "epoch": 0.8271947447303805,
+      "grad_norm": 0.16658318042755127,
+      "learning_rate": 9.715353370383069e-05,
+      "loss": 0.9044,
+      "step": 25940
+    },
+    {
+      "epoch": 0.8275136324500143,
+      "grad_norm": 0.16725113987922668,
+      "learning_rate": 9.674675028013712e-05,
+      "loss": 0.8774,
+      "step": 25950
+    },
+    {
+      "epoch": 0.8278325201696483,
+      "grad_norm": 0.17133763432502747,
+      "learning_rate": 9.634167006524597e-05,
+      "loss": 0.905,
+      "step": 25960
+    },
+    {
+      "epoch": 0.8281514078892822,
+      "grad_norm": 0.17014974355697632,
+      "learning_rate": 9.593828592779434e-05,
+      "loss": 0.9,
+      "step": 25970
+    },
+    {
+      "epoch": 0.828470295608916,
+      "grad_norm": 0.17587527632713318,
+      "learning_rate": 9.553659076627844e-05,
+      "loss": 0.9114,
+      "step": 25980
+    },
+    {
+      "epoch": 0.82878918332855,
+      "grad_norm": 0.17234353721141815,
+      "learning_rate": 9.513657750892853e-05,
+      "loss": 0.8956,
+      "step": 25990
+    },
+    {
+      "epoch": 0.829108071048184,
+      "grad_norm": 0.17450740933418274,
+      "learning_rate": 9.473823911358465e-05,
+      "loss": 0.8936,
+      "step": 26000
+    },
+    {
+      "epoch": 0.8294269587678178,
+      "grad_norm": 0.16306303441524506,
+      "learning_rate": 9.434156856757232e-05,
+      "loss": 0.8906,
+      "step": 26010
+    },
+    {
+      "epoch": 0.8297458464874518,
+      "grad_norm": 0.16995976865291595,
+      "learning_rate": 9.394655888757933e-05,
+      "loss": 0.891,
+      "step": 26020
+    },
+    {
+      "epoch": 0.8300647342070857,
+      "grad_norm": 0.16628731787204742,
+      "learning_rate": 9.355320311953263e-05,
+      "loss": 0.9054,
+      "step": 26030
+    },
+    {
+      "epoch": 0.8303836219267196,
+      "grad_norm": 0.17312276363372803,
+      "learning_rate": 9.316149433847609e-05,
+      "loss": 0.8978,
+      "step": 26040
+    },
+    {
+      "epoch": 0.8307025096463535,
+      "grad_norm": 0.16996102035045624,
+      "learning_rate": 9.27714256484484e-05,
+      "loss": 0.9096,
+      "step": 26050
+    },
+    {
+      "epoch": 0.8310213973659875,
+      "grad_norm": 0.1690961867570877,
+      "learning_rate": 9.238299018236176e-05,
+      "loss": 0.8966,
+      "step": 26060
+    },
+    {
+      "epoch": 0.8313402850856213,
+      "grad_norm": 0.16808012127876282,
+      "learning_rate": 9.199618110188106e-05,
+      "loss": 0.8875,
+      "step": 26070
+    },
+    {
+      "epoch": 0.8316591728052553,
+      "grad_norm": 0.16871201992034912,
+      "learning_rate": 9.161099159730329e-05,
+      "loss": 0.898,
+      "step": 26080
+    },
+    {
+      "epoch": 0.8319780605248892,
+      "grad_norm": 0.16951216757297516,
+      "learning_rate": 9.122741488743787e-05,
+      "loss": 0.8942,
+      "step": 26090
+    },
+    {
+      "epoch": 0.832296948244523,
+      "grad_norm": 0.17077015340328217,
+      "learning_rate": 9.084544421948714e-05,
+      "loss": 0.9001,
+      "step": 26100
+    },
+    {
+      "epoch": 0.832615835964157,
+      "grad_norm": 0.16758744418621063,
+      "learning_rate": 9.046507286892751e-05,
+      "loss": 0.93,
+      "step": 26110
+    },
+    {
+      "epoch": 0.832934723683791,
+      "grad_norm": 0.1701851636171341,
+      "learning_rate": 9.008629413939108e-05,
+      "loss": 0.9008,
+      "step": 26120
+    },
+    {
+      "epoch": 0.8332536114034248,
+      "grad_norm": 0.16817468404769897,
+      "learning_rate": 8.970910136254777e-05,
+      "loss": 0.8946,
+      "step": 26130
+    },
+    {
+      "epoch": 0.8335724991230588,
+      "grad_norm": 0.16986632347106934,
+      "learning_rate": 8.93334878979879e-05,
+      "loss": 0.8882,
+      "step": 26140
+    },
+    {
+      "epoch": 0.8338913868426927,
+      "grad_norm": 0.16999934613704681,
+      "learning_rate": 8.895944713310525e-05,
+      "loss": 0.8989,
+      "step": 26150
+    },
+    {
+      "epoch": 0.8342102745623267,
+      "grad_norm": 0.1713329702615738,
+      "learning_rate": 8.858697248298071e-05,
+      "loss": 0.9081,
+      "step": 26160
+    },
+    {
+      "epoch": 0.8345291622819605,
+      "grad_norm": 0.1673143059015274,
+      "learning_rate": 8.821605739026645e-05,
+      "loss": 0.8731,
+      "step": 26170
+    },
+    {
+      "epoch": 0.8348480500015945,
+      "grad_norm": 0.1767398566007614,
+      "learning_rate": 8.784669532507018e-05,
+      "loss": 0.9198,
+      "step": 26180
+    },
+    {
+      "epoch": 0.8351669377212284,
+      "grad_norm": 0.1685178577899933,
+      "learning_rate": 8.747887978484048e-05,
+      "loss": 0.8949,
+      "step": 26190
+    },
+    {
+      "epoch": 0.8354858254408623,
+      "grad_norm": 0.1668165922164917,
+      "learning_rate": 8.71126042942522e-05,
+      "loss": 0.8824,
+      "step": 26200
+    },
+    {
+      "epoch": 0.8358047131604962,
+      "grad_norm": 0.16967454552650452,
+      "learning_rate": 8.674786240509246e-05,
+      "loss": 0.8866,
+      "step": 26210
+    },
+    {
+      "epoch": 0.8361236008801302,
+      "grad_norm": 0.16545966267585754,
+      "learning_rate": 8.638464769614718e-05,
+      "loss": 0.8944,
+      "step": 26220
+    },
+    {
+      "epoch": 0.836442488599764,
+      "grad_norm": 0.1696653962135315,
+      "learning_rate": 8.602295377308798e-05,
+      "loss": 0.8903,
+      "step": 26230
+    },
+    {
+      "epoch": 0.836761376319398,
+      "grad_norm": 0.1657346487045288,
+      "learning_rate": 8.566277426835967e-05,
+      "loss": 0.8936,
+      "step": 26240
+    },
+    {
+      "epoch": 0.8370802640390319,
+      "grad_norm": 0.17045268416404724,
+      "learning_rate": 8.53041028410681e-05,
+      "loss": 0.8925,
+      "step": 26250
+    },
+    {
+      "epoch": 0.8373991517586657,
+      "grad_norm": 0.16796588897705078,
+      "learning_rate": 8.494693317686852e-05,
+      "loss": 0.8732,
+      "step": 26260
+    },
+    {
+      "epoch": 0.8377180394782997,
+      "grad_norm": 0.16764356195926666,
+      "learning_rate": 8.459125898785451e-05,
+      "loss": 0.9136,
+      "step": 26270
+    },
+    {
+      "epoch": 0.8380369271979337,
+      "grad_norm": 0.17094580829143524,
+      "learning_rate": 8.423707401244714e-05,
+      "loss": 0.9105,
+      "step": 26280
+    },
+    {
+      "epoch": 0.8383558149175675,
+      "grad_norm": 0.16578027606010437,
+      "learning_rate": 8.388437201528488e-05,
+      "loss": 0.8921,
+      "step": 26290
+    },
+    {
+      "epoch": 0.8386747026372015,
+      "grad_norm": 0.16390374302864075,
+      "learning_rate": 8.353314678711372e-05,
+      "loss": 0.9015,
+      "step": 26300
+    },
+    {
+      "epoch": 0.8389935903568354,
+      "grad_norm": 0.16992898285388947,
+      "learning_rate": 8.318339214467789e-05,
+      "loss": 0.8958,
+      "step": 26310
+    },
+    {
+      "epoch": 0.8393124780764692,
+      "grad_norm": 0.16880562901496887,
+      "learning_rate": 8.283510193061105e-05,
+      "loss": 0.8929,
+      "step": 26320
+    },
+    {
+      "epoch": 0.8396313657961032,
+      "grad_norm": 0.17104868590831757,
+      "learning_rate": 8.248827001332778e-05,
+      "loss": 0.8978,
+      "step": 26330
+    },
+    {
+      "epoch": 0.8399502535157372,
+      "grad_norm": 0.1668756604194641,
+      "learning_rate": 8.214289028691584e-05,
+      "loss": 0.893,
+      "step": 26340
+    },
+    {
+      "epoch": 0.840269141235371,
+      "grad_norm": 0.164683997631073,
+      "learning_rate": 8.179895667102844e-05,
+      "loss": 0.8812,
+      "step": 26350
+    },
+    {
+      "epoch": 0.840588028955005,
+      "grad_norm": 0.1663207858800888,
+      "learning_rate": 8.145646311077731e-05,
+      "loss": 0.9042,
+      "step": 26360
+    },
+    {
+      "epoch": 0.8409069166746389,
+      "grad_norm": 0.17519667744636536,
+      "learning_rate": 8.11154035766261e-05,
+      "loss": 0.8994,
+      "step": 26370
+    },
+    {
+      "epoch": 0.8412258043942727,
+      "grad_norm": 0.17041118443012238,
+      "learning_rate": 8.077577206428427e-05,
+      "loss": 0.894,
+      "step": 26380
+    },
+    {
+      "epoch": 0.8415446921139067,
+      "grad_norm": 0.16244420409202576,
+      "learning_rate": 8.043756259460127e-05,
+      "loss": 0.8898,
+      "step": 26390
+    },
+    {
+      "epoch": 0.8418635798335407,
+      "grad_norm": 0.17468824982643127,
+      "learning_rate": 8.010076921346141e-05,
+      "loss": 0.911,
+      "step": 26400
+    },
+    {
+      "epoch": 0.8421824675531745,
+      "grad_norm": 0.16985976696014404,
+      "learning_rate": 7.976538599167896e-05,
+      "loss": 0.9037,
+      "step": 26410
+    },
+    {
+      "epoch": 0.8425013552728084,
+      "grad_norm": 0.16972625255584717,
+      "learning_rate": 7.943140702489378e-05,
+      "loss": 0.8783,
+      "step": 26420
+    },
+    {
+      "epoch": 0.8428202429924424,
+      "grad_norm": 0.1677105873823166,
+      "learning_rate": 7.909882643346739e-05,
+      "loss": 0.8946,
+      "step": 26430
+    },
+    {
+      "epoch": 0.8431391307120762,
+      "grad_norm": 0.16455358266830444,
+      "learning_rate": 7.876763836237944e-05,
+      "loss": 0.8986,
+      "step": 26440
+    },
+    {
+      "epoch": 0.8434580184317102,
+      "grad_norm": 0.17376911640167236,
+      "learning_rate": 7.843783698112465e-05,
+      "loss": 0.9102,
+      "step": 26450
+    },
+    {
+      "epoch": 0.8437769061513442,
+      "grad_norm": 0.16827256977558136,
+      "learning_rate": 7.810941648361018e-05,
+      "loss": 0.9022,
+      "step": 26460
+    },
+    {
+      "epoch": 0.844095793870978,
+      "grad_norm": 0.17023880779743195,
+      "learning_rate": 7.778237108805339e-05,
+      "loss": 0.8833,
+      "step": 26470
+    },
+    {
+      "epoch": 0.844414681590612,
+      "grad_norm": 0.16073589026927948,
+      "learning_rate": 7.745669503688002e-05,
+      "loss": 0.8917,
+      "step": 26480
+    },
+    {
+      "epoch": 0.8447335693102459,
+      "grad_norm": 0.1697019338607788,
+      "learning_rate": 7.71323825966229e-05,
+      "loss": 0.8898,
+      "step": 26490
+    },
+    {
+      "epoch": 0.8450524570298797,
+      "grad_norm": 0.1616843193769455,
+      "learning_rate": 7.680942805782095e-05,
+      "loss": 0.8913,
+      "step": 26500
+    },
+    {
+      "epoch": 0.8453713447495137,
+      "grad_norm": 0.16631615161895752,
+      "learning_rate": 7.648782573491877e-05,
+      "loss": 0.8904,
+      "step": 26510
+    },
+    {
+      "epoch": 0.8456902324691477,
+      "grad_norm": 0.16899070143699646,
+      "learning_rate": 7.616756996616643e-05,
+      "loss": 0.8829,
+      "step": 26520
+    },
+    {
+      "epoch": 0.8460091201887815,
+      "grad_norm": 0.17207767069339752,
+      "learning_rate": 7.58486551135198e-05,
+      "loss": 0.8887,
+      "step": 26530
+    },
+    {
+      "epoch": 0.8463280079084154,
+      "grad_norm": 0.1661059707403183,
+      "learning_rate": 7.553107556254135e-05,
+      "loss": 0.886,
+      "step": 26540
+    },
+    {
+      "epoch": 0.8466468956280494,
+      "grad_norm": 0.1708707958459854,
+      "learning_rate": 7.521482572230134e-05,
+      "loss": 0.8922,
+      "step": 26550
+    },
+    {
+      "epoch": 0.8469657833476832,
+      "grad_norm": 0.17209689319133759,
+      "learning_rate": 7.48999000252793e-05,
+      "loss": 0.8907,
+      "step": 26560
+    },
+    {
+      "epoch": 0.8472846710673172,
+      "grad_norm": 0.16356642544269562,
+      "learning_rate": 7.458629292726607e-05,
+      "loss": 0.9037,
+      "step": 26570
+    },
+    {
+      "epoch": 0.8476035587869512,
+      "grad_norm": 0.1679961085319519,
+      "learning_rate": 7.427399890726616e-05,
+      "loss": 0.8862,
+      "step": 26580
+    },
+    {
+      "epoch": 0.847922446506585,
+      "grad_norm": 0.16855265200138092,
+      "learning_rate": 7.396301246740063e-05,
+      "loss": 0.8844,
+      "step": 26590
+    },
+    {
+      "epoch": 0.848241334226219,
+      "grad_norm": 0.1739872395992279,
+      "learning_rate": 7.36533281328102e-05,
+      "loss": 0.898,
+      "step": 26600
+    },
+    {
+      "epoch": 0.8485602219458529,
+      "grad_norm": 0.17627912759780884,
+      "learning_rate": 7.334494045155892e-05,
+      "loss": 0.8887,
+      "step": 26610
+    },
+    {
+      "epoch": 0.8488791096654867,
+      "grad_norm": 0.1743604987859726,
+      "learning_rate": 7.303784399453824e-05,
+      "loss": 0.8938,
+      "step": 26620
+    },
+    {
+      "epoch": 0.8491979973851207,
+      "grad_norm": 0.17387619614601135,
+      "learning_rate": 7.273203335537129e-05,
+      "loss": 0.8926,
+      "step": 26630
+    },
+    {
+      "epoch": 0.8495168851047546,
+      "grad_norm": 0.16775065660476685,
+      "learning_rate": 7.242750315031787e-05,
+      "loss": 0.9088,
+      "step": 26640
+    },
+    {
+      "epoch": 0.8498357728243885,
+      "grad_norm": 0.16406485438346863,
+      "learning_rate": 7.21242480181795e-05,
+      "loss": 0.8927,
+      "step": 26650
+    },
+    {
+      "epoch": 0.8501546605440224,
+      "grad_norm": 0.16974839568138123,
+      "learning_rate": 7.182226262020522e-05,
+      "loss": 0.884,
+      "step": 26660
+    },
+    {
+      "epoch": 0.8504735482636564,
+      "grad_norm": 0.1713690608739853,
+      "learning_rate": 7.15215416399974e-05,
+      "loss": 0.8959,
+      "step": 26670
+    },
+    {
+      "epoch": 0.8507924359832902,
+      "grad_norm": 0.16833071410655975,
+      "learning_rate": 7.122207978341839e-05,
+      "loss": 0.8971,
+      "step": 26680
+    },
+    {
+      "epoch": 0.8511113237029242,
+      "grad_norm": 0.16310231387615204,
+      "learning_rate": 7.092387177849706e-05,
+      "loss": 0.8876,
+      "step": 26690
+    },
+    {
+      "epoch": 0.8514302114225581,
+      "grad_norm": 0.16681894659996033,
+      "learning_rate": 7.062691237533617e-05,
+      "loss": 0.8831,
+      "step": 26700
+    },
+    {
+      "epoch": 0.851749099142192,
+      "grad_norm": 0.17143681645393372,
+      "learning_rate": 7.033119634601985e-05,
+      "loss": 0.8951,
+      "step": 26710
+    },
+    {
+      "epoch": 0.8520679868618259,
+      "grad_norm": 0.17085224390029907,
+      "learning_rate": 7.003671848452163e-05,
+      "loss": 0.8917,
+      "step": 26720
+    },
+    {
+      "epoch": 0.8523868745814599,
+      "grad_norm": 0.17123526334762573,
+      "learning_rate": 6.974347360661275e-05,
+      "loss": 0.8968,
+      "step": 26730
+    },
+    {
+      "epoch": 0.8527057623010937,
+      "grad_norm": 0.16923531889915466,
+      "learning_rate": 6.945145654977087e-05,
+      "loss": 0.903,
+      "step": 26740
+    },
+    {
+      "epoch": 0.8530246500207277,
+      "grad_norm": 0.16685211658477783,
+      "learning_rate": 6.916066217308926e-05,
+      "loss": 0.9017,
+      "step": 26750
+    },
+    {
+      "epoch": 0.8533435377403616,
+      "grad_norm": 0.17287422716617584,
+      "learning_rate": 6.887108535718623e-05,
+      "loss": 0.8875,
+      "step": 26760
+    },
+    {
+      "epoch": 0.8536624254599955,
+      "grad_norm": 0.1733153909444809,
+      "learning_rate": 6.858272100411499e-05,
+      "loss": 0.882,
+      "step": 26770
+    },
+    {
+      "epoch": 0.8539813131796294,
+      "grad_norm": 0.17159032821655273,
+      "learning_rate": 6.829556403727401e-05,
+      "loss": 0.9034,
+      "step": 26780
+    },
+    {
+      "epoch": 0.8543002008992634,
+      "grad_norm": 0.17427712678909302,
+      "learning_rate": 6.800960940131751e-05,
+      "loss": 0.8958,
+      "step": 26790
+    },
+    {
+      "epoch": 0.8546190886188972,
+      "grad_norm": 0.17055897414684296,
+      "learning_rate": 6.772485206206656e-05,
+      "loss": 0.8752,
+      "step": 26800
+    },
+    {
+      "epoch": 0.8549379763385312,
+      "grad_norm": 0.16725292801856995,
+      "learning_rate": 6.74412870064204e-05,
+      "loss": 0.8948,
+      "step": 26810
+    },
+    {
+      "epoch": 0.8552568640581651,
+      "grad_norm": 0.17402830719947815,
+      "learning_rate": 6.71589092422682e-05,
+      "loss": 0.9024,
+      "step": 26820
+    },
+    {
+      "epoch": 0.855575751777799,
+      "grad_norm": 0.17112645506858826,
+      "learning_rate": 6.687771379840115e-05,
+      "loss": 0.9049,
+      "step": 26830
+    },
+    {
+      "epoch": 0.8558946394974329,
+      "grad_norm": 0.16811510920524597,
+      "learning_rate": 6.659769572442513e-05,
+      "loss": 0.8944,
+      "step": 26840
+    },
+    {
+      "epoch": 0.8562135272170669,
+      "grad_norm": 0.17514300346374512,
+      "learning_rate": 6.631885009067319e-05,
+      "loss": 0.8874,
+      "step": 26850
+    },
+    {
+      "epoch": 0.8565324149367007,
+      "grad_norm": 0.17014634609222412,
+      "learning_rate": 6.60411719881191e-05,
+      "loss": 0.8843,
+      "step": 26860
+    },
+    {
+      "epoch": 0.8568513026563347,
+      "grad_norm": 0.1698083132505417,
+      "learning_rate": 6.576465652829075e-05,
+      "loss": 0.8958,
+      "step": 26870
+    },
+    {
+      "epoch": 0.8571701903759686,
+      "grad_norm": 0.17299139499664307,
+      "learning_rate": 6.548929884318418e-05,
+      "loss": 0.8875,
+      "step": 26880
+    },
+    {
+      "epoch": 0.8574890780956025,
+      "grad_norm": 0.1718268096446991,
+      "learning_rate": 6.521509408517782e-05,
+      "loss": 0.8985,
+      "step": 26890
+    },
+    {
+      "epoch": 0.8578079658152364,
+      "grad_norm": 0.17264573276042938,
+      "learning_rate": 6.494203742694715e-05,
+      "loss": 0.901,
+      "step": 26900
+    },
+    {
+      "epoch": 0.8581268535348704,
+      "grad_norm": 0.17113256454467773,
+      "learning_rate": 6.46701240613798e-05,
+      "loss": 0.8991,
+      "step": 26910
+    },
+    {
+      "epoch": 0.8584457412545042,
+      "grad_norm": 0.1725773960351944,
+      "learning_rate": 6.439934920149081e-05,
+      "loss": 0.8862,
+      "step": 26920
+    },
+    {
+      "epoch": 0.8587646289741382,
+      "grad_norm": 0.16512632369995117,
+      "learning_rate": 6.412970808033838e-05,
+      "loss": 0.9074,
+      "step": 26930
+    },
+    {
+      "epoch": 0.8590835166937721,
+      "grad_norm": 0.16803482174873352,
+      "learning_rate": 6.386119595094003e-05,
+      "loss": 0.8861,
+      "step": 26940
+    },
+    {
+      "epoch": 0.859402404413406,
+      "grad_norm": 0.17119905352592468,
+      "learning_rate": 6.359380808618895e-05,
+      "loss": 0.8833,
+      "step": 26950
+    },
+    {
+      "epoch": 0.8597212921330399,
+      "grad_norm": 0.17140822112560272,
+      "learning_rate": 6.332753977877079e-05,
+      "loss": 0.8966,
+      "step": 26960
+    },
+    {
+      "epoch": 0.8600401798526739,
+      "grad_norm": 0.16653090715408325,
+      "learning_rate": 6.306238634108082e-05,
+      "loss": 0.8692,
+      "step": 26970
+    },
+    {
+      "epoch": 0.8603590675723078,
+      "grad_norm": 0.17240291833877563,
+      "learning_rate": 6.279834310514136e-05,
+      "loss": 0.8787,
+      "step": 26980
+    },
+    {
+      "epoch": 0.8606779552919417,
+      "grad_norm": 0.1726362258195877,
+      "learning_rate": 6.253540542251968e-05,
+      "loss": 0.8907,
+      "step": 26990
+    },
+    {
+      "epoch": 0.8609968430115756,
+      "grad_norm": 0.16658638417720795,
+      "learning_rate": 6.227356866424601e-05,
+      "loss": 0.8746,
+      "step": 27000
+    },
+    {
+      "epoch": 0.8613157307312096,
+      "grad_norm": 0.1688665747642517,
+      "learning_rate": 6.201282822073233e-05,
+      "loss": 0.8718,
+      "step": 27010
+    },
+    {
+      "epoch": 0.8616346184508434,
+      "grad_norm": 0.16762253642082214,
+      "learning_rate": 6.175317950169087e-05,
+      "loss": 0.8908,
+      "step": 27020
+    },
+    {
+      "epoch": 0.8619535061704774,
+      "grad_norm": 0.1679273396730423,
+      "learning_rate": 6.149461793605354e-05,
+      "loss": 0.896,
+      "step": 27030
+    },
+    {
+      "epoch": 0.8622723938901113,
+      "grad_norm": 0.16647253930568695,
+      "learning_rate": 6.123713897189136e-05,
+      "loss": 0.9041,
+      "step": 27040
+    },
+    {
+      "epoch": 0.8625912816097452,
+      "grad_norm": 0.17113405466079712,
+      "learning_rate": 6.0980738076334414e-05,
+      "loss": 0.9049,
+      "step": 27050
+    },
+    {
+      "epoch": 0.8629101693293791,
+      "grad_norm": 0.16743695735931396,
+      "learning_rate": 6.0725410735491895e-05,
+      "loss": 0.8993,
+      "step": 27060
+    },
+    {
+      "epoch": 0.8632290570490131,
+      "grad_norm": 0.1676100492477417,
+      "learning_rate": 6.0471152454372806e-05,
+      "loss": 0.8842,
+      "step": 27070
+    },
+    {
+      "epoch": 0.8635479447686469,
+      "grad_norm": 0.16340462863445282,
+      "learning_rate": 6.02179587568067e-05,
+      "loss": 0.8992,
+      "step": 27080
+    },
+    {
+      "epoch": 0.8638668324882809,
+      "grad_norm": 0.17687806487083435,
+      "learning_rate": 5.9965825185364964e-05,
+      "loss": 0.9011,
+      "step": 27090
+    },
+    {
+      "epoch": 0.8641857202079148,
+      "grad_norm": 0.16743621230125427,
+      "learning_rate": 5.971474730128228e-05,
+      "loss": 0.8925,
+      "step": 27100
+    },
+    {
+      "epoch": 0.8645046079275487,
+      "grad_norm": 0.16600142419338226,
+      "learning_rate": 5.946472068437854e-05,
+      "loss": 0.8834,
+      "step": 27110
+    },
+    {
+      "epoch": 0.8648234956471826,
+      "grad_norm": 0.16654567420482635,
+      "learning_rate": 5.9215740932980986e-05,
+      "loss": 0.8979,
+      "step": 27120
+    },
+    {
+      "epoch": 0.8651423833668166,
+      "grad_norm": 0.16914024949073792,
+      "learning_rate": 5.8967803663846734e-05,
+      "loss": 0.8792,
+      "step": 27130
+    },
+    {
+      "epoch": 0.8654612710864504,
+      "grad_norm": 0.1687769889831543,
+      "learning_rate": 5.8720904512085626e-05,
+      "loss": 0.8741,
+      "step": 27140
+    },
+    {
+      "epoch": 0.8657801588060844,
+      "grad_norm": 0.16539452970027924,
+      "learning_rate": 5.8475039131083355e-05,
+      "loss": 0.8891,
+      "step": 27150
+    },
+    {
+      "epoch": 0.8660990465257183,
+      "grad_norm": 0.1684424728155136,
+      "learning_rate": 5.823020319242495e-05,
+      "loss": 0.8984,
+      "step": 27160
+    },
+    {
+      "epoch": 0.8664179342453522,
+      "grad_norm": 0.17129109799861908,
+      "learning_rate": 5.7986392385818584e-05,
+      "loss": 0.8829,
+      "step": 27170
+    },
+    {
+      "epoch": 0.8667368219649861,
+      "grad_norm": 0.16870862245559692,
+      "learning_rate": 5.774360241901975e-05,
+      "loss": 0.8918,
+      "step": 27180
+    },
+    {
+      "epoch": 0.8670557096846201,
+      "grad_norm": 0.1671636551618576,
+      "learning_rate": 5.7501829017755564e-05,
+      "loss": 0.8901,
+      "step": 27190
+    },
+    {
+      "epoch": 0.8673745974042539,
+      "grad_norm": 0.1682850569486618,
+      "learning_rate": 5.7261067925649635e-05,
+      "loss": 0.8859,
+      "step": 27200
+    },
+    {
+      "epoch": 0.8676934851238879,
+      "grad_norm": 0.1720380336046219,
+      "learning_rate": 5.7021314904147045e-05,
+      "loss": 0.8873,
+      "step": 27210
+    },
+    {
+      "epoch": 0.8680123728435218,
+      "grad_norm": 0.16618755459785461,
+      "learning_rate": 5.678256573243984e-05,
+      "loss": 0.8861,
+      "step": 27220
+    },
+    {
+      "epoch": 0.8683312605631557,
+      "grad_norm": 0.16924364864826202,
+      "learning_rate": 5.6544816207392587e-05,
+      "loss": 0.8817,
+      "step": 27230
+    },
+    {
+      "epoch": 0.8686501482827896,
+      "grad_norm": 0.17504048347473145,
+      "learning_rate": 5.630806214346851e-05,
+      "loss": 0.8845,
+      "step": 27240
+    },
+    {
+      "epoch": 0.8689690360024236,
+      "grad_norm": 0.17469248175621033,
+      "learning_rate": 5.6072299372655695e-05,
+      "loss": 0.8939,
+      "step": 27250
+    },
+    {
+      "epoch": 0.8692879237220574,
+      "grad_norm": 0.17130273580551147,
+      "learning_rate": 5.58375237443938e-05,
+      "loss": 0.8864,
+      "step": 27260
+    },
+    {
+      "epoch": 0.8696068114416914,
+      "grad_norm": 0.16825518012046814,
+      "learning_rate": 5.5603731125500924e-05,
+      "loss": 0.8862,
+      "step": 27270
+    },
+    {
+      "epoch": 0.8699256991613253,
+      "grad_norm": 0.16963137686252594,
+      "learning_rate": 5.537091740010087e-05,
+      "loss": 0.8836,
+      "step": 27280
+    },
+    {
+      "epoch": 0.8702445868809592,
+      "grad_norm": 0.16663995385169983,
+      "learning_rate": 5.513907846955069e-05,
+      "loss": 0.889,
+      "step": 27290
+    },
+    {
+      "epoch": 0.8705634746005931,
+      "grad_norm": 0.17626075446605682,
+      "learning_rate": 5.490821025236851e-05,
+      "loss": 0.8988,
+      "step": 27300
+    },
+    {
+      "epoch": 0.8708823623202271,
+      "grad_norm": 0.17636112868785858,
+      "learning_rate": 5.467830868416169e-05,
+      "loss": 0.9077,
+      "step": 27310
+    },
+    {
+      "epoch": 0.8712012500398609,
+      "grad_norm": 0.16531747579574585,
+      "learning_rate": 5.444936971755526e-05,
+      "loss": 0.8831,
+      "step": 27320
+    },
+    {
+      "epoch": 0.8715201377594949,
+      "grad_norm": 0.16421984136104584,
+      "learning_rate": 5.42213893221207e-05,
+      "loss": 0.8964,
+      "step": 27330
+    },
+    {
+      "epoch": 0.8718390254791288,
+      "grad_norm": 0.16629725694656372,
+      "learning_rate": 5.399436348430491e-05,
+      "loss": 0.898,
+      "step": 27340
+    },
+    {
+      "epoch": 0.8721579131987627,
+      "grad_norm": 0.16994720697402954,
+      "learning_rate": 5.3768288207359675e-05,
+      "loss": 0.8807,
+      "step": 27350
+    },
+    {
+      "epoch": 0.8724768009183966,
+      "grad_norm": 0.16791296005249023,
+      "learning_rate": 5.354315951127119e-05,
+      "loss": 0.8975,
+      "step": 27360
+    },
+    {
+      "epoch": 0.8727956886380306,
+      "grad_norm": 0.16906745731830597,
+      "learning_rate": 5.331897343269001e-05,
+      "loss": 0.8791,
+      "step": 27370
+    },
+    {
+      "epoch": 0.8731145763576644,
+      "grad_norm": 0.16941961646080017,
+      "learning_rate": 5.309572602486132e-05,
+      "loss": 0.8884,
+      "step": 27380
+    },
+    {
+      "epoch": 0.8734334640772984,
+      "grad_norm": 0.1706288754940033,
+      "learning_rate": 5.287341335755546e-05,
+      "loss": 0.8908,
+      "step": 27390
+    },
+    {
+      "epoch": 0.8737523517969323,
+      "grad_norm": 0.17212943732738495,
+      "learning_rate": 5.265203151699865e-05,
+      "loss": 0.8844,
+      "step": 27400
+    },
+    {
+      "epoch": 0.8740712395165662,
+      "grad_norm": 0.169947549700737,
+      "learning_rate": 5.243157660580418e-05,
+      "loss": 0.8755,
+      "step": 27410
+    },
+    {
+      "epoch": 0.8743901272362001,
+      "grad_norm": 0.16829125583171844,
+      "learning_rate": 5.221204474290376e-05,
+      "loss": 0.8802,
+      "step": 27420
+    },
+    {
+      "epoch": 0.8747090149558341,
+      "grad_norm": 0.16881629824638367,
+      "learning_rate": 5.199343206347918e-05,
+      "loss": 0.8895,
+      "step": 27430
+    },
+    {
+      "epoch": 0.8750279026754679,
+      "grad_norm": 0.17132100462913513,
+      "learning_rate": 5.1775734718894304e-05,
+      "loss": 0.8841,
+      "step": 27440
+    },
+    {
+      "epoch": 0.8753467903951019,
+      "grad_norm": 0.16899465024471283,
+      "learning_rate": 5.155894887662728e-05,
+      "loss": 0.8875,
+      "step": 27450
+    },
+    {
+      "epoch": 0.8756656781147358,
+      "grad_norm": 0.16898700594902039,
+      "learning_rate": 5.134307072020311e-05,
+      "loss": 0.8845,
+      "step": 27460
+    },
+    {
+      "epoch": 0.8759845658343697,
+      "grad_norm": 0.16705811023712158,
+      "learning_rate": 5.1128096449126425e-05,
+      "loss": 0.8938,
+      "step": 27470
+    },
+    {
+      "epoch": 0.8763034535540036,
+      "grad_norm": 0.1732359230518341,
+      "learning_rate": 5.091402227881458e-05,
+      "loss": 0.8851,
+      "step": 27480
+    },
+    {
+      "epoch": 0.8766223412736376,
+      "grad_norm": 0.16625288128852844,
+      "learning_rate": 5.070084444053106e-05,
+      "loss": 0.8773,
+      "step": 27490
+    },
+    {
+      "epoch": 0.8769412289932714,
+      "grad_norm": 0.1704675406217575,
+      "learning_rate": 5.0488559181319116e-05,
+      "loss": 0.8947,
+      "step": 27500
+    },
+    {
+      "epoch": 0.8772601167129054,
+      "grad_norm": 0.1694217324256897,
+      "learning_rate": 5.027716276393563e-05,
+      "loss": 0.8796,
+      "step": 27510
+    },
+    {
+      "epoch": 0.8775790044325393,
+      "grad_norm": 0.1724761426448822,
+      "learning_rate": 5.00666514667855e-05,
+      "loss": 0.8954,
+      "step": 27520
+    },
+    {
+      "epoch": 0.8778978921521732,
+      "grad_norm": 0.17092296481132507,
+      "learning_rate": 4.9857021583855885e-05,
+      "loss": 0.8679,
+      "step": 27530
+    },
+    {
+      "epoch": 0.8782167798718071,
+      "grad_norm": 0.16874085366725922,
+      "learning_rate": 4.9648269424651144e-05,
+      "loss": 0.8892,
+      "step": 27540
+    },
+    {
+      "epoch": 0.8785356675914411,
+      "grad_norm": 0.1716938316822052,
+      "learning_rate": 4.9440391314127776e-05,
+      "loss": 0.8908,
+      "step": 27550
+    },
+    {
+      "epoch": 0.8788545553110749,
+      "grad_norm": 0.16745954751968384,
+      "learning_rate": 4.9233383592629755e-05,
+      "loss": 0.8868,
+      "step": 27560
+    },
+    {
+      "epoch": 0.8791734430307089,
+      "grad_norm": 0.16559866070747375,
+      "learning_rate": 4.90272426158241e-05,
+      "loss": 0.8863,
+      "step": 27570
+    },
+    {
+      "epoch": 0.8794923307503428,
+      "grad_norm": 0.17054715752601624,
+      "learning_rate": 4.88219647546367e-05,
+      "loss": 0.8916,
+      "step": 27580
+    },
+    {
+      "epoch": 0.8798112184699767,
+      "grad_norm": 0.16883596777915955,
+      "learning_rate": 4.861754639518844e-05,
+      "loss": 0.8855,
+      "step": 27590
+    },
+    {
+      "epoch": 0.8801301061896106,
+      "grad_norm": 0.17326515913009644,
+      "learning_rate": 4.841398393873158e-05,
+      "loss": 0.8869,
+      "step": 27600
+    },
+    {
+      "epoch": 0.8804489939092446,
+      "grad_norm": 0.16945582628250122,
+      "learning_rate": 4.821127380158639e-05,
+      "loss": 0.8828,
+      "step": 27610
+    },
+    {
+      "epoch": 0.8807678816288784,
+      "grad_norm": 0.17122694849967957,
+      "learning_rate": 4.800941241507807e-05,
+      "loss": 0.8858,
+      "step": 27620
+    },
+    {
+      "epoch": 0.8810867693485124,
+      "grad_norm": 0.1702810525894165,
+      "learning_rate": 4.7808396225473926e-05,
+      "loss": 0.888,
+      "step": 27630
+    },
+    {
+      "epoch": 0.8814056570681463,
+      "grad_norm": 0.1752856969833374,
+      "learning_rate": 4.760822169392078e-05,
+      "loss": 0.8836,
+      "step": 27640
+    },
+    {
+      "epoch": 0.8817245447877802,
+      "grad_norm": 0.17062492668628693,
+      "learning_rate": 4.74088852963827e-05,
+      "loss": 0.8794,
+      "step": 27650
+    },
+    {
+      "epoch": 0.8820434325074141,
+      "grad_norm": 0.17067866027355194,
+      "learning_rate": 4.7210383523578954e-05,
+      "loss": 0.8948,
+      "step": 27660
+    },
+    {
+      "epoch": 0.8823623202270481,
+      "grad_norm": 0.17074808478355408,
+      "learning_rate": 4.701271288092223e-05,
+      "loss": 0.8772,
+      "step": 27670
+    },
+    {
+      "epoch": 0.8826812079466819,
+      "grad_norm": 0.17069657146930695,
+      "learning_rate": 4.681586988845707e-05,
+      "loss": 0.8873,
+      "step": 27680
+    },
+    {
+      "epoch": 0.8830000956663159,
+      "grad_norm": 0.17014001309871674,
+      "learning_rate": 4.661985108079873e-05,
+      "loss": 0.906,
+      "step": 27690
+    },
+    {
+      "epoch": 0.8833189833859498,
+      "grad_norm": 0.1751696914434433,
+      "learning_rate": 4.6424653007072e-05,
+      "loss": 0.8868,
+      "step": 27700
+    },
+    {
+      "epoch": 0.8836378711055837,
+      "grad_norm": 0.17144422233104706,
+      "learning_rate": 4.62302722308506e-05,
+      "loss": 0.8726,
+      "step": 27710
+    },
+    {
+      "epoch": 0.8839567588252176,
+      "grad_norm": 0.16718114912509918,
+      "learning_rate": 4.603670533009658e-05,
+      "loss": 0.88,
+      "step": 27720
+    },
+    {
+      "epoch": 0.8842756465448516,
+      "grad_norm": 0.16777946054935455,
+      "learning_rate": 4.5843948897100144e-05,
+      "loss": 0.8843,
+      "step": 27730
+    },
+    {
+      "epoch": 0.8845945342644854,
+      "grad_norm": 0.1685672551393509,
+      "learning_rate": 4.565199953841964e-05,
+      "loss": 0.8636,
+      "step": 27740
+    },
+    {
+      "epoch": 0.8849134219841194,
+      "grad_norm": 0.16482628881931305,
+      "learning_rate": 4.5460853874821776e-05,
+      "loss": 0.8987,
+      "step": 27750
+    },
+    {
+      "epoch": 0.8852323097037533,
+      "grad_norm": 0.1724732220172882,
+      "learning_rate": 4.52705085412222e-05,
+      "loss": 0.8861,
+      "step": 27760
+    },
+    {
+      "epoch": 0.8855511974233872,
+      "grad_norm": 0.16705913841724396,
+      "learning_rate": 4.50809601866262e-05,
+      "loss": 0.8876,
+      "step": 27770
+    },
+    {
+      "epoch": 0.8858700851430211,
+      "grad_norm": 0.1683725118637085,
+      "learning_rate": 4.4892205474069755e-05,
+      "loss": 0.8732,
+      "step": 27780
+    },
+    {
+      "epoch": 0.8861889728626551,
+      "grad_norm": 0.16657312214374542,
+      "learning_rate": 4.4704241080560745e-05,
+      "loss": 0.879,
+      "step": 27790
+    },
+    {
+      "epoch": 0.886507860582289,
+      "grad_norm": 0.17320038378238678,
+      "learning_rate": 4.4517063697020485e-05,
+      "loss": 0.8927,
+      "step": 27800
+    },
+    {
+      "epoch": 0.8868267483019229,
+      "grad_norm": 0.168224036693573,
+      "learning_rate": 4.433067002822545e-05,
+      "loss": 0.8843,
+      "step": 27810
+    },
+    {
+      "epoch": 0.8871456360215568,
+      "grad_norm": 0.17128513753414154,
+      "learning_rate": 4.414505679274928e-05,
+      "loss": 0.8674,
+      "step": 27820
+    },
+    {
+      "epoch": 0.8874645237411908,
+      "grad_norm": 0.1696639060974121,
+      "learning_rate": 4.396022072290497e-05,
+      "loss": 0.8786,
+      "step": 27830
+    },
+    {
+      "epoch": 0.8877834114608246,
+      "grad_norm": 0.16941528022289276,
+      "learning_rate": 4.3776158564687404e-05,
+      "loss": 0.883,
+      "step": 27840
+    },
+    {
+      "epoch": 0.8881022991804586,
+      "grad_norm": 0.17108596861362457,
+      "learning_rate": 4.359286707771605e-05,
+      "loss": 0.8974,
+      "step": 27850
+    },
+    {
+      "epoch": 0.8884211869000925,
+      "grad_norm": 0.17025145888328552,
+      "learning_rate": 4.341034303517787e-05,
+      "loss": 0.8835,
+      "step": 27860
+    },
+    {
+      "epoch": 0.8887400746197264,
+      "grad_norm": 0.1636083573102951,
+      "learning_rate": 4.322858322377056e-05,
+      "loss": 0.8814,
+      "step": 27870
+    },
+    {
+      "epoch": 0.8890589623393603,
+      "grad_norm": 0.17729361355304718,
+      "learning_rate": 4.3047584443645955e-05,
+      "loss": 0.889,
+      "step": 27880
+    },
+    {
+      "epoch": 0.8893778500589943,
+      "grad_norm": 0.17000319063663483,
+      "learning_rate": 4.2867343508353714e-05,
+      "loss": 0.8673,
+      "step": 27890
+    },
+    {
+      "epoch": 0.8896967377786281,
+      "grad_norm": 0.16710565984249115,
+      "learning_rate": 4.2687857244785214e-05,
+      "loss": 0.8925,
+      "step": 27900
+    },
+    {
+      "epoch": 0.8900156254982621,
+      "grad_norm": 0.16732816398143768,
+      "learning_rate": 4.2509122493117696e-05,
+      "loss": 0.891,
+      "step": 27910
+    },
+    {
+      "epoch": 0.890334513217896,
+      "grad_norm": 0.1728038638830185,
+      "learning_rate": 4.2331136106758614e-05,
+      "loss": 0.8982,
+      "step": 27920
+    },
+    {
+      "epoch": 0.8906534009375299,
+      "grad_norm": 0.17565219104290009,
+      "learning_rate": 4.2153894952290276e-05,
+      "loss": 0.8811,
+      "step": 27930
+    },
+    {
+      "epoch": 0.8909722886571638,
+      "grad_norm": 0.16937175393104553,
+      "learning_rate": 4.1977395909414635e-05,
+      "loss": 0.8812,
+      "step": 27940
+    },
+    {
+      "epoch": 0.8912911763767978,
+      "grad_norm": 0.17110460996627808,
+      "learning_rate": 4.180163587089841e-05,
+      "loss": 0.8875,
+      "step": 27950
+    },
+    {
+      "epoch": 0.8916100640964316,
+      "grad_norm": 0.17079512774944305,
+      "learning_rate": 4.162661174251835e-05,
+      "loss": 0.893,
+      "step": 27960
+    },
+    {
+      "epoch": 0.8919289518160656,
+      "grad_norm": 0.17112374305725098,
+      "learning_rate": 4.1452320443006753e-05,
+      "loss": 0.8969,
+      "step": 27970
+    },
+    {
+      "epoch": 0.8922478395356995,
+      "grad_norm": 0.17021846771240234,
+      "learning_rate": 4.1278758903997235e-05,
+      "loss": 0.9055,
+      "step": 27980
+    },
+    {
+      "epoch": 0.8925667272553334,
+      "grad_norm": 0.1693822592496872,
+      "learning_rate": 4.110592406997074e-05,
+      "loss": 0.881,
+      "step": 27990
+    },
+    {
+      "epoch": 0.8928856149749673,
+      "grad_norm": 0.1631503701210022,
+      "learning_rate": 4.0933812898201676e-05,
+      "loss": 0.8829,
+      "step": 28000
+    },
+    {
+      "epoch": 0.8932045026946013,
+      "grad_norm": 0.17165639996528625,
+      "learning_rate": 4.0762422358704414e-05,
+      "loss": 0.8823,
+      "step": 28010
+    },
+    {
+      "epoch": 0.8935233904142351,
+      "grad_norm": 0.16423417627811432,
+      "learning_rate": 4.059174943417997e-05,
+      "loss": 0.8846,
+      "step": 28020
+    },
+    {
+      "epoch": 0.8938422781338691,
+      "grad_norm": 0.1652715504169464,
+      "learning_rate": 4.0421791119962764e-05,
+      "loss": 0.8891,
+      "step": 28030
+    },
+    {
+      "epoch": 0.894161165853503,
+      "grad_norm": 0.16647374629974365,
+      "learning_rate": 4.0252544423967835e-05,
+      "loss": 0.8773,
+      "step": 28040
+    },
+    {
+      "epoch": 0.8944800535731369,
+      "grad_norm": 0.1734417974948883,
+      "learning_rate": 4.008400636663814e-05,
+      "loss": 0.8669,
+      "step": 28050
+    },
+    {
+      "epoch": 0.8947989412927708,
+      "grad_norm": 0.17314991354942322,
+      "learning_rate": 3.991617398089205e-05,
+      "loss": 0.8803,
+      "step": 28060
+    },
+    {
+      "epoch": 0.8951178290124048,
+      "grad_norm": 0.17280423641204834,
+      "learning_rate": 3.974904431207121e-05,
+      "loss": 0.8789,
+      "step": 28070
+    },
+    {
+      "epoch": 0.8954367167320386,
+      "grad_norm": 0.1620936095714569,
+      "learning_rate": 3.958261441788843e-05,
+      "loss": 0.8818,
+      "step": 28080
+    },
+    {
+      "epoch": 0.8957556044516726,
+      "grad_norm": 0.17496642470359802,
+      "learning_rate": 3.941688136837593e-05,
+      "loss": 0.8794,
+      "step": 28090
+    },
+    {
+      "epoch": 0.8960744921713065,
+      "grad_norm": 0.1700027585029602,
+      "learning_rate": 3.925184224583376e-05,
+      "loss": 0.8868,
+      "step": 28100
+    },
+    {
+      "epoch": 0.8963933798909404,
+      "grad_norm": 0.17198674380779266,
+      "learning_rate": 3.908749414477842e-05,
+      "loss": 0.875,
+      "step": 28110
+    },
+    {
+      "epoch": 0.8967122676105743,
+      "grad_norm": 0.17227719724178314,
+      "learning_rate": 3.8923834171891724e-05,
+      "loss": 0.8971,
+      "step": 28120
+    },
+    {
+      "epoch": 0.8970311553302083,
+      "grad_norm": 0.1696770191192627,
+      "learning_rate": 3.876085944596985e-05,
+      "loss": 0.8776,
+      "step": 28130
+    },
+    {
+      "epoch": 0.8973500430498421,
+      "grad_norm": 0.17306533455848694,
+      "learning_rate": 3.8598567097872634e-05,
+      "loss": 0.9021,
+      "step": 28140
+    },
+    {
+      "epoch": 0.8976689307694761,
+      "grad_norm": 0.1734350621700287,
+      "learning_rate": 3.843695427047303e-05,
+      "loss": 0.8853,
+      "step": 28150
+    },
+    {
+      "epoch": 0.89798781848911,
+      "grad_norm": 0.1695108264684677,
+      "learning_rate": 3.8276018118606846e-05,
+      "loss": 0.8854,
+      "step": 28160
+    },
+    {
+      "epoch": 0.8983067062087439,
+      "grad_norm": 0.17414987087249756,
+      "learning_rate": 3.8115755809022624e-05,
+      "loss": 0.8736,
+      "step": 28170
+    },
+    {
+      "epoch": 0.8986255939283778,
+      "grad_norm": 0.17102502286434174,
+      "learning_rate": 3.7956164520331766e-05,
+      "loss": 0.8808,
+      "step": 28180
+    },
+    {
+      "epoch": 0.8989444816480118,
+      "grad_norm": 0.1671244353055954,
+      "learning_rate": 3.7797241442958946e-05,
+      "loss": 0.8792,
+      "step": 28190
+    },
+    {
+      "epoch": 0.8992633693676456,
+      "grad_norm": 0.16684648394584656,
+      "learning_rate": 3.7638983779092486e-05,
+      "loss": 0.8803,
+      "step": 28200
+    },
+    {
+      "epoch": 0.8995822570872796,
+      "grad_norm": 0.16940031945705414,
+      "learning_rate": 3.748138874263523e-05,
+      "loss": 0.8861,
+      "step": 28210
+    },
+    {
+      "epoch": 0.8999011448069135,
+      "grad_norm": 0.17539231479167938,
+      "learning_rate": 3.732445355915546e-05,
+      "loss": 0.8952,
+      "step": 28220
+    },
+    {
+      "epoch": 0.9002200325265474,
+      "grad_norm": 0.1667565107345581,
+      "learning_rate": 3.716817546583803e-05,
+      "loss": 0.8709,
+      "step": 28230
+    },
+    {
+      "epoch": 0.9005389202461813,
+      "grad_norm": 0.17119689285755157,
+      "learning_rate": 3.7012551711435764e-05,
+      "loss": 0.8894,
+      "step": 28240
+    },
+    {
+      "epoch": 0.9008578079658153,
+      "grad_norm": 0.16709868609905243,
+      "learning_rate": 3.6857579556221e-05,
+      "loss": 0.8902,
+      "step": 28250
+    },
+    {
+      "epoch": 0.9011766956854491,
+      "grad_norm": 0.16595764458179474,
+      "learning_rate": 3.670325627193734e-05,
+      "loss": 0.8678,
+      "step": 28260
+    },
+    {
+      "epoch": 0.9014955834050831,
+      "grad_norm": 0.1742243766784668,
+      "learning_rate": 3.654957914175167e-05,
+      "loss": 0.889,
+      "step": 28270
+    },
+    {
+      "epoch": 0.901814471124717,
+      "grad_norm": 0.16573461890220642,
+      "learning_rate": 3.639654546020629e-05,
+      "loss": 0.8755,
+      "step": 28280
+    },
+    {
+      "epoch": 0.9021333588443509,
+      "grad_norm": 0.16762784123420715,
+      "learning_rate": 3.6244152533171275e-05,
+      "loss": 0.8882,
+      "step": 28290
+    },
+    {
+      "epoch": 0.9024522465639848,
+      "grad_norm": 0.17437873780727386,
+      "learning_rate": 3.609239767779709e-05,
+      "loss": 0.9081,
+      "step": 28300
+    },
+    {
+      "epoch": 0.9027711342836188,
+      "grad_norm": 0.1660221666097641,
+      "learning_rate": 3.594127822246733e-05,
+      "loss": 0.8893,
+      "step": 28310
+    },
+    {
+      "epoch": 0.9030900220032526,
+      "grad_norm": 0.1724633127450943,
+      "learning_rate": 3.579079150675168e-05,
+      "loss": 0.8775,
+      "step": 28320
+    },
+    {
+      "epoch": 0.9034089097228866,
+      "grad_norm": 0.16792918741703033,
+      "learning_rate": 3.564093488135911e-05,
+      "loss": 0.8816,
+      "step": 28330
+    },
+    {
+      "epoch": 0.9037277974425205,
+      "grad_norm": 0.16467350721359253,
+      "learning_rate": 3.54917057080912e-05,
+      "loss": 0.8676,
+      "step": 28340
+    },
+    {
+      "epoch": 0.9040466851621544,
+      "grad_norm": 0.17283858358860016,
+      "learning_rate": 3.534310135979569e-05,
+      "loss": 0.8891,
+      "step": 28350
+    },
+    {
+      "epoch": 0.9043655728817883,
+      "grad_norm": 0.16493968665599823,
+      "learning_rate": 3.51951192203203e-05,
+      "loss": 0.886,
+      "step": 28360
+    },
+    {
+      "epoch": 0.9046844606014223,
+      "grad_norm": 0.16881853342056274,
+      "learning_rate": 3.504775668446659e-05,
+      "loss": 0.8881,
+      "step": 28370
+    },
+    {
+      "epoch": 0.9050033483210561,
+      "grad_norm": 0.17092576622962952,
+      "learning_rate": 3.4901011157944126e-05,
+      "loss": 0.8742,
+      "step": 28380
+    },
+    {
+      "epoch": 0.9053222360406901,
+      "grad_norm": 0.16808760166168213,
+      "learning_rate": 3.475488005732481e-05,
+      "loss": 0.8764,
+      "step": 28390
+    },
+    {
+      "epoch": 0.905641123760324,
+      "grad_norm": 0.17145703732967377,
+      "learning_rate": 3.460936080999741e-05,
+      "loss": 0.8797,
+      "step": 28400
+    },
+    {
+      "epoch": 0.9059600114799579,
+      "grad_norm": 0.17452701926231384,
+      "learning_rate": 3.4464450854122266e-05,
+      "loss": 0.8841,
+      "step": 28410
+    },
+    {
+      "epoch": 0.9062788991995918,
+      "grad_norm": 0.17020602524280548,
+      "learning_rate": 3.4320147638586154e-05,
+      "loss": 0.8796,
+      "step": 28420
+    },
+    {
+      "epoch": 0.9065977869192258,
+      "grad_norm": 0.167311429977417,
+      "learning_rate": 3.4176448622957427e-05,
+      "loss": 0.9003,
+      "step": 28430
+    },
+    {
+      "epoch": 0.9069166746388596,
+      "grad_norm": 0.17277421057224274,
+      "learning_rate": 3.4033351277441255e-05,
+      "loss": 0.8943,
+      "step": 28440
+    },
+    {
+      "epoch": 0.9072355623584936,
+      "grad_norm": 0.16550636291503906,
+      "learning_rate": 3.38908530828351e-05,
+      "loss": 0.891,
+      "step": 28450
+    },
+    {
+      "epoch": 0.9075544500781275,
+      "grad_norm": 0.17455977201461792,
+      "learning_rate": 3.374895153048438e-05,
+      "loss": 0.8742,
+      "step": 28460
+    },
+    {
+      "epoch": 0.9078733377977614,
+      "grad_norm": 0.1689627766609192,
+      "learning_rate": 3.360764412223827e-05,
+      "loss": 0.8864,
+      "step": 28470
+    },
+    {
+      "epoch": 0.9081922255173953,
+      "grad_norm": 0.16884039342403412,
+      "learning_rate": 3.3466928370405766e-05,
+      "loss": 0.8854,
+      "step": 28480
+    },
+    {
+      "epoch": 0.9085111132370293,
+      "grad_norm": 0.17706649005413055,
+      "learning_rate": 3.332680179771184e-05,
+      "loss": 0.8888,
+      "step": 28490
+    },
+    {
+      "epoch": 0.9088300009566631,
+      "grad_norm": 0.17109733819961548,
+      "learning_rate": 3.318726193725388e-05,
+      "loss": 0.8859,
+      "step": 28500
+    },
+    {
+      "epoch": 0.9091488886762971,
+      "grad_norm": 0.16820432245731354,
+      "learning_rate": 3.304830633245822e-05,
+      "loss": 0.8813,
+      "step": 28510
+    },
+    {
+      "epoch": 0.909467776395931,
+      "grad_norm": 0.16424791514873505,
+      "learning_rate": 3.290993253703689e-05,
+      "loss": 0.8874,
+      "step": 28520
+    },
+    {
+      "epoch": 0.9097866641155649,
+      "grad_norm": 0.1689070761203766,
+      "learning_rate": 3.277213811494463e-05,
+      "loss": 0.8583,
+      "step": 28530
+    },
+    {
+      "epoch": 0.9101055518351988,
+      "grad_norm": 0.1675632745027542,
+      "learning_rate": 3.263492064033587e-05,
+      "loss": 0.8841,
+      "step": 28540
+    },
+    {
+      "epoch": 0.9104244395548328,
+      "grad_norm": 0.16971516609191895,
+      "learning_rate": 3.249827769752215e-05,
+      "loss": 0.8744,
+      "step": 28550
+    },
+    {
+      "epoch": 0.9107433272744666,
+      "grad_norm": 0.17137958109378815,
+      "learning_rate": 3.2362206880929494e-05,
+      "loss": 0.8926,
+      "step": 28560
+    },
+    {
+      "epoch": 0.9110622149941006,
+      "grad_norm": 0.16915251314640045,
+      "learning_rate": 3.222670579505612e-05,
+      "loss": 0.8867,
+      "step": 28570
+    },
+    {
+      "epoch": 0.9113811027137345,
+      "grad_norm": 0.1734655350446701,
+      "learning_rate": 3.2091772054430256e-05,
+      "loss": 0.8892,
+      "step": 28580
+    },
+    {
+      "epoch": 0.9116999904333684,
+      "grad_norm": 0.17015081644058228,
+      "learning_rate": 3.195740328356815e-05,
+      "loss": 0.893,
+      "step": 28590
+    },
+    {
+      "epoch": 0.9120188781530023,
+      "grad_norm": 0.17893768846988678,
+      "learning_rate": 3.182359711693219e-05,
+      "loss": 0.899,
+      "step": 28600
+    },
+    {
+      "epoch": 0.9123377658726363,
+      "grad_norm": 0.17800496518611908,
+      "learning_rate": 3.169035119888938e-05,
+      "loss": 0.8932,
+      "step": 28610
+    },
+    {
+      "epoch": 0.9126566535922701,
+      "grad_norm": 0.167316734790802,
+      "learning_rate": 3.155766318366973e-05,
+      "loss": 0.8821,
+      "step": 28620
+    },
+    {
+      "epoch": 0.9129755413119041,
+      "grad_norm": 0.16948860883712769,
+      "learning_rate": 3.142553073532508e-05,
+      "loss": 0.8715,
+      "step": 28630
+    },
+    {
+      "epoch": 0.913294429031538,
+      "grad_norm": 0.16441291570663452,
+      "learning_rate": 3.129395152768789e-05,
+      "loss": 0.8813,
+      "step": 28640
+    },
+    {
+      "epoch": 0.913613316751172,
+      "grad_norm": 0.17731501162052155,
+      "learning_rate": 3.116292324433033e-05,
+      "loss": 0.8748,
+      "step": 28650
+    },
+    {
+      "epoch": 0.9139322044708058,
+      "grad_norm": 0.1751125305891037,
+      "learning_rate": 3.103244357852349e-05,
+      "loss": 0.8809,
+      "step": 28660
+    },
+    {
+      "epoch": 0.9142510921904398,
+      "grad_norm": 0.17110154032707214,
+      "learning_rate": 3.090251023319679e-05,
+      "loss": 0.8851,
+      "step": 28670
+    },
+    {
+      "epoch": 0.9145699799100737,
+      "grad_norm": 0.17188861966133118,
+      "learning_rate": 3.0773120920897486e-05,
+      "loss": 0.8927,
+      "step": 28680
+    },
+    {
+      "epoch": 0.9148888676297076,
+      "grad_norm": 0.16816741228103638,
+      "learning_rate": 3.064427336375045e-05,
+      "loss": 0.8914,
+      "step": 28690
+    },
+    {
+      "epoch": 0.9152077553493415,
+      "grad_norm": 0.1723993867635727,
+      "learning_rate": 3.0515965293418096e-05,
+      "loss": 0.8826,
+      "step": 28700
+    },
+    {
+      "epoch": 0.9155266430689755,
+      "grad_norm": 0.17036281526088715,
+      "learning_rate": 3.038819445106034e-05,
+      "loss": 0.8818,
+      "step": 28710
+    },
+    {
+      "epoch": 0.9158455307886093,
+      "grad_norm": 0.17272977530956268,
+      "learning_rate": 3.0260958587294935e-05,
+      "loss": 0.8868,
+      "step": 28720
+    },
+    {
+      "epoch": 0.9161644185082433,
+      "grad_norm": 0.16877420246601105,
+      "learning_rate": 3.013425546215782e-05,
+      "loss": 0.899,
+      "step": 28730
+    },
+    {
+      "epoch": 0.9164833062278772,
+      "grad_norm": 0.17606976628303528,
+      "learning_rate": 3.0008082845063706e-05,
+      "loss": 0.9043,
+      "step": 28740
+    },
+    {
+      "epoch": 0.9168021939475111,
+      "grad_norm": 0.1669413298368454,
+      "learning_rate": 2.9882438514766802e-05,
+      "loss": 0.8902,
+      "step": 28750
+    },
+    {
+      "epoch": 0.917121081667145,
+      "grad_norm": 0.1702776402235031,
+      "learning_rate": 2.9757320259321707e-05,
+      "loss": 0.8857,
+      "step": 28760
+    },
+    {
+      "epoch": 0.917439969386779,
+      "grad_norm": 0.17731963098049164,
+      "learning_rate": 2.963272587604447e-05,
+      "loss": 0.8812,
+      "step": 28770
+    },
+    {
+      "epoch": 0.9177588571064128,
+      "grad_norm": 0.17650321125984192,
+      "learning_rate": 2.9508653171473813e-05,
+      "loss": 0.8824,
+      "step": 28780
+    },
+    {
+      "epoch": 0.9180777448260468,
+      "grad_norm": 0.1683824211359024,
+      "learning_rate": 2.9385099961332523e-05,
+      "loss": 0.8816,
+      "step": 28790
+    },
+    {
+      "epoch": 0.9183966325456807,
+      "grad_norm": 0.16492652893066406,
+      "learning_rate": 2.926206407048898e-05,
+      "loss": 0.8801,
+      "step": 28800
+    },
+    {
+      "epoch": 0.9187155202653146,
+      "grad_norm": 0.171652689576149,
+      "learning_rate": 2.913954333291889e-05,
+      "loss": 0.878,
+      "step": 28810
+    },
+    {
+      "epoch": 0.9190344079849485,
+      "grad_norm": 0.17132626473903656,
+      "learning_rate": 2.901753559166712e-05,
+      "loss": 0.8644,
+      "step": 28820
+    },
+    {
+      "epoch": 0.9193532957045825,
+      "grad_norm": 0.16985774040222168,
+      "learning_rate": 2.8896038698809757e-05,
+      "loss": 0.876,
+      "step": 28830
+    },
+    {
+      "epoch": 0.9196721834242163,
+      "grad_norm": 0.16665631532669067,
+      "learning_rate": 2.8775050515416274e-05,
+      "loss": 0.8824,
+      "step": 28840
+    },
+    {
+      "epoch": 0.9199910711438503,
+      "grad_norm": 0.16637495160102844,
+      "learning_rate": 2.865456891151188e-05,
+      "loss": 0.8828,
+      "step": 28850
+    },
+    {
+      "epoch": 0.9203099588634842,
+      "grad_norm": 0.17059068381786346,
+      "learning_rate": 2.8534591766040024e-05,
+      "loss": 0.8874,
+      "step": 28860
+    },
+    {
+      "epoch": 0.920628846583118,
+      "grad_norm": 0.17028462886810303,
+      "learning_rate": 2.8415116966825072e-05,
+      "loss": 0.8691,
+      "step": 28870
+    },
+    {
+      "epoch": 0.920947734302752,
+      "grad_norm": 0.17274853587150574,
+      "learning_rate": 2.8296142410535074e-05,
+      "loss": 0.8799,
+      "step": 28880
+    },
+    {
+      "epoch": 0.921266622022386,
+      "grad_norm": 0.17280860245227814,
+      "learning_rate": 2.8177666002644782e-05,
+      "loss": 0.8886,
+      "step": 28890
+    },
+    {
+      "epoch": 0.9215855097420198,
+      "grad_norm": 0.17315393686294556,
+      "learning_rate": 2.8059685657398748e-05,
+      "loss": 0.8912,
+      "step": 28900
+    },
+    {
+      "epoch": 0.9219043974616538,
+      "grad_norm": 0.1712070256471634,
+      "learning_rate": 2.7942199297774628e-05,
+      "loss": 0.8801,
+      "step": 28910
+    },
+    {
+      "epoch": 0.9222232851812877,
+      "grad_norm": 0.1668015420436859,
+      "learning_rate": 2.7825204855446604e-05,
+      "loss": 0.8856,
+      "step": 28920
+    },
+    {
+      "epoch": 0.9225421729009216,
+      "grad_norm": 0.17411622405052185,
+      "learning_rate": 2.7708700270748953e-05,
+      "loss": 0.8919,
+      "step": 28930
+    },
+    {
+      "epoch": 0.9228610606205555,
+      "grad_norm": 0.1741609424352646,
+      "learning_rate": 2.759268349263983e-05,
+      "loss": 0.8824,
+      "step": 28940
+    },
+    {
+      "epoch": 0.9231799483401895,
+      "grad_norm": 0.16620464622974396,
+      "learning_rate": 2.7477152478665126e-05,
+      "loss": 0.8814,
+      "step": 28950
+    },
+    {
+      "epoch": 0.9234988360598233,
+      "grad_norm": 0.1636183261871338,
+      "learning_rate": 2.736210519492252e-05,
+      "loss": 0.8813,
+      "step": 28960
+    },
+    {
+      "epoch": 0.9238177237794573,
+      "grad_norm": 0.17570005357265472,
+      "learning_rate": 2.7247539616025663e-05,
+      "loss": 0.8769,
+      "step": 28970
+    },
+    {
+      "epoch": 0.9241366114990912,
+      "grad_norm": 0.17470939457416534,
+      "learning_rate": 2.7133453725068553e-05,
+      "loss": 0.8717,
+      "step": 28980
+    },
+    {
+      "epoch": 0.924455499218725,
+      "grad_norm": 0.1725778877735138,
+      "learning_rate": 2.701984551358999e-05,
+      "loss": 0.8794,
+      "step": 28990
+    },
+    {
+      "epoch": 0.924774386938359,
+      "grad_norm": 0.17439262568950653,
+      "learning_rate": 2.6906712981538236e-05,
+      "loss": 0.871,
+      "step": 29000
+    },
+    {
+      "epoch": 0.925093274657993,
+      "grad_norm": 0.17004485428333282,
+      "learning_rate": 2.6794054137235806e-05,
+      "loss": 0.8829,
+      "step": 29010
+    },
+    {
+      "epoch": 0.9254121623776268,
+      "grad_norm": 0.16694343090057373,
+      "learning_rate": 2.668186699734439e-05,
+      "loss": 0.8817,
+      "step": 29020
+    },
+    {
+      "epoch": 0.9257310500972608,
+      "grad_norm": 0.16977915167808533,
+      "learning_rate": 2.657014958682998e-05,
+      "loss": 0.8775,
+      "step": 29030
+    },
+    {
+      "epoch": 0.9260499378168947,
+      "grad_norm": 0.16832669079303741,
+      "learning_rate": 2.6458899938928035e-05,
+      "loss": 0.8709,
+      "step": 29040
+    },
+    {
+      "epoch": 0.9263688255365286,
+      "grad_norm": 0.17108400166034698,
+      "learning_rate": 2.634811609510889e-05,
+      "loss": 0.8769,
+      "step": 29050
+    },
+    {
+      "epoch": 0.9266877132561625,
+      "grad_norm": 0.16579735279083252,
+      "learning_rate": 2.6237796105043294e-05,
+      "loss": 0.8902,
+      "step": 29060
+    },
+    {
+      "epoch": 0.9270066009757965,
+      "grad_norm": 0.17090383172035217,
+      "learning_rate": 2.6127938026568044e-05,
+      "loss": 0.8801,
+      "step": 29070
+    },
+    {
+      "epoch": 0.9273254886954303,
+      "grad_norm": 0.16820332407951355,
+      "learning_rate": 2.6018539925651806e-05,
+      "loss": 0.8622,
+      "step": 29080
+    },
+    {
+      "epoch": 0.9276443764150643,
+      "grad_norm": 0.16625086963176727,
+      "learning_rate": 2.5909599876361067e-05,
+      "loss": 0.8704,
+      "step": 29090
+    },
+    {
+      "epoch": 0.9279632641346982,
+      "grad_norm": 0.16523200273513794,
+      "learning_rate": 2.5801115960826234e-05,
+      "loss": 0.8886,
+      "step": 29100
+    },
+    {
+      "epoch": 0.928282151854332,
+      "grad_norm": 0.17791950702667236,
+      "learning_rate": 2.5693086269207865e-05,
+      "loss": 0.9059,
+      "step": 29110
+    },
+    {
+      "epoch": 0.928601039573966,
+      "grad_norm": 0.17191381752490997,
+      "learning_rate": 2.5585508899663037e-05,
+      "loss": 0.8945,
+      "step": 29120
+    },
+    {
+      "epoch": 0.9289199272936,
+      "grad_norm": 0.16965806484222412,
+      "learning_rate": 2.5478381958311885e-05,
+      "loss": 0.8772,
+      "step": 29130
+    },
+    {
+      "epoch": 0.9292388150132338,
+      "grad_norm": 0.17314006388187408,
+      "learning_rate": 2.5371703559204248e-05,
+      "loss": 0.8806,
+      "step": 29140
+    },
+    {
+      "epoch": 0.9295577027328678,
+      "grad_norm": 0.16447074711322784,
+      "learning_rate": 2.526547182428646e-05,
+      "loss": 0.8664,
+      "step": 29150
+    },
+    {
+      "epoch": 0.9298765904525017,
+      "grad_norm": 0.1670355200767517,
+      "learning_rate": 2.5159684883368308e-05,
+      "loss": 0.8647,
+      "step": 29160
+    },
+    {
+      "epoch": 0.9301954781721355,
+      "grad_norm": 0.17165569961071014,
+      "learning_rate": 2.505434087409009e-05,
+      "loss": 0.887,
+      "step": 29170
+    },
+    {
+      "epoch": 0.9305143658917695,
+      "grad_norm": 0.1715584695339203,
+      "learning_rate": 2.4949437941889832e-05,
+      "loss": 0.8798,
+      "step": 29180
+    },
+    {
+      "epoch": 0.9308332536114035,
+      "grad_norm": 0.16674351692199707,
+      "learning_rate": 2.4844974239970637e-05,
+      "loss": 0.8571,
+      "step": 29190
+    },
+    {
+      "epoch": 0.9311521413310373,
+      "grad_norm": 0.17173631489276886,
+      "learning_rate": 2.4740947929268202e-05,
+      "loss": 0.8847,
+      "step": 29200
+    },
+    {
+      "epoch": 0.9314710290506713,
+      "grad_norm": 0.16858477890491486,
+      "learning_rate": 2.4637357178418395e-05,
+      "loss": 0.8984,
+      "step": 29210
+    },
+    {
+      "epoch": 0.9317899167703052,
+      "grad_norm": 0.1788715422153473,
+      "learning_rate": 2.4534200163725028e-05,
+      "loss": 0.8812,
+      "step": 29220
+    },
+    {
+      "epoch": 0.932108804489939,
+      "grad_norm": 0.16826432943344116,
+      "learning_rate": 2.443147506912777e-05,
+      "loss": 0.8722,
+      "step": 29230
+    },
+    {
+      "epoch": 0.932427692209573,
+      "grad_norm": 0.17240197956562042,
+      "learning_rate": 2.4329180086170172e-05,
+      "loss": 0.873,
+      "step": 29240
+    },
+    {
+      "epoch": 0.932746579929207,
+      "grad_norm": 0.17108172178268433,
+      "learning_rate": 2.4227313413967803e-05,
+      "loss": 0.8774,
+      "step": 29250
+    },
+    {
+      "epoch": 0.9330654676488408,
+      "grad_norm": 0.17302483320236206,
+      "learning_rate": 2.412587325917658e-05,
+      "loss": 0.8797,
+      "step": 29260
+    },
+    {
+      "epoch": 0.9333843553684747,
+      "grad_norm": 0.17489032447338104,
+      "learning_rate": 2.4024857835961166e-05,
+      "loss": 0.8738,
+      "step": 29270
+    },
+    {
+      "epoch": 0.9337032430881087,
+      "grad_norm": 0.1717793345451355,
+      "learning_rate": 2.392426536596356e-05,
+      "loss": 0.8737,
+      "step": 29280
+    },
+    {
+      "epoch": 0.9340221308077425,
+      "grad_norm": 0.1737363338470459,
+      "learning_rate": 2.3824094078271775e-05,
+      "loss": 0.8741,
+      "step": 29290
+    },
+    {
+      "epoch": 0.9343410185273765,
+      "grad_norm": 0.167171448469162,
+      "learning_rate": 2.3724342209388646e-05,
+      "loss": 0.8704,
+      "step": 29300
+    },
+    {
+      "epoch": 0.9346599062470105,
+      "grad_norm": 0.17019569873809814,
+      "learning_rate": 2.3625008003200812e-05,
+      "loss": 0.8873,
+      "step": 29310
+    },
+    {
+      "epoch": 0.9349787939666443,
+      "grad_norm": 0.16639681160449982,
+      "learning_rate": 2.352608971094778e-05,
+      "loss": 0.8801,
+      "step": 29320
+    },
+    {
+      "epoch": 0.9352976816862782,
+      "grad_norm": 0.16961655020713806,
+      "learning_rate": 2.3427585591191153e-05,
+      "loss": 0.8958,
+      "step": 29330
+    },
+    {
+      "epoch": 0.9356165694059122,
+      "grad_norm": 0.16953252255916595,
+      "learning_rate": 2.3329493909783962e-05,
+      "loss": 0.8683,
+      "step": 29340
+    },
+    {
+      "epoch": 0.935935457125546,
+      "grad_norm": 0.17069625854492188,
+      "learning_rate": 2.3231812939840138e-05,
+      "loss": 0.8698,
+      "step": 29350
+    },
+    {
+      "epoch": 0.93625434484518,
+      "grad_norm": 0.17173562943935394,
+      "learning_rate": 2.3134540961704098e-05,
+      "loss": 0.8596,
+      "step": 29360
+    },
+    {
+      "epoch": 0.936573232564814,
+      "grad_norm": 0.17173653841018677,
+      "learning_rate": 2.3037676262920523e-05,
+      "loss": 0.8856,
+      "step": 29370
+    },
+    {
+      "epoch": 0.9368921202844478,
+      "grad_norm": 0.16764011979103088,
+      "learning_rate": 2.2941217138204138e-05,
+      "loss": 0.8833,
+      "step": 29380
+    },
+    {
+      "epoch": 0.9372110080040817,
+      "grad_norm": 0.17342248558998108,
+      "learning_rate": 2.2845161889409744e-05,
+      "loss": 0.8802,
+      "step": 29390
+    },
+    {
+      "epoch": 0.9375298957237157,
+      "grad_norm": 0.16736873984336853,
+      "learning_rate": 2.2749508825502283e-05,
+      "loss": 0.8853,
+      "step": 29400
+    },
+    {
+      "epoch": 0.9378487834433495,
+      "grad_norm": 0.16693644225597382,
+      "learning_rate": 2.2654256262527105e-05,
+      "loss": 0.8825,
+      "step": 29410
+    },
+    {
+      "epoch": 0.9381676711629835,
+      "grad_norm": 0.17182965576648712,
+      "learning_rate": 2.2559402523580303e-05,
+      "loss": 0.8868,
+      "step": 29420
+    },
+    {
+      "epoch": 0.9384865588826174,
+      "grad_norm": 0.17486432194709778,
+      "learning_rate": 2.2464945938779194e-05,
+      "loss": 0.8742,
+      "step": 29430
+    },
+    {
+      "epoch": 0.9388054466022513,
+      "grad_norm": 0.17658363282680511,
+      "learning_rate": 2.2370884845232913e-05,
+      "loss": 0.8701,
+      "step": 29440
+    },
+    {
+      "epoch": 0.9391243343218852,
+      "grad_norm": 0.16929444670677185,
+      "learning_rate": 2.2277217587013156e-05,
+      "loss": 0.8716,
+      "step": 29450
+    },
+    {
+      "epoch": 0.9394432220415192,
+      "grad_norm": 0.1668825000524521,
+      "learning_rate": 2.2183942515125016e-05,
+      "loss": 0.8824,
+      "step": 29460
+    },
+    {
+      "epoch": 0.9397621097611532,
+      "grad_norm": 0.16626104712486267,
+      "learning_rate": 2.2091057987477944e-05,
+      "loss": 0.8784,
+      "step": 29470
+    },
+    {
+      "epoch": 0.940080997480787,
+      "grad_norm": 0.16839087009429932,
+      "learning_rate": 2.1998562368856864e-05,
+      "loss": 0.8781,
+      "step": 29480
+    },
+    {
+      "epoch": 0.940399885200421,
+      "grad_norm": 0.1720423549413681,
+      "learning_rate": 2.190645403089337e-05,
+      "loss": 0.8778,
+      "step": 29490
+    },
+    {
+      "epoch": 0.9407187729200549,
+      "grad_norm": 0.17319394648075104,
+      "learning_rate": 2.181473135203705e-05,
+      "loss": 0.8832,
+      "step": 29500
+    },
+    {
+      "epoch": 0.9410376606396887,
+      "grad_norm": 0.16802969574928284,
+      "learning_rate": 2.172339271752697e-05,
+      "loss": 0.8736,
+      "step": 29510
+    },
+    {
+      "epoch": 0.9413565483593227,
+      "grad_norm": 0.17481163144111633,
+      "learning_rate": 2.1632436519363203e-05,
+      "loss": 0.8721,
+      "step": 29520
+    },
+    {
+      "epoch": 0.9416754360789567,
+      "grad_norm": 0.1747370958328247,
+      "learning_rate": 2.1541861156278552e-05,
+      "loss": 0.8597,
+      "step": 29530
+    },
+    {
+      "epoch": 0.9419943237985905,
+      "grad_norm": 0.17296218872070312,
+      "learning_rate": 2.145166503371037e-05,
+      "loss": 0.8858,
+      "step": 29540
+    },
+    {
+      "epoch": 0.9423132115182244,
+      "grad_norm": 0.1767471730709076,
+      "learning_rate": 2.1361846563772446e-05,
+      "loss": 0.8844,
+      "step": 29550
+    },
+    {
+      "epoch": 0.9426320992378584,
+      "grad_norm": 0.16958440840244293,
+      "learning_rate": 2.127240416522709e-05,
+      "loss": 0.8722,
+      "step": 29560
+    },
+    {
+      "epoch": 0.9429509869574922,
+      "grad_norm": 0.17238831520080566,
+      "learning_rate": 2.1183336263457258e-05,
+      "loss": 0.8849,
+      "step": 29570
+    },
+    {
+      "epoch": 0.9432698746771262,
+      "grad_norm": 0.17094658315181732,
+      "learning_rate": 2.109464129043888e-05,
+      "loss": 0.8672,
+      "step": 29580
+    },
+    {
+      "epoch": 0.9435887623967602,
+      "grad_norm": 0.17316597700119019,
+      "learning_rate": 2.100631768471321e-05,
+      "loss": 0.8698,
+      "step": 29590
+    },
+    {
+      "epoch": 0.943907650116394,
+      "grad_norm": 0.17091426253318787,
+      "learning_rate": 2.0918363891359365e-05,
+      "loss": 0.8863,
+      "step": 29600
+    },
+    {
+      "epoch": 0.944226537836028,
+      "grad_norm": 0.17296843230724335,
+      "learning_rate": 2.0830778361966933e-05,
+      "loss": 0.8859,
+      "step": 29610
+    },
+    {
+      "epoch": 0.9445454255556619,
+      "grad_norm": 0.16335801780223846,
+      "learning_rate": 2.0743559554608734e-05,
+      "loss": 0.8889,
+      "step": 29620
+    },
+    {
+      "epoch": 0.9448643132752957,
+      "grad_norm": 0.17202697694301605,
+      "learning_rate": 2.0656705933813646e-05,
+      "loss": 0.8884,
+      "step": 29630
+    },
+    {
+      "epoch": 0.9451832009949297,
+      "grad_norm": 0.1737438142299652,
+      "learning_rate": 2.05702159705396e-05,
+      "loss": 0.8747,
+      "step": 29640
+    },
+    {
+      "epoch": 0.9455020887145636,
+      "grad_norm": 0.16886982321739197,
+      "learning_rate": 2.0484088142146646e-05,
+      "loss": 0.8689,
+      "step": 29650
+    },
+    {
+      "epoch": 0.9458209764341975,
+      "grad_norm": 0.1703483760356903,
+      "learning_rate": 2.039832093237016e-05,
+      "loss": 0.8752,
+      "step": 29660
+    },
+    {
+      "epoch": 0.9461398641538314,
+      "grad_norm": 0.17107826471328735,
+      "learning_rate": 2.0312912831294133e-05,
+      "loss": 0.8907,
+      "step": 29670
+    },
+    {
+      "epoch": 0.9464587518734654,
+      "grad_norm": 0.17108580470085144,
+      "learning_rate": 2.022786233532461e-05,
+      "loss": 0.8825,
+      "step": 29680
+    },
+    {
+      "epoch": 0.9467776395930992,
+      "grad_norm": 0.17031945288181305,
+      "learning_rate": 2.014316794716319e-05,
+      "loss": 0.8757,
+      "step": 29690
+    },
+    {
+      "epoch": 0.9470965273127332,
+      "grad_norm": 0.1722126007080078,
+      "learning_rate": 2.0058828175780694e-05,
+      "loss": 0.8735,
+      "step": 29700
+    },
+    {
+      "epoch": 0.9474154150323671,
+      "grad_norm": 0.16917821764945984,
+      "learning_rate": 1.9974841536390925e-05,
+      "loss": 0.8595,
+      "step": 29710
+    },
+    {
+      "epoch": 0.947734302752001,
+      "grad_norm": 0.1698334962129593,
+      "learning_rate": 1.989120655042449e-05,
+      "loss": 0.8678,
+      "step": 29720
+    },
+    {
+      "epoch": 0.9480531904716349,
+      "grad_norm": 0.17627465724945068,
+      "learning_rate": 1.9807921745502785e-05,
+      "loss": 0.8899,
+      "step": 29730
+    },
+    {
+      "epoch": 0.9483720781912689,
+      "grad_norm": 0.17068272829055786,
+      "learning_rate": 1.972498565541209e-05,
+      "loss": 0.8645,
+      "step": 29740
+    },
+    {
+      "epoch": 0.9486909659109027,
+      "grad_norm": 0.17030967772006989,
+      "learning_rate": 1.964239682007775e-05,
+      "loss": 0.8811,
+      "step": 29750
+    },
+    {
+      "epoch": 0.9490098536305367,
+      "grad_norm": 0.1698002964258194,
+      "learning_rate": 1.956015378553845e-05,
+      "loss": 0.8813,
+      "step": 29760
+    },
+    {
+      "epoch": 0.9493287413501706,
+      "grad_norm": 0.17307348549365997,
+      "learning_rate": 1.947825510392065e-05,
+      "loss": 0.8724,
+      "step": 29770
+    },
+    {
+      "epoch": 0.9496476290698045,
+      "grad_norm": 0.1735437661409378,
+      "learning_rate": 1.939669933341307e-05,
+      "loss": 0.8746,
+      "step": 29780
+    },
+    {
+      "epoch": 0.9499665167894384,
+      "grad_norm": 0.1740484982728958,
+      "learning_rate": 1.9315485038241333e-05,
+      "loss": 0.8795,
+      "step": 29790
+    },
+    {
+      "epoch": 0.9502854045090724,
+      "grad_norm": 0.1707371026277542,
+      "learning_rate": 1.9234610788642647e-05,
+      "loss": 0.8771,
+      "step": 29800
+    },
+    {
+      "epoch": 0.9506042922287062,
+      "grad_norm": 0.1706804782152176,
+      "learning_rate": 1.9154075160840683e-05,
+      "loss": 0.8634,
+      "step": 29810
+    },
+    {
+      "epoch": 0.9509231799483402,
+      "grad_norm": 0.17211967706680298,
+      "learning_rate": 1.907387673702047e-05,
+      "loss": 0.8853,
+      "step": 29820
+    },
+    {
+      "epoch": 0.9512420676679741,
+      "grad_norm": 0.16938042640686035,
+      "learning_rate": 1.8994014105303468e-05,
+      "loss": 0.8897,
+      "step": 29830
+    },
+    {
+      "epoch": 0.951560955387608,
+      "grad_norm": 0.17498505115509033,
+      "learning_rate": 1.8914485859722682e-05,
+      "loss": 0.8851,
+      "step": 29840
+    },
+    {
+      "epoch": 0.9518798431072419,
+      "grad_norm": 0.1684788018465042,
+      "learning_rate": 1.8835290600197926e-05,
+      "loss": 0.8822,
+      "step": 29850
+    },
+    {
+      "epoch": 0.9521987308268759,
+      "grad_norm": 0.1729750782251358,
+      "learning_rate": 1.8756426932511182e-05,
+      "loss": 0.8712,
+      "step": 29860
+    },
+    {
+      "epoch": 0.9525176185465097,
+      "grad_norm": 0.1740708351135254,
+      "learning_rate": 1.8677893468282023e-05,
+      "loss": 0.8855,
+      "step": 29870
+    },
+    {
+      "epoch": 0.9528365062661437,
+      "grad_norm": 0.1664377897977829,
+      "learning_rate": 1.8599688824943227e-05,
+      "loss": 0.8728,
+      "step": 29880
+    },
+    {
+      "epoch": 0.9531553939857776,
+      "grad_norm": 0.17212225496768951,
+      "learning_rate": 1.852181162571638e-05,
+      "loss": 0.8684,
+      "step": 29890
+    },
+    {
+      "epoch": 0.9534742817054115,
+      "grad_norm": 0.17010776698589325,
+      "learning_rate": 1.8444260499587665e-05,
+      "loss": 0.8701,
+      "step": 29900
+    },
+    {
+      "epoch": 0.9537931694250454,
+      "grad_norm": 0.1705661565065384,
+      "learning_rate": 1.8367034081283716e-05,
+      "loss": 0.8809,
+      "step": 29910
+    },
+    {
+      "epoch": 0.9541120571446794,
+      "grad_norm": 0.1687018722295761,
+      "learning_rate": 1.829013101124761e-05,
+      "loss": 0.8812,
+      "step": 29920
+    },
+    {
+      "epoch": 0.9544309448643132,
+      "grad_norm": 0.16721801459789276,
+      "learning_rate": 1.8213549935614886e-05,
+      "loss": 0.8739,
+      "step": 29930
+    },
+    {
+      "epoch": 0.9547498325839472,
+      "grad_norm": 0.1781189888715744,
+      "learning_rate": 1.8137289506189752e-05,
+      "loss": 0.8778,
+      "step": 29940
+    },
+    {
+      "epoch": 0.9550687203035811,
+      "grad_norm": 0.17269571125507355,
+      "learning_rate": 1.806134838042133e-05,
+      "loss": 0.9006,
+      "step": 29950
+    },
+    {
+      "epoch": 0.955387608023215,
+      "grad_norm": 0.17672085762023926,
+      "learning_rate": 1.7985725221380017e-05,
+      "loss": 0.8752,
+      "step": 29960
+    },
+    {
+      "epoch": 0.9557064957428489,
+      "grad_norm": 0.1679825484752655,
+      "learning_rate": 1.7910418697733964e-05,
+      "loss": 0.8853,
+      "step": 29970
+    },
+    {
+      "epoch": 0.9560253834624829,
+      "grad_norm": 0.16887645423412323,
+      "learning_rate": 1.7835427483725635e-05,
+      "loss": 0.8723,
+      "step": 29980
+    },
+    {
+      "epoch": 0.9563442711821167,
+      "grad_norm": 0.17646433413028717,
+      "learning_rate": 1.7760750259148444e-05,
+      "loss": 0.8949,
+      "step": 29990
+    },
+    {
+      "epoch": 0.9566631589017507,
+      "grad_norm": 0.17680183053016663,
+      "learning_rate": 1.7686385709323553e-05,
+      "loss": 0.8886,
+      "step": 30000
+    },
+    {
+      "epoch": 0.9569820466213846,
+      "grad_norm": 0.1752249002456665,
+      "learning_rate": 1.7612332525076688e-05,
+      "loss": 0.8928,
+      "step": 30010
+    },
+    {
+      "epoch": 0.9573009343410185,
+      "grad_norm": 0.17468760907649994,
+      "learning_rate": 1.7538589402715118e-05,
+      "loss": 0.8648,
+      "step": 30020
+    },
+    {
+      "epoch": 0.9576198220606524,
+      "grad_norm": 0.16802415251731873,
+      "learning_rate": 1.7465155044004687e-05,
+      "loss": 0.8745,
+      "step": 30030
+    },
+    {
+      "epoch": 0.9579387097802864,
+      "grad_norm": 0.1705239713191986,
+      "learning_rate": 1.7392028156146985e-05,
+      "loss": 0.8756,
+      "step": 30040
+    },
+    {
+      "epoch": 0.9582575974999202,
+      "grad_norm": 0.17179515957832336,
+      "learning_rate": 1.7319207451756553e-05,
+      "loss": 0.8656,
+      "step": 30050
+    },
+    {
+      "epoch": 0.9585764852195542,
+      "grad_norm": 0.17518065869808197,
+      "learning_rate": 1.724669164883824e-05,
+      "loss": 0.8944,
+      "step": 30060
+    },
+    {
+      "epoch": 0.9588953729391881,
+      "grad_norm": 0.1713200956583023,
+      "learning_rate": 1.7174479470764624e-05,
+      "loss": 0.8925,
+      "step": 30070
+    },
+    {
+      "epoch": 0.959214260658822,
+      "grad_norm": 0.17048819363117218,
+      "learning_rate": 1.7102569646253555e-05,
+      "loss": 0.874,
+      "step": 30080
+    },
+    {
+      "epoch": 0.9595331483784559,
+      "grad_norm": 0.1709250658750534,
+      "learning_rate": 1.7030960909345757e-05,
+      "loss": 0.8777,
+      "step": 30090
+    },
+    {
+      "epoch": 0.9598520360980899,
+      "grad_norm": 0.1756034642457962,
+      "learning_rate": 1.6959651999382535e-05,
+      "loss": 0.8848,
+      "step": 30100
+    },
+    {
+      "epoch": 0.9601709238177237,
+      "grad_norm": 0.171064093708992,
+      "learning_rate": 1.688864166098361e-05,
+      "loss": 0.8829,
+      "step": 30110
+    },
+    {
+      "epoch": 0.9604898115373577,
+      "grad_norm": 0.1718042641878128,
+      "learning_rate": 1.6817928644024988e-05,
+      "loss": 0.8746,
+      "step": 30120
+    },
+    {
+      "epoch": 0.9608086992569916,
+      "grad_norm": 0.16927538812160492,
+      "learning_rate": 1.6747511703616973e-05,
+      "loss": 0.8656,
+      "step": 30130
+    },
+    {
+      "epoch": 0.9611275869766255,
+      "grad_norm": 0.17567414045333862,
+      "learning_rate": 1.6677389600082237e-05,
+      "loss": 0.8831,
+      "step": 30140
+    },
+    {
+      "epoch": 0.9614464746962594,
+      "grad_norm": 0.16869021952152252,
+      "learning_rate": 1.6607561098934e-05,
+      "loss": 0.8792,
+      "step": 30150
+    },
+    {
+      "epoch": 0.9617653624158934,
+      "grad_norm": 0.17464403808116913,
+      "learning_rate": 1.6538024970854306e-05,
+      "loss": 0.8719,
+      "step": 30160
+    },
+    {
+      "epoch": 0.9620842501355272,
+      "grad_norm": 0.1718774139881134,
+      "learning_rate": 1.646877999167236e-05,
+      "loss": 0.8693,
+      "step": 30170
+    },
+    {
+      "epoch": 0.9624031378551612,
+      "grad_norm": 0.16808542609214783,
+      "learning_rate": 1.639982494234301e-05,
+      "loss": 0.8805,
+      "step": 30180
+    },
+    {
+      "epoch": 0.9627220255747951,
+      "grad_norm": 0.17283496260643005,
+      "learning_rate": 1.633115860892524e-05,
+      "loss": 0.8868,
+      "step": 30190
+    },
+    {
+      "epoch": 0.963040913294429,
+      "grad_norm": 0.17048844695091248,
+      "learning_rate": 1.6262779782560837e-05,
+      "loss": 0.8721,
+      "step": 30200
+    },
+    {
+      "epoch": 0.9633598010140629,
+      "grad_norm": 0.17138278484344482,
+      "learning_rate": 1.6194687259453118e-05,
+      "loss": 0.8829,
+      "step": 30210
+    },
+    {
+      "epoch": 0.9636786887336969,
+      "grad_norm": 0.1714802086353302,
+      "learning_rate": 1.6126879840845686e-05,
+      "loss": 0.8881,
+      "step": 30220
+    },
+    {
+      "epoch": 0.9639975764533307,
+      "grad_norm": 0.17534974217414856,
+      "learning_rate": 1.6059356333001364e-05,
+      "loss": 0.8867,
+      "step": 30230
+    },
+    {
+      "epoch": 0.9643164641729647,
+      "grad_norm": 0.17431078851222992,
+      "learning_rate": 1.599211554718118e-05,
+      "loss": 0.8848,
+      "step": 30240
+    },
+    {
+      "epoch": 0.9646353518925986,
+      "grad_norm": 0.16717074811458588,
+      "learning_rate": 1.5925156299623422e-05,
+      "loss": 0.8569,
+      "step": 30250
+    },
+    {
+      "epoch": 0.9649542396122325,
+      "grad_norm": 0.17699192464351654,
+      "learning_rate": 1.5858477411522817e-05,
+      "loss": 0.8696,
+      "step": 30260
+    },
+    {
+      "epoch": 0.9652731273318664,
+      "grad_norm": 0.1727645993232727,
+      "learning_rate": 1.5792077709009763e-05,
+      "loss": 0.8755,
+      "step": 30270
+    },
+    {
+      "epoch": 0.9655920150515004,
+      "grad_norm": 0.1745349019765854,
+      "learning_rate": 1.5725956023129666e-05,
+      "loss": 0.8632,
+      "step": 30280
+    },
+    {
+      "epoch": 0.9659109027711343,
+      "grad_norm": 0.17108358442783356,
+      "learning_rate": 1.566011118982237e-05,
+      "loss": 0.8732,
+      "step": 30290
+    },
+    {
+      "epoch": 0.9662297904907682,
+      "grad_norm": 0.17055270075798035,
+      "learning_rate": 1.559454204990166e-05,
+      "loss": 0.887,
+      "step": 30300
+    },
+    {
+      "epoch": 0.9665486782104021,
+      "grad_norm": 0.17470329999923706,
+      "learning_rate": 1.5529247449034844e-05,
+      "loss": 0.8752,
+      "step": 30310
+    },
+    {
+      "epoch": 0.9668675659300361,
+      "grad_norm": 0.17025306820869446,
+      "learning_rate": 1.5464226237722444e-05,
+      "loss": 0.878,
+      "step": 30320
+    },
+    {
+      "epoch": 0.9671864536496699,
+      "grad_norm": 0.1691192239522934,
+      "learning_rate": 1.5399477271277955e-05,
+      "loss": 0.8757,
+      "step": 30330
+    },
+    {
+      "epoch": 0.9675053413693039,
+      "grad_norm": 0.17150309681892395,
+      "learning_rate": 1.5334999409807695e-05,
+      "loss": 0.8994,
+      "step": 30340
+    },
+    {
+      "epoch": 0.9678242290889378,
+      "grad_norm": 0.1817106157541275,
+      "learning_rate": 1.5270791518190734e-05,
+      "loss": 0.8946,
+      "step": 30350
+    },
+    {
+      "epoch": 0.9681431168085717,
+      "grad_norm": 0.17197778820991516,
+      "learning_rate": 1.5206852466058905e-05,
+      "loss": 0.881,
+      "step": 30360
+    },
+    {
+      "epoch": 0.9684620045282056,
+      "grad_norm": 0.17225344479084015,
+      "learning_rate": 1.5143181127776916e-05,
+      "loss": 0.8959,
+      "step": 30370
+    },
+    {
+      "epoch": 0.9687808922478396,
+      "grad_norm": 0.1728435456752777,
+      "learning_rate": 1.5079776382422542e-05,
+      "loss": 0.8694,
+      "step": 30380
+    },
+    {
+      "epoch": 0.9690997799674734,
+      "grad_norm": 0.17408286035060883,
+      "learning_rate": 1.5016637113766854e-05,
+      "loss": 0.8748,
+      "step": 30390
+    },
+    {
+      "epoch": 0.9694186676871074,
+      "grad_norm": 0.16729173064231873,
+      "learning_rate": 1.495376221025461e-05,
+      "loss": 0.888,
+      "step": 30400
+    },
+    {
+      "epoch": 0.9697375554067413,
+      "grad_norm": 0.17172005772590637,
+      "learning_rate": 1.4891150564984654e-05,
+      "loss": 0.8711,
+      "step": 30410
+    },
+    {
+      "epoch": 0.9700564431263752,
+      "grad_norm": 0.17462344467639923,
+      "learning_rate": 1.482880107569045e-05,
+      "loss": 0.8854,
+      "step": 30420
+    },
+    {
+      "epoch": 0.9703753308460091,
+      "grad_norm": 0.17063672840595245,
+      "learning_rate": 1.4766712644720671e-05,
+      "loss": 0.8837,
+      "step": 30430
+    },
+    {
+      "epoch": 0.9706942185656431,
+      "grad_norm": 0.170479878783226,
+      "learning_rate": 1.4704884179019873e-05,
+      "loss": 0.8683,
+      "step": 30440
+    },
+    {
+      "epoch": 0.9710131062852769,
+      "grad_norm": 0.165556862950325,
+      "learning_rate": 1.464331459010925e-05,
+      "loss": 0.8834,
+      "step": 30450
+    },
+    {
+      "epoch": 0.9713319940049109,
+      "grad_norm": 0.16775330901145935,
+      "learning_rate": 1.4582002794067476e-05,
+      "loss": 0.8865,
+      "step": 30460
+    },
+    {
+      "epoch": 0.9716508817245448,
+      "grad_norm": 0.16928617656230927,
+      "learning_rate": 1.4520947711511627e-05,
+      "loss": 0.8902,
+      "step": 30470
+    },
+    {
+      "epoch": 0.9719697694441787,
+      "grad_norm": 0.17188507318496704,
+      "learning_rate": 1.446014826757816e-05,
+      "loss": 0.8896,
+      "step": 30480
+    },
+    {
+      "epoch": 0.9722886571638126,
+      "grad_norm": 0.17368733882904053,
+      "learning_rate": 1.4399603391904017e-05,
+      "loss": 0.8858,
+      "step": 30490
+    },
+    {
+      "epoch": 0.9726075448834466,
+      "grad_norm": 0.1663048416376114,
+      "learning_rate": 1.4339312018607758e-05,
+      "loss": 0.881,
+      "step": 30500
+    },
+    {
+      "epoch": 0.9729264326030804,
+      "grad_norm": 0.17066152393817902,
+      "learning_rate": 1.4279273086270809e-05,
+      "loss": 0.8813,
+      "step": 30510
+    },
+    {
+      "epoch": 0.9732453203227144,
+      "grad_norm": 0.17235307395458221,
+      "learning_rate": 1.4219485537918775e-05,
+      "loss": 0.8846,
+      "step": 30520
+    },
+    {
+      "epoch": 0.9735642080423483,
+      "grad_norm": 0.17622828483581543,
+      "learning_rate": 1.4159948321002827e-05,
+      "loss": 0.8801,
+      "step": 30530
+    },
+    {
+      "epoch": 0.9738830957619822,
+      "grad_norm": 0.17217357456684113,
+      "learning_rate": 1.4100660387381168e-05,
+      "loss": 0.8739,
+      "step": 30540
+    },
+    {
+      "epoch": 0.9742019834816161,
+      "grad_norm": 0.1750701367855072,
+      "learning_rate": 1.4041620693300609e-05,
+      "loss": 0.887,
+      "step": 30550
+    },
+    {
+      "epoch": 0.9745208712012501,
+      "grad_norm": 0.17289851605892181,
+      "learning_rate": 1.3982828199378157e-05,
+      "loss": 0.8878,
+      "step": 30560
+    },
+    {
+      "epoch": 0.9748397589208839,
+      "grad_norm": 0.1716613620519638,
+      "learning_rate": 1.3924281870582728e-05,
+      "loss": 0.8808,
+      "step": 30570
+    },
+    {
+      "epoch": 0.9751586466405179,
+      "grad_norm": 0.17046386003494263,
+      "learning_rate": 1.3865980676216942e-05,
+      "loss": 0.8796,
+      "step": 30580
+    },
+    {
+      "epoch": 0.9754775343601518,
+      "grad_norm": 0.16988524794578552,
+      "learning_rate": 1.3807923589898957e-05,
+      "loss": 0.8868,
+      "step": 30590
+    },
+    {
+      "epoch": 0.9757964220797857,
+      "grad_norm": 0.1751394271850586,
+      "learning_rate": 1.3750109589544415e-05,
+      "loss": 0.9025,
+      "step": 30600
+    },
+    {
+      "epoch": 0.9761153097994196,
+      "grad_norm": 0.16379600763320923,
+      "learning_rate": 1.3692537657348436e-05,
+      "loss": 0.8785,
+      "step": 30610
+    },
+    {
+      "epoch": 0.9764341975190536,
+      "grad_norm": 0.1647310107946396,
+      "learning_rate": 1.3635206779767707e-05,
+      "loss": 0.8775,
+      "step": 30620
+    },
+    {
+      "epoch": 0.9767530852386874,
+      "grad_norm": 0.1679423302412033,
+      "learning_rate": 1.3578115947502643e-05,
+      "loss": 0.8741,
+      "step": 30630
+    },
+    {
+      "epoch": 0.9770719729583214,
+      "grad_norm": 0.17554236948490143,
+      "learning_rate": 1.3521264155479603e-05,
+      "loss": 0.8844,
+      "step": 30640
+    },
+    {
+      "epoch": 0.9773908606779553,
+      "grad_norm": 0.17117299139499664,
+      "learning_rate": 1.3464650402833215e-05,
+      "loss": 0.8864,
+      "step": 30650
+    },
+    {
+      "epoch": 0.9777097483975892,
+      "grad_norm": 0.17434369027614594,
+      "learning_rate": 1.3408273692888739e-05,
+      "loss": 0.8729,
+      "step": 30660
+    },
+    {
+      "epoch": 0.9780286361172231,
+      "grad_norm": 0.17379888892173767,
+      "learning_rate": 1.3352133033144537e-05,
+      "loss": 0.8756,
+      "step": 30670
+    },
+    {
+      "epoch": 0.9783475238368571,
+      "grad_norm": 0.17442287504673004,
+      "learning_rate": 1.3296227435254582e-05,
+      "loss": 0.8775,
+      "step": 30680
+    },
+    {
+      "epoch": 0.9786664115564909,
+      "grad_norm": 0.17200325429439545,
+      "learning_rate": 1.3240555915011072e-05,
+      "loss": 0.8863,
+      "step": 30690
+    },
+    {
+      "epoch": 0.9789852992761249,
+      "grad_norm": 0.16841526329517365,
+      "learning_rate": 1.31851174923271e-05,
+      "loss": 0.8805,
+      "step": 30700
+    },
+    {
+      "epoch": 0.9793041869957588,
+      "grad_norm": 0.16977955400943756,
+      "learning_rate": 1.3129911191219392e-05,
+      "loss": 0.889,
+      "step": 30710
+    },
+    {
+      "epoch": 0.9796230747153927,
+      "grad_norm": 0.1685485690832138,
+      "learning_rate": 1.307493603979115e-05,
+      "loss": 0.8745,
+      "step": 30720
+    },
+    {
+      "epoch": 0.9799419624350266,
+      "grad_norm": 0.16957922279834747,
+      "learning_rate": 1.3020191070214908e-05,
+      "loss": 0.8651,
+      "step": 30730
+    },
+    {
+      "epoch": 0.9802608501546606,
+      "grad_norm": 0.16795560717582703,
+      "learning_rate": 1.2965675318715508e-05,
+      "loss": 0.8608,
+      "step": 30740
+    },
+    {
+      "epoch": 0.9805797378742944,
+      "grad_norm": 0.17047218978405,
+      "learning_rate": 1.2911387825553141e-05,
+      "loss": 0.8793,
+      "step": 30750
+    },
+    {
+      "epoch": 0.9808986255939284,
+      "grad_norm": 0.1741809844970703,
+      "learning_rate": 1.2857327635006445e-05,
+      "loss": 0.8855,
+      "step": 30760
+    },
+    {
+      "epoch": 0.9812175133135623,
+      "grad_norm": 0.17339977622032166,
+      "learning_rate": 1.2803493795355673e-05,
+      "loss": 0.8727,
+      "step": 30770
+    },
+    {
+      "epoch": 0.9815364010331962,
+      "grad_norm": 0.16938899457454681,
+      "learning_rate": 1.2749885358865947e-05,
+      "loss": 0.8811,
+      "step": 30780
+    },
+    {
+      "epoch": 0.9818552887528301,
+      "grad_norm": 0.17674767971038818,
+      "learning_rate": 1.269650138177057e-05,
+      "loss": 0.8734,
+      "step": 30790
+    },
+    {
+      "epoch": 0.9821741764724641,
+      "grad_norm": 0.17623662948608398,
+      "learning_rate": 1.2643340924254416e-05,
+      "loss": 0.869,
+      "step": 30800
+    },
+    {
+      "epoch": 0.9824930641920979,
+      "grad_norm": 0.16907668113708496,
+      "learning_rate": 1.2590403050437372e-05,
+      "loss": 0.8666,
+      "step": 30810
+    },
+    {
+      "epoch": 0.9828119519117319,
+      "grad_norm": 0.17110703885555267,
+      "learning_rate": 1.2537686828357875e-05,
+      "loss": 0.8691,
+      "step": 30820
+    },
+    {
+      "epoch": 0.9831308396313658,
+      "grad_norm": 0.16781097650527954,
+      "learning_rate": 1.2485191329956501e-05,
+      "loss": 0.871,
+      "step": 30830
+    },
+    {
+      "epoch": 0.9834497273509997,
+      "grad_norm": 0.16678723692893982,
+      "learning_rate": 1.2432915631059624e-05,
+      "loss": 0.8652,
+      "step": 30840
+    },
+    {
+      "epoch": 0.9837686150706336,
+      "grad_norm": 0.17238172888755798,
+      "learning_rate": 1.2380858811363149e-05,
+      "loss": 0.8761,
+      "step": 30850
+    },
+    {
+      "epoch": 0.9840875027902676,
+      "grad_norm": 0.1715608835220337,
+      "learning_rate": 1.2329019954416307e-05,
+      "loss": 0.8715,
+      "step": 30860
+    },
+    {
+      "epoch": 0.9844063905099014,
+      "grad_norm": 0.16758796572685242,
+      "learning_rate": 1.2277398147605525e-05,
+      "loss": 0.8706,
+      "step": 30870
+    },
+    {
+      "epoch": 0.9847252782295354,
+      "grad_norm": 0.17507082223892212,
+      "learning_rate": 1.2225992482138354e-05,
+      "loss": 0.8715,
+      "step": 30880
+    },
+    {
+      "epoch": 0.9850441659491693,
+      "grad_norm": 0.17221605777740479,
+      "learning_rate": 1.2174802053027486e-05,
+      "loss": 0.8773,
+      "step": 30890
+    },
+    {
+      "epoch": 0.9853630536688032,
+      "grad_norm": 0.1673084944486618,
+      "learning_rate": 1.2123825959074799e-05,
+      "loss": 0.8777,
+      "step": 30900
+    },
+    {
+      "epoch": 0.9856819413884371,
+      "grad_norm": 0.17324191331863403,
+      "learning_rate": 1.2073063302855502e-05,
+      "loss": 0.8656,
+      "step": 30910
+    },
+    {
+      "epoch": 0.9860008291080711,
+      "grad_norm": 0.1690828800201416,
+      "learning_rate": 1.2022513190702337e-05,
+      "loss": 0.879,
+      "step": 30920
+    },
+    {
+      "epoch": 0.9863197168277049,
+      "grad_norm": 0.17445652186870575,
+      "learning_rate": 1.1972174732689848e-05,
+      "loss": 0.8715,
+      "step": 30930
+    },
+    {
+      "epoch": 0.9866386045473389,
+      "grad_norm": 0.1705475151538849,
+      "learning_rate": 1.1922047042618712e-05,
+      "loss": 0.8645,
+      "step": 30940
+    },
+    {
+      "epoch": 0.9869574922669728,
+      "grad_norm": 0.17571979761123657,
+      "learning_rate": 1.1872129238000134e-05,
+      "loss": 0.8752,
+      "step": 30950
+    },
+    {
+      "epoch": 0.9872763799866067,
+      "grad_norm": 0.17660577595233917,
+      "learning_rate": 1.1822420440040315e-05,
+      "loss": 0.8826,
+      "step": 30960
+    },
+    {
+      "epoch": 0.9875952677062406,
+      "grad_norm": 0.1759461760520935,
+      "learning_rate": 1.1772919773624978e-05,
+      "loss": 0.8774,
+      "step": 30970
+    },
+    {
+      "epoch": 0.9879141554258746,
+      "grad_norm": 0.17794868350028992,
+      "learning_rate": 1.1723626367303966e-05,
+      "loss": 0.8807,
+      "step": 30980
+    },
+    {
+      "epoch": 0.9882330431455084,
+      "grad_norm": 0.1735115349292755,
+      "learning_rate": 1.1674539353275899e-05,
+      "loss": 0.8959,
+      "step": 30990
+    },
+    {
+      "epoch": 0.9885519308651424,
+      "grad_norm": 0.16572922468185425,
+      "learning_rate": 1.1625657867372895e-05,
+      "loss": 0.8712,
+      "step": 31000
+    },
+    {
+      "epoch": 0.9888708185847763,
+      "grad_norm": 0.17459288239479065,
+      "learning_rate": 1.1576981049045353e-05,
+      "loss": 0.9071,
+      "step": 31010
+    },
+    {
+      "epoch": 0.9891897063044102,
+      "grad_norm": 0.17562910914421082,
+      "learning_rate": 1.1528508041346812e-05,
+      "loss": 0.875,
+      "step": 31020
+    },
+    {
+      "epoch": 0.9895085940240441,
+      "grad_norm": 0.16303937137126923,
+      "learning_rate": 1.1480237990918854e-05,
+      "loss": 0.8776,
+      "step": 31030
+    },
+    {
+      "epoch": 0.9898274817436781,
+      "grad_norm": 0.17524918913841248,
+      "learning_rate": 1.1432170047976097e-05,
+      "loss": 0.8894,
+      "step": 31040
+    },
+    {
+      "epoch": 0.9901463694633119,
+      "grad_norm": 0.1762991100549698,
+      "learning_rate": 1.1384303366291208e-05,
+      "loss": 0.8783,
+      "step": 31050
+    },
+    {
+      "epoch": 0.9904652571829459,
+      "grad_norm": 0.1695641428232193,
+      "learning_rate": 1.1336637103180043e-05,
+      "loss": 0.8692,
+      "step": 31060
+    },
+    {
+      "epoch": 0.9907841449025798,
+      "grad_norm": 0.17893773317337036,
+      "learning_rate": 1.1289170419486774e-05,
+      "loss": 0.8776,
+      "step": 31070
+    },
+    {
+      "epoch": 0.9911030326222137,
+      "grad_norm": 0.17522774636745453,
+      "learning_rate": 1.1241902479569133e-05,
+      "loss": 0.871,
+      "step": 31080
+    },
+    {
+      "epoch": 0.9914219203418476,
+      "grad_norm": 0.17226657271385193,
+      "learning_rate": 1.1194832451283707e-05,
+      "loss": 0.8735,
+      "step": 31090
+    },
+    {
+      "epoch": 0.9917408080614816,
+      "grad_norm": 0.17155729234218597,
+      "learning_rate": 1.1147959505971274e-05,
+      "loss": 0.877,
+      "step": 31100
+    },
+    {
+      "epoch": 0.9920596957811154,
+      "grad_norm": 0.16582393646240234,
+      "learning_rate": 1.110128281844223e-05,
+      "loss": 0.859,
+      "step": 31110
+    },
+    {
+      "epoch": 0.9923785835007494,
+      "grad_norm": 0.1682422012090683,
+      "learning_rate": 1.1054801566962041e-05,
+      "loss": 0.8771,
+      "step": 31120
+    },
+    {
+      "epoch": 0.9926974712203833,
+      "grad_norm": 0.17062459886074066,
+      "learning_rate": 1.1008514933236803e-05,
+      "loss": 0.8733,
+      "step": 31130
+    },
+    {
+      "epoch": 0.9930163589400173,
+      "grad_norm": 0.17315512895584106,
+      "learning_rate": 1.096242210239881e-05,
+      "loss": 0.8878,
+      "step": 31140
+    },
+    {
+      "epoch": 0.9933352466596511,
+      "grad_norm": 0.1766100823879242,
+      "learning_rate": 1.0916522262992225e-05,
+      "loss": 0.8758,
+      "step": 31150
+    },
+    {
+      "epoch": 0.9936541343792851,
+      "grad_norm": 0.17147080600261688,
+      "learning_rate": 1.0870814606958793e-05,
+      "loss": 0.8713,
+      "step": 31160
+    },
+    {
+      "epoch": 0.993973022098919,
+      "grad_norm": 0.1682249754667282,
+      "learning_rate": 1.0825298329623609e-05,
+      "loss": 0.8689,
+      "step": 31170
+    },
+    {
+      "epoch": 0.9942919098185529,
+      "grad_norm": 0.16981013119220734,
+      "learning_rate": 1.0779972629680952e-05,
+      "loss": 0.8667,
+      "step": 31180
+    },
+    {
+      "epoch": 0.9946107975381868,
+      "grad_norm": 0.17520365118980408,
+      "learning_rate": 1.0734836709180184e-05,
+      "loss": 0.869,
+      "step": 31190
+    },
+    {
+      "epoch": 0.9949296852578208,
+      "grad_norm": 0.17918860912322998,
+      "learning_rate": 1.0689889773511703e-05,
+      "loss": 0.8717,
+      "step": 31200
+    },
+    {
+      "epoch": 0.9952485729774546,
+      "grad_norm": 0.17659881711006165,
+      "learning_rate": 1.0645131031392937e-05,
+      "loss": 0.894,
+      "step": 31210
+    },
+    {
+      "epoch": 0.9955674606970886,
+      "grad_norm": 0.16800250113010406,
+      "learning_rate": 1.060055969485445e-05,
+      "loss": 0.8641,
+      "step": 31220
+    },
+    {
+      "epoch": 0.9958863484167225,
+      "grad_norm": 0.16605263948440552,
+      "learning_rate": 1.0556174979226025e-05,
+      "loss": 0.8679,
+      "step": 31230
+    },
+    {
+      "epoch": 0.9962052361363564,
+      "grad_norm": 0.1731094866991043,
+      "learning_rate": 1.0511976103122883e-05,
+      "loss": 0.8639,
+      "step": 31240
+    },
+    {
+      "epoch": 0.9965241238559903,
+      "grad_norm": 0.17676420509815216,
+      "learning_rate": 1.0467962288431913e-05,
+      "loss": 0.8796,
+      "step": 31250
+    },
+    {
+      "epoch": 0.9968430115756243,
+      "grad_norm": 0.1687876135110855,
+      "learning_rate": 1.0424132760297974e-05,
+      "loss": 0.8637,
+      "step": 31260
+    },
+    {
+      "epoch": 0.9971618992952581,
+      "grad_norm": 0.17514613270759583,
+      "learning_rate": 1.0380486747110261e-05,
+      "loss": 0.8864,
+      "step": 31270
+    },
+    {
+      "epoch": 0.9974807870148921,
+      "grad_norm": 0.17455033957958221,
+      "learning_rate": 1.033702348048871e-05,
+      "loss": 0.8689,
+      "step": 31280
+    },
+    {
+      "epoch": 0.997799674734526,
+      "grad_norm": 0.16740122437477112,
+      "learning_rate": 1.0293742195270484e-05,
+      "loss": 0.8631,
+      "step": 31290
+    },
+    {
+      "epoch": 0.9981185624541599,
+      "grad_norm": 0.17466361820697784,
+      "learning_rate": 1.0250642129496489e-05,
+      "loss": 0.8783,
+      "step": 31300
+    },
+    {
+      "epoch": 0.9984374501737938,
+      "grad_norm": 0.16998808085918427,
+      "learning_rate": 1.0207722524397968e-05,
+      "loss": 0.8581,
+      "step": 31310
+    },
+    {
+      "epoch": 0.9987563378934278,
+      "grad_norm": 0.16853420436382294,
+      "learning_rate": 1.0164982624383143e-05,
+      "loss": 0.8747,
+      "step": 31320
+    },
+    {
+      "epoch": 0.9990752256130616,
+      "grad_norm": 0.17115938663482666,
+      "learning_rate": 1.0122421677023911e-05,
+      "loss": 0.8842,
+      "step": 31330
+    },
+    {
+      "epoch": 0.9993941133326956,
+      "grad_norm": 0.16913767158985138,
+      "learning_rate": 1.0080038933042593e-05,
+      "loss": 0.8605,
+      "step": 31340
+    },
+    {
+      "epoch": 0.9997130010523295,
+      "grad_norm": 0.17436672747135162,
+      "learning_rate": 1.003783364629875e-05,
+      "loss": 0.8836,
+      "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": 1.1294004204869434e+18,
+  "train_batch_size": 512,
+  "trial_name": null,
+  "trial_params": null
+}