diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.3044371717786742,
+  "epoch": 0.6088743435573484,
   "eval_steps": 5000,
   "global_step": 5000,
   "is_hyper_param_search": false,
@@ -9,3516 +9,3516 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0006088743435573484,
-      "grad_norm": 1040.0,
-      "learning_rate": 7.9999238911415e-06,
-      "loss": 29.6613,
+      "epoch": 0.0012177486871146968,
+      "grad_norm": 469.75,
+      "learning_rate": 9.99980972785375e-06,
+      "loss": 26.2289,
       "step": 10
     },
     {
-      "epoch": 0.0012177486871146968,
-      "grad_norm": 547.5,
-      "learning_rate": 7.999847782282999e-06,
-      "loss": 27.1939,
+      "epoch": 0.0024354973742293936,
+      "grad_norm": 405.0,
+      "learning_rate": 9.999619455707499e-06,
+      "loss": 23.5619,
       "step": 20
     },
     {
-      "epoch": 0.001826623030672045,
-      "grad_norm": 661.0,
-      "learning_rate": 7.999771673424498e-06,
-      "loss": 27.143,
+      "epoch": 0.00365324606134409,
+      "grad_norm": 326.75,
+      "learning_rate": 9.999429183561249e-06,
+      "loss": 23.5219,
       "step": 30
     },
     {
-      "epoch": 0.0024354973742293936,
-      "grad_norm": 610.5,
-      "learning_rate": 7.999695564565998e-06,
-      "loss": 26.7644,
+      "epoch": 0.004870994748458787,
+      "grad_norm": 312.0,
+      "learning_rate": 9.999238911414998e-06,
+      "loss": 23.3801,
       "step": 40
     },
     {
-      "epoch": 0.0030443717177867416,
-      "grad_norm": 559.0,
-      "learning_rate": 7.999619455707497e-06,
-      "loss": 26.845,
+      "epoch": 0.006088743435573483,
+      "grad_norm": 379.75,
+      "learning_rate": 9.999048639268747e-06,
+      "loss": 23.3185,
       "step": 50
     },
     {
-      "epoch": 0.00365324606134409,
-      "grad_norm": 978.5,
-      "learning_rate": 7.999543346848999e-06,
-      "loss": 26.7081,
+      "epoch": 0.00730649212268818,
+      "grad_norm": 507.5,
+      "learning_rate": 9.998858367122496e-06,
+      "loss": 23.4722,
       "step": 60
     },
     {
-      "epoch": 0.004262120404901438,
-      "grad_norm": 1142.0,
-      "learning_rate": 7.999467237990496e-06,
-      "loss": 26.5517,
+      "epoch": 0.008524240809802876,
+      "grad_norm": 302.5,
+      "learning_rate": 9.998668094976245e-06,
+      "loss": 23.1901,
       "step": 70
     },
     {
-      "epoch": 0.004870994748458787,
-      "grad_norm": 652.0,
-      "learning_rate": 7.999391129131998e-06,
-      "loss": 26.164,
+      "epoch": 0.009741989496917574,
+      "grad_norm": 325.5,
+      "learning_rate": 9.998477822829995e-06,
+      "loss": 23.3673,
       "step": 80
     },
     {
-      "epoch": 0.005479869092016135,
-      "grad_norm": 522.0,
-      "learning_rate": 7.999315020273497e-06,
-      "loss": 26.3264,
+      "epoch": 0.01095973818403227,
+      "grad_norm": 375.25,
+      "learning_rate": 9.998287550683744e-06,
+      "loss": 23.4508,
       "step": 90
     },
     {
-      "epoch": 0.006088743435573483,
-      "grad_norm": 506.0,
-      "learning_rate": 7.999238911414997e-06,
-      "loss": 26.0035,
+      "epoch": 0.012177486871146967,
+      "grad_norm": 466.25,
+      "learning_rate": 9.998097278537493e-06,
+      "loss": 23.3656,
       "step": 100
     },
     {
-      "epoch": 0.006697617779130832,
-      "grad_norm": 680.0,
-      "learning_rate": 7.999162802556496e-06,
-      "loss": 26.426,
+      "epoch": 0.013395235558261664,
+      "grad_norm": 405.0,
+      "learning_rate": 9.997907006391242e-06,
+      "loss": 23.2151,
       "step": 110
     },
     {
-      "epoch": 0.00730649212268818,
-      "grad_norm": 391.25,
-      "learning_rate": 7.999086693697996e-06,
-      "loss": 26.5592,
+      "epoch": 0.01461298424537636,
+      "grad_norm": 327.5,
+      "learning_rate": 9.99771673424499e-06,
+      "loss": 23.481,
       "step": 120
     },
     {
-      "epoch": 0.007915366466245528,
-      "grad_norm": 529.0,
-      "learning_rate": 7.999010584839495e-06,
-      "loss": 26.5778,
+      "epoch": 0.015830732932491057,
+      "grad_norm": 598.5,
+      "learning_rate": 9.997526462098741e-06,
+      "loss": 23.4568,
       "step": 130
     },
     {
-      "epoch": 0.008524240809802876,
-      "grad_norm": 416.5,
-      "learning_rate": 7.998934475980995e-06,
-      "loss": 26.117,
+      "epoch": 0.017048481619605753,
+      "grad_norm": 479.0,
+      "learning_rate": 9.99733618995249e-06,
+      "loss": 23.4055,
       "step": 140
     },
     {
-      "epoch": 0.009133115153360225,
-      "grad_norm": 635.5,
-      "learning_rate": 7.998858367122495e-06,
-      "loss": 26.3732,
+      "epoch": 0.01826623030672045,
+      "grad_norm": 449.75,
+      "learning_rate": 9.997145917806239e-06,
+      "loss": 23.4494,
       "step": 150
     },
     {
-      "epoch": 0.009741989496917574,
-      "grad_norm": 547.5,
-      "learning_rate": 7.998782258263996e-06,
-      "loss": 26.0139,
+      "epoch": 0.01948397899383515,
+      "grad_norm": 643.5,
+      "learning_rate": 9.996955645659988e-06,
+      "loss": 23.3568,
       "step": 160
     },
     {
-      "epoch": 0.010350863840474922,
-      "grad_norm": 511.25,
-      "learning_rate": 7.998706149405494e-06,
-      "loss": 26.1682,
+      "epoch": 0.020701727680949845,
+      "grad_norm": 333.0,
+      "learning_rate": 9.996765373513737e-06,
+      "loss": 23.4843,
       "step": 170
     },
     {
-      "epoch": 0.01095973818403227,
-      "grad_norm": 435.5,
-      "learning_rate": 7.998630040546995e-06,
-      "loss": 26.065,
+      "epoch": 0.02191947636806454,
+      "grad_norm": 416.0,
+      "learning_rate": 9.996575101367487e-06,
+      "loss": 23.1901,
       "step": 180
     },
     {
-      "epoch": 0.011568612527589619,
-      "grad_norm": 620.5,
-      "learning_rate": 7.998553931688493e-06,
-      "loss": 25.868,
+      "epoch": 0.023137225055179237,
+      "grad_norm": 512.0,
+      "learning_rate": 9.996384829221236e-06,
+      "loss": 23.3191,
       "step": 190
     },
     {
-      "epoch": 0.012177486871146967,
-      "grad_norm": 1011.5,
-      "learning_rate": 7.998477822829994e-06,
-      "loss": 26.3943,
+      "epoch": 0.024354973742293933,
+      "grad_norm": 602.0,
+      "learning_rate": 9.996194557074985e-06,
+      "loss": 23.602,
       "step": 200
     },
     {
-      "epoch": 0.012786361214704315,
-      "grad_norm": 625.5,
-      "learning_rate": 7.998401713971494e-06,
-      "loss": 25.6513,
+      "epoch": 0.02557272242940863,
+      "grad_norm": 361.75,
+      "learning_rate": 9.996004284928734e-06,
+      "loss": 23.1068,
       "step": 210
     },
     {
-      "epoch": 0.013395235558261664,
-      "grad_norm": 464.25,
-      "learning_rate": 7.998325605112993e-06,
-      "loss": 25.8677,
+      "epoch": 0.02679047111652333,
+      "grad_norm": 597.5,
+      "learning_rate": 9.995814012782483e-06,
+      "loss": 23.5984,
       "step": 220
     },
     {
-      "epoch": 0.014004109901819013,
-      "grad_norm": 418.25,
-      "learning_rate": 7.998249496254493e-06,
-      "loss": 25.698,
+      "epoch": 0.028008219803638025,
+      "grad_norm": 881.0,
+      "learning_rate": 9.995623740636233e-06,
+      "loss": 23.2802,
       "step": 230
     },
     {
-      "epoch": 0.01461298424537636,
-      "grad_norm": 439.25,
-      "learning_rate": 7.998173387395992e-06,
-      "loss": 26.0902,
+      "epoch": 0.02922596849075272,
+      "grad_norm": 485.0,
+      "learning_rate": 9.995433468489982e-06,
+      "loss": 23.0281,
       "step": 240
     },
     {
-      "epoch": 0.015221858588933709,
-      "grad_norm": 643.0,
-      "learning_rate": 7.998097278537492e-06,
-      "loss": 25.6914,
+      "epoch": 0.030443717177867417,
+      "grad_norm": 466.0,
+      "learning_rate": 9.995243196343731e-06,
+      "loss": 23.307,
       "step": 250
     },
     {
-      "epoch": 0.015830732932491057,
-      "grad_norm": 627.0,
-      "learning_rate": 7.998021169678991e-06,
-      "loss": 25.5016,
+      "epoch": 0.03166146586498211,
+      "grad_norm": 473.25,
+      "learning_rate": 9.99505292419748e-06,
+      "loss": 23.3621,
       "step": 260
     },
     {
-      "epoch": 0.016439607276048405,
-      "grad_norm": 472.75,
-      "learning_rate": 7.99794506082049e-06,
-      "loss": 25.5562,
+      "epoch": 0.03287921455209681,
+      "grad_norm": 839.0,
+      "learning_rate": 9.994862652051229e-06,
+      "loss": 23.326,
       "step": 270
     },
     {
-      "epoch": 0.017048481619605753,
-      "grad_norm": 560.5,
-      "learning_rate": 7.99786895196199e-06,
-      "loss": 25.5837,
+      "epoch": 0.034096963239211506,
+      "grad_norm": 734.0,
+      "learning_rate": 9.99467237990498e-06,
+      "loss": 23.3765,
       "step": 280
     },
     {
-      "epoch": 0.0176573559631631,
-      "grad_norm": 661.0,
-      "learning_rate": 7.99779284310349e-06,
-      "loss": 25.7263,
+      "epoch": 0.0353147119263262,
+      "grad_norm": 499.75,
+      "learning_rate": 9.994482107758728e-06,
+      "loss": 23.1504,
       "step": 290
     },
     {
-      "epoch": 0.01826623030672045,
-      "grad_norm": 444.5,
-      "learning_rate": 7.99771673424499e-06,
-      "loss": 26.0324,
+      "epoch": 0.0365324606134409,
+      "grad_norm": 449.5,
+      "learning_rate": 9.994291835612477e-06,
+      "loss": 23.0619,
       "step": 300
     },
     {
-      "epoch": 0.0188751046502778,
-      "grad_norm": 521.0,
-      "learning_rate": 7.99764062538649e-06,
-      "loss": 25.6377,
+      "epoch": 0.0377502093005556,
+      "grad_norm": 318.25,
+      "learning_rate": 9.994101563466226e-06,
+      "loss": 23.0764,
       "step": 310
     },
     {
-      "epoch": 0.01948397899383515,
-      "grad_norm": 643.0,
-      "learning_rate": 7.99756451652799e-06,
-      "loss": 25.5232,
+      "epoch": 0.0389679579876703,
+      "grad_norm": 382.0,
+      "learning_rate": 9.993911291319975e-06,
+      "loss": 23.3104,
       "step": 320
     },
     {
-      "epoch": 0.020092853337392497,
-      "grad_norm": 463.75,
-      "learning_rate": 7.99748840766949e-06,
-      "loss": 25.2923,
+      "epoch": 0.04018570667478499,
+      "grad_norm": 605.0,
+      "learning_rate": 9.993721019173725e-06,
+      "loss": 23.2775,
       "step": 330
     },
     {
-      "epoch": 0.020701727680949845,
-      "grad_norm": 411.75,
-      "learning_rate": 7.99741229881099e-06,
-      "loss": 25.5332,
+      "epoch": 0.04140345536189969,
+      "grad_norm": 399.5,
+      "learning_rate": 9.993530747027474e-06,
+      "loss": 22.8484,
       "step": 340
     },
     {
-      "epoch": 0.021310602024507193,
-      "grad_norm": 419.25,
-      "learning_rate": 7.997336189952489e-06,
-      "loss": 25.5767,
+      "epoch": 0.042621204049014386,
+      "grad_norm": 425.0,
+      "learning_rate": 9.993340474881223e-06,
+      "loss": 23.3281,
       "step": 350
     },
     {
-      "epoch": 0.02191947636806454,
-      "grad_norm": 697.5,
-      "learning_rate": 7.997260081093988e-06,
-      "loss": 25.3458,
+      "epoch": 0.04383895273612908,
+      "grad_norm": 313.25,
+      "learning_rate": 9.993150202734972e-06,
+      "loss": 23.0878,
       "step": 360
     },
     {
-      "epoch": 0.02252835071162189,
-      "grad_norm": 633.5,
-      "learning_rate": 7.997183972235488e-06,
-      "loss": 25.5739,
+      "epoch": 0.04505670142324378,
+      "grad_norm": 670.0,
+      "learning_rate": 9.992959930588721e-06,
+      "loss": 23.2712,
       "step": 370
     },
     {
-      "epoch": 0.023137225055179237,
-      "grad_norm": 488.25,
-      "learning_rate": 7.997107863376988e-06,
-      "loss": 25.6053,
+      "epoch": 0.046274450110358474,
+      "grad_norm": 478.5,
+      "learning_rate": 9.992769658442471e-06,
+      "loss": 23.266,
       "step": 380
     },
     {
-      "epoch": 0.023746099398736585,
-      "grad_norm": 515.0,
-      "learning_rate": 7.997031754518487e-06,
-      "loss": 25.4469,
+      "epoch": 0.04749219879747317,
+      "grad_norm": 531.5,
+      "learning_rate": 9.99257938629622e-06,
+      "loss": 23.3419,
       "step": 390
     },
     {
-      "epoch": 0.024354973742293933,
-      "grad_norm": 428.0,
-      "learning_rate": 7.996955645659987e-06,
-      "loss": 25.6501,
+      "epoch": 0.048709947484587866,
+      "grad_norm": 317.5,
+      "learning_rate": 9.99238911414997e-06,
+      "loss": 23.0632,
       "step": 400
     },
     {
-      "epoch": 0.02496384808585128,
-      "grad_norm": 570.0,
-      "learning_rate": 7.996879536801488e-06,
-      "loss": 25.5624,
+      "epoch": 0.04992769617170256,
+      "grad_norm": 439.75,
+      "learning_rate": 9.992198842003718e-06,
+      "loss": 23.3886,
       "step": 410
     },
     {
-      "epoch": 0.02557272242940863,
-      "grad_norm": 498.75,
-      "learning_rate": 7.996803427942986e-06,
-      "loss": 25.5509,
+      "epoch": 0.05114544485881726,
+      "grad_norm": 438.5,
+      "learning_rate": 9.992008569857467e-06,
+      "loss": 23.1328,
       "step": 420
     },
     {
-      "epoch": 0.02618159677296598,
-      "grad_norm": 513.0,
-      "learning_rate": 7.996727319084487e-06,
-      "loss": 25.487,
+      "epoch": 0.05236319354593196,
+      "grad_norm": 325.75,
+      "learning_rate": 9.991818297711218e-06,
+      "loss": 23.0403,
       "step": 430
     },
     {
-      "epoch": 0.02679047111652333,
-      "grad_norm": 815.5,
-      "learning_rate": 7.996651210225985e-06,
-      "loss": 25.7721,
+      "epoch": 0.05358094223304666,
+      "grad_norm": 338.75,
+      "learning_rate": 9.991628025564966e-06,
+      "loss": 23.3389,
       "step": 440
     },
     {
-      "epoch": 0.027399345460080677,
-      "grad_norm": 640.5,
-      "learning_rate": 7.996575101367486e-06,
-      "loss": 25.3603,
+      "epoch": 0.054798690920161354,
+      "grad_norm": 342.5,
+      "learning_rate": 9.991437753418715e-06,
+      "loss": 23.4204,
       "step": 450
     },
     {
-      "epoch": 0.028008219803638025,
-      "grad_norm": 688.0,
-      "learning_rate": 7.996498992508986e-06,
-      "loss": 25.2232,
+      "epoch": 0.05601643960727605,
+      "grad_norm": 411.0,
+      "learning_rate": 9.991247481272464e-06,
+      "loss": 23.1465,
       "step": 460
     },
     {
-      "epoch": 0.028617094147195373,
-      "grad_norm": 595.0,
-      "learning_rate": 7.996422883650485e-06,
-      "loss": 25.2017,
+      "epoch": 0.057234188294390746,
+      "grad_norm": 393.25,
+      "learning_rate": 9.991057209126213e-06,
+      "loss": 23.0713,
       "step": 470
     },
     {
-      "epoch": 0.02922596849075272,
-      "grad_norm": 472.0,
-      "learning_rate": 7.996346774791985e-06,
-      "loss": 25.3047,
+      "epoch": 0.05845193698150544,
+      "grad_norm": 360.5,
+      "learning_rate": 9.990866936979964e-06,
+      "loss": 23.1229,
       "step": 480
     },
     {
-      "epoch": 0.02983484283431007,
-      "grad_norm": 643.5,
-      "learning_rate": 7.996270665933484e-06,
-      "loss": 25.2193,
+      "epoch": 0.05966968566862014,
+      "grad_norm": 421.75,
+      "learning_rate": 9.990676664833712e-06,
+      "loss": 23.5678,
       "step": 490
     },
     {
-      "epoch": 0.030443717177867417,
-      "grad_norm": 445.25,
-      "learning_rate": 7.996194557074984e-06,
-      "loss": 25.2767,
+      "epoch": 0.060887434355734835,
+      "grad_norm": 402.75,
+      "learning_rate": 9.990486392687461e-06,
+      "loss": 23.4222,
       "step": 500
     },
     {
-      "epoch": 0.031052591521424765,
-      "grad_norm": 642.0,
-      "learning_rate": 7.996118448216483e-06,
-      "loss": 25.1364,
+      "epoch": 0.06210518304284953,
+      "grad_norm": 476.0,
+      "learning_rate": 9.99029612054121e-06,
+      "loss": 23.4822,
       "step": 510
     },
     {
-      "epoch": 0.03166146586498211,
-      "grad_norm": 587.5,
-      "learning_rate": 7.996042339357983e-06,
-      "loss": 25.345,
+      "epoch": 0.06332293172996423,
+      "grad_norm": 464.5,
+      "learning_rate": 9.990105848394959e-06,
+      "loss": 23.2711,
       "step": 520
     },
     {
-      "epoch": 0.032270340208539465,
-      "grad_norm": 807.5,
-      "learning_rate": 7.995966230499482e-06,
-      "loss": 25.1371,
+      "epoch": 0.06454068041707893,
+      "grad_norm": 490.25,
+      "learning_rate": 9.98991557624871e-06,
+      "loss": 23.3672,
       "step": 530
     },
     {
-      "epoch": 0.03287921455209681,
-      "grad_norm": 639.5,
-      "learning_rate": 7.995890121640982e-06,
-      "loss": 25.2102,
+      "epoch": 0.06575842910419362,
+      "grad_norm": 313.5,
+      "learning_rate": 9.989725304102459e-06,
+      "loss": 23.0706,
       "step": 540
     },
     {
-      "epoch": 0.03348808889565416,
-      "grad_norm": 463.75,
-      "learning_rate": 7.995814012782483e-06,
-      "loss": 25.1115,
+      "epoch": 0.06697617779130832,
+      "grad_norm": 336.25,
+      "learning_rate": 9.989535031956207e-06,
+      "loss": 23.1567,
       "step": 550
     },
     {
-      "epoch": 0.034096963239211506,
-      "grad_norm": 696.5,
-      "learning_rate": 7.995737903923983e-06,
-      "loss": 25.3349,
+      "epoch": 0.06819392647842301,
+      "grad_norm": 722.0,
+      "learning_rate": 9.989344759809956e-06,
+      "loss": 23.3509,
       "step": 560
     },
     {
-      "epoch": 0.03470583758276886,
-      "grad_norm": 544.0,
-      "learning_rate": 7.995661795065482e-06,
-      "loss": 25.0044,
+      "epoch": 0.06941167516553771,
+      "grad_norm": 372.0,
+      "learning_rate": 9.989154487663705e-06,
+      "loss": 23.3139,
       "step": 570
     },
     {
-      "epoch": 0.0353147119263262,
-      "grad_norm": 610.0,
-      "learning_rate": 7.995585686206982e-06,
-      "loss": 24.9601,
+      "epoch": 0.0706294238526524,
+      "grad_norm": 320.0,
+      "learning_rate": 9.988964215517456e-06,
+      "loss": 23.0371,
       "step": 580
     },
     {
-      "epoch": 0.03592358626988355,
-      "grad_norm": 445.25,
-      "learning_rate": 7.995509577348481e-06,
-      "loss": 24.8361,
+      "epoch": 0.0718471725397671,
+      "grad_norm": 499.0,
+      "learning_rate": 9.988773943371205e-06,
+      "loss": 23.176,
       "step": 590
     },
     {
-      "epoch": 0.0365324606134409,
-      "grad_norm": 562.5,
-      "learning_rate": 7.995433468489981e-06,
-      "loss": 24.9502,
+      "epoch": 0.0730649212268818,
+      "grad_norm": 367.5,
+      "learning_rate": 9.988583671224953e-06,
+      "loss": 23.1789,
       "step": 600
     },
     {
-      "epoch": 0.03714133495699825,
-      "grad_norm": 668.0,
-      "learning_rate": 7.99535735963148e-06,
-      "loss": 25.0697,
+      "epoch": 0.0742826699139965,
+      "grad_norm": 473.0,
+      "learning_rate": 9.988393399078702e-06,
+      "loss": 23.4324,
       "step": 610
     },
     {
-      "epoch": 0.0377502093005556,
-      "grad_norm": 510.75,
-      "learning_rate": 7.99528125077298e-06,
-      "loss": 24.9929,
+      "epoch": 0.0755004186011112,
+      "grad_norm": 328.0,
+      "learning_rate": 9.988203126932451e-06,
+      "loss": 23.2466,
       "step": 620
     },
     {
-      "epoch": 0.038359083644112946,
-      "grad_norm": 512.0,
-      "learning_rate": 7.99520514191448e-06,
-      "loss": 25.3818,
+      "epoch": 0.07671816728822589,
+      "grad_norm": 391.75,
+      "learning_rate": 9.988012854786202e-06,
+      "loss": 22.8369,
       "step": 630
     },
     {
-      "epoch": 0.0389679579876703,
-      "grad_norm": 408.25,
-      "learning_rate": 7.99512903305598e-06,
-      "loss": 24.6347,
+      "epoch": 0.0779359159753406,
+      "grad_norm": 334.0,
+      "learning_rate": 9.98782258263995e-06,
+      "loss": 23.5017,
       "step": 640
     },
     {
-      "epoch": 0.03957683233122764,
-      "grad_norm": 444.75,
-      "learning_rate": 7.995052924197479e-06,
-      "loss": 24.6194,
+      "epoch": 0.07915366466245528,
+      "grad_norm": 367.5,
+      "learning_rate": 9.9876323104937e-06,
+      "loss": 22.9409,
       "step": 650
     },
     {
-      "epoch": 0.04018570667478499,
-      "grad_norm": 706.5,
-      "learning_rate": 7.99497681533898e-06,
-      "loss": 25.3,
+      "epoch": 0.08037141334956999,
+      "grad_norm": 318.75,
+      "learning_rate": 9.987442038347448e-06,
+      "loss": 23.3296,
       "step": 660
     },
     {
-      "epoch": 0.04079458101834234,
-      "grad_norm": 545.5,
-      "learning_rate": 7.994900706480478e-06,
-      "loss": 24.8079,
+      "epoch": 0.08158916203668468,
+      "grad_norm": 385.75,
+      "learning_rate": 9.987251766201197e-06,
+      "loss": 23.0806,
       "step": 670
     },
     {
-      "epoch": 0.04140345536189969,
-      "grad_norm": 568.5,
-      "learning_rate": 7.994824597621979e-06,
-      "loss": 24.6425,
+      "epoch": 0.08280691072379938,
+      "grad_norm": 339.25,
+      "learning_rate": 9.987061494054948e-06,
+      "loss": 23.2532,
       "step": 680
     },
     {
-      "epoch": 0.042012329705457034,
-      "grad_norm": 559.0,
-      "learning_rate": 7.994748488763477e-06,
-      "loss": 25.3957,
+      "epoch": 0.08402465941091407,
+      "grad_norm": 356.25,
+      "learning_rate": 9.986871221908697e-06,
+      "loss": 23.0229,
       "step": 690
     },
     {
-      "epoch": 0.042621204049014386,
-      "grad_norm": 505.75,
-      "learning_rate": 7.994672379904978e-06,
-      "loss": 25.0813,
+      "epoch": 0.08524240809802877,
+      "grad_norm": 536.0,
+      "learning_rate": 9.986680949762446e-06,
+      "loss": 23.0866,
       "step": 700
     },
     {
-      "epoch": 0.04323007839257173,
-      "grad_norm": 644.5,
-      "learning_rate": 7.994596271046478e-06,
-      "loss": 24.7936,
+      "epoch": 0.08646015678514346,
+      "grad_norm": 369.5,
+      "learning_rate": 9.986490677616194e-06,
+      "loss": 23.3175,
       "step": 710
     },
     {
-      "epoch": 0.04383895273612908,
-      "grad_norm": 481.75,
-      "learning_rate": 7.994520162187977e-06,
-      "loss": 24.94,
+      "epoch": 0.08767790547225816,
+      "grad_norm": 388.5,
+      "learning_rate": 9.986300405469945e-06,
+      "loss": 23.0809,
       "step": 720
     },
     {
-      "epoch": 0.044447827079686426,
-      "grad_norm": 673.0,
-      "learning_rate": 7.994444053329477e-06,
-      "loss": 24.5047,
+      "epoch": 0.08889565415937285,
+      "grad_norm": 360.0,
+      "learning_rate": 9.986110133323694e-06,
+      "loss": 23.3686,
       "step": 730
     },
     {
-      "epoch": 0.04505670142324378,
-      "grad_norm": 473.5,
-      "learning_rate": 7.994367944470976e-06,
-      "loss": 24.8956,
+      "epoch": 0.09011340284648756,
+      "grad_norm": 321.75,
+      "learning_rate": 9.985919861177443e-06,
+      "loss": 23.0928,
       "step": 740
     },
     {
-      "epoch": 0.04566557576680113,
-      "grad_norm": 532.0,
-      "learning_rate": 7.994291835612476e-06,
-      "loss": 24.5649,
+      "epoch": 0.09133115153360226,
+      "grad_norm": 393.5,
+      "learning_rate": 9.985729589031192e-06,
+      "loss": 23.2937,
       "step": 750
     },
     {
-      "epoch": 0.046274450110358474,
-      "grad_norm": 395.5,
-      "learning_rate": 7.994215726753975e-06,
-      "loss": 24.8702,
+      "epoch": 0.09254890022071695,
+      "grad_norm": 676.5,
+      "learning_rate": 9.98553931688494e-06,
+      "loss": 23.462,
       "step": 760
     },
     {
-      "epoch": 0.046883324453915826,
-      "grad_norm": 437.5,
-      "learning_rate": 7.994139617895475e-06,
-      "loss": 24.6667,
+      "epoch": 0.09376664890783165,
+      "grad_norm": 332.0,
+      "learning_rate": 9.985349044738691e-06,
+      "loss": 22.8664,
       "step": 770
     },
     {
-      "epoch": 0.04749219879747317,
-      "grad_norm": 467.0,
-      "learning_rate": 7.994063509036976e-06,
-      "loss": 24.1927,
+      "epoch": 0.09498439759494634,
+      "grad_norm": 316.25,
+      "learning_rate": 9.98515877259244e-06,
+      "loss": 23.0369,
       "step": 780
     },
     {
-      "epoch": 0.04810107314103052,
-      "grad_norm": 475.5,
-      "learning_rate": 7.993987400178474e-06,
-      "loss": 25.1248,
+      "epoch": 0.09620214628206104,
+      "grad_norm": 318.5,
+      "learning_rate": 9.984968500446189e-06,
+      "loss": 23.437,
       "step": 790
     },
     {
-      "epoch": 0.048709947484587866,
-      "grad_norm": 656.0,
-      "learning_rate": 7.993911291319975e-06,
-      "loss": 24.3436,
+      "epoch": 0.09741989496917573,
+      "grad_norm": 612.0,
+      "learning_rate": 9.984778228299938e-06,
+      "loss": 22.9744,
       "step": 800
     },
     {
-      "epoch": 0.04931882182814522,
-      "grad_norm": 392.0,
-      "learning_rate": 7.993835182461475e-06,
-      "loss": 24.6863,
+      "epoch": 0.09863764365629044,
+      "grad_norm": 397.0,
+      "learning_rate": 9.984587956153687e-06,
+      "loss": 22.9221,
       "step": 810
     },
     {
-      "epoch": 0.04992769617170256,
-      "grad_norm": 549.5,
-      "learning_rate": 7.993759073602974e-06,
-      "loss": 24.9492,
+      "epoch": 0.09985539234340512,
+      "grad_norm": 531.5,
+      "learning_rate": 9.984397684007437e-06,
+      "loss": 23.1225,
       "step": 820
     },
     {
-      "epoch": 0.050536570515259914,
-      "grad_norm": 472.5,
-      "learning_rate": 7.993682964744474e-06,
-      "loss": 24.5096,
+      "epoch": 0.10107314103051983,
+      "grad_norm": 485.75,
+      "learning_rate": 9.984207411861186e-06,
+      "loss": 22.9605,
       "step": 830
     },
     {
-      "epoch": 0.05114544485881726,
-      "grad_norm": 654.5,
-      "learning_rate": 7.993606855885974e-06,
-      "loss": 24.7159,
+      "epoch": 0.10229088971763452,
+      "grad_norm": 439.25,
+      "learning_rate": 9.984017139714935e-06,
+      "loss": 22.8921,
       "step": 840
     },
     {
-      "epoch": 0.05175431920237461,
-      "grad_norm": 578.5,
-      "learning_rate": 7.993530747027473e-06,
-      "loss": 24.7425,
+      "epoch": 0.10350863840474922,
+      "grad_norm": 614.5,
+      "learning_rate": 9.983826867568684e-06,
+      "loss": 23.1956,
       "step": 850
     },
     {
-      "epoch": 0.05236319354593196,
-      "grad_norm": 539.0,
-      "learning_rate": 7.993454638168973e-06,
-      "loss": 24.9575,
+      "epoch": 0.10472638709186392,
+      "grad_norm": 384.5,
+      "learning_rate": 9.983636595422434e-06,
+      "loss": 23.2236,
       "step": 860
     },
     {
-      "epoch": 0.052972067889489306,
-      "grad_norm": 798.0,
-      "learning_rate": 7.993378529310472e-06,
-      "loss": 24.9244,
+      "epoch": 0.10594413577897861,
+      "grad_norm": 442.75,
+      "learning_rate": 9.983446323276183e-06,
+      "loss": 22.8962,
       "step": 870
     },
     {
-      "epoch": 0.05358094223304666,
-      "grad_norm": 388.5,
-      "learning_rate": 7.993302420451972e-06,
-      "loss": 24.9071,
+      "epoch": 0.10716188446609332,
+      "grad_norm": 438.0,
+      "learning_rate": 9.983256051129932e-06,
+      "loss": 23.0436,
       "step": 880
     },
     {
-      "epoch": 0.054189816576604,
-      "grad_norm": 604.0,
-      "learning_rate": 7.993226311593471e-06,
-      "loss": 24.7988,
+      "epoch": 0.108379633153208,
+      "grad_norm": 402.75,
+      "learning_rate": 9.983065778983681e-06,
+      "loss": 23.3182,
       "step": 890
     },
     {
-      "epoch": 0.054798690920161354,
-      "grad_norm": 668.0,
-      "learning_rate": 7.993150202734971e-06,
-      "loss": 24.9017,
+      "epoch": 0.10959738184032271,
+      "grad_norm": 758.5,
+      "learning_rate": 9.98287550683743e-06,
+      "loss": 22.9257,
       "step": 900
     },
     {
-      "epoch": 0.0554075652637187,
-      "grad_norm": 607.5,
-      "learning_rate": 7.993074093876472e-06,
-      "loss": 24.6064,
+      "epoch": 0.1108151305274374,
+      "grad_norm": 880.5,
+      "learning_rate": 9.98268523469118e-06,
+      "loss": 23.1712,
       "step": 910
     },
     {
-      "epoch": 0.05601643960727605,
-      "grad_norm": 387.75,
-      "learning_rate": 7.99299798501797e-06,
-      "loss": 24.7402,
+      "epoch": 0.1120328792145521,
+      "grad_norm": 281.25,
+      "learning_rate": 9.98249496254493e-06,
+      "loss": 22.8172,
       "step": 920
     },
     {
-      "epoch": 0.056625313950833395,
-      "grad_norm": 582.0,
-      "learning_rate": 7.992921876159471e-06,
-      "loss": 24.5883,
+      "epoch": 0.11325062790166679,
+      "grad_norm": 413.0,
+      "learning_rate": 9.982304690398678e-06,
+      "loss": 23.3211,
       "step": 930
     },
     {
-      "epoch": 0.057234188294390746,
-      "grad_norm": 510.75,
-      "learning_rate": 7.99284576730097e-06,
-      "loss": 24.2875,
+      "epoch": 0.11446837658878149,
+      "grad_norm": 769.5,
+      "learning_rate": 9.982114418252427e-06,
+      "loss": 22.8698,
       "step": 940
     },
     {
-      "epoch": 0.05784306263794809,
-      "grad_norm": 480.0,
-      "learning_rate": 7.99276965844247e-06,
-      "loss": 24.6182,
+      "epoch": 0.11568612527589618,
+      "grad_norm": 406.75,
+      "learning_rate": 9.981924146106178e-06,
+      "loss": 22.8464,
       "step": 950
     },
     {
-      "epoch": 0.05845193698150544,
-      "grad_norm": 494.0,
-      "learning_rate": 7.99269354958397e-06,
-      "loss": 24.3947,
+      "epoch": 0.11690387396301088,
+      "grad_norm": 372.75,
+      "learning_rate": 9.981733873959926e-06,
+      "loss": 23.105,
       "step": 960
     },
     {
-      "epoch": 0.05906081132506279,
-      "grad_norm": 717.0,
-      "learning_rate": 7.99261744072547e-06,
-      "loss": 24.9133,
+      "epoch": 0.11812162265012557,
+      "grad_norm": 348.0,
+      "learning_rate": 9.981543601813675e-06,
+      "loss": 23.2349,
       "step": 970
     },
     {
-      "epoch": 0.05966968566862014,
-      "grad_norm": 412.0,
-      "learning_rate": 7.992541331866969e-06,
-      "loss": 24.6561,
+      "epoch": 0.11933937133724028,
+      "grad_norm": 261.75,
+      "learning_rate": 9.981353329667424e-06,
+      "loss": 23.0146,
       "step": 980
     },
     {
-      "epoch": 0.06027856001217749,
-      "grad_norm": 524.0,
-      "learning_rate": 7.992465223008469e-06,
-      "loss": 24.5113,
+      "epoch": 0.12055712002435498,
+      "grad_norm": 516.5,
+      "learning_rate": 9.981163057521173e-06,
+      "loss": 22.9435,
       "step": 990
     },
     {
-      "epoch": 0.060887434355734835,
-      "grad_norm": 512.5,
-      "learning_rate": 7.992389114149968e-06,
-      "loss": 24.9458,
+      "epoch": 0.12177486871146967,
+      "grad_norm": 293.0,
+      "learning_rate": 9.980972785374924e-06,
+      "loss": 23.3143,
       "step": 1000
     },
     {
-      "epoch": 0.061496308699292186,
-      "grad_norm": 686.0,
-      "learning_rate": 7.99231300529147e-06,
-      "loss": 24.7602,
+      "epoch": 0.12299261739858437,
+      "grad_norm": 834.0,
+      "learning_rate": 9.980782513228672e-06,
+      "loss": 23.1654,
       "step": 1010
     },
     {
-      "epoch": 0.06210518304284953,
-      "grad_norm": 448.25,
-      "learning_rate": 7.992236896432967e-06,
-      "loss": 24.7965,
+      "epoch": 0.12421036608569906,
+      "grad_norm": 313.25,
+      "learning_rate": 9.980592241082421e-06,
+      "loss": 22.9153,
       "step": 1020
     },
     {
-      "epoch": 0.06271405738640688,
-      "grad_norm": 674.0,
-      "learning_rate": 7.992160787574468e-06,
-      "loss": 24.6205,
+      "epoch": 0.12542811477281376,
+      "grad_norm": 432.0,
+      "learning_rate": 9.98040196893617e-06,
+      "loss": 22.8277,
       "step": 1030
     },
     {
-      "epoch": 0.06332293172996423,
-      "grad_norm": 418.75,
-      "learning_rate": 7.992084678715966e-06,
-      "loss": 24.7116,
+      "epoch": 0.12664586345992845,
+      "grad_norm": 883.0,
+      "learning_rate": 9.980211696789919e-06,
+      "loss": 22.8528,
       "step": 1040
     },
     {
-      "epoch": 0.06393180607352157,
-      "grad_norm": 504.75,
-      "learning_rate": 7.992008569857467e-06,
-      "loss": 24.7183,
+      "epoch": 0.12786361214704314,
+      "grad_norm": 351.75,
+      "learning_rate": 9.98002142464367e-06,
+      "loss": 22.4458,
       "step": 1050
     },
     {
-      "epoch": 0.06454068041707893,
-      "grad_norm": 633.5,
-      "learning_rate": 7.991932460998967e-06,
-      "loss": 24.7446,
+      "epoch": 0.12908136083415786,
+      "grad_norm": 322.0,
+      "learning_rate": 9.979831152497418e-06,
+      "loss": 22.9115,
       "step": 1060
     },
     {
-      "epoch": 0.06514955476063627,
-      "grad_norm": 604.0,
-      "learning_rate": 7.991856352140467e-06,
-      "loss": 24.362,
+      "epoch": 0.13029910952127255,
+      "grad_norm": 345.5,
+      "learning_rate": 9.979640880351167e-06,
+      "loss": 23.1693,
       "step": 1070
     },
     {
-      "epoch": 0.06575842910419362,
-      "grad_norm": 507.25,
-      "learning_rate": 7.991780243281966e-06,
-      "loss": 24.3108,
+      "epoch": 0.13151685820838724,
+      "grad_norm": 595.5,
+      "learning_rate": 9.979450608204916e-06,
+      "loss": 23.0466,
       "step": 1080
     },
     {
-      "epoch": 0.06636730344775096,
-      "grad_norm": 382.0,
-      "learning_rate": 7.991704134423466e-06,
-      "loss": 24.6816,
+      "epoch": 0.13273460689550193,
+      "grad_norm": 890.5,
+      "learning_rate": 9.979260336058667e-06,
+      "loss": 23.039,
       "step": 1090
     },
     {
-      "epoch": 0.06697617779130832,
-      "grad_norm": 436.25,
-      "learning_rate": 7.991628025564965e-06,
-      "loss": 24.4788,
+      "epoch": 0.13395235558261664,
+      "grad_norm": 331.75,
+      "learning_rate": 9.979070063912416e-06,
+      "loss": 22.9233,
       "step": 1100
     },
     {
-      "epoch": 0.06758505213486567,
-      "grad_norm": 421.5,
-      "learning_rate": 7.991551916706465e-06,
-      "loss": 24.6223,
+      "epoch": 0.13517010426973133,
+      "grad_norm": 599.5,
+      "learning_rate": 9.978879791766165e-06,
+      "loss": 22.7559,
       "step": 1110
     },
     {
-      "epoch": 0.06819392647842301,
-      "grad_norm": 438.5,
-      "learning_rate": 7.991475807847964e-06,
-      "loss": 24.7549,
+      "epoch": 0.13638785295684602,
+      "grad_norm": 346.0,
+      "learning_rate": 9.978689519619913e-06,
+      "loss": 23.0462,
       "step": 1120
     },
     {
-      "epoch": 0.06880280082198037,
-      "grad_norm": 501.5,
-      "learning_rate": 7.991399698989464e-06,
-      "loss": 24.5644,
+      "epoch": 0.13760560164396074,
+      "grad_norm": 637.5,
+      "learning_rate": 9.978499247473662e-06,
+      "loss": 22.9928,
       "step": 1130
     },
     {
-      "epoch": 0.06941167516553771,
-      "grad_norm": 597.0,
-      "learning_rate": 7.991323590130963e-06,
-      "loss": 24.3507,
+      "epoch": 0.13882335033107543,
+      "grad_norm": 255.875,
+      "learning_rate": 9.978308975327413e-06,
+      "loss": 22.8982,
       "step": 1140
     },
     {
-      "epoch": 0.07002054950909506,
-      "grad_norm": 446.75,
-      "learning_rate": 7.991247481272463e-06,
-      "loss": 24.456,
+      "epoch": 0.14004109901819012,
+      "grad_norm": 487.0,
+      "learning_rate": 9.978118703181162e-06,
+      "loss": 23.2006,
       "step": 1150
     },
     {
-      "epoch": 0.0706294238526524,
-      "grad_norm": 480.0,
-      "learning_rate": 7.991171372413964e-06,
-      "loss": 24.1804,
+      "epoch": 0.1412588477053048,
+      "grad_norm": 427.25,
+      "learning_rate": 9.97792843103491e-06,
+      "loss": 23.2043,
       "step": 1160
     },
     {
-      "epoch": 0.07123829819620976,
-      "grad_norm": 483.25,
-      "learning_rate": 7.991095263555464e-06,
-      "loss": 24.2766,
+      "epoch": 0.14247659639241952,
+      "grad_norm": 925.0,
+      "learning_rate": 9.97773815888866e-06,
+      "loss": 22.6858,
       "step": 1170
     },
     {
-      "epoch": 0.0718471725397671,
-      "grad_norm": 876.0,
-      "learning_rate": 7.991019154696963e-06,
-      "loss": 24.5967,
+      "epoch": 0.1436943450795342,
+      "grad_norm": 317.0,
+      "learning_rate": 9.977547886742408e-06,
+      "loss": 22.8387,
       "step": 1180
     },
     {
-      "epoch": 0.07245604688332445,
-      "grad_norm": 576.5,
-      "learning_rate": 7.990943045838463e-06,
-      "loss": 24.5191,
+      "epoch": 0.1449120937666489,
+      "grad_norm": 462.75,
+      "learning_rate": 9.977357614596159e-06,
+      "loss": 22.6462,
       "step": 1190
     },
     {
-      "epoch": 0.0730649212268818,
-      "grad_norm": 398.25,
-      "learning_rate": 7.990866936979962e-06,
-      "loss": 24.2136,
+      "epoch": 0.1461298424537636,
+      "grad_norm": 527.0,
+      "learning_rate": 9.977167342449908e-06,
+      "loss": 23.044,
       "step": 1200
     },
     {
-      "epoch": 0.07367379557043915,
-      "grad_norm": 626.5,
-      "learning_rate": 7.990790828121462e-06,
-      "loss": 24.7589,
+      "epoch": 0.1473475911408783,
+      "grad_norm": 310.75,
+      "learning_rate": 9.976977070303657e-06,
+      "loss": 23.0414,
       "step": 1210
     },
     {
-      "epoch": 0.0742826699139965,
-      "grad_norm": 478.75,
-      "learning_rate": 7.990714719262962e-06,
-      "loss": 24.5127,
+      "epoch": 0.148565339827993,
+      "grad_norm": 364.75,
+      "learning_rate": 9.976786798157406e-06,
+      "loss": 22.6991,
       "step": 1220
     },
     {
-      "epoch": 0.07489154425755384,
-      "grad_norm": 526.0,
-      "learning_rate": 7.990638610404461e-06,
-      "loss": 24.5955,
+      "epoch": 0.1497830885151077,
+      "grad_norm": 349.0,
+      "learning_rate": 9.976596526011154e-06,
+      "loss": 22.9886,
       "step": 1230
     },
     {
-      "epoch": 0.0755004186011112,
-      "grad_norm": 501.75,
-      "learning_rate": 7.99056250154596e-06,
-      "loss": 24.5776,
+      "epoch": 0.1510008372022224,
+      "grad_norm": 354.75,
+      "learning_rate": 9.976406253864905e-06,
+      "loss": 23.014,
       "step": 1240
     },
     {
-      "epoch": 0.07610929294466855,
-      "grad_norm": 440.5,
-      "learning_rate": 7.99048639268746e-06,
-      "loss": 24.0178,
+      "epoch": 0.1522185858893371,
+      "grad_norm": 353.25,
+      "learning_rate": 9.976215981718654e-06,
+      "loss": 22.8259,
       "step": 1250
     },
     {
-      "epoch": 0.07671816728822589,
-      "grad_norm": 479.75,
-      "learning_rate": 7.99041028382896e-06,
-      "loss": 24.2665,
+      "epoch": 0.15343633457645178,
+      "grad_norm": 304.5,
+      "learning_rate": 9.976025709572403e-06,
+      "loss": 22.9747,
       "step": 1260
     },
     {
-      "epoch": 0.07732704163178324,
-      "grad_norm": 483.5,
-      "learning_rate": 7.99033417497046e-06,
-      "loss": 24.4262,
+      "epoch": 0.15465408326356647,
+      "grad_norm": 305.5,
+      "learning_rate": 9.975835437426152e-06,
+      "loss": 22.9474,
       "step": 1270
     },
     {
-      "epoch": 0.0779359159753406,
-      "grad_norm": 407.25,
-      "learning_rate": 7.99025806611196e-06,
-      "loss": 24.3286,
+      "epoch": 0.1558718319506812,
+      "grad_norm": 372.25,
+      "learning_rate": 9.9756451652799e-06,
+      "loss": 23.1447,
       "step": 1280
     },
     {
-      "epoch": 0.07854479031889794,
-      "grad_norm": 450.5,
-      "learning_rate": 7.990181957253458e-06,
-      "loss": 24.3012,
+      "epoch": 0.15708958063779588,
+      "grad_norm": 313.5,
+      "learning_rate": 9.975454893133651e-06,
+      "loss": 22.9684,
       "step": 1290
     },
     {
-      "epoch": 0.07915366466245528,
-      "grad_norm": 677.5,
-      "learning_rate": 7.99010584839496e-06,
-      "loss": 24.7612,
+      "epoch": 0.15830732932491057,
+      "grad_norm": 352.5,
+      "learning_rate": 9.9752646209874e-06,
+      "loss": 22.856,
       "step": 1300
     },
     {
-      "epoch": 0.07976253900601263,
-      "grad_norm": 843.5,
-      "learning_rate": 7.990029739536457e-06,
-      "loss": 24.9348,
+      "epoch": 0.15952507801202526,
+      "grad_norm": 464.0,
+      "learning_rate": 9.975074348841149e-06,
+      "loss": 22.9834,
       "step": 1310
     },
     {
-      "epoch": 0.08037141334956999,
-      "grad_norm": 539.0,
-      "learning_rate": 7.989953630677959e-06,
-      "loss": 24.5834,
+      "epoch": 0.16074282669913997,
+      "grad_norm": 501.75,
+      "learning_rate": 9.974884076694898e-06,
+      "loss": 22.9727,
       "step": 1320
     },
     {
-      "epoch": 0.08098028769312733,
-      "grad_norm": 495.25,
-      "learning_rate": 7.989877521819458e-06,
-      "loss": 24.1917,
+      "epoch": 0.16196057538625466,
+      "grad_norm": 321.75,
+      "learning_rate": 9.974693804548646e-06,
+      "loss": 22.6032,
       "step": 1330
     },
     {
-      "epoch": 0.08158916203668468,
-      "grad_norm": 636.5,
-      "learning_rate": 7.989801412960958e-06,
-      "loss": 24.5528,
+      "epoch": 0.16317832407336935,
+      "grad_norm": 412.25,
+      "learning_rate": 9.974503532402397e-06,
+      "loss": 23.0908,
       "step": 1340
     },
     {
-      "epoch": 0.08219803638024203,
-      "grad_norm": 637.5,
-      "learning_rate": 7.989725304102457e-06,
-      "loss": 24.2443,
+      "epoch": 0.16439607276048407,
+      "grad_norm": 368.75,
+      "learning_rate": 9.974313260256146e-06,
+      "loss": 22.9596,
       "step": 1350
     },
     {
-      "epoch": 0.08280691072379938,
-      "grad_norm": 475.25,
-      "learning_rate": 7.989649195243957e-06,
-      "loss": 24.4378,
+      "epoch": 0.16561382144759876,
+      "grad_norm": 527.5,
+      "learning_rate": 9.974122988109895e-06,
+      "loss": 22.9562,
       "step": 1360
     },
     {
-      "epoch": 0.08341578506735672,
-      "grad_norm": 462.0,
-      "learning_rate": 7.989573086385456e-06,
-      "loss": 24.1952,
+      "epoch": 0.16683157013471345,
+      "grad_norm": 470.0,
+      "learning_rate": 9.973932715963644e-06,
+      "loss": 23.0617,
       "step": 1370
     },
     {
-      "epoch": 0.08402465941091407,
-      "grad_norm": 612.0,
-      "learning_rate": 7.989496977526956e-06,
-      "loss": 23.8663,
+      "epoch": 0.16804931882182814,
+      "grad_norm": 428.75,
+      "learning_rate": 9.973742443817393e-06,
+      "loss": 22.7603,
       "step": 1380
     },
     {
-      "epoch": 0.08463353375447143,
-      "grad_norm": 378.5,
-      "learning_rate": 7.989420868668456e-06,
-      "loss": 24.0396,
+      "epoch": 0.16926706750894285,
+      "grad_norm": 380.25,
+      "learning_rate": 9.973552171671143e-06,
+      "loss": 22.7434,
       "step": 1390
     },
     {
-      "epoch": 0.08524240809802877,
-      "grad_norm": 525.0,
-      "learning_rate": 7.989344759809955e-06,
-      "loss": 24.4567,
+      "epoch": 0.17048481619605754,
+      "grad_norm": 467.0,
+      "learning_rate": 9.973361899524892e-06,
+      "loss": 23.0467,
       "step": 1400
     },
     {
-      "epoch": 0.08585128244158612,
-      "grad_norm": 436.0,
-      "learning_rate": 7.989268650951455e-06,
-      "loss": 24.6169,
+      "epoch": 0.17170256488317223,
+      "grad_norm": 307.0,
+      "learning_rate": 9.97317162737864e-06,
+      "loss": 22.8621,
       "step": 1410
     },
     {
-      "epoch": 0.08646015678514346,
-      "grad_norm": 735.5,
-      "learning_rate": 7.989192542092956e-06,
-      "loss": 24.3255,
+      "epoch": 0.17292031357028692,
+      "grad_norm": 846.0,
+      "learning_rate": 9.97298135523239e-06,
+      "loss": 22.9859,
       "step": 1420
     },
     {
-      "epoch": 0.08706903112870082,
-      "grad_norm": 490.75,
-      "learning_rate": 7.989116433234455e-06,
-      "loss": 24.4478,
+      "epoch": 0.17413806225740164,
+      "grad_norm": 398.75,
+      "learning_rate": 9.972791083086139e-06,
+      "loss": 22.7079,
       "step": 1430
     },
     {
-      "epoch": 0.08767790547225816,
-      "grad_norm": 503.5,
-      "learning_rate": 7.989040324375955e-06,
-      "loss": 23.8608,
+      "epoch": 0.17535581094451633,
+      "grad_norm": 673.5,
+      "learning_rate": 9.972600810939889e-06,
+      "loss": 22.8856,
       "step": 1440
     },
     {
-      "epoch": 0.08828677981581551,
-      "grad_norm": 504.5,
-      "learning_rate": 7.988964215517455e-06,
-      "loss": 24.5023,
+      "epoch": 0.17657355963163102,
+      "grad_norm": 614.0,
+      "learning_rate": 9.972410538793638e-06,
+      "loss": 22.9402,
       "step": 1450
     },
     {
-      "epoch": 0.08889565415937285,
-      "grad_norm": 475.0,
-      "learning_rate": 7.988888106658954e-06,
-      "loss": 24.4643,
+      "epoch": 0.1777913083187457,
+      "grad_norm": 453.75,
+      "learning_rate": 9.972220266647387e-06,
+      "loss": 22.5813,
       "step": 1460
     },
     {
-      "epoch": 0.08950452850293021,
-      "grad_norm": 429.75,
-      "learning_rate": 7.988811997800454e-06,
-      "loss": 24.4112,
+      "epoch": 0.17900905700586042,
+      "grad_norm": 265.75,
+      "learning_rate": 9.972029994501136e-06,
+      "loss": 23.0499,
       "step": 1470
     },
     {
-      "epoch": 0.09011340284648756,
-      "grad_norm": 474.25,
-      "learning_rate": 7.988735888941953e-06,
-      "loss": 24.0898,
+      "epoch": 0.1802268056929751,
+      "grad_norm": 368.75,
+      "learning_rate": 9.971839722354885e-06,
+      "loss": 23.0325,
       "step": 1480
     },
     {
-      "epoch": 0.0907222771900449,
-      "grad_norm": 535.0,
-      "learning_rate": 7.988659780083453e-06,
-      "loss": 24.243,
+      "epoch": 0.1814445543800898,
+      "grad_norm": 324.25,
+      "learning_rate": 9.971649450208635e-06,
+      "loss": 22.9105,
       "step": 1490
     },
     {
-      "epoch": 0.09133115153360226,
-      "grad_norm": 396.75,
-      "learning_rate": 7.988583671224952e-06,
-      "loss": 24.9429,
+      "epoch": 0.18266230306720452,
+      "grad_norm": 335.25,
+      "learning_rate": 9.971459178062384e-06,
+      "loss": 23.0539,
       "step": 1500
     },
     {
-      "epoch": 0.0919400258771596,
-      "grad_norm": 553.0,
-      "learning_rate": 7.988507562366452e-06,
-      "loss": 24.6321,
+      "epoch": 0.1838800517543192,
+      "grad_norm": 533.0,
+      "learning_rate": 9.971268905916133e-06,
+      "loss": 22.8024,
       "step": 1510
     },
     {
-      "epoch": 0.09254890022071695,
-      "grad_norm": 550.0,
-      "learning_rate": 7.988431453507951e-06,
-      "loss": 24.4815,
+      "epoch": 0.1850978004414339,
+      "grad_norm": 259.0,
+      "learning_rate": 9.971078633769882e-06,
+      "loss": 23.0791,
       "step": 1520
     },
     {
-      "epoch": 0.09315777456427429,
-      "grad_norm": 547.5,
-      "learning_rate": 7.988355344649453e-06,
-      "loss": 24.044,
+      "epoch": 0.18631554912854859,
+      "grad_norm": 435.25,
+      "learning_rate": 9.97088836162363e-06,
+      "loss": 22.7367,
       "step": 1530
     },
     {
-      "epoch": 0.09376664890783165,
-      "grad_norm": 494.25,
-      "learning_rate": 7.98827923579095e-06,
-      "loss": 23.9887,
+      "epoch": 0.1875332978156633,
+      "grad_norm": 573.0,
+      "learning_rate": 9.970698089477381e-06,
+      "loss": 22.9542,
       "step": 1540
     },
     {
-      "epoch": 0.094375523251389,
-      "grad_norm": 438.25,
-      "learning_rate": 7.988203126932452e-06,
-      "loss": 24.1213,
+      "epoch": 0.188751046502778,
+      "grad_norm": 420.0,
+      "learning_rate": 9.97050781733113e-06,
+      "loss": 22.7956,
       "step": 1550
     },
     {
-      "epoch": 0.09498439759494634,
-      "grad_norm": 879.5,
-      "learning_rate": 7.98812701807395e-06,
-      "loss": 24.2047,
+      "epoch": 0.18996879518989268,
+      "grad_norm": 425.0,
+      "learning_rate": 9.970317545184879e-06,
+      "loss": 22.7087,
       "step": 1560
     },
     {
-      "epoch": 0.09559327193850368,
-      "grad_norm": 554.5,
-      "learning_rate": 7.98805090921545e-06,
-      "loss": 24.3719,
+      "epoch": 0.19118654387700737,
+      "grad_norm": 352.0,
+      "learning_rate": 9.970127273038628e-06,
+      "loss": 22.9419,
       "step": 1570
     },
     {
-      "epoch": 0.09620214628206104,
-      "grad_norm": 867.0,
-      "learning_rate": 7.98797480035695e-06,
-      "loss": 24.7275,
+      "epoch": 0.1924042925641221,
+      "grad_norm": 446.25,
+      "learning_rate": 9.969937000892377e-06,
+      "loss": 22.9542,
       "step": 1580
     },
     {
-      "epoch": 0.09681102062561839,
-      "grad_norm": 423.5,
-      "learning_rate": 7.98789869149845e-06,
-      "loss": 24.1115,
+      "epoch": 0.19362204125123678,
+      "grad_norm": 357.25,
+      "learning_rate": 9.969746728746127e-06,
+      "loss": 23.0573,
       "step": 1590
     },
     {
-      "epoch": 0.09741989496917573,
-      "grad_norm": 475.75,
-      "learning_rate": 7.98782258263995e-06,
-      "loss": 24.2442,
+      "epoch": 0.19483978993835147,
+      "grad_norm": 290.75,
+      "learning_rate": 9.969556456599876e-06,
+      "loss": 22.9406,
       "step": 1600
     },
     {
-      "epoch": 0.09802876931273309,
-      "grad_norm": 589.5,
-      "learning_rate": 7.987746473781449e-06,
-      "loss": 24.1252,
+      "epoch": 0.19605753862546618,
+      "grad_norm": 452.5,
+      "learning_rate": 9.969366184453625e-06,
+      "loss": 23.0729,
       "step": 1610
     },
     {
-      "epoch": 0.09863764365629044,
-      "grad_norm": 568.5,
-      "learning_rate": 7.987670364922949e-06,
-      "loss": 24.2393,
+      "epoch": 0.19727528731258087,
+      "grad_norm": 304.75,
+      "learning_rate": 9.969175912307374e-06,
+      "loss": 22.5812,
       "step": 1620
     },
     {
-      "epoch": 0.09924651799984778,
-      "grad_norm": 540.0,
-      "learning_rate": 7.987594256064448e-06,
-      "loss": 24.3589,
+      "epoch": 0.19849303599969556,
+      "grad_norm": 420.5,
+      "learning_rate": 9.968985640161123e-06,
+      "loss": 22.8672,
       "step": 1630
     },
     {
-      "epoch": 0.09985539234340512,
-      "grad_norm": 473.5,
-      "learning_rate": 7.987518147205948e-06,
-      "loss": 23.7216,
+      "epoch": 0.19971078468681025,
+      "grad_norm": 570.5,
+      "learning_rate": 9.968795368014873e-06,
+      "loss": 22.9808,
       "step": 1640
     },
     {
-      "epoch": 0.10046426668696248,
-      "grad_norm": 484.75,
-      "learning_rate": 7.987442038347449e-06,
-      "loss": 24.146,
+      "epoch": 0.20092853337392497,
+      "grad_norm": 497.25,
+      "learning_rate": 9.968605095868622e-06,
+      "loss": 22.8251,
       "step": 1650
     },
     {
-      "epoch": 0.10107314103051983,
-      "grad_norm": 521.0,
-      "learning_rate": 7.987365929488947e-06,
-      "loss": 23.83,
+      "epoch": 0.20214628206103966,
+      "grad_norm": 371.25,
+      "learning_rate": 9.968414823722371e-06,
+      "loss": 22.6911,
       "step": 1660
     },
     {
-      "epoch": 0.10168201537407717,
-      "grad_norm": 650.0,
-      "learning_rate": 7.987289820630448e-06,
-      "loss": 24.0998,
+      "epoch": 0.20336403074815435,
+      "grad_norm": 411.5,
+      "learning_rate": 9.96822455157612e-06,
+      "loss": 22.7651,
       "step": 1670
     },
     {
-      "epoch": 0.10229088971763452,
-      "grad_norm": 817.0,
-      "learning_rate": 7.987213711771948e-06,
-      "loss": 24.3715,
+      "epoch": 0.20458177943526903,
+      "grad_norm": 538.0,
+      "learning_rate": 9.968034279429869e-06,
+      "loss": 22.7192,
       "step": 1680
     },
     {
-      "epoch": 0.10289976406119188,
-      "grad_norm": 456.25,
-      "learning_rate": 7.987137602913447e-06,
-      "loss": 24.1391,
+      "epoch": 0.20579952812238375,
+      "grad_norm": 296.25,
+      "learning_rate": 9.96784400728362e-06,
+      "loss": 22.857,
       "step": 1690
     },
     {
-      "epoch": 0.10350863840474922,
-      "grad_norm": 491.25,
-      "learning_rate": 7.987061494054947e-06,
-      "loss": 24.3116,
+      "epoch": 0.20701727680949844,
+      "grad_norm": 442.25,
+      "learning_rate": 9.967653735137368e-06,
+      "loss": 22.7056,
       "step": 1700
     },
     {
-      "epoch": 0.10411751274830656,
-      "grad_norm": 486.0,
-      "learning_rate": 7.986985385196446e-06,
-      "loss": 24.544,
+      "epoch": 0.20823502549661313,
+      "grad_norm": 286.75,
+      "learning_rate": 9.967463462991117e-06,
+      "loss": 22.7335,
       "step": 1710
     },
     {
-      "epoch": 0.10472638709186392,
-      "grad_norm": 448.0,
-      "learning_rate": 7.986909276337946e-06,
-      "loss": 24.0699,
+      "epoch": 0.20945277418372785,
+      "grad_norm": 431.0,
+      "learning_rate": 9.967273190844866e-06,
+      "loss": 23.0438,
       "step": 1720
     },
     {
-      "epoch": 0.10533526143542127,
-      "grad_norm": 1032.0,
-      "learning_rate": 7.986833167479445e-06,
-      "loss": 24.0863,
+      "epoch": 0.21067052287084254,
+      "grad_norm": 418.5,
+      "learning_rate": 9.967082918698615e-06,
+      "loss": 22.9182,
       "step": 1730
     },
     {
-      "epoch": 0.10594413577897861,
-      "grad_norm": 666.5,
-      "learning_rate": 7.986757058620945e-06,
-      "loss": 23.7643,
+      "epoch": 0.21188827155795723,
+      "grad_norm": 573.0,
+      "learning_rate": 9.966892646552365e-06,
+      "loss": 22.9267,
       "step": 1740
     },
     {
-      "epoch": 0.10655301012253596,
-      "grad_norm": 818.0,
-      "learning_rate": 7.986680949762444e-06,
-      "loss": 23.8848,
+      "epoch": 0.21310602024507191,
+      "grad_norm": 338.25,
+      "learning_rate": 9.966702374406114e-06,
+      "loss": 22.9226,
       "step": 1750
     },
     {
-      "epoch": 0.10716188446609332,
-      "grad_norm": 592.5,
-      "learning_rate": 7.986604840903944e-06,
-      "loss": 23.5618,
+      "epoch": 0.21432376893218663,
+      "grad_norm": 380.0,
+      "learning_rate": 9.966512102259863e-06,
+      "loss": 22.5836,
       "step": 1760
     },
     {
-      "epoch": 0.10777075880965066,
-      "grad_norm": 452.0,
-      "learning_rate": 7.986528732045443e-06,
-      "loss": 23.7178,
+      "epoch": 0.21554151761930132,
+      "grad_norm": 365.75,
+      "learning_rate": 9.966321830113612e-06,
+      "loss": 22.8506,
       "step": 1770
     },
     {
-      "epoch": 0.108379633153208,
-      "grad_norm": 539.0,
-      "learning_rate": 7.986452623186945e-06,
-      "loss": 24.337,
+      "epoch": 0.216759266306416,
+      "grad_norm": 461.25,
+      "learning_rate": 9.966131557967361e-06,
+      "loss": 22.8096,
       "step": 1780
     },
     {
-      "epoch": 0.10898850749676535,
-      "grad_norm": 635.5,
-      "learning_rate": 7.986376514328443e-06,
-      "loss": 23.9829,
+      "epoch": 0.2179770149935307,
+      "grad_norm": 326.0,
+      "learning_rate": 9.96594128582111e-06,
+      "loss": 22.7708,
       "step": 1790
     },
     {
-      "epoch": 0.10959738184032271,
-      "grad_norm": 587.0,
-      "learning_rate": 7.986300405469944e-06,
-      "loss": 24.2237,
+      "epoch": 0.21919476368064542,
+      "grad_norm": 336.5,
+      "learning_rate": 9.96575101367486e-06,
+      "loss": 23.11,
       "step": 1800
     },
     {
-      "epoch": 0.11020625618388005,
-      "grad_norm": 507.75,
-      "learning_rate": 7.986224296611443e-06,
-      "loss": 24.1134,
+      "epoch": 0.2204125123677601,
+      "grad_norm": 400.25,
+      "learning_rate": 9.96556074152861e-06,
+      "loss": 22.8898,
       "step": 1810
     },
     {
-      "epoch": 0.1108151305274374,
-      "grad_norm": 391.75,
-      "learning_rate": 7.986148187752943e-06,
-      "loss": 24.2022,
+      "epoch": 0.2216302610548748,
+      "grad_norm": 372.0,
+      "learning_rate": 9.965370469382358e-06,
+      "loss": 22.9877,
       "step": 1820
     },
     {
-      "epoch": 0.11142400487099474,
-      "grad_norm": 654.5,
-      "learning_rate": 7.986072078894442e-06,
-      "loss": 23.9758,
+      "epoch": 0.22284800974198948,
+      "grad_norm": 330.75,
+      "learning_rate": 9.965180197236107e-06,
+      "loss": 23.0434,
       "step": 1830
     },
     {
-      "epoch": 0.1120328792145521,
-      "grad_norm": 389.25,
-      "learning_rate": 7.985995970035942e-06,
-      "loss": 23.7702,
+      "epoch": 0.2240657584291042,
+      "grad_norm": 302.75,
+      "learning_rate": 9.964989925089856e-06,
+      "loss": 23.0615,
       "step": 1840
     },
     {
-      "epoch": 0.11264175355810944,
-      "grad_norm": 525.5,
-      "learning_rate": 7.985919861177442e-06,
-      "loss": 23.9939,
+      "epoch": 0.2252835071162189,
+      "grad_norm": 489.25,
+      "learning_rate": 9.964799652943606e-06,
+      "loss": 22.9214,
       "step": 1850
     },
     {
-      "epoch": 0.11325062790166679,
-      "grad_norm": 944.5,
-      "learning_rate": 7.985843752318941e-06,
-      "loss": 23.9027,
+      "epoch": 0.22650125580333358,
+      "grad_norm": 355.75,
+      "learning_rate": 9.964609380797355e-06,
+      "loss": 22.929,
       "step": 1860
     },
     {
-      "epoch": 0.11385950224522415,
-      "grad_norm": 728.0,
-      "learning_rate": 7.98576764346044e-06,
-      "loss": 24.0871,
+      "epoch": 0.2277190044904483,
+      "grad_norm": 318.0,
+      "learning_rate": 9.964419108651104e-06,
+      "loss": 22.7594,
       "step": 1870
     },
     {
-      "epoch": 0.11446837658878149,
-      "grad_norm": 504.5,
-      "learning_rate": 7.985691534601942e-06,
-      "loss": 24.2737,
+      "epoch": 0.22893675317756298,
+      "grad_norm": 463.0,
+      "learning_rate": 9.964228836504853e-06,
+      "loss": 22.6917,
       "step": 1880
     },
     {
-      "epoch": 0.11507725093233884,
-      "grad_norm": 532.0,
-      "learning_rate": 7.98561542574344e-06,
-      "loss": 23.9491,
+      "epoch": 0.23015450186467767,
+      "grad_norm": 697.5,
+      "learning_rate": 9.964038564358602e-06,
+      "loss": 22.8248,
       "step": 1890
     },
     {
-      "epoch": 0.11568612527589618,
-      "grad_norm": 852.0,
-      "learning_rate": 7.985539316884941e-06,
-      "loss": 23.9135,
+      "epoch": 0.23137225055179236,
+      "grad_norm": 908.5,
+      "learning_rate": 9.963848292212352e-06,
+      "loss": 22.9424,
       "step": 1900
     },
     {
-      "epoch": 0.11629499961945354,
-      "grad_norm": 410.5,
-      "learning_rate": 7.985463208026439e-06,
-      "loss": 23.7667,
+      "epoch": 0.23258999923890708,
+      "grad_norm": 454.0,
+      "learning_rate": 9.963658020066101e-06,
+      "loss": 22.5561,
       "step": 1910
     },
     {
-      "epoch": 0.11690387396301088,
-      "grad_norm": 482.25,
-      "learning_rate": 7.98538709916794e-06,
-      "loss": 24.0657,
+      "epoch": 0.23380774792602177,
+      "grad_norm": 354.25,
+      "learning_rate": 9.96346774791985e-06,
+      "loss": 22.566,
       "step": 1920
     },
     {
-      "epoch": 0.11751274830656823,
-      "grad_norm": 513.0,
-      "learning_rate": 7.98531099030944e-06,
-      "loss": 24.1884,
+      "epoch": 0.23502549661313646,
+      "grad_norm": 351.5,
+      "learning_rate": 9.963277475773599e-06,
+      "loss": 22.8295,
       "step": 1930
     },
     {
-      "epoch": 0.11812162265012557,
-      "grad_norm": 404.75,
-      "learning_rate": 7.98523488145094e-06,
-      "loss": 23.8655,
+      "epoch": 0.23624324530025115,
+      "grad_norm": 306.5,
+      "learning_rate": 9.963087203627348e-06,
+      "loss": 22.791,
       "step": 1940
     },
     {
-      "epoch": 0.11873049699368293,
-      "grad_norm": 415.25,
-      "learning_rate": 7.985158772592439e-06,
-      "loss": 23.8893,
+      "epoch": 0.23746099398736586,
+      "grad_norm": 381.25,
+      "learning_rate": 9.962896931481099e-06,
+      "loss": 22.6901,
       "step": 1950
     },
     {
-      "epoch": 0.11933937133724028,
-      "grad_norm": 576.5,
-      "learning_rate": 7.985082663733938e-06,
-      "loss": 24.1121,
+      "epoch": 0.23867874267448055,
+      "grad_norm": 350.75,
+      "learning_rate": 9.962706659334847e-06,
+      "loss": 22.5385,
       "step": 1960
     },
     {
-      "epoch": 0.11994824568079762,
-      "grad_norm": 441.0,
-      "learning_rate": 7.985006554875438e-06,
-      "loss": 23.9494,
+      "epoch": 0.23989649136159524,
+      "grad_norm": 258.75,
+      "learning_rate": 9.962516387188596e-06,
+      "loss": 22.6354,
       "step": 1970
     },
     {
-      "epoch": 0.12055712002435498,
-      "grad_norm": 465.25,
-      "learning_rate": 7.984930446016937e-06,
-      "loss": 24.1016,
+      "epoch": 0.24111424004870996,
+      "grad_norm": 284.5,
+      "learning_rate": 9.962326115042345e-06,
+      "loss": 22.6697,
       "step": 1980
     },
     {
-      "epoch": 0.12116599436791232,
-      "grad_norm": 500.5,
-      "learning_rate": 7.984854337158437e-06,
-      "loss": 24.2461,
+      "epoch": 0.24233198873582465,
+      "grad_norm": 320.5,
+      "learning_rate": 9.962135842896094e-06,
+      "loss": 22.739,
       "step": 1990
     },
     {
-      "epoch": 0.12177486871146967,
-      "grad_norm": 538.0,
-      "learning_rate": 7.984778228299936e-06,
-      "loss": 24.4163,
+      "epoch": 0.24354973742293934,
+      "grad_norm": 541.0,
+      "learning_rate": 9.961945570749845e-06,
+      "loss": 22.7237,
       "step": 2000
     },
     {
-      "epoch": 0.12238374305502701,
-      "grad_norm": 428.25,
-      "learning_rate": 7.984702119441436e-06,
-      "loss": 23.7818,
+      "epoch": 0.24476748611005403,
+      "grad_norm": 392.0,
+      "learning_rate": 9.961755298603593e-06,
+      "loss": 22.7622,
       "step": 2010
     },
     {
-      "epoch": 0.12299261739858437,
-      "grad_norm": 441.5,
-      "learning_rate": 7.984626010582936e-06,
-      "loss": 23.9228,
+      "epoch": 0.24598523479716874,
+      "grad_norm": 703.5,
+      "learning_rate": 9.961565026457342e-06,
+      "loss": 22.6726,
       "step": 2020
     },
     {
-      "epoch": 0.12360149174214172,
-      "grad_norm": 537.0,
-      "learning_rate": 7.984549901724437e-06,
-      "loss": 23.7794,
+      "epoch": 0.24720298348428343,
+      "grad_norm": 384.25,
+      "learning_rate": 9.961374754311091e-06,
+      "loss": 22.7774,
       "step": 2030
     },
     {
-      "epoch": 0.12421036608569906,
-      "grad_norm": 434.5,
-      "learning_rate": 7.984473792865936e-06,
-      "loss": 23.7708,
+      "epoch": 0.24842073217139812,
+      "grad_norm": 415.5,
+      "learning_rate": 9.96118448216484e-06,
+      "loss": 23.0255,
       "step": 2040
     },
     {
-      "epoch": 0.1248192404292564,
-      "grad_norm": 494.75,
-      "learning_rate": 7.984397684007436e-06,
-      "loss": 23.8304,
+      "epoch": 0.2496384808585128,
+      "grad_norm": 519.5,
+      "learning_rate": 9.96099421001859e-06,
+      "loss": 22.9514,
       "step": 2050
     },
     {
-      "epoch": 0.12542811477281376,
-      "grad_norm": 446.5,
-      "learning_rate": 7.984321575148935e-06,
-      "loss": 23.8985,
+      "epoch": 0.25085622954562753,
+      "grad_norm": 437.75,
+      "learning_rate": 9.96080393787234e-06,
+      "loss": 22.6842,
       "step": 2060
     },
     {
-      "epoch": 0.1260369891163711,
-      "grad_norm": 368.75,
-      "learning_rate": 7.984245466290435e-06,
-      "loss": 23.742,
+      "epoch": 0.2520739782327422,
+      "grad_norm": 554.5,
+      "learning_rate": 9.960613665726088e-06,
+      "loss": 22.5928,
       "step": 2070
     },
     {
-      "epoch": 0.12664586345992845,
-      "grad_norm": 393.0,
-      "learning_rate": 7.984169357431935e-06,
-      "loss": 23.7806,
+      "epoch": 0.2532917269198569,
+      "grad_norm": 702.5,
+      "learning_rate": 9.960423393579837e-06,
+      "loss": 22.7435,
       "step": 2080
     },
     {
-      "epoch": 0.1272547378034858,
-      "grad_norm": 440.75,
-      "learning_rate": 7.984093248573434e-06,
-      "loss": 23.8871,
+      "epoch": 0.2545094756069716,
+      "grad_norm": 519.5,
+      "learning_rate": 9.960233121433586e-06,
+      "loss": 22.9607,
       "step": 2090
     },
     {
-      "epoch": 0.12786361214704314,
-      "grad_norm": 556.0,
-      "learning_rate": 7.984017139714934e-06,
-      "loss": 23.8715,
+      "epoch": 0.2557272242940863,
+      "grad_norm": 476.0,
+      "learning_rate": 9.960042849287337e-06,
+      "loss": 22.789,
       "step": 2100
     },
     {
-      "epoch": 0.1284724864906005,
-      "grad_norm": 496.5,
-      "learning_rate": 7.983941030856433e-06,
-      "loss": 23.5876,
+      "epoch": 0.256944972981201,
+      "grad_norm": 492.5,
+      "learning_rate": 9.959852577141086e-06,
+      "loss": 22.5661,
       "step": 2110
     },
     {
-      "epoch": 0.12908136083415786,
-      "grad_norm": 502.75,
-      "learning_rate": 7.983864921997933e-06,
-      "loss": 23.7611,
+      "epoch": 0.2581627216683157,
+      "grad_norm": 514.0,
+      "learning_rate": 9.959662304994834e-06,
+      "loss": 22.9395,
       "step": 2120
     },
     {
-      "epoch": 0.1296902351777152,
-      "grad_norm": 830.5,
-      "learning_rate": 7.983788813139434e-06,
-      "loss": 23.8297,
+      "epoch": 0.2593804703554304,
+      "grad_norm": 577.0,
+      "learning_rate": 9.959472032848583e-06,
+      "loss": 22.6245,
       "step": 2130
     },
     {
-      "epoch": 0.13029910952127255,
-      "grad_norm": 553.5,
-      "learning_rate": 7.983712704280932e-06,
-      "loss": 23.8472,
+      "epoch": 0.2605982190425451,
+      "grad_norm": 326.75,
+      "learning_rate": 9.959281760702332e-06,
+      "loss": 22.5902,
       "step": 2140
     },
     {
-      "epoch": 0.1309079838648299,
-      "grad_norm": 620.5,
-      "learning_rate": 7.983636595422433e-06,
-      "loss": 24.02,
+      "epoch": 0.2618159677296598,
+      "grad_norm": 367.25,
+      "learning_rate": 9.959091488556083e-06,
+      "loss": 22.6378,
       "step": 2150
     },
     {
-      "epoch": 0.13151685820838724,
-      "grad_norm": 799.5,
-      "learning_rate": 7.983560486563931e-06,
-      "loss": 23.7774,
+      "epoch": 0.2630337164167745,
+      "grad_norm": 365.0,
+      "learning_rate": 9.958901216409832e-06,
+      "loss": 22.8368,
       "step": 2160
     },
     {
-      "epoch": 0.1321257325519446,
-      "grad_norm": 459.25,
-      "learning_rate": 7.983484377705432e-06,
-      "loss": 23.7076,
+      "epoch": 0.2642514651038892,
+      "grad_norm": 293.25,
+      "learning_rate": 9.95871094426358e-06,
+      "loss": 22.6077,
       "step": 2170
     },
     {
-      "epoch": 0.13273460689550193,
-      "grad_norm": 484.25,
-      "learning_rate": 7.983408268846932e-06,
-      "loss": 24.0992,
+      "epoch": 0.26546921379100386,
+      "grad_norm": 734.5,
+      "learning_rate": 9.95852067211733e-06,
+      "loss": 22.5928,
       "step": 2180
     },
     {
-      "epoch": 0.13334348123905929,
-      "grad_norm": 392.5,
-      "learning_rate": 7.983332159988431e-06,
-      "loss": 23.6718,
+      "epoch": 0.26668696247811857,
+      "grad_norm": 480.75,
+      "learning_rate": 9.958330399971078e-06,
+      "loss": 22.7898,
       "step": 2190
     },
     {
-      "epoch": 0.13395235558261664,
-      "grad_norm": 460.25,
-      "learning_rate": 7.98325605112993e-06,
-      "loss": 24.0149,
+      "epoch": 0.2679047111652333,
+      "grad_norm": 564.5,
+      "learning_rate": 9.958140127824829e-06,
+      "loss": 23.1092,
       "step": 2200
     },
     {
-      "epoch": 0.13456122992617398,
-      "grad_norm": 403.0,
-      "learning_rate": 7.98317994227143e-06,
-      "loss": 23.6711,
+      "epoch": 0.26912245985234795,
+      "grad_norm": 371.0,
+      "learning_rate": 9.957949855678578e-06,
+      "loss": 22.576,
       "step": 2210
     },
     {
-      "epoch": 0.13517010426973133,
-      "grad_norm": 525.0,
-      "learning_rate": 7.98310383341293e-06,
-      "loss": 24.3896,
+      "epoch": 0.27034020853946267,
+      "grad_norm": 304.5,
+      "learning_rate": 9.957759583532327e-06,
+      "loss": 22.851,
       "step": 2220
     },
     {
-      "epoch": 0.1357789786132887,
-      "grad_norm": 585.5,
-      "learning_rate": 7.98302772455443e-06,
-      "loss": 23.6051,
+      "epoch": 0.2715579572265774,
+      "grad_norm": 324.25,
+      "learning_rate": 9.957569311386075e-06,
+      "loss": 22.7072,
       "step": 2230
     },
     {
-      "epoch": 0.13638785295684602,
-      "grad_norm": 563.0,
-      "learning_rate": 7.982951615695929e-06,
-      "loss": 23.8391,
+      "epoch": 0.27277570591369205,
+      "grad_norm": 293.5,
+      "learning_rate": 9.957379039239824e-06,
+      "loss": 22.6171,
       "step": 2240
     },
     {
-      "epoch": 0.13699672730040338,
-      "grad_norm": 456.0,
-      "learning_rate": 7.982875506837429e-06,
-      "loss": 23.7537,
+      "epoch": 0.27399345460080676,
+      "grad_norm": 600.0,
+      "learning_rate": 9.957188767093575e-06,
+      "loss": 22.5538,
       "step": 2250
     },
     {
-      "epoch": 0.13760560164396074,
-      "grad_norm": 476.0,
-      "learning_rate": 7.982799397978928e-06,
-      "loss": 23.8925,
+      "epoch": 0.2752112032879215,
+      "grad_norm": 322.75,
+      "learning_rate": 9.956998494947324e-06,
+      "loss": 23.0903,
       "step": 2260
     },
     {
-      "epoch": 0.13821447598751807,
-      "grad_norm": 430.75,
-      "learning_rate": 7.98272328912043e-06,
-      "loss": 23.7052,
+      "epoch": 0.27642895197503614,
+      "grad_norm": 437.0,
+      "learning_rate": 9.956808222801073e-06,
+      "loss": 22.719,
       "step": 2270
     },
     {
-      "epoch": 0.13882335033107543,
-      "grad_norm": 739.5,
-      "learning_rate": 7.982647180261929e-06,
-      "loss": 24.0857,
+      "epoch": 0.27764670066215086,
+      "grad_norm": 248.75,
+      "learning_rate": 9.956617950654821e-06,
+      "loss": 22.6685,
       "step": 2280
     },
     {
-      "epoch": 0.13943222467463276,
-      "grad_norm": 560.5,
-      "learning_rate": 7.982571071403428e-06,
-      "loss": 24.0643,
+      "epoch": 0.2788644493492655,
+      "grad_norm": 407.0,
+      "learning_rate": 9.95642767850857e-06,
+      "loss": 22.9552,
       "step": 2290
     },
     {
-      "epoch": 0.14004109901819012,
-      "grad_norm": 457.0,
-      "learning_rate": 7.982494962544928e-06,
-      "loss": 23.7443,
+      "epoch": 0.28008219803638024,
+      "grad_norm": 501.25,
+      "learning_rate": 9.956237406362321e-06,
+      "loss": 22.716,
       "step": 2300
     },
     {
-      "epoch": 0.14064997336174748,
-      "grad_norm": 705.0,
-      "learning_rate": 7.982418853686428e-06,
-      "loss": 23.7807,
+      "epoch": 0.28129994672349495,
+      "grad_norm": 466.25,
+      "learning_rate": 9.95604713421607e-06,
+      "loss": 22.9931,
       "step": 2310
     },
     {
-      "epoch": 0.1412588477053048,
-      "grad_norm": 491.75,
-      "learning_rate": 7.982342744827927e-06,
-      "loss": 23.6736,
+      "epoch": 0.2825176954106096,
+      "grad_norm": 345.5,
+      "learning_rate": 9.955856862069819e-06,
+      "loss": 22.9603,
       "step": 2320
     },
     {
-      "epoch": 0.14186772204886217,
-      "grad_norm": 507.25,
-      "learning_rate": 7.982266635969427e-06,
-      "loss": 23.3962,
+      "epoch": 0.28373544409772433,
+      "grad_norm": 375.75,
+      "learning_rate": 9.955666589923568e-06,
+      "loss": 22.9,
       "step": 2330
     },
     {
-      "epoch": 0.14247659639241952,
-      "grad_norm": 466.75,
-      "learning_rate": 7.982190527110926e-06,
-      "loss": 23.7991,
+      "epoch": 0.28495319278483905,
+      "grad_norm": 360.75,
+      "learning_rate": 9.955476317777318e-06,
+      "loss": 22.6346,
       "step": 2340
     },
     {
-      "epoch": 0.14308547073597686,
-      "grad_norm": 612.0,
-      "learning_rate": 7.982114418252426e-06,
-      "loss": 23.9104,
+      "epoch": 0.2861709414719537,
+      "grad_norm": 332.75,
+      "learning_rate": 9.955286045631067e-06,
+      "loss": 22.8878,
       "step": 2350
     },
     {
-      "epoch": 0.1436943450795342,
-      "grad_norm": 421.5,
-      "learning_rate": 7.982038309393925e-06,
-      "loss": 23.8134,
+      "epoch": 0.2873886901590684,
+      "grad_norm": 562.5,
+      "learning_rate": 9.955095773484816e-06,
+      "loss": 22.8146,
       "step": 2360
     },
     {
-      "epoch": 0.14430321942309157,
-      "grad_norm": 523.0,
-      "learning_rate": 7.981962200535425e-06,
-      "loss": 23.702,
+      "epoch": 0.28860643884618314,
+      "grad_norm": 595.5,
+      "learning_rate": 9.954905501338565e-06,
+      "loss": 22.6711,
       "step": 2370
     },
     {
-      "epoch": 0.1449120937666489,
-      "grad_norm": 736.5,
-      "learning_rate": 7.981886091676926e-06,
-      "loss": 23.6135,
+      "epoch": 0.2898241875332978,
+      "grad_norm": 425.25,
+      "learning_rate": 9.954715229192314e-06,
+      "loss": 22.8447,
       "step": 2380
     },
     {
-      "epoch": 0.14552096811020626,
-      "grad_norm": 848.0,
-      "learning_rate": 7.981809982818424e-06,
-      "loss": 23.8981,
+      "epoch": 0.2910419362204125,
+      "grad_norm": 633.0,
+      "learning_rate": 9.954524957046064e-06,
+      "loss": 22.6317,
       "step": 2390
     },
     {
-      "epoch": 0.1461298424537636,
-      "grad_norm": 489.75,
-      "learning_rate": 7.981733873959925e-06,
-      "loss": 23.5943,
+      "epoch": 0.2922596849075272,
+      "grad_norm": 336.25,
+      "learning_rate": 9.954334684899813e-06,
+      "loss": 22.7242,
       "step": 2400
     },
     {
-      "epoch": 0.14673871679732095,
-      "grad_norm": 519.5,
-      "learning_rate": 7.981657765101423e-06,
-      "loss": 23.3523,
+      "epoch": 0.2934774335946419,
+      "grad_norm": 445.0,
+      "learning_rate": 9.954144412753562e-06,
+      "loss": 22.8296,
       "step": 2410
     },
     {
-      "epoch": 0.1473475911408783,
-      "grad_norm": 467.25,
-      "learning_rate": 7.981581656242924e-06,
-      "loss": 23.8777,
+      "epoch": 0.2946951822817566,
+      "grad_norm": 339.0,
+      "learning_rate": 9.95395414060731e-06,
+      "loss": 22.6007,
       "step": 2420
     },
     {
-      "epoch": 0.14795646548443564,
-      "grad_norm": 443.0,
-      "learning_rate": 7.981505547384424e-06,
-      "loss": 23.6681,
+      "epoch": 0.2959129309688713,
+      "grad_norm": 393.75,
+      "learning_rate": 9.95376386846106e-06,
+      "loss": 22.9184,
       "step": 2430
     },
     {
-      "epoch": 0.148565339827993,
-      "grad_norm": 418.0,
-      "learning_rate": 7.981429438525923e-06,
-      "loss": 23.6622,
+      "epoch": 0.297130679655986,
+      "grad_norm": 611.0,
+      "learning_rate": 9.95357359631481e-06,
+      "loss": 22.8158,
       "step": 2440
     },
     {
-      "epoch": 0.14917421417155036,
-      "grad_norm": 469.75,
-      "learning_rate": 7.981353329667423e-06,
-      "loss": 23.9613,
+      "epoch": 0.2983484283431007,
+      "grad_norm": 264.5,
+      "learning_rate": 9.953383324168559e-06,
+      "loss": 22.4974,
       "step": 2450
     },
     {
-      "epoch": 0.1497830885151077,
-      "grad_norm": 475.5,
-      "learning_rate": 7.981277220808922e-06,
-      "loss": 24.2091,
+      "epoch": 0.2995661770302154,
+      "grad_norm": 269.25,
+      "learning_rate": 9.953193052022308e-06,
+      "loss": 22.5819,
       "step": 2460
     },
     {
-      "epoch": 0.15039196285866505,
-      "grad_norm": 469.5,
-      "learning_rate": 7.981201111950422e-06,
-      "loss": 23.5705,
+      "epoch": 0.3007839257173301,
+      "grad_norm": 363.25,
+      "learning_rate": 9.953002779876057e-06,
+      "loss": 22.5765,
       "step": 2470
     },
     {
-      "epoch": 0.1510008372022224,
-      "grad_norm": 524.0,
-      "learning_rate": 7.981125003091922e-06,
-      "loss": 23.7235,
+      "epoch": 0.3020016744044448,
+      "grad_norm": 340.25,
+      "learning_rate": 9.952812507729807e-06,
+      "loss": 22.8563,
       "step": 2480
     },
     {
-      "epoch": 0.15160971154577974,
-      "grad_norm": 470.0,
-      "learning_rate": 7.981048894233421e-06,
-      "loss": 23.7459,
+      "epoch": 0.30321942309155947,
+      "grad_norm": 342.0,
+      "learning_rate": 9.952622235583556e-06,
+      "loss": 23.1226,
       "step": 2490
     },
     {
-      "epoch": 0.1522185858893371,
-      "grad_norm": 483.25,
-      "learning_rate": 7.980972785374922e-06,
-      "loss": 23.2756,
+      "epoch": 0.3044371717786742,
+      "grad_norm": 340.5,
+      "learning_rate": 9.952431963437305e-06,
+      "loss": 22.5687,
       "step": 2500
     },
     {
-      "epoch": 0.15282746023289442,
-      "grad_norm": 397.0,
-      "learning_rate": 7.98089667651642e-06,
-      "loss": 23.3925,
+      "epoch": 0.30565492046578885,
+      "grad_norm": 392.5,
+      "learning_rate": 9.952241691291054e-06,
+      "loss": 22.8702,
       "step": 2510
     },
     {
-      "epoch": 0.15343633457645178,
-      "grad_norm": 465.25,
-      "learning_rate": 7.980820567657921e-06,
-      "loss": 23.8293,
+      "epoch": 0.30687266915290357,
+      "grad_norm": 364.25,
+      "learning_rate": 9.952051419144803e-06,
+      "loss": 22.6302,
       "step": 2520
     },
     {
-      "epoch": 0.15404520892000914,
-      "grad_norm": 519.5,
-      "learning_rate": 7.980744458799421e-06,
-      "loss": 24.0543,
+      "epoch": 0.3080904178400183,
+      "grad_norm": 549.0,
+      "learning_rate": 9.951861146998553e-06,
+      "loss": 23.2274,
       "step": 2530
     },
     {
-      "epoch": 0.15465408326356647,
-      "grad_norm": 450.25,
-      "learning_rate": 7.98066834994092e-06,
-      "loss": 23.3999,
+      "epoch": 0.30930816652713294,
+      "grad_norm": 630.5,
+      "learning_rate": 9.951670874852302e-06,
+      "loss": 22.7162,
       "step": 2540
     },
     {
-      "epoch": 0.15526295760712383,
-      "grad_norm": 621.5,
-      "learning_rate": 7.98059224108242e-06,
-      "loss": 23.8093,
+      "epoch": 0.31052591521424766,
+      "grad_norm": 472.0,
+      "learning_rate": 9.951480602706051e-06,
+      "loss": 22.4515,
       "step": 2550
     },
     {
-      "epoch": 0.1558718319506812,
-      "grad_norm": 370.75,
-      "learning_rate": 7.98051613222392e-06,
-      "loss": 24.1084,
+      "epoch": 0.3117436639013624,
+      "grad_norm": 550.5,
+      "learning_rate": 9.9512903305598e-06,
+      "loss": 22.7794,
       "step": 2560
     },
     {
-      "epoch": 0.15648070629423852,
-      "grad_norm": 705.0,
-      "learning_rate": 7.98044002336542e-06,
-      "loss": 23.6988,
+      "epoch": 0.31296141258847704,
+      "grad_norm": 411.25,
+      "learning_rate": 9.95110005841355e-06,
+      "loss": 22.6228,
       "step": 2570
     },
     {
-      "epoch": 0.15708958063779588,
-      "grad_norm": 492.25,
-      "learning_rate": 7.980363914506919e-06,
-      "loss": 23.6,
+      "epoch": 0.31417916127559176,
+      "grad_norm": 640.0,
+      "learning_rate": 9.9509097862673e-06,
+      "loss": 22.6809,
       "step": 2580
     },
     {
-      "epoch": 0.15769845498135324,
-      "grad_norm": 408.5,
-      "learning_rate": 7.980287805648418e-06,
-      "loss": 23.6126,
+      "epoch": 0.3153969099627065,
+      "grad_norm": 538.5,
+      "learning_rate": 9.950719514121048e-06,
+      "loss": 22.4981,
       "step": 2590
     },
     {
-      "epoch": 0.15830732932491057,
-      "grad_norm": 499.5,
-      "learning_rate": 7.980211696789918e-06,
-      "loss": 23.6962,
+      "epoch": 0.31661465864982113,
+      "grad_norm": 685.5,
+      "learning_rate": 9.950529241974797e-06,
+      "loss": 22.6211,
       "step": 2600
     },
     {
-      "epoch": 0.15891620366846793,
-      "grad_norm": 447.5,
-      "learning_rate": 7.980135587931417e-06,
-      "loss": 23.7325,
+      "epoch": 0.31783240733693585,
+      "grad_norm": 310.0,
+      "learning_rate": 9.950338969828546e-06,
+      "loss": 22.239,
       "step": 2610
     },
     {
-      "epoch": 0.15952507801202526,
-      "grad_norm": 390.75,
-      "learning_rate": 7.980059479072917e-06,
-      "loss": 23.474,
+      "epoch": 0.3190501560240505,
+      "grad_norm": 295.0,
+      "learning_rate": 9.950148697682297e-06,
+      "loss": 22.8624,
       "step": 2620
     },
     {
-      "epoch": 0.16013395235558261,
-      "grad_norm": 571.5,
-      "learning_rate": 7.979983370214418e-06,
-      "loss": 23.73,
+      "epoch": 0.32026790471116523,
+      "grad_norm": 528.5,
+      "learning_rate": 9.949958425536046e-06,
+      "loss": 22.7837,
       "step": 2630
     },
     {
-      "epoch": 0.16074282669913997,
-      "grad_norm": 501.5,
-      "learning_rate": 7.979907261355916e-06,
-      "loss": 23.5425,
+      "epoch": 0.32148565339827995,
+      "grad_norm": 433.5,
+      "learning_rate": 9.949768153389794e-06,
+      "loss": 22.6294,
       "step": 2640
     },
     {
-      "epoch": 0.1613517010426973,
-      "grad_norm": 925.0,
-      "learning_rate": 7.979831152497417e-06,
-      "loss": 23.9183,
+      "epoch": 0.3227034020853946,
+      "grad_norm": 513.0,
+      "learning_rate": 9.949577881243543e-06,
+      "loss": 23.0151,
       "step": 2650
     },
     {
-      "epoch": 0.16196057538625466,
-      "grad_norm": 386.75,
-      "learning_rate": 7.979755043638917e-06,
-      "loss": 23.5385,
+      "epoch": 0.3239211507725093,
+      "grad_norm": 499.75,
+      "learning_rate": 9.949387609097292e-06,
+      "loss": 22.5745,
       "step": 2660
     },
     {
-      "epoch": 0.16256944972981202,
-      "grad_norm": 629.5,
-      "learning_rate": 7.979678934780416e-06,
-      "loss": 23.3944,
+      "epoch": 0.32513889945962404,
+      "grad_norm": 429.75,
+      "learning_rate": 9.949197336951043e-06,
+      "loss": 22.5822,
       "step": 2670
     },
     {
-      "epoch": 0.16317832407336935,
-      "grad_norm": 797.5,
-      "learning_rate": 7.979602825921916e-06,
-      "loss": 23.7524,
+      "epoch": 0.3263566481467387,
+      "grad_norm": 472.5,
+      "learning_rate": 9.949007064804792e-06,
+      "loss": 22.6606,
       "step": 2680
     },
     {
-      "epoch": 0.1637871984169267,
-      "grad_norm": 679.0,
-      "learning_rate": 7.979526717063415e-06,
-      "loss": 23.3417,
+      "epoch": 0.3275743968338534,
+      "grad_norm": 418.75,
+      "learning_rate": 9.94881679265854e-06,
+      "loss": 22.6881,
       "step": 2690
     },
     {
-      "epoch": 0.16439607276048407,
-      "grad_norm": 599.5,
-      "learning_rate": 7.979450608204915e-06,
-      "loss": 24.1741,
+      "epoch": 0.32879214552096814,
+      "grad_norm": 672.0,
+      "learning_rate": 9.94862652051229e-06,
+      "loss": 22.7059,
       "step": 2700
     },
     {
-      "epoch": 0.1650049471040414,
-      "grad_norm": 1185.0,
-      "learning_rate": 7.979374499346415e-06,
-      "loss": 23.5533,
+      "epoch": 0.3300098942080828,
+      "grad_norm": 332.75,
+      "learning_rate": 9.94843624836604e-06,
+      "loss": 22.6674,
       "step": 2710
     },
     {
-      "epoch": 0.16561382144759876,
-      "grad_norm": 403.0,
-      "learning_rate": 7.979298390487914e-06,
-      "loss": 23.7548,
+      "epoch": 0.3312276428951975,
+      "grad_norm": 681.5,
+      "learning_rate": 9.948245976219789e-06,
+      "loss": 22.8335,
       "step": 2720
     },
     {
-      "epoch": 0.1662226957911561,
-      "grad_norm": 597.0,
-      "learning_rate": 7.979222281629415e-06,
-      "loss": 24.4425,
+      "epoch": 0.3324453915823122,
+      "grad_norm": 340.75,
+      "learning_rate": 9.948055704073538e-06,
+      "loss": 22.8004,
       "step": 2730
     },
     {
-      "epoch": 0.16683157013471345,
-      "grad_norm": 648.5,
-      "learning_rate": 7.979146172770913e-06,
-      "loss": 23.3017,
+      "epoch": 0.3336631402694269,
+      "grad_norm": 726.5,
+      "learning_rate": 9.947865431927287e-06,
+      "loss": 22.6123,
       "step": 2740
     },
     {
-      "epoch": 0.1674404444782708,
-      "grad_norm": 910.5,
-      "learning_rate": 7.979070063912414e-06,
-      "loss": 23.7393,
+      "epoch": 0.3348808889565416,
+      "grad_norm": 467.5,
+      "learning_rate": 9.947675159781035e-06,
+      "loss": 22.7943,
       "step": 2750
     },
     {
-      "epoch": 0.16804931882182814,
-      "grad_norm": 901.0,
-      "learning_rate": 7.978993955053912e-06,
-      "loss": 23.8384,
+      "epoch": 0.3360986376436563,
+      "grad_norm": 473.75,
+      "learning_rate": 9.947484887634786e-06,
+      "loss": 22.5676,
       "step": 2760
     },
     {
-      "epoch": 0.1686581931653855,
-      "grad_norm": 569.0,
-      "learning_rate": 7.978917846195414e-06,
-      "loss": 23.6584,
+      "epoch": 0.337316386330771,
+      "grad_norm": 430.0,
+      "learning_rate": 9.947294615488535e-06,
+      "loss": 22.7151,
       "step": 2770
     },
     {
-      "epoch": 0.16926706750894285,
-      "grad_norm": 438.25,
-      "learning_rate": 7.978841737336913e-06,
-      "loss": 23.7324,
+      "epoch": 0.3385341350178857,
+      "grad_norm": 285.5,
+      "learning_rate": 9.947104343342284e-06,
+      "loss": 22.8741,
       "step": 2780
     },
     {
-      "epoch": 0.16987594185250018,
-      "grad_norm": 535.0,
-      "learning_rate": 7.978765628478413e-06,
-      "loss": 23.9481,
+      "epoch": 0.33975188370500037,
+      "grad_norm": 396.5,
+      "learning_rate": 9.946914071196033e-06,
+      "loss": 22.5662,
       "step": 2790
     },
     {
-      "epoch": 0.17048481619605754,
-      "grad_norm": 678.0,
-      "learning_rate": 7.978689519619912e-06,
-      "loss": 23.4555,
+      "epoch": 0.3409696323921151,
+      "grad_norm": 307.25,
+      "learning_rate": 9.946723799049781e-06,
+      "loss": 22.8248,
       "step": 2800
     },
     {
-      "epoch": 0.17109369053961487,
-      "grad_norm": 645.5,
-      "learning_rate": 7.978613410761412e-06,
-      "loss": 23.6862,
+      "epoch": 0.34218738107922975,
+      "grad_norm": 676.0,
+      "learning_rate": 9.946533526903532e-06,
+      "loss": 22.8945,
       "step": 2810
     },
     {
-      "epoch": 0.17170256488317223,
-      "grad_norm": 445.25,
-      "learning_rate": 7.978537301902911e-06,
-      "loss": 23.776,
+      "epoch": 0.34340512976634446,
+      "grad_norm": 429.5,
+      "learning_rate": 9.946343254757281e-06,
+      "loss": 22.7002,
       "step": 2820
     },
     {
-      "epoch": 0.1723114392267296,
-      "grad_norm": 480.25,
-      "learning_rate": 7.978461193044411e-06,
-      "loss": 23.5924,
+      "epoch": 0.3446228784534592,
+      "grad_norm": 327.25,
+      "learning_rate": 9.94615298261103e-06,
+      "loss": 22.7292,
       "step": 2830
     },
     {
-      "epoch": 0.17292031357028692,
-      "grad_norm": 436.25,
-      "learning_rate": 7.97838508418591e-06,
-      "loss": 23.6267,
+      "epoch": 0.34584062714057384,
+      "grad_norm": 431.75,
+      "learning_rate": 9.945962710464779e-06,
+      "loss": 22.413,
       "step": 2840
     },
     {
-      "epoch": 0.17352918791384428,
-      "grad_norm": 579.0,
-      "learning_rate": 7.97830897532741e-06,
-      "loss": 23.7811,
+      "epoch": 0.34705837582768856,
+      "grad_norm": 291.25,
+      "learning_rate": 9.94577243831853e-06,
+      "loss": 22.6488,
       "step": 2850
     },
     {
-      "epoch": 0.17413806225740164,
-      "grad_norm": 535.5,
-      "learning_rate": 7.97823286646891e-06,
-      "loss": 23.8594,
+      "epoch": 0.3482761245148033,
+      "grad_norm": 408.25,
+      "learning_rate": 9.945582166172278e-06,
+      "loss": 22.8355,
       "step": 2860
     },
     {
-      "epoch": 0.17474693660095897,
-      "grad_norm": 498.0,
-      "learning_rate": 7.978156757610409e-06,
-      "loss": 23.2003,
+      "epoch": 0.34949387320191794,
+      "grad_norm": 403.5,
+      "learning_rate": 9.945391894026027e-06,
+      "loss": 22.7133,
       "step": 2870
     },
     {
-      "epoch": 0.17535581094451633,
-      "grad_norm": 623.0,
-      "learning_rate": 7.97808064875191e-06,
-      "loss": 23.4851,
+      "epoch": 0.35071162188903265,
+      "grad_norm": 371.5,
+      "learning_rate": 9.945201621879776e-06,
+      "loss": 22.5127,
       "step": 2880
     },
     {
-      "epoch": 0.17596468528807369,
-      "grad_norm": 467.0,
-      "learning_rate": 7.97800453989341e-06,
-      "loss": 23.7073,
+      "epoch": 0.35192937057614737,
+      "grad_norm": 490.25,
+      "learning_rate": 9.945011349733525e-06,
+      "loss": 22.7957,
       "step": 2890
     },
     {
-      "epoch": 0.17657355963163102,
-      "grad_norm": 474.75,
-      "learning_rate": 7.97792843103491e-06,
-      "loss": 23.5397,
+      "epoch": 0.35314711926326203,
+      "grad_norm": 472.75,
+      "learning_rate": 9.944821077587275e-06,
+      "loss": 22.3429,
       "step": 2900
     },
     {
-      "epoch": 0.17718243397518837,
-      "grad_norm": 399.5,
-      "learning_rate": 7.977852322176409e-06,
-      "loss": 23.4913,
+      "epoch": 0.35436486795037675,
+      "grad_norm": 540.5,
+      "learning_rate": 9.944630805441024e-06,
+      "loss": 22.7809,
       "step": 2910
     },
     {
-      "epoch": 0.1777913083187457,
-      "grad_norm": 454.5,
-      "learning_rate": 7.977776213317908e-06,
-      "loss": 23.2491,
+      "epoch": 0.3555826166374914,
+      "grad_norm": 330.25,
+      "learning_rate": 9.944440533294773e-06,
+      "loss": 22.3547,
       "step": 2920
     },
     {
-      "epoch": 0.17840018266230306,
-      "grad_norm": 628.5,
-      "learning_rate": 7.977700104459408e-06,
-      "loss": 23.9576,
+      "epoch": 0.35680036532460613,
+      "grad_norm": 766.5,
+      "learning_rate": 9.944250261148522e-06,
+      "loss": 22.3245,
       "step": 2930
     },
     {
-      "epoch": 0.17900905700586042,
-      "grad_norm": 503.0,
-      "learning_rate": 7.977623995600908e-06,
-      "loss": 23.7708,
+      "epoch": 0.35801811401172084,
+      "grad_norm": 316.0,
+      "learning_rate": 9.94405998900227e-06,
+      "loss": 22.2698,
       "step": 2940
     },
     {
-      "epoch": 0.17961793134941775,
-      "grad_norm": 866.0,
-      "learning_rate": 7.977547886742407e-06,
-      "loss": 23.7517,
+      "epoch": 0.3592358626988355,
+      "grad_norm": 320.25,
+      "learning_rate": 9.94386971685602e-06,
+      "loss": 22.6861,
       "step": 2950
     },
     {
-      "epoch": 0.1802268056929751,
-      "grad_norm": 515.5,
-      "learning_rate": 7.977471777883907e-06,
-      "loss": 23.4176,
+      "epoch": 0.3604536113859502,
+      "grad_norm": 377.0,
+      "learning_rate": 9.94367944470977e-06,
+      "loss": 22.624,
       "step": 2960
     },
     {
-      "epoch": 0.18083568003653247,
-      "grad_norm": 581.0,
-      "learning_rate": 7.977395669025406e-06,
-      "loss": 23.7534,
+      "epoch": 0.36167136007306494,
+      "grad_norm": 396.0,
+      "learning_rate": 9.943489172563519e-06,
+      "loss": 23.0753,
       "step": 2970
     },
     {
-      "epoch": 0.1814445543800898,
-      "grad_norm": 594.0,
-      "learning_rate": 7.977319560166907e-06,
-      "loss": 23.2326,
+      "epoch": 0.3628891087601796,
+      "grad_norm": 373.25,
+      "learning_rate": 9.943298900417268e-06,
+      "loss": 22.5845,
       "step": 2980
     },
     {
-      "epoch": 0.18205342872364716,
-      "grad_norm": 472.25,
-      "learning_rate": 7.977243451308405e-06,
-      "loss": 23.8834,
+      "epoch": 0.3641068574472943,
+      "grad_norm": 312.0,
+      "learning_rate": 9.943108628271017e-06,
+      "loss": 22.4724,
       "step": 2990
     },
     {
-      "epoch": 0.18266230306720452,
-      "grad_norm": 773.0,
-      "learning_rate": 7.977167342449907e-06,
-      "loss": 23.5239,
+      "epoch": 0.36532460613440904,
+      "grad_norm": 433.75,
+      "learning_rate": 9.942918356124766e-06,
+      "loss": 22.6247,
       "step": 3000
     },
     {
-      "epoch": 0.18327117741076185,
-      "grad_norm": 663.5,
-      "learning_rate": 7.977091233591404e-06,
-      "loss": 23.4971,
+      "epoch": 0.3665423548215237,
+      "grad_norm": 422.0,
+      "learning_rate": 9.942728083978516e-06,
+      "loss": 22.4805,
       "step": 3010
     },
     {
-      "epoch": 0.1838800517543192,
-      "grad_norm": 632.0,
-      "learning_rate": 7.977015124732906e-06,
-      "loss": 23.6645,
+      "epoch": 0.3677601035086384,
+      "grad_norm": 409.75,
+      "learning_rate": 9.942537811832265e-06,
+      "loss": 22.3236,
       "step": 3020
     },
     {
-      "epoch": 0.18448892609787654,
-      "grad_norm": 452.75,
-      "learning_rate": 7.976939015874405e-06,
-      "loss": 23.514,
+      "epoch": 0.3689778521957531,
+      "grad_norm": 488.5,
+      "learning_rate": 9.942347539686014e-06,
+      "loss": 22.579,
       "step": 3030
     },
     {
-      "epoch": 0.1850978004414339,
-      "grad_norm": 816.5,
-      "learning_rate": 7.976862907015905e-06,
-      "loss": 23.4402,
+      "epoch": 0.3701956008828678,
+      "grad_norm": 335.25,
+      "learning_rate": 9.942157267539763e-06,
+      "loss": 22.6445,
       "step": 3040
     },
     {
-      "epoch": 0.18570667478499125,
-      "grad_norm": 821.0,
-      "learning_rate": 7.976786798157404e-06,
-      "loss": 23.4769,
+      "epoch": 0.3714133495699825,
+      "grad_norm": 454.0,
+      "learning_rate": 9.941966995393512e-06,
+      "loss": 22.8879,
       "step": 3050
     },
     {
-      "epoch": 0.18631554912854859,
-      "grad_norm": 604.0,
-      "learning_rate": 7.976710689298904e-06,
-      "loss": 23.7916,
+      "epoch": 0.37263109825709717,
+      "grad_norm": 745.5,
+      "learning_rate": 9.941776723247262e-06,
+      "loss": 22.7475,
       "step": 3060
     },
     {
-      "epoch": 0.18692442347210594,
-      "grad_norm": 743.0,
-      "learning_rate": 7.976634580440403e-06,
-      "loss": 23.257,
+      "epoch": 0.3738488469442119,
+      "grad_norm": 779.0,
+      "learning_rate": 9.941586451101011e-06,
+      "loss": 22.8279,
       "step": 3070
     },
     {
-      "epoch": 0.1875332978156633,
-      "grad_norm": 985.0,
-      "learning_rate": 7.976558471581903e-06,
-      "loss": 23.2621,
+      "epoch": 0.3750665956313266,
+      "grad_norm": 693.0,
+      "learning_rate": 9.94139617895476e-06,
+      "loss": 22.6676,
       "step": 3080
     },
     {
-      "epoch": 0.18814217215922063,
-      "grad_norm": 676.0,
-      "learning_rate": 7.976482362723403e-06,
-      "loss": 23.3334,
+      "epoch": 0.37628434431844127,
+      "grad_norm": 499.0,
+      "learning_rate": 9.941205906808509e-06,
+      "loss": 22.7062,
       "step": 3090
     },
     {
-      "epoch": 0.188751046502778,
-      "grad_norm": 514.0,
-      "learning_rate": 7.976406253864902e-06,
-      "loss": 23.5823,
+      "epoch": 0.377502093005556,
+      "grad_norm": 345.25,
+      "learning_rate": 9.941015634662258e-06,
+      "loss": 22.8102,
       "step": 3100
     },
     {
-      "epoch": 0.18935992084633535,
-      "grad_norm": 907.5,
-      "learning_rate": 7.976330145006402e-06,
-      "loss": 23.5722,
+      "epoch": 0.3787198416926707,
+      "grad_norm": 655.0,
+      "learning_rate": 9.940825362516008e-06,
+      "loss": 22.2855,
       "step": 3110
     },
     {
-      "epoch": 0.18996879518989268,
-      "grad_norm": 578.5,
-      "learning_rate": 7.976254036147903e-06,
-      "loss": 23.8378,
+      "epoch": 0.37993759037978536,
+      "grad_norm": 422.25,
+      "learning_rate": 9.940635090369757e-06,
+      "loss": 22.5699,
       "step": 3120
     },
     {
-      "epoch": 0.19057766953345004,
-      "grad_norm": 577.0,
-      "learning_rate": 7.976177927289402e-06,
-      "loss": 23.5676,
+      "epoch": 0.3811553390669001,
+      "grad_norm": 325.5,
+      "learning_rate": 9.940444818223506e-06,
+      "loss": 22.5509,
       "step": 3130
     },
     {
-      "epoch": 0.19118654387700737,
-      "grad_norm": 729.5,
-      "learning_rate": 7.976101818430902e-06,
-      "loss": 23.6104,
+      "epoch": 0.38237308775401474,
+      "grad_norm": 568.5,
+      "learning_rate": 9.940254546077255e-06,
+      "loss": 22.5317,
       "step": 3140
     },
     {
-      "epoch": 0.19179541822056473,
-      "grad_norm": 560.5,
-      "learning_rate": 7.976025709572402e-06,
-      "loss": 23.8956,
+      "epoch": 0.38359083644112946,
+      "grad_norm": 328.0,
+      "learning_rate": 9.940064273931004e-06,
+      "loss": 22.5232,
       "step": 3150
     },
     {
-      "epoch": 0.1924042925641221,
-      "grad_norm": 439.75,
-      "learning_rate": 7.975949600713901e-06,
-      "loss": 23.664,
+      "epoch": 0.3848085851282442,
+      "grad_norm": 519.5,
+      "learning_rate": 9.939874001784754e-06,
+      "loss": 22.5387,
       "step": 3160
     },
     {
-      "epoch": 0.19301316690767942,
-      "grad_norm": 544.5,
-      "learning_rate": 7.9758734918554e-06,
-      "loss": 23.6346,
+      "epoch": 0.38602633381535884,
+      "grad_norm": 381.0,
+      "learning_rate": 9.939683729638503e-06,
+      "loss": 22.8314,
       "step": 3170
     },
     {
-      "epoch": 0.19362204125123678,
-      "grad_norm": 482.75,
-      "learning_rate": 7.9757973829969e-06,
-      "loss": 23.4919,
+      "epoch": 0.38724408250247355,
+      "grad_norm": 381.5,
+      "learning_rate": 9.939493457492252e-06,
+      "loss": 22.4892,
       "step": 3180
     },
     {
-      "epoch": 0.19423091559479413,
-      "grad_norm": 634.0,
-      "learning_rate": 7.9757212741384e-06,
-      "loss": 23.8638,
+      "epoch": 0.38846183118958827,
+      "grad_norm": 555.5,
+      "learning_rate": 9.939303185346001e-06,
+      "loss": 22.4801,
       "step": 3190
     },
     {
-      "epoch": 0.19483978993835147,
-      "grad_norm": 412.0,
-      "learning_rate": 7.9756451652799e-06,
-      "loss": 23.4453,
+      "epoch": 0.38967957987670293,
+      "grad_norm": 305.0,
+      "learning_rate": 9.93911291319975e-06,
+      "loss": 22.6149,
       "step": 3200
     },
     {
-      "epoch": 0.19544866428190882,
-      "grad_norm": 830.5,
-      "learning_rate": 7.975569056421399e-06,
-      "loss": 23.3403,
+      "epoch": 0.39089732856381765,
+      "grad_norm": 326.5,
+      "learning_rate": 9.9389226410535e-06,
+      "loss": 22.4393,
       "step": 3210
     },
     {
-      "epoch": 0.19605753862546618,
-      "grad_norm": 483.5,
-      "learning_rate": 7.975492947562898e-06,
-      "loss": 23.6017,
+      "epoch": 0.39211507725093236,
+      "grad_norm": 380.25,
+      "learning_rate": 9.93873236890725e-06,
+      "loss": 22.4641,
       "step": 3220
     },
     {
-      "epoch": 0.1966664129690235,
-      "grad_norm": 572.5,
-      "learning_rate": 7.9754168387044e-06,
-      "loss": 23.0888,
+      "epoch": 0.393332825938047,
+      "grad_norm": 615.5,
+      "learning_rate": 9.938542096760998e-06,
+      "loss": 22.9049,
       "step": 3230
     },
     {
-      "epoch": 0.19727528731258087,
-      "grad_norm": 383.75,
-      "learning_rate": 7.975340729845897e-06,
-      "loss": 23.4226,
+      "epoch": 0.39455057462516174,
+      "grad_norm": 297.75,
+      "learning_rate": 9.938351824614747e-06,
+      "loss": 22.5782,
       "step": 3240
     },
     {
-      "epoch": 0.1978841616561382,
-      "grad_norm": 598.5,
-      "learning_rate": 7.975264620987399e-06,
-      "loss": 23.4296,
+      "epoch": 0.3957683233122764,
+      "grad_norm": 511.0,
+      "learning_rate": 9.938161552468496e-06,
+      "loss": 22.6692,
       "step": 3250
     },
     {
-      "epoch": 0.19849303599969556,
-      "grad_norm": 679.0,
-      "learning_rate": 7.975188512128897e-06,
-      "loss": 23.4003,
+      "epoch": 0.3969860719993911,
+      "grad_norm": 311.75,
+      "learning_rate": 9.937971280322246e-06,
+      "loss": 22.4024,
       "step": 3260
     },
     {
-      "epoch": 0.19910191034325292,
-      "grad_norm": 487.0,
-      "learning_rate": 7.975112403270398e-06,
-      "loss": 23.4645,
+      "epoch": 0.39820382068650584,
+      "grad_norm": 350.75,
+      "learning_rate": 9.937781008175995e-06,
+      "loss": 22.6302,
       "step": 3270
     },
     {
-      "epoch": 0.19971078468681025,
-      "grad_norm": 824.5,
-      "learning_rate": 7.975036294411897e-06,
-      "loss": 23.2812,
+      "epoch": 0.3994215693736205,
+      "grad_norm": 343.0,
+      "learning_rate": 9.937590736029744e-06,
+      "loss": 22.7398,
       "step": 3280
     },
     {
-      "epoch": 0.2003196590303676,
-      "grad_norm": 796.0,
-      "learning_rate": 7.974960185553397e-06,
-      "loss": 23.2575,
+      "epoch": 0.4006393180607352,
+      "grad_norm": 329.75,
+      "learning_rate": 9.937400463883493e-06,
+      "loss": 22.581,
       "step": 3290
     },
     {
-      "epoch": 0.20092853337392497,
-      "grad_norm": 444.75,
-      "learning_rate": 7.974884076694896e-06,
-      "loss": 23.474,
+      "epoch": 0.40185706674784993,
+      "grad_norm": 373.5,
+      "learning_rate": 9.937210191737242e-06,
+      "loss": 22.6314,
       "step": 3300
     },
     {
-      "epoch": 0.2015374077174823,
-      "grad_norm": 449.5,
-      "learning_rate": 7.974807967836396e-06,
-      "loss": 23.4397,
+      "epoch": 0.4030748154349646,
+      "grad_norm": 417.25,
+      "learning_rate": 9.937019919590993e-06,
+      "loss": 22.5158,
       "step": 3310
     },
     {
-      "epoch": 0.20214628206103966,
-      "grad_norm": 722.5,
-      "learning_rate": 7.974731858977896e-06,
-      "loss": 23.2618,
+      "epoch": 0.4042925641220793,
+      "grad_norm": 291.25,
+      "learning_rate": 9.936829647444741e-06,
+      "loss": 22.4056,
       "step": 3320
     },
     {
-      "epoch": 0.20275515640459701,
-      "grad_norm": 471.75,
-      "learning_rate": 7.974655750119395e-06,
-      "loss": 23.6437,
+      "epoch": 0.40551031280919403,
+      "grad_norm": 356.5,
+      "learning_rate": 9.93663937529849e-06,
+      "loss": 22.748,
       "step": 3330
     },
     {
-      "epoch": 0.20336403074815435,
-      "grad_norm": 506.75,
-      "learning_rate": 7.974579641260895e-06,
-      "loss": 23.1774,
+      "epoch": 0.4067280614963087,
+      "grad_norm": 562.0,
+      "learning_rate": 9.936449103152239e-06,
+      "loss": 22.7601,
       "step": 3340
     },
     {
-      "epoch": 0.2039729050917117,
-      "grad_norm": 694.5,
-      "learning_rate": 7.974503532402396e-06,
-      "loss": 23.0922,
+      "epoch": 0.4079458101834234,
+      "grad_norm": 394.0,
+      "learning_rate": 9.936258831005988e-06,
+      "loss": 22.4228,
       "step": 3350
     },
     {
-      "epoch": 0.20458177943526903,
-      "grad_norm": 450.5,
-      "learning_rate": 7.974427423543894e-06,
-      "loss": 23.4778,
+      "epoch": 0.40916355887053807,
+      "grad_norm": 458.75,
+      "learning_rate": 9.936068558859739e-06,
+      "loss": 22.6764,
       "step": 3360
     },
     {
-      "epoch": 0.2051906537788264,
-      "grad_norm": 399.75,
-      "learning_rate": 7.974351314685395e-06,
-      "loss": 23.4635,
+      "epoch": 0.4103813075576528,
+      "grad_norm": 401.25,
+      "learning_rate": 9.935878286713487e-06,
+      "loss": 22.561,
       "step": 3370
     },
     {
-      "epoch": 0.20579952812238375,
-      "grad_norm": 515.5,
-      "learning_rate": 7.974275205826895e-06,
-      "loss": 23.1266,
+      "epoch": 0.4115990562447675,
+      "grad_norm": 289.0,
+      "learning_rate": 9.935688014567236e-06,
+      "loss": 22.6726,
       "step": 3380
     },
     {
-      "epoch": 0.20640840246594108,
-      "grad_norm": 499.25,
-      "learning_rate": 7.974199096968394e-06,
-      "loss": 23.3548,
+      "epoch": 0.41281680493188216,
+      "grad_norm": 531.0,
+      "learning_rate": 9.935497742420985e-06,
+      "loss": 22.6173,
       "step": 3390
     },
     {
-      "epoch": 0.20701727680949844,
-      "grad_norm": 432.0,
-      "learning_rate": 7.974122988109894e-06,
-      "loss": 23.5029,
+      "epoch": 0.4140345536189969,
+      "grad_norm": 299.5,
+      "learning_rate": 9.935307470274734e-06,
+      "loss": 22.4596,
       "step": 3400
     },
     {
-      "epoch": 0.2076261511530558,
-      "grad_norm": 581.5,
-      "learning_rate": 7.974046879251393e-06,
-      "loss": 23.5039,
+      "epoch": 0.4152523023061116,
+      "grad_norm": 329.75,
+      "learning_rate": 9.935117198128485e-06,
+      "loss": 22.6521,
       "step": 3410
     },
     {
-      "epoch": 0.20823502549661313,
-      "grad_norm": 537.5,
-      "learning_rate": 7.973970770392893e-06,
-      "loss": 23.0694,
+      "epoch": 0.41647005099322626,
+      "grad_norm": 337.75,
+      "learning_rate": 9.934926925982233e-06,
+      "loss": 22.5495,
       "step": 3420
     },
     {
-      "epoch": 0.2088438998401705,
-      "grad_norm": 490.25,
-      "learning_rate": 7.973894661534392e-06,
-      "loss": 23.3756,
+      "epoch": 0.417687799680341,
+      "grad_norm": 406.5,
+      "learning_rate": 9.934736653835982e-06,
+      "loss": 22.6828,
       "step": 3430
     },
     {
-      "epoch": 0.20945277418372785,
-      "grad_norm": 610.5,
-      "learning_rate": 7.973818552675892e-06,
-      "loss": 23.9835,
+      "epoch": 0.4189055483674557,
+      "grad_norm": 393.5,
+      "learning_rate": 9.934546381689731e-06,
+      "loss": 22.5028,
       "step": 3440
     },
     {
-      "epoch": 0.21006164852728518,
-      "grad_norm": 435.25,
-      "learning_rate": 7.973742443817391e-06,
-      "loss": 23.0766,
+      "epoch": 0.42012329705457035,
+      "grad_norm": 281.0,
+      "learning_rate": 9.93435610954348e-06,
+      "loss": 22.5415,
       "step": 3450
     },
     {
-      "epoch": 0.21067052287084254,
-      "grad_norm": 422.5,
-      "learning_rate": 7.973666334958891e-06,
-      "loss": 23.0507,
+      "epoch": 0.42134104574168507,
+      "grad_norm": 379.75,
+      "learning_rate": 9.93416583739723e-06,
+      "loss": 22.653,
       "step": 3460
     },
     {
-      "epoch": 0.21127939721439987,
-      "grad_norm": 570.0,
-      "learning_rate": 7.97359022610039e-06,
-      "loss": 23.6335,
+      "epoch": 0.42255879442879973,
+      "grad_norm": 341.0,
+      "learning_rate": 9.93397556525098e-06,
+      "loss": 22.5459,
       "step": 3470
     },
     {
-      "epoch": 0.21188827155795723,
-      "grad_norm": 384.5,
-      "learning_rate": 7.973514117241892e-06,
-      "loss": 23.7522,
+      "epoch": 0.42377654311591445,
+      "grad_norm": 456.25,
+      "learning_rate": 9.933785293104728e-06,
+      "loss": 22.6737,
       "step": 3480
     },
     {
-      "epoch": 0.21249714590151458,
-      "grad_norm": 883.5,
-      "learning_rate": 7.97343800838339e-06,
-      "loss": 23.1334,
+      "epoch": 0.42499429180302917,
+      "grad_norm": 422.75,
+      "learning_rate": 9.933595020958477e-06,
+      "loss": 22.3798,
       "step": 3490
     },
     {
-      "epoch": 0.21310602024507191,
-      "grad_norm": 445.75,
-      "learning_rate": 7.97336189952489e-06,
-      "loss": 23.0855,
+      "epoch": 0.42621204049014383,
+      "grad_norm": 411.25,
+      "learning_rate": 9.933404748812226e-06,
+      "loss": 22.5427,
       "step": 3500
     },
     {
-      "epoch": 0.21371489458862927,
-      "grad_norm": 566.5,
-      "learning_rate": 7.97328579066639e-06,
-      "loss": 23.5073,
+      "epoch": 0.42742978917725855,
+      "grad_norm": 354.5,
+      "learning_rate": 9.933214476665977e-06,
+      "loss": 22.4864,
       "step": 3510
     },
     {
-      "epoch": 0.21432376893218663,
-      "grad_norm": 375.0,
-      "learning_rate": 7.97320968180789e-06,
-      "loss": 23.1962,
+      "epoch": 0.42864753786437326,
+      "grad_norm": 281.5,
+      "learning_rate": 9.933024204519726e-06,
+      "loss": 22.3149,
       "step": 3520
     },
     {
-      "epoch": 0.21493264327574396,
-      "grad_norm": 546.5,
-      "learning_rate": 7.97313357294939e-06,
-      "loss": 23.5628,
+      "epoch": 0.4298652865514879,
+      "grad_norm": 276.25,
+      "learning_rate": 9.932833932373474e-06,
+      "loss": 22.5575,
       "step": 3530
     },
     {
-      "epoch": 0.21554151761930132,
-      "grad_norm": 445.75,
-      "learning_rate": 7.973057464090889e-06,
-      "loss": 23.6678,
+      "epoch": 0.43108303523860264,
+      "grad_norm": 385.25,
+      "learning_rate": 9.932643660227223e-06,
+      "loss": 22.2717,
       "step": 3540
     },
     {
-      "epoch": 0.21615039196285865,
-      "grad_norm": 452.75,
-      "learning_rate": 7.972981355232389e-06,
-      "loss": 23.2279,
+      "epoch": 0.4323007839257173,
+      "grad_norm": 341.5,
+      "learning_rate": 9.932453388080972e-06,
+      "loss": 22.1493,
       "step": 3550
     },
     {
-      "epoch": 0.216759266306416,
-      "grad_norm": 414.75,
-      "learning_rate": 7.972905246373888e-06,
-      "loss": 23.5895,
+      "epoch": 0.433518532612832,
+      "grad_norm": 325.5,
+      "learning_rate": 9.932263115934723e-06,
+      "loss": 22.6066,
       "step": 3560
     },
     {
-      "epoch": 0.21736814064997337,
-      "grad_norm": 381.0,
-      "learning_rate": 7.972829137515388e-06,
-      "loss": 23.2508,
+      "epoch": 0.43473628129994674,
+      "grad_norm": 298.75,
+      "learning_rate": 9.932072843788472e-06,
+      "loss": 22.8055,
       "step": 3570
     },
     {
-      "epoch": 0.2179770149935307,
-      "grad_norm": 573.5,
-      "learning_rate": 7.972753028656887e-06,
-      "loss": 23.1862,
+      "epoch": 0.4359540299870614,
+      "grad_norm": 777.0,
+      "learning_rate": 9.93188257164222e-06,
+      "loss": 22.5982,
       "step": 3580
     },
     {
-      "epoch": 0.21858588933708806,
-      "grad_norm": 563.5,
-      "learning_rate": 7.972676919798387e-06,
-      "loss": 23.4532,
+      "epoch": 0.4371717786741761,
+      "grad_norm": 612.0,
+      "learning_rate": 9.93169229949597e-06,
+      "loss": 22.5007,
       "step": 3590
     },
     {
-      "epoch": 0.21919476368064542,
-      "grad_norm": 563.0,
-      "learning_rate": 7.972600810939888e-06,
-      "loss": 23.3432,
+      "epoch": 0.43838952736129083,
+      "grad_norm": 396.75,
+      "learning_rate": 9.931502027349718e-06,
+      "loss": 22.478,
       "step": 3600
     },
     {
-      "epoch": 0.21980363802420275,
-      "grad_norm": 449.25,
-      "learning_rate": 7.972524702081386e-06,
-      "loss": 23.3239,
+      "epoch": 0.4396072760484055,
+      "grad_norm": 405.5,
+      "learning_rate": 9.931311755203469e-06,
+      "loss": 22.6416,
       "step": 3610
     },
     {
-      "epoch": 0.2204125123677601,
-      "grad_norm": 733.5,
-      "learning_rate": 7.972448593222887e-06,
-      "loss": 23.6231,
+      "epoch": 0.4408250247355202,
+      "grad_norm": 424.5,
+      "learning_rate": 9.931121483057218e-06,
+      "loss": 22.4448,
       "step": 3620
     },
     {
-      "epoch": 0.22102138671131746,
-      "grad_norm": 388.25,
-      "learning_rate": 7.972372484364385e-06,
-      "loss": 23.2343,
+      "epoch": 0.4420427734226349,
+      "grad_norm": 330.0,
+      "learning_rate": 9.930931210910967e-06,
+      "loss": 22.7654,
       "step": 3630
     },
     {
-      "epoch": 0.2216302610548748,
-      "grad_norm": 386.0,
-      "learning_rate": 7.972296375505886e-06,
-      "loss": 23.293,
+      "epoch": 0.4432605221097496,
+      "grad_norm": 471.75,
+      "learning_rate": 9.930740938764715e-06,
+      "loss": 22.5143,
       "step": 3640
     },
     {
-      "epoch": 0.22223913539843215,
-      "grad_norm": 539.0,
-      "learning_rate": 7.972220266647386e-06,
-      "loss": 23.8215,
+      "epoch": 0.4444782707968643,
+      "grad_norm": 286.25,
+      "learning_rate": 9.930550666618464e-06,
+      "loss": 22.3018,
       "step": 3650
     },
     {
-      "epoch": 0.22284800974198948,
-      "grad_norm": 627.5,
-      "learning_rate": 7.972144157788885e-06,
-      "loss": 23.6097,
+      "epoch": 0.44569601948397897,
+      "grad_norm": 441.75,
+      "learning_rate": 9.930360394472215e-06,
+      "loss": 22.6912,
       "step": 3660
     },
     {
-      "epoch": 0.22345688408554684,
-      "grad_norm": 552.5,
-      "learning_rate": 7.972068048930385e-06,
-      "loss": 23.3981,
+      "epoch": 0.4469137681710937,
+      "grad_norm": 269.75,
+      "learning_rate": 9.930170122325964e-06,
+      "loss": 22.6082,
       "step": 3670
     },
     {
-      "epoch": 0.2240657584291042,
-      "grad_norm": 736.0,
-      "learning_rate": 7.971991940071884e-06,
-      "loss": 23.202,
+      "epoch": 0.4481315168582084,
+      "grad_norm": 559.5,
+      "learning_rate": 9.929979850179713e-06,
+      "loss": 22.7375,
       "step": 3680
     },
     {
-      "epoch": 0.22467463277266153,
-      "grad_norm": 517.5,
-      "learning_rate": 7.971915831213384e-06,
-      "loss": 22.9725,
+      "epoch": 0.44934926554532306,
+      "grad_norm": 508.0,
+      "learning_rate": 9.929789578033461e-06,
+      "loss": 22.7566,
       "step": 3690
     },
     {
-      "epoch": 0.2252835071162189,
-      "grad_norm": 720.5,
-      "learning_rate": 7.971839722354883e-06,
-      "loss": 23.3352,
+      "epoch": 0.4505670142324378,
+      "grad_norm": 277.25,
+      "learning_rate": 9.92959930588721e-06,
+      "loss": 22.2571,
       "step": 3700
     },
     {
-      "epoch": 0.22589238145977625,
-      "grad_norm": 673.0,
-      "learning_rate": 7.971763613496383e-06,
-      "loss": 23.7148,
+      "epoch": 0.4517847629195525,
+      "grad_norm": 330.5,
+      "learning_rate": 9.929409033740961e-06,
+      "loss": 22.5837,
       "step": 3710
     },
     {
-      "epoch": 0.22650125580333358,
-      "grad_norm": 584.0,
-      "learning_rate": 7.971687504637883e-06,
-      "loss": 23.3811,
+      "epoch": 0.45300251160666716,
+      "grad_norm": 335.75,
+      "learning_rate": 9.92921876159471e-06,
+      "loss": 22.1115,
       "step": 3720
     },
     {
-      "epoch": 0.22711013014689094,
-      "grad_norm": 528.0,
-      "learning_rate": 7.971611395779382e-06,
-      "loss": 23.4013,
+      "epoch": 0.4542202602937819,
+      "grad_norm": 457.0,
+      "learning_rate": 9.929028489448459e-06,
+      "loss": 22.6568,
       "step": 3730
     },
     {
-      "epoch": 0.2277190044904483,
-      "grad_norm": 431.0,
-      "learning_rate": 7.971535286920883e-06,
-      "loss": 23.374,
+      "epoch": 0.4554380089808966,
+      "grad_norm": 383.75,
+      "learning_rate": 9.928838217302208e-06,
+      "loss": 22.6078,
       "step": 3740
     },
     {
-      "epoch": 0.22832787883400563,
-      "grad_norm": 388.0,
-      "learning_rate": 7.971459178062383e-06,
-      "loss": 23.2945,
+      "epoch": 0.45665575766801125,
+      "grad_norm": 312.25,
+      "learning_rate": 9.928647945155956e-06,
+      "loss": 22.5092,
       "step": 3750
     },
     {
-      "epoch": 0.22893675317756298,
-      "grad_norm": 376.5,
-      "learning_rate": 7.971383069203882e-06,
-      "loss": 23.2406,
+      "epoch": 0.45787350635512597,
+      "grad_norm": 272.25,
+      "learning_rate": 9.928457673009707e-06,
+      "loss": 22.6988,
       "step": 3760
     },
     {
-      "epoch": 0.22954562752112032,
-      "grad_norm": 450.0,
-      "learning_rate": 7.971306960345382e-06,
-      "loss": 23.1831,
+      "epoch": 0.45909125504224063,
+      "grad_norm": 329.5,
+      "learning_rate": 9.928267400863456e-06,
+      "loss": 22.5299,
       "step": 3770
     },
     {
-      "epoch": 0.23015450186467767,
-      "grad_norm": 422.25,
-      "learning_rate": 7.971230851486882e-06,
-      "loss": 23.0082,
+      "epoch": 0.46030900372935535,
+      "grad_norm": 454.0,
+      "learning_rate": 9.928077128717205e-06,
+      "loss": 22.7585,
       "step": 3780
     },
     {
-      "epoch": 0.23076337620823503,
-      "grad_norm": 396.75,
-      "learning_rate": 7.971154742628381e-06,
-      "loss": 23.6475,
+      "epoch": 0.46152675241647007,
+      "grad_norm": 297.75,
+      "learning_rate": 9.927886856570954e-06,
+      "loss": 22.5894,
       "step": 3790
     },
     {
-      "epoch": 0.23137225055179236,
-      "grad_norm": 440.5,
-      "learning_rate": 7.97107863376988e-06,
-      "loss": 23.6128,
+      "epoch": 0.4627445011035847,
+      "grad_norm": 330.0,
+      "learning_rate": 9.927696584424702e-06,
+      "loss": 22.7329,
       "step": 3800
     },
     {
-      "epoch": 0.23198112489534972,
-      "grad_norm": 421.75,
-      "learning_rate": 7.97100252491138e-06,
-      "loss": 23.2097,
+      "epoch": 0.46396224979069944,
+      "grad_norm": 377.25,
+      "learning_rate": 9.927506312278453e-06,
+      "loss": 22.8173,
       "step": 3810
     },
     {
-      "epoch": 0.23258999923890708,
-      "grad_norm": 662.5,
-      "learning_rate": 7.97092641605288e-06,
-      "loss": 23.3367,
+      "epoch": 0.46517999847781416,
+      "grad_norm": 339.75,
+      "learning_rate": 9.927316040132202e-06,
+      "loss": 22.5607,
       "step": 3820
     },
     {
-      "epoch": 0.2331988735824644,
-      "grad_norm": 808.0,
-      "learning_rate": 7.97085030719438e-06,
-      "loss": 23.7226,
+      "epoch": 0.4663977471649288,
+      "grad_norm": 314.25,
+      "learning_rate": 9.92712576798595e-06,
+      "loss": 22.5022,
       "step": 3830
     },
     {
-      "epoch": 0.23380774792602177,
-      "grad_norm": 615.5,
-      "learning_rate": 7.970774198335879e-06,
-      "loss": 23.1694,
+      "epoch": 0.46761549585204354,
+      "grad_norm": 693.0,
+      "learning_rate": 9.9269354958397e-06,
+      "loss": 22.6175,
       "step": 3840
     },
     {
-      "epoch": 0.23441662226957913,
-      "grad_norm": 466.25,
-      "learning_rate": 7.97069808947738e-06,
-      "loss": 23.564,
+      "epoch": 0.46883324453915826,
+      "grad_norm": 620.5,
+      "learning_rate": 9.926745223693449e-06,
+      "loss": 22.5456,
       "step": 3850
     },
     {
-      "epoch": 0.23502549661313646,
-      "grad_norm": 597.5,
-      "learning_rate": 7.970621980618878e-06,
-      "loss": 23.5629,
+      "epoch": 0.4700509932262729,
+      "grad_norm": 356.75,
+      "learning_rate": 9.926554951547199e-06,
+      "loss": 22.3967,
       "step": 3860
     },
     {
-      "epoch": 0.23563437095669382,
-      "grad_norm": 515.5,
-      "learning_rate": 7.97054587176038e-06,
-      "loss": 23.4269,
+      "epoch": 0.47126874191338763,
+      "grad_norm": 380.75,
+      "learning_rate": 9.926364679400948e-06,
+      "loss": 22.576,
       "step": 3870
     },
     {
-      "epoch": 0.23624324530025115,
-      "grad_norm": 487.0,
-      "learning_rate": 7.970469762901877e-06,
-      "loss": 23.0315,
+      "epoch": 0.4724864906005023,
+      "grad_norm": 678.5,
+      "learning_rate": 9.926174407254697e-06,
+      "loss": 22.8256,
       "step": 3880
     },
     {
-      "epoch": 0.2368521196438085,
-      "grad_norm": 372.25,
-      "learning_rate": 7.970393654043378e-06,
-      "loss": 23.3436,
+      "epoch": 0.473704239287617,
+      "grad_norm": 692.5,
+      "learning_rate": 9.925984135108446e-06,
+      "loss": 22.5717,
       "step": 3890
     },
     {
-      "epoch": 0.23746099398736586,
-      "grad_norm": 640.0,
-      "learning_rate": 7.970317545184878e-06,
-      "loss": 23.0826,
+      "epoch": 0.47492198797473173,
+      "grad_norm": 263.5,
+      "learning_rate": 9.925793862962195e-06,
+      "loss": 22.6193,
       "step": 3900
     },
     {
-      "epoch": 0.2380698683309232,
-      "grad_norm": 474.25,
-      "learning_rate": 7.970241436326377e-06,
-      "loss": 23.1085,
+      "epoch": 0.4761397366618464,
+      "grad_norm": 335.0,
+      "learning_rate": 9.925603590815945e-06,
+      "loss": 22.4029,
       "step": 3910
     },
     {
-      "epoch": 0.23867874267448055,
-      "grad_norm": 407.75,
-      "learning_rate": 7.970165327467877e-06,
-      "loss": 23.3948,
+      "epoch": 0.4773574853489611,
+      "grad_norm": 347.0,
+      "learning_rate": 9.925413318669694e-06,
+      "loss": 22.3616,
       "step": 3920
     },
     {
-      "epoch": 0.2392876170180379,
-      "grad_norm": 364.5,
-      "learning_rate": 7.970089218609376e-06,
-      "loss": 23.6434,
+      "epoch": 0.4785752340360758,
+      "grad_norm": 319.0,
+      "learning_rate": 9.925223046523443e-06,
+      "loss": 22.3703,
       "step": 3930
     },
     {
-      "epoch": 0.23989649136159524,
-      "grad_norm": 483.75,
-      "learning_rate": 7.970013109750876e-06,
-      "loss": 23.1049,
+      "epoch": 0.4797929827231905,
+      "grad_norm": 419.0,
+      "learning_rate": 9.925032774377192e-06,
+      "loss": 22.7403,
       "step": 3940
     },
     {
-      "epoch": 0.2405053657051526,
-      "grad_norm": 547.5,
-      "learning_rate": 7.969937000892376e-06,
-      "loss": 23.1653,
+      "epoch": 0.4810107314103052,
+      "grad_norm": 435.5,
+      "learning_rate": 9.92484250223094e-06,
+      "loss": 22.5668,
       "step": 3950
     },
     {
-      "epoch": 0.24111424004870996,
-      "grad_norm": 612.0,
-      "learning_rate": 7.969860892033875e-06,
-      "loss": 23.2834,
+      "epoch": 0.4822284800974199,
+      "grad_norm": 308.5,
+      "learning_rate": 9.924652230084691e-06,
+      "loss": 22.4282,
       "step": 3960
     },
     {
-      "epoch": 0.2417231143922673,
-      "grad_norm": 510.75,
-      "learning_rate": 7.969784783175375e-06,
-      "loss": 23.5791,
+      "epoch": 0.4834462287845346,
+      "grad_norm": 765.5,
+      "learning_rate": 9.92446195793844e-06,
+      "loss": 22.7642,
       "step": 3970
     },
     {
-      "epoch": 0.24233198873582465,
-      "grad_norm": 417.25,
-      "learning_rate": 7.969708674316874e-06,
-      "loss": 23.2259,
+      "epoch": 0.4846639774716493,
+      "grad_norm": 405.0,
+      "learning_rate": 9.924271685792189e-06,
+      "loss": 22.6432,
       "step": 3980
     },
     {
-      "epoch": 0.24294086307938198,
-      "grad_norm": 454.0,
-      "learning_rate": 7.969632565458375e-06,
-      "loss": 23.3012,
+      "epoch": 0.48588172615876396,
+      "grad_norm": 371.75,
+      "learning_rate": 9.924081413645938e-06,
+      "loss": 22.7205,
       "step": 3990
     },
     {
-      "epoch": 0.24354973742293934,
-      "grad_norm": 502.5,
-      "learning_rate": 7.969556456599875e-06,
-      "loss": 23.4953,
+      "epoch": 0.4870994748458787,
+      "grad_norm": 363.0,
+      "learning_rate": 9.923891141499687e-06,
+      "loss": 22.3163,
       "step": 4000
     },
     {
-      "epoch": 0.2441586117664967,
-      "grad_norm": 484.5,
-      "learning_rate": 7.969480347741375e-06,
-      "loss": 23.1314,
+      "epoch": 0.4883172235329934,
+      "grad_norm": 505.25,
+      "learning_rate": 9.923700869353437e-06,
+      "loss": 22.7384,
       "step": 4010
     },
     {
-      "epoch": 0.24476748611005403,
-      "grad_norm": 449.25,
-      "learning_rate": 7.969404238882874e-06,
-      "loss": 23.573,
+      "epoch": 0.48953497222010806,
+      "grad_norm": 407.25,
+      "learning_rate": 9.923510597207186e-06,
+      "loss": 22.4285,
       "step": 4020
     },
     {
-      "epoch": 0.2453763604536114,
-      "grad_norm": 391.5,
-      "learning_rate": 7.969328130024374e-06,
-      "loss": 23.5152,
+      "epoch": 0.4907527209072228,
+      "grad_norm": 483.5,
+      "learning_rate": 9.923320325060935e-06,
+      "loss": 22.5651,
       "step": 4030
     },
     {
-      "epoch": 0.24598523479716874,
-      "grad_norm": 460.0,
-      "learning_rate": 7.969252021165873e-06,
-      "loss": 23.2815,
+      "epoch": 0.4919704695943375,
+      "grad_norm": 442.0,
+      "learning_rate": 9.923130052914684e-06,
+      "loss": 22.8026,
       "step": 4040
     },
     {
-      "epoch": 0.24659410914072608,
-      "grad_norm": 440.5,
-      "learning_rate": 7.969175912307373e-06,
-      "loss": 23.8162,
+      "epoch": 0.49318821828145215,
+      "grad_norm": 561.5,
+      "learning_rate": 9.922939780768433e-06,
+      "loss": 22.578,
       "step": 4050
     },
     {
-      "epoch": 0.24720298348428343,
-      "grad_norm": 495.5,
-      "learning_rate": 7.969099803448872e-06,
-      "loss": 23.1437,
+      "epoch": 0.49440596696856687,
+      "grad_norm": 375.5,
+      "learning_rate": 9.922749508622183e-06,
+      "loss": 22.505,
       "step": 4060
     },
     {
-      "epoch": 0.2478118578278408,
-      "grad_norm": 576.5,
-      "learning_rate": 7.969023694590372e-06,
-      "loss": 23.1463,
+      "epoch": 0.4956237156556816,
+      "grad_norm": 365.5,
+      "learning_rate": 9.922559236475932e-06,
+      "loss": 22.562,
       "step": 4070
     },
     {
-      "epoch": 0.24842073217139812,
-      "grad_norm": 423.75,
-      "learning_rate": 7.968947585731871e-06,
-      "loss": 23.3878,
+      "epoch": 0.49684146434279625,
+      "grad_norm": 306.0,
+      "learning_rate": 9.922368964329681e-06,
+      "loss": 22.6089,
       "step": 4080
     },
     {
-      "epoch": 0.24902960651495548,
-      "grad_norm": 465.25,
-      "learning_rate": 7.968871476873371e-06,
-      "loss": 23.4622,
+      "epoch": 0.49805921302991096,
+      "grad_norm": 342.25,
+      "learning_rate": 9.92217869218343e-06,
+      "loss": 22.2594,
       "step": 4090
     },
     {
-      "epoch": 0.2496384808585128,
-      "grad_norm": 444.0,
-      "learning_rate": 7.968795368014872e-06,
-      "loss": 23.1397,
+      "epoch": 0.4992769617170256,
+      "grad_norm": 264.75,
+      "learning_rate": 9.92198842003718e-06,
+      "loss": 22.5169,
       "step": 4100
     },
     {
-      "epoch": 0.2502473552020702,
-      "grad_norm": 484.5,
-      "learning_rate": 7.96871925915637e-06,
-      "loss": 23.2921,
+      "epoch": 0.5004947104041404,
+      "grad_norm": 355.75,
+      "learning_rate": 9.92179814789093e-06,
+      "loss": 22.2688,
       "step": 4110
     },
     {
-      "epoch": 0.25085622954562753,
-      "grad_norm": 427.5,
-      "learning_rate": 7.968643150297871e-06,
-      "loss": 23.2433,
+      "epoch": 0.5017124590912551,
+      "grad_norm": 298.75,
+      "learning_rate": 9.921607875744678e-06,
+      "loss": 22.6661,
       "step": 4120
     },
     {
-      "epoch": 0.25146510388918486,
-      "grad_norm": 377.0,
-      "learning_rate": 7.968567041439369e-06,
-      "loss": 22.8399,
+      "epoch": 0.5029302077783697,
+      "grad_norm": 327.0,
+      "learning_rate": 9.921417603598427e-06,
+      "loss": 22.6486,
       "step": 4130
     },
     {
-      "epoch": 0.2520739782327422,
-      "grad_norm": 557.5,
-      "learning_rate": 7.96849093258087e-06,
-      "loss": 23.5973,
+      "epoch": 0.5041479564654844,
+      "grad_norm": 316.5,
+      "learning_rate": 9.921227331452176e-06,
+      "loss": 22.554,
       "step": 4140
     },
     {
-      "epoch": 0.2526828525762996,
-      "grad_norm": 412.5,
-      "learning_rate": 7.96841482372237e-06,
-      "loss": 23.1057,
+      "epoch": 0.5053657051525992,
+      "grad_norm": 528.5,
+      "learning_rate": 9.921037059305927e-06,
+      "loss": 22.5602,
       "step": 4150
     },
     {
-      "epoch": 0.2532917269198569,
-      "grad_norm": 470.5,
-      "learning_rate": 7.96833871486387e-06,
-      "loss": 23.4342,
+      "epoch": 0.5065834538397138,
+      "grad_norm": 579.5,
+      "learning_rate": 9.920846787159675e-06,
+      "loss": 22.4661,
       "step": 4160
     },
     {
-      "epoch": 0.25390060126341424,
-      "grad_norm": 491.25,
-      "learning_rate": 7.968262606005369e-06,
-      "loss": 23.3106,
+      "epoch": 0.5078012025268285,
+      "grad_norm": 307.5,
+      "learning_rate": 9.920656515013424e-06,
+      "loss": 22.8724,
       "step": 4170
     },
     {
-      "epoch": 0.2545094756069716,
-      "grad_norm": 563.5,
-      "learning_rate": 7.968186497146869e-06,
-      "loss": 23.4464,
+      "epoch": 0.5090189512139432,
+      "grad_norm": 531.0,
+      "learning_rate": 9.920466242867173e-06,
+      "loss": 22.5746,
       "step": 4180
     },
     {
-      "epoch": 0.25511834995052896,
-      "grad_norm": 739.0,
-      "learning_rate": 7.968110388288368e-06,
-      "loss": 23.0911,
+      "epoch": 0.5102366999010579,
+      "grad_norm": 415.0,
+      "learning_rate": 9.920275970720922e-06,
+      "loss": 22.6794,
       "step": 4190
     },
     {
-      "epoch": 0.2557272242940863,
-      "grad_norm": 345.0,
-      "learning_rate": 7.968034279429868e-06,
-      "loss": 23.2429,
+      "epoch": 0.5114544485881726,
+      "grad_norm": 349.25,
+      "learning_rate": 9.920085698574673e-06,
+      "loss": 22.4776,
       "step": 4200
     },
     {
-      "epoch": 0.2563360986376437,
-      "grad_norm": 435.25,
-      "learning_rate": 7.967958170571367e-06,
-      "loss": 23.5266,
+      "epoch": 0.5126721972752873,
+      "grad_norm": 708.5,
+      "learning_rate": 9.919895426428421e-06,
+      "loss": 22.7027,
       "step": 4210
     },
     {
-      "epoch": 0.256944972981201,
-      "grad_norm": 575.5,
-      "learning_rate": 7.967882061712868e-06,
-      "loss": 23.529,
+      "epoch": 0.513889945962402,
+      "grad_norm": 647.0,
+      "learning_rate": 9.91970515428217e-06,
+      "loss": 22.1834,
       "step": 4220
     },
     {
-      "epoch": 0.25755384732475833,
-      "grad_norm": 465.75,
-      "learning_rate": 7.967805952854366e-06,
-      "loss": 23.1344,
+      "epoch": 0.5151076946495167,
+      "grad_norm": 248.375,
+      "learning_rate": 9.91951488213592e-06,
+      "loss": 22.3468,
       "step": 4230
     },
     {
-      "epoch": 0.2581627216683157,
-      "grad_norm": 512.5,
-      "learning_rate": 7.967729843995868e-06,
-      "loss": 23.103,
+      "epoch": 0.5163254433366314,
+      "grad_norm": 301.75,
+      "learning_rate": 9.91932460998967e-06,
+      "loss": 22.2697,
       "step": 4240
     },
     {
-      "epoch": 0.25877159601187305,
-      "grad_norm": 639.5,
-      "learning_rate": 7.967653735137367e-06,
-      "loss": 23.2826,
+      "epoch": 0.5175431920237461,
+      "grad_norm": 387.75,
+      "learning_rate": 9.919134337843419e-06,
+      "loss": 22.3615,
       "step": 4250
     },
     {
-      "epoch": 0.2593804703554304,
-      "grad_norm": 457.75,
-      "learning_rate": 7.967577626278867e-06,
-      "loss": 23.3039,
+      "epoch": 0.5187609407108608,
+      "grad_norm": 316.75,
+      "learning_rate": 9.918944065697168e-06,
+      "loss": 22.314,
       "step": 4260
     },
     {
-      "epoch": 0.25998934469898777,
-      "grad_norm": 405.25,
-      "learning_rate": 7.967501517420366e-06,
-      "loss": 23.2088,
+      "epoch": 0.5199786893979755,
+      "grad_norm": 310.0,
+      "learning_rate": 9.918753793550916e-06,
+      "loss": 22.5011,
       "step": 4270
     },
     {
-      "epoch": 0.2605982190425451,
-      "grad_norm": 672.5,
-      "learning_rate": 7.967425408561866e-06,
-      "loss": 23.4113,
+      "epoch": 0.5211964380850902,
+      "grad_norm": 431.5,
+      "learning_rate": 9.918563521404665e-06,
+      "loss": 22.4555,
       "step": 4280
     },
     {
-      "epoch": 0.26120709338610243,
-      "grad_norm": 788.0,
-      "learning_rate": 7.967349299703365e-06,
-      "loss": 23.2271,
+      "epoch": 0.5224141867722049,
+      "grad_norm": 377.0,
+      "learning_rate": 9.918373249258416e-06,
+      "loss": 22.8352,
       "step": 4290
     },
     {
-      "epoch": 0.2618159677296598,
-      "grad_norm": 1123.0,
-      "learning_rate": 7.967273190844865e-06,
-      "loss": 23.6025,
+      "epoch": 0.5236319354593196,
+      "grad_norm": 284.25,
+      "learning_rate": 9.918182977112165e-06,
+      "loss": 22.6719,
       "step": 4300
     },
     {
-      "epoch": 0.26242484207321715,
-      "grad_norm": 754.0,
-      "learning_rate": 7.967197081986364e-06,
-      "loss": 23.528,
+      "epoch": 0.5248496841464343,
+      "grad_norm": 281.75,
+      "learning_rate": 9.917992704965914e-06,
+      "loss": 22.5164,
       "step": 4310
     },
     {
-      "epoch": 0.2630337164167745,
-      "grad_norm": 577.0,
-      "learning_rate": 7.967120973127864e-06,
-      "loss": 22.7315,
+      "epoch": 0.526067432833549,
+      "grad_norm": 830.0,
+      "learning_rate": 9.917802432819662e-06,
+      "loss": 22.4646,
       "step": 4320
     },
     {
-      "epoch": 0.26364259076033186,
-      "grad_norm": 702.5,
-      "learning_rate": 7.967044864269363e-06,
-      "loss": 22.9606,
+      "epoch": 0.5272851815206637,
+      "grad_norm": 496.5,
+      "learning_rate": 9.917612160673413e-06,
+      "loss": 22.6187,
       "step": 4330
     },
     {
-      "epoch": 0.2642514651038892,
-      "grad_norm": 462.75,
-      "learning_rate": 7.966968755410863e-06,
-      "loss": 22.9621,
+      "epoch": 0.5285029302077784,
+      "grad_norm": 495.0,
+      "learning_rate": 9.917421888527162e-06,
+      "loss": 22.6751,
       "step": 4340
     },
     {
-      "epoch": 0.2648603394474465,
-      "grad_norm": 572.5,
-      "learning_rate": 7.966892646552364e-06,
-      "loss": 23.2063,
+      "epoch": 0.529720678894893,
+      "grad_norm": 428.0,
+      "learning_rate": 9.91723161638091e-06,
+      "loss": 22.1483,
       "step": 4350
     },
     {
-      "epoch": 0.26546921379100386,
-      "grad_norm": 578.5,
-      "learning_rate": 7.966816537693862e-06,
-      "loss": 23.5121,
+      "epoch": 0.5309384275820077,
+      "grad_norm": 286.5,
+      "learning_rate": 9.91704134423466e-06,
+      "loss": 22.5045,
       "step": 4360
     },
     {
-      "epoch": 0.26607808813456124,
-      "grad_norm": 470.5,
-      "learning_rate": 7.966740428835363e-06,
-      "loss": 23.4099,
+      "epoch": 0.5321561762691225,
+      "grad_norm": 601.0,
+      "learning_rate": 9.916851072088408e-06,
+      "loss": 22.6367,
       "step": 4370
     },
     {
-      "epoch": 0.26668696247811857,
-      "grad_norm": 445.0,
-      "learning_rate": 7.966664319976863e-06,
-      "loss": 23.2292,
+      "epoch": 0.5333739249562371,
+      "grad_norm": 403.0,
+      "learning_rate": 9.916660799942159e-06,
+      "loss": 22.7377,
       "step": 4380
     },
     {
-      "epoch": 0.2672958368216759,
-      "grad_norm": 450.0,
-      "learning_rate": 7.966588211118362e-06,
-      "loss": 23.3068,
+      "epoch": 0.5345916736433518,
+      "grad_norm": 295.75,
+      "learning_rate": 9.916470527795908e-06,
+      "loss": 22.26,
       "step": 4390
     },
     {
-      "epoch": 0.2679047111652333,
-      "grad_norm": 435.0,
-      "learning_rate": 7.966512102259862e-06,
-      "loss": 23.0844,
+      "epoch": 0.5358094223304666,
+      "grad_norm": 460.5,
+      "learning_rate": 9.916280255649657e-06,
+      "loss": 22.4242,
       "step": 4400
     },
     {
-      "epoch": 0.2685135855087906,
-      "grad_norm": 457.75,
-      "learning_rate": 7.966435993401362e-06,
-      "loss": 23.0199,
+      "epoch": 0.5370271710175812,
+      "grad_norm": 422.75,
+      "learning_rate": 9.916089983503406e-06,
+      "loss": 22.2663,
       "step": 4410
     },
     {
-      "epoch": 0.26912245985234795,
-      "grad_norm": 783.5,
-      "learning_rate": 7.966359884542861e-06,
-      "loss": 23.145,
+      "epoch": 0.5382449197046959,
+      "grad_norm": 498.0,
+      "learning_rate": 9.915899711357155e-06,
+      "loss": 22.7889,
       "step": 4420
     },
     {
-      "epoch": 0.26973133419590534,
-      "grad_norm": 460.75,
-      "learning_rate": 7.96628377568436e-06,
-      "loss": 23.0592,
+      "epoch": 0.5394626683918107,
+      "grad_norm": 316.75,
+      "learning_rate": 9.915709439210905e-06,
+      "loss": 22.488,
       "step": 4430
     },
     {
-      "epoch": 0.27034020853946267,
-      "grad_norm": 515.5,
-      "learning_rate": 7.96620766682586e-06,
-      "loss": 22.9736,
+      "epoch": 0.5406804170789253,
+      "grad_norm": 291.5,
+      "learning_rate": 9.915519167064654e-06,
+      "loss": 22.4644,
       "step": 4440
     },
     {
-      "epoch": 0.27094908288302,
-      "grad_norm": 446.0,
-      "learning_rate": 7.966131557967361e-06,
-      "loss": 23.3037,
+      "epoch": 0.54189816576604,
+      "grad_norm": 389.0,
+      "learning_rate": 9.915328894918403e-06,
+      "loss": 22.3433,
       "step": 4450
     },
     {
-      "epoch": 0.2715579572265774,
-      "grad_norm": 448.25,
-      "learning_rate": 7.96605544910886e-06,
-      "loss": 23.4446,
+      "epoch": 0.5431159144531548,
+      "grad_norm": 302.5,
+      "learning_rate": 9.915138622772152e-06,
+      "loss": 22.3713,
       "step": 4460
     },
     {
-      "epoch": 0.2721668315701347,
-      "grad_norm": 386.75,
-      "learning_rate": 7.96597934025036e-06,
-      "loss": 23.1174,
+      "epoch": 0.5443336631402694,
+      "grad_norm": 495.25,
+      "learning_rate": 9.914948350625902e-06,
+      "loss": 22.7835,
       "step": 4470
     },
     {
-      "epoch": 0.27277570591369205,
-      "grad_norm": 462.25,
-      "learning_rate": 7.965903231391858e-06,
-      "loss": 23.0895,
+      "epoch": 0.5455514118273841,
+      "grad_norm": 469.5,
+      "learning_rate": 9.914758078479651e-06,
+      "loss": 22.1895,
       "step": 4480
     },
     {
-      "epoch": 0.27338458025724943,
-      "grad_norm": 681.0,
-      "learning_rate": 7.96582712253336e-06,
-      "loss": 23.1573,
+      "epoch": 0.5467691605144989,
+      "grad_norm": 295.75,
+      "learning_rate": 9.9145678063334e-06,
+      "loss": 22.8347,
       "step": 4490
     },
     {
-      "epoch": 0.27399345460080676,
-      "grad_norm": 451.5,
-      "learning_rate": 7.96575101367486e-06,
-      "loss": 22.9773,
+      "epoch": 0.5479869092016135,
+      "grad_norm": 479.25,
+      "learning_rate": 9.914377534187149e-06,
+      "loss": 22.5627,
       "step": 4500
     },
     {
-      "epoch": 0.2746023289443641,
-      "grad_norm": 499.75,
-      "learning_rate": 7.965674904816359e-06,
-      "loss": 23.3175,
+      "epoch": 0.5492046578887282,
+      "grad_norm": 311.0,
+      "learning_rate": 9.914187262040898e-06,
+      "loss": 22.2682,
       "step": 4510
     },
     {
-      "epoch": 0.2752112032879215,
-      "grad_norm": 609.5,
-      "learning_rate": 7.965598795957858e-06,
-      "loss": 23.4505,
+      "epoch": 0.550422406575843,
+      "grad_norm": 896.5,
+      "learning_rate": 9.913996989894648e-06,
+      "loss": 22.2863,
       "step": 4520
     },
     {
-      "epoch": 0.2758200776314788,
-      "grad_norm": 423.25,
-      "learning_rate": 7.965522687099358e-06,
-      "loss": 23.1152,
+      "epoch": 0.5516401552629576,
+      "grad_norm": 618.5,
+      "learning_rate": 9.913806717748397e-06,
+      "loss": 22.7516,
       "step": 4530
     },
     {
-      "epoch": 0.27642895197503614,
-      "grad_norm": 360.5,
-      "learning_rate": 7.965446578240857e-06,
-      "loss": 23.0735,
+      "epoch": 0.5528579039500723,
+      "grad_norm": 747.0,
+      "learning_rate": 9.913616445602146e-06,
+      "loss": 22.3653,
       "step": 4540
     },
     {
-      "epoch": 0.2770378263185935,
-      "grad_norm": 491.25,
-      "learning_rate": 7.965370469382357e-06,
-      "loss": 23.3136,
+      "epoch": 0.554075652637187,
+      "grad_norm": 272.5,
+      "learning_rate": 9.913426173455895e-06,
+      "loss": 22.181,
       "step": 4550
     },
     {
-      "epoch": 0.27764670066215086,
-      "grad_norm": 464.0,
-      "learning_rate": 7.965294360523856e-06,
-      "loss": 23.1639,
+      "epoch": 0.5552934013243017,
+      "grad_norm": 505.5,
+      "learning_rate": 9.913235901309644e-06,
+      "loss": 22.4785,
       "step": 4560
     },
     {
-      "epoch": 0.2782555750057082,
-      "grad_norm": 521.5,
-      "learning_rate": 7.965218251665356e-06,
-      "loss": 23.3381,
+      "epoch": 0.5565111500114164,
+      "grad_norm": 636.0,
+      "learning_rate": 9.913045629163394e-06,
+      "loss": 22.508,
       "step": 4570
     },
     {
-      "epoch": 0.2788644493492655,
-      "grad_norm": 434.25,
-      "learning_rate": 7.965142142806856e-06,
-      "loss": 23.8722,
+      "epoch": 0.557728898698531,
+      "grad_norm": 339.0,
+      "learning_rate": 9.912855357017143e-06,
+      "loss": 22.3734,
       "step": 4580
     },
     {
-      "epoch": 0.2794733236928229,
-      "grad_norm": 613.5,
-      "learning_rate": 7.965066033948355e-06,
-      "loss": 22.7006,
+      "epoch": 0.5589466473856458,
+      "grad_norm": 447.5,
+      "learning_rate": 9.912665084870892e-06,
+      "loss": 22.115,
       "step": 4590
     },
     {
-      "epoch": 0.28008219803638024,
-      "grad_norm": 499.75,
-      "learning_rate": 7.964989925089856e-06,
-      "loss": 22.9187,
+      "epoch": 0.5601643960727605,
+      "grad_norm": 328.25,
+      "learning_rate": 9.912474812724641e-06,
+      "loss": 22.4416,
       "step": 4600
     },
     {
-      "epoch": 0.28069107237993757,
-      "grad_norm": 968.5,
-      "learning_rate": 7.964913816231356e-06,
-      "loss": 22.9807,
+      "epoch": 0.5613821447598751,
+      "grad_norm": 770.5,
+      "learning_rate": 9.91228454057839e-06,
+      "loss": 22.5458,
       "step": 4610
     },
     {
-      "epoch": 0.28129994672349495,
-      "grad_norm": 834.0,
-      "learning_rate": 7.964837707372855e-06,
-      "loss": 23.3458,
+      "epoch": 0.5625998934469899,
+      "grad_norm": 465.5,
+      "learning_rate": 9.91209426843214e-06,
+      "loss": 22.454,
       "step": 4620
     },
     {
-      "epoch": 0.2819088210670523,
-      "grad_norm": 442.75,
-      "learning_rate": 7.964761598514355e-06,
-      "loss": 23.0276,
+      "epoch": 0.5638176421341046,
+      "grad_norm": 348.0,
+      "learning_rate": 9.91190399628589e-06,
+      "loss": 22.1964,
       "step": 4630
     },
     {
-      "epoch": 0.2825176954106096,
-      "grad_norm": 648.5,
-      "learning_rate": 7.964685489655855e-06,
-      "loss": 23.0723,
+      "epoch": 0.5650353908212192,
+      "grad_norm": 997.5,
+      "learning_rate": 9.911713724139638e-06,
+      "loss": 22.3406,
       "step": 4640
     },
     {
-      "epoch": 0.283126569754167,
-      "grad_norm": 423.25,
-      "learning_rate": 7.964609380797354e-06,
-      "loss": 23.0752,
+      "epoch": 0.566253139508334,
+      "grad_norm": 500.75,
+      "learning_rate": 9.911523451993387e-06,
+      "loss": 22.1292,
       "step": 4650
     },
     {
-      "epoch": 0.28373544409772433,
-      "grad_norm": 472.5,
-      "learning_rate": 7.964533271938854e-06,
-      "loss": 23.1104,
+      "epoch": 0.5674708881954487,
+      "grad_norm": 355.0,
+      "learning_rate": 9.911333179847136e-06,
+      "loss": 22.3442,
       "step": 4660
     },
     {
-      "epoch": 0.28434431844128166,
-      "grad_norm": 595.5,
-      "learning_rate": 7.964457163080353e-06,
-      "loss": 23.3532,
+      "epoch": 0.5686886368825633,
+      "grad_norm": 433.75,
+      "learning_rate": 9.911142907700886e-06,
+      "loss": 22.1724,
       "step": 4670
     },
     {
-      "epoch": 0.28495319278483905,
-      "grad_norm": 347.0,
-      "learning_rate": 7.964381054221853e-06,
-      "loss": 23.0569,
+      "epoch": 0.5699063855696781,
+      "grad_norm": 310.0,
+      "learning_rate": 9.910952635554635e-06,
+      "loss": 22.4883,
       "step": 4680
     },
     {
-      "epoch": 0.2855620671283964,
-      "grad_norm": 417.75,
-      "learning_rate": 7.964304945363352e-06,
-      "loss": 23.4733,
+      "epoch": 0.5711241342567928,
+      "grad_norm": 398.0,
+      "learning_rate": 9.910762363408384e-06,
+      "loss": 22.3938,
       "step": 4690
     },
     {
-      "epoch": 0.2861709414719537,
-      "grad_norm": 461.5,
-      "learning_rate": 7.964228836504854e-06,
-      "loss": 23.3129,
+      "epoch": 0.5723418829439074,
+      "grad_norm": 308.75,
+      "learning_rate": 9.910572091262133e-06,
+      "loss": 22.2949,
       "step": 4700
     },
     {
-      "epoch": 0.2867798158155111,
-      "grad_norm": 498.25,
-      "learning_rate": 7.964152727646351e-06,
-      "loss": 23.4051,
+      "epoch": 0.5735596316310222,
+      "grad_norm": 715.0,
+      "learning_rate": 9.910381819115882e-06,
+      "loss": 22.4441,
       "step": 4710
     },
     {
-      "epoch": 0.2873886901590684,
-      "grad_norm": 506.5,
-      "learning_rate": 7.964076618787853e-06,
-      "loss": 23.1157,
+      "epoch": 0.5747773803181369,
+      "grad_norm": 526.0,
+      "learning_rate": 9.910191546969633e-06,
+      "loss": 22.2936,
       "step": 4720
     },
     {
-      "epoch": 0.28799756450262576,
-      "grad_norm": 687.5,
-      "learning_rate": 7.96400050992935e-06,
-      "loss": 23.1517,
+      "epoch": 0.5759951290052515,
+      "grad_norm": 285.5,
+      "learning_rate": 9.910001274823381e-06,
+      "loss": 22.4871,
       "step": 4730
     },
     {
-      "epoch": 0.28860643884618314,
-      "grad_norm": 490.25,
-      "learning_rate": 7.963924401070852e-06,
-      "loss": 22.461,
+      "epoch": 0.5772128776923663,
+      "grad_norm": 820.5,
+      "learning_rate": 9.90981100267713e-06,
+      "loss": 22.5226,
       "step": 4740
     },
     {
-      "epoch": 0.2892153131897405,
-      "grad_norm": 352.75,
-      "learning_rate": 7.963848292212351e-06,
-      "loss": 22.8928,
+      "epoch": 0.578430626379481,
+      "grad_norm": 763.5,
+      "learning_rate": 9.909620730530879e-06,
+      "loss": 22.6878,
       "step": 4750
     },
     {
-      "epoch": 0.2898241875332978,
-      "grad_norm": 523.5,
-      "learning_rate": 7.963772183353851e-06,
-      "loss": 23.0327,
+      "epoch": 0.5796483750665956,
+      "grad_norm": 271.0,
+      "learning_rate": 9.909430458384628e-06,
+      "loss": 22.1892,
       "step": 4760
     },
     {
-      "epoch": 0.29043306187685514,
-      "grad_norm": 426.5,
-      "learning_rate": 7.96369607449535e-06,
-      "loss": 22.9028,
+      "epoch": 0.5808661237537103,
+      "grad_norm": 344.5,
+      "learning_rate": 9.909240186238379e-06,
+      "loss": 22.5667,
       "step": 4770
     },
     {
-      "epoch": 0.2910419362204125,
-      "grad_norm": 664.0,
-      "learning_rate": 7.96361996563685e-06,
-      "loss": 23.0223,
+      "epoch": 0.582083872440825,
+      "grad_norm": 279.75,
+      "learning_rate": 9.909049914092127e-06,
+      "loss": 22.5652,
       "step": 4780
     },
     {
-      "epoch": 0.29165081056396985,
-      "grad_norm": 534.5,
-      "learning_rate": 7.96354385677835e-06,
-      "loss": 23.1204,
+      "epoch": 0.5833016211279397,
+      "grad_norm": 529.0,
+      "learning_rate": 9.908859641945876e-06,
+      "loss": 22.8018,
       "step": 4790
     },
     {
-      "epoch": 0.2922596849075272,
-      "grad_norm": 502.25,
-      "learning_rate": 7.963467747919849e-06,
-      "loss": 23.1276,
+      "epoch": 0.5845193698150544,
+      "grad_norm": 402.5,
+      "learning_rate": 9.908669369799625e-06,
+      "loss": 22.7684,
       "step": 4800
     },
     {
-      "epoch": 0.29286855925108457,
-      "grad_norm": 462.75,
-      "learning_rate": 7.963391639061349e-06,
-      "loss": 23.1368,
+      "epoch": 0.5857371185021691,
+      "grad_norm": 631.0,
+      "learning_rate": 9.908479097653374e-06,
+      "loss": 22.4754,
       "step": 4810
     },
     {
-      "epoch": 0.2934774335946419,
-      "grad_norm": 427.75,
-      "learning_rate": 7.963315530202848e-06,
-      "loss": 22.6855,
+      "epoch": 0.5869548671892838,
+      "grad_norm": 260.0,
+      "learning_rate": 9.908288825507125e-06,
+      "loss": 22.5717,
       "step": 4820
     },
     {
-      "epoch": 0.29408630793819923,
-      "grad_norm": 469.75,
-      "learning_rate": 7.963239421344348e-06,
-      "loss": 22.8931,
+      "epoch": 0.5881726158763985,
+      "grad_norm": 549.0,
+      "learning_rate": 9.908098553360874e-06,
+      "loss": 22.4253,
       "step": 4830
     },
     {
-      "epoch": 0.2946951822817566,
-      "grad_norm": 651.0,
-      "learning_rate": 7.963163312485849e-06,
-      "loss": 23.0368,
+      "epoch": 0.5893903645635132,
+      "grad_norm": 767.0,
+      "learning_rate": 9.907908281214622e-06,
+      "loss": 22.3855,
       "step": 4840
     },
     {
-      "epoch": 0.29530405662531395,
-      "grad_norm": 584.0,
-      "learning_rate": 7.963087203627348e-06,
-      "loss": 23.5478,
+      "epoch": 0.5906081132506279,
+      "grad_norm": 874.0,
+      "learning_rate": 9.907718009068371e-06,
+      "loss": 21.8794,
       "step": 4850
     },
     {
-      "epoch": 0.2959129309688713,
-      "grad_norm": 846.5,
-      "learning_rate": 7.963011094768848e-06,
-      "loss": 23.2312,
+      "epoch": 0.5918258619377426,
+      "grad_norm": 373.25,
+      "learning_rate": 9.90752773692212e-06,
+      "loss": 22.1089,
       "step": 4860
     },
     {
-      "epoch": 0.29652180531242867,
-      "grad_norm": 697.0,
-      "learning_rate": 7.962934985910348e-06,
-      "loss": 22.9755,
+      "epoch": 0.5930436106248573,
+      "grad_norm": 459.75,
+      "learning_rate": 9.90733746477587e-06,
+      "loss": 21.9316,
       "step": 4870
     },
     {
-      "epoch": 0.297130679655986,
-      "grad_norm": 686.5,
-      "learning_rate": 7.962858877051847e-06,
-      "loss": 23.1155,
+      "epoch": 0.594261359311972,
+      "grad_norm": 693.0,
+      "learning_rate": 9.90714719262962e-06,
+      "loss": 21.9847,
       "step": 4880
     },
     {
-      "epoch": 0.2977395539995433,
-      "grad_norm": 549.5,
-      "learning_rate": 7.962782768193347e-06,
-      "loss": 23.0959,
+      "epoch": 0.5954791079990867,
+      "grad_norm": 802.5,
+      "learning_rate": 9.906956920483368e-06,
+      "loss": 22.4072,
       "step": 4890
     },
     {
-      "epoch": 0.2983484283431007,
-      "grad_norm": 430.5,
-      "learning_rate": 7.962706659334846e-06,
-      "loss": 22.8863,
+      "epoch": 0.5966968566862014,
+      "grad_norm": 467.5,
+      "learning_rate": 9.906766648337117e-06,
+      "loss": 22.2952,
       "step": 4900
     },
     {
-      "epoch": 0.29895730268665804,
-      "grad_norm": 403.0,
-      "learning_rate": 7.962630550476346e-06,
-      "loss": 23.0113,
+      "epoch": 0.5979146053733161,
+      "grad_norm": 507.0,
+      "learning_rate": 9.906576376190866e-06,
+      "loss": 22.2678,
       "step": 4910
     },
     {
-      "epoch": 0.2995661770302154,
-      "grad_norm": 716.0,
-      "learning_rate": 7.962554441617845e-06,
-      "loss": 23.1476,
+      "epoch": 0.5991323540604307,
+      "grad_norm": 430.5,
+      "learning_rate": 9.906386104044617e-06,
+      "loss": 22.4595,
       "step": 4920
     },
     {
-      "epoch": 0.30017505137377276,
-      "grad_norm": 402.0,
-      "learning_rate": 7.962478332759345e-06,
-      "loss": 22.8651,
+      "epoch": 0.6003501027475455,
+      "grad_norm": 357.5,
+      "learning_rate": 9.906195831898366e-06,
+      "loss": 22.149,
       "step": 4930
     },
     {
-      "epoch": 0.3007839257173301,
-      "grad_norm": 439.0,
-      "learning_rate": 7.962402223900844e-06,
-      "loss": 22.6472,
+      "epoch": 0.6015678514346602,
+      "grad_norm": 518.0,
+      "learning_rate": 9.906005559752114e-06,
+      "loss": 22.2035,
       "step": 4940
     },
     {
-      "epoch": 0.3013928000608874,
-      "grad_norm": 448.0,
-      "learning_rate": 7.962326115042346e-06,
-      "loss": 23.111,
+      "epoch": 0.6027856001217748,
+      "grad_norm": 688.5,
+      "learning_rate": 9.905815287605863e-06,
+      "loss": 22.3291,
       "step": 4950
     },
     {
-      "epoch": 0.3020016744044448,
-      "grad_norm": 375.25,
-      "learning_rate": 7.962250006183844e-06,
-      "loss": 23.1311,
+      "epoch": 0.6040033488088896,
+      "grad_norm": 508.75,
+      "learning_rate": 9.905625015459612e-06,
+      "loss": 22.1793,
       "step": 4960
     },
     {
-      "epoch": 0.30261054874800214,
-      "grad_norm": 589.0,
-      "learning_rate": 7.962173897325345e-06,
-      "loss": 22.864,
+      "epoch": 0.6052210974960043,
+      "grad_norm": 575.0,
+      "learning_rate": 9.905434743313363e-06,
+      "loss": 22.5242,
       "step": 4970
     },
     {
-      "epoch": 0.30321942309155947,
-      "grad_norm": 468.75,
-      "learning_rate": 7.962097788466843e-06,
-      "loss": 22.9987,
+      "epoch": 0.6064388461831189,
+      "grad_norm": 760.5,
+      "learning_rate": 9.905244471167112e-06,
+      "loss": 22.1918,
       "step": 4980
     },
     {
-      "epoch": 0.3038282974351168,
-      "grad_norm": 608.5,
-      "learning_rate": 7.962021679608344e-06,
-      "loss": 23.2671,
+      "epoch": 0.6076565948702336,
+      "grad_norm": 335.75,
+      "learning_rate": 9.90505419902086e-06,
+      "loss": 22.12,
       "step": 4990
     },
     {
-      "epoch": 0.3044371717786742,
-      "grad_norm": 388.5,
-      "learning_rate": 7.961945570749843e-06,
-      "loss": 23.0071,
+      "epoch": 0.6088743435573484,
+      "grad_norm": 343.5,
+      "learning_rate": 9.90486392687461e-06,
+      "loss": 22.0701,
       "step": 5000
     },
     {
-      "epoch": 0.3044371717786742,
-      "eval_loss": 2.88149094581604,
-      "eval_runtime": 95.0877,
-      "eval_samples_per_second": 2267.548,
-      "eval_steps_per_second": 35.43,
+      "epoch": 0.6088743435573484,
+      "eval_loss": 2.798462390899658,
+      "eval_runtime": 95.0278,
+      "eval_samples_per_second": 2268.978,
+      "eval_steps_per_second": 35.453,
       "step": 5000
     }
   ],
   "logging_steps": 10,
-  "max_steps": 16423,
+  "max_steps": 8211,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 5000,
@@ -3534,8 +3534,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 8.72761379520512e+17,
-  "train_batch_size": 8,
+  "total_flos": 1.745522759041024e+18,
+  "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null
 }