diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16133 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.04182968498155174,
+  "eval_steps": 500,
+  "global_step": 23000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.818681955719641e-05,
+      "grad_norm": 2.1063554286956787,
+      "learning_rate": 0.0002,
+      "loss": 1.9357,
+      "step": 10
+    },
+    {
+      "epoch": 3.637363911439282e-05,
+      "grad_norm": 0.9359453320503235,
+      "learning_rate": 0.0002,
+      "loss": 0.2208,
+      "step": 20
+    },
+    {
+      "epoch": 5.4560458671589234e-05,
+      "grad_norm": 0.5420117378234863,
+      "learning_rate": 0.0002,
+      "loss": 0.1459,
+      "step": 30
+    },
+    {
+      "epoch": 7.274727822878565e-05,
+      "grad_norm": 0.05442357063293457,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 40
+    },
+    {
+      "epoch": 9.093409778598205e-05,
+      "grad_norm": 0.0005907653248868883,
+      "learning_rate": 0.0002,
+      "loss": 0.0005,
+      "step": 50
+    },
+    {
+      "epoch": 0.00010912091734317847,
+      "grad_norm": 0.26516178250312805,
+      "learning_rate": 0.0002,
+      "loss": 1.0686,
+      "step": 60
+    },
+    {
+      "epoch": 0.00012730773690037487,
+      "grad_norm": 0.44067099690437317,
+      "learning_rate": 0.0002,
+      "loss": 0.2613,
+      "step": 70
+    },
+    {
+      "epoch": 0.0001454945564575713,
+      "grad_norm": 0.09356075525283813,
+      "learning_rate": 0.0002,
+      "loss": 0.1415,
+      "step": 80
+    },
+    {
+      "epoch": 0.0001636813760147677,
+      "grad_norm": 0.017799921333789825,
+      "learning_rate": 0.0002,
+      "loss": 0.1013,
+      "step": 90
+    },
+    {
+      "epoch": 0.0001818681955719641,
+      "grad_norm": 0.0018534553237259388,
+      "learning_rate": 0.0002,
+      "loss": 0.0001,
+      "step": 100
+    },
+    {
+      "epoch": 0.00020005501512916052,
+      "grad_norm": 0.35472020506858826,
+      "learning_rate": 0.0002,
+      "loss": 0.73,
+      "step": 110
+    },
+    {
+      "epoch": 0.00021824183468635694,
+      "grad_norm": 0.3880878686904907,
+      "learning_rate": 0.0002,
+      "loss": 0.1424,
+      "step": 120
+    },
+    {
+      "epoch": 0.00023642865424355333,
+      "grad_norm": 0.19027432799339294,
+      "learning_rate": 0.0002,
+      "loss": 0.1173,
+      "step": 130
+    },
+    {
+      "epoch": 0.00025461547380074975,
+      "grad_norm": 0.019047321751713753,
+      "learning_rate": 0.0002,
+      "loss": 0.0977,
+      "step": 140
+    },
+    {
+      "epoch": 0.00027280229335794617,
+      "grad_norm": 0.0003795044613070786,
+      "learning_rate": 0.0002,
+      "loss": 0.0007,
+      "step": 150
+    },
+    {
+      "epoch": 0.0002909891129151426,
+      "grad_norm": 0.08740618824958801,
+      "learning_rate": 0.0002,
+      "loss": 0.801,
+      "step": 160
+    },
+    {
+      "epoch": 0.000309175932472339,
+      "grad_norm": 0.2661634087562561,
+      "learning_rate": 0.0002,
+      "loss": 0.1274,
+      "step": 170
+    },
+    {
+      "epoch": 0.0003273627520295354,
+      "grad_norm": 0.05828547850251198,
+      "learning_rate": 0.0002,
+      "loss": 0.1184,
+      "step": 180
+    },
+    {
+      "epoch": 0.0003455495715867318,
+      "grad_norm": 0.02175055630505085,
+      "learning_rate": 0.0002,
+      "loss": 0.0752,
+      "step": 190
+    },
+    {
+      "epoch": 0.0003637363911439282,
+      "grad_norm": 0.0009504792396910489,
+      "learning_rate": 0.0002,
+      "loss": 0.0005,
+      "step": 200
+    },
+    {
+      "epoch": 0.0003819232107011246,
+      "grad_norm": 0.25059741735458374,
+      "learning_rate": 0.0002,
+      "loss": 0.5125,
+      "step": 210
+    },
+    {
+      "epoch": 0.00040011003025832104,
+      "grad_norm": 0.13256193697452545,
+      "learning_rate": 0.0002,
+      "loss": 0.1014,
+      "step": 220
+    },
+    {
+      "epoch": 0.00041829684981551746,
+      "grad_norm": 0.09446375072002411,
+      "learning_rate": 0.0002,
+      "loss": 0.0896,
+      "step": 230
+    },
+    {
+      "epoch": 0.0004364836693727139,
+      "grad_norm": 0.019389621913433075,
+      "learning_rate": 0.0002,
+      "loss": 0.0726,
+      "step": 240
+    },
+    {
+      "epoch": 0.0004546704889299103,
+      "grad_norm": 0.0032304900232702494,
+      "learning_rate": 0.0002,
+      "loss": 0.0023,
+      "step": 250
+    },
+    {
+      "epoch": 0.00047285730848710666,
+      "grad_norm": 2.5549609661102295,
+      "learning_rate": 0.0002,
+      "loss": 0.3884,
+      "step": 260
+    },
+    {
+      "epoch": 0.0004910441280443031,
+      "grad_norm": 0.44937047362327576,
+      "learning_rate": 0.0002,
+      "loss": 0.1071,
+      "step": 270
+    },
+    {
+      "epoch": 0.0005092309476014995,
+      "grad_norm": 0.1509999781847,
+      "learning_rate": 0.0002,
+      "loss": 0.0979,
+      "step": 280
+    },
+    {
+      "epoch": 0.0005274177671586959,
+      "grad_norm": 0.006468054372817278,
+      "learning_rate": 0.0002,
+      "loss": 0.0611,
+      "step": 290
+    },
+    {
+      "epoch": 0.0005456045867158923,
+      "grad_norm": 0.0002916739322245121,
+      "learning_rate": 0.0002,
+      "loss": 0.001,
+      "step": 300
+    },
+    {
+      "epoch": 0.0005637914062730887,
+      "grad_norm": 0.23081810772418976,
+      "learning_rate": 0.0002,
+      "loss": 0.5894,
+      "step": 310
+    },
+    {
+      "epoch": 0.0005819782258302852,
+      "grad_norm": 0.22755394876003265,
+      "learning_rate": 0.0002,
+      "loss": 0.114,
+      "step": 320
+    },
+    {
+      "epoch": 0.0006001650453874816,
+      "grad_norm": 0.49973106384277344,
+      "learning_rate": 0.0002,
+      "loss": 0.093,
+      "step": 330
+    },
+    {
+      "epoch": 0.000618351864944678,
+      "grad_norm": 0.08789435774087906,
+      "learning_rate": 0.0002,
+      "loss": 0.0745,
+      "step": 340
+    },
+    {
+      "epoch": 0.0006365386845018744,
+      "grad_norm": 0.0058497479185462,
+      "learning_rate": 0.0002,
+      "loss": 0.0007,
+      "step": 350
+    },
+    {
+      "epoch": 0.0006547255040590708,
+      "grad_norm": 0.30569636821746826,
+      "learning_rate": 0.0002,
+      "loss": 0.5169,
+      "step": 360
+    },
+    {
+      "epoch": 0.0006729123236162671,
+      "grad_norm": 0.2783024311065674,
+      "learning_rate": 0.0002,
+      "loss": 0.13,
+      "step": 370
+    },
+    {
+      "epoch": 0.0006910991431734636,
+      "grad_norm": 0.13052967190742493,
+      "learning_rate": 0.0002,
+      "loss": 0.0907,
+      "step": 380
+    },
+    {
+      "epoch": 0.00070928596273066,
+      "grad_norm": 0.15066476166248322,
+      "learning_rate": 0.0002,
+      "loss": 0.0996,
+      "step": 390
+    },
+    {
+      "epoch": 0.0007274727822878564,
+      "grad_norm": 0.0005865198327228427,
+      "learning_rate": 0.0002,
+      "loss": 0.0021,
+      "step": 400
+    },
+    {
+      "epoch": 0.0007456596018450528,
+      "grad_norm": 0.31872233748435974,
+      "learning_rate": 0.0002,
+      "loss": 0.4507,
+      "step": 410
+    },
+    {
+      "epoch": 0.0007638464214022492,
+      "grad_norm": 0.08874880522489548,
+      "learning_rate": 0.0002,
+      "loss": 0.136,
+      "step": 420
+    },
+    {
+      "epoch": 0.0007820332409594457,
+      "grad_norm": 0.10985178500413895,
+      "learning_rate": 0.0002,
+      "loss": 0.0992,
+      "step": 430
+    },
+    {
+      "epoch": 0.0008002200605166421,
+      "grad_norm": 0.10776215046644211,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 440
+    },
+    {
+      "epoch": 0.0008184068800738385,
+      "grad_norm": 0.006612936966121197,
+      "learning_rate": 0.0002,
+      "loss": 0.0009,
+      "step": 450
+    },
+    {
+      "epoch": 0.0008365936996310349,
+      "grad_norm": 0.2757071256637573,
+      "learning_rate": 0.0002,
+      "loss": 0.6376,
+      "step": 460
+    },
+    {
+      "epoch": 0.0008547805191882313,
+      "grad_norm": 0.24748466908931732,
+      "learning_rate": 0.0002,
+      "loss": 0.1241,
+      "step": 470
+    },
+    {
+      "epoch": 0.0008729673387454278,
+      "grad_norm": 0.1035066694021225,
+      "learning_rate": 0.0002,
+      "loss": 0.1008,
+      "step": 480
+    },
+    {
+      "epoch": 0.0008911541583026242,
+      "grad_norm": 0.06515783071517944,
+      "learning_rate": 0.0002,
+      "loss": 0.0711,
+      "step": 490
+    },
+    {
+      "epoch": 0.0009093409778598206,
+      "grad_norm": 0.011224807240068913,
+      "learning_rate": 0.0002,
+      "loss": 0.0004,
+      "step": 500
+    },
+    {
+      "epoch": 0.000927527797417017,
+      "grad_norm": 0.2669332027435303,
+      "learning_rate": 0.0002,
+      "loss": 0.5618,
+      "step": 510
+    },
+    {
+      "epoch": 0.0009457146169742133,
+      "grad_norm": 0.26048392057418823,
+      "learning_rate": 0.0002,
+      "loss": 0.1259,
+      "step": 520
+    },
+    {
+      "epoch": 0.0009639014365314097,
+      "grad_norm": 0.22928836941719055,
+      "learning_rate": 0.0002,
+      "loss": 0.0956,
+      "step": 530
+    },
+    {
+      "epoch": 0.0009820882560886062,
+      "grad_norm": 0.084063321352005,
+      "learning_rate": 0.0002,
+      "loss": 0.0708,
+      "step": 540
+    },
+    {
+      "epoch": 0.0010002750756458027,
+      "grad_norm": 0.004612344317138195,
+      "learning_rate": 0.0002,
+      "loss": 0.0007,
+      "step": 550
+    },
+    {
+      "epoch": 0.001018461895202999,
+      "grad_norm": 0.3866584599018097,
+      "learning_rate": 0.0002,
+      "loss": 0.5406,
+      "step": 560
+    },
+    {
+      "epoch": 0.0010366487147601955,
+      "grad_norm": 0.32303065061569214,
+      "learning_rate": 0.0002,
+      "loss": 0.1001,
+      "step": 570
+    },
+    {
+      "epoch": 0.0010548355343173918,
+      "grad_norm": 0.09439560770988464,
+      "learning_rate": 0.0002,
+      "loss": 0.1051,
+      "step": 580
+    },
+    {
+      "epoch": 0.0010730223538745881,
+      "grad_norm": 0.028145521879196167,
+      "learning_rate": 0.0002,
+      "loss": 0.0638,
+      "step": 590
+    },
+    {
+      "epoch": 0.0010912091734317847,
+      "grad_norm": 0.00048497263924218714,
+      "learning_rate": 0.0002,
+      "loss": 0.002,
+      "step": 600
+    },
+    {
+      "epoch": 0.001109395992988981,
+      "grad_norm": 0.32391539216041565,
+      "learning_rate": 0.0002,
+      "loss": 0.5483,
+      "step": 610
+    },
+    {
+      "epoch": 0.0011275828125461775,
+      "grad_norm": 0.02977031283080578,
+      "learning_rate": 0.0002,
+      "loss": 0.1264,
+      "step": 620
+    },
+    {
+      "epoch": 0.0011457696321033738,
+      "grad_norm": 0.07332426309585571,
+      "learning_rate": 0.0002,
+      "loss": 0.1018,
+      "step": 630
+    },
+    {
+      "epoch": 0.0011639564516605703,
+      "grad_norm": 0.05653443560004234,
+      "learning_rate": 0.0002,
+      "loss": 0.0666,
+      "step": 640
+    },
+    {
+      "epoch": 0.0011821432712177666,
+      "grad_norm": 0.0010635281214490533,
+      "learning_rate": 0.0002,
+      "loss": 0.0009,
+      "step": 650
+    },
+    {
+      "epoch": 0.0012003300907749632,
+      "grad_norm": 0.04933600872755051,
+      "learning_rate": 0.0002,
+      "loss": 0.3902,
+      "step": 660
+    },
+    {
+      "epoch": 0.0012185169103321595,
+      "grad_norm": 0.14713574945926666,
+      "learning_rate": 0.0002,
+      "loss": 0.0905,
+      "step": 670
+    },
+    {
+      "epoch": 0.001236703729889356,
+      "grad_norm": 0.05463952198624611,
+      "learning_rate": 0.0002,
+      "loss": 0.0909,
+      "step": 680
+    },
+    {
+      "epoch": 0.0012548905494465523,
+      "grad_norm": 0.10299955308437347,
+      "learning_rate": 0.0002,
+      "loss": 0.07,
+      "step": 690
+    },
+    {
+      "epoch": 0.0012730773690037488,
+      "grad_norm": 0.022791124880313873,
+      "learning_rate": 0.0002,
+      "loss": 0.0027,
+      "step": 700
+    },
+    {
+      "epoch": 0.0012912641885609452,
+      "grad_norm": 0.27977490425109863,
+      "learning_rate": 0.0002,
+      "loss": 0.4421,
+      "step": 710
+    },
+    {
+      "epoch": 0.0013094510081181417,
+      "grad_norm": 0.2346329241991043,
+      "learning_rate": 0.0002,
+      "loss": 0.1263,
+      "step": 720
+    },
+    {
+      "epoch": 0.001327637827675338,
+      "grad_norm": 0.09294597059488297,
+      "learning_rate": 0.0002,
+      "loss": 0.096,
+      "step": 730
+    },
+    {
+      "epoch": 0.0013458246472325343,
+      "grad_norm": 0.10317150503396988,
+      "learning_rate": 0.0002,
+      "loss": 0.0727,
+      "step": 740
+    },
+    {
+      "epoch": 0.0013640114667897308,
+      "grad_norm": 0.001372635131701827,
+      "learning_rate": 0.0002,
+      "loss": 0.001,
+      "step": 750
+    },
+    {
+      "epoch": 0.0013821982863469271,
+      "grad_norm": 0.10563486814498901,
+      "learning_rate": 0.0002,
+      "loss": 0.596,
+      "step": 760
+    },
+    {
+      "epoch": 0.0014003851059041237,
+      "grad_norm": 0.14429838955402374,
+      "learning_rate": 0.0002,
+      "loss": 0.1178,
+      "step": 770
+    },
+    {
+      "epoch": 0.00141857192546132,
+      "grad_norm": 0.0848163515329361,
+      "learning_rate": 0.0002,
+      "loss": 0.1008,
+      "step": 780
+    },
+    {
+      "epoch": 0.0014367587450185165,
+      "grad_norm": 0.07259710133075714,
+      "learning_rate": 0.0002,
+      "loss": 0.069,
+      "step": 790
+    },
+    {
+      "epoch": 0.0014549455645757128,
+      "grad_norm": 0.0019098519114777446,
+      "learning_rate": 0.0002,
+      "loss": 0.0023,
+      "step": 800
+    },
+    {
+      "epoch": 0.0014731323841329093,
+      "grad_norm": 0.2433256059885025,
+      "learning_rate": 0.0002,
+      "loss": 0.2937,
+      "step": 810
+    },
+    {
+      "epoch": 0.0014913192036901056,
+      "grad_norm": 0.04093409329652786,
+      "learning_rate": 0.0002,
+      "loss": 0.1133,
+      "step": 820
+    },
+    {
+      "epoch": 0.0015095060232473022,
+      "grad_norm": 0.0480966717004776,
+      "learning_rate": 0.0002,
+      "loss": 0.0969,
+      "step": 830
+    },
+    {
+      "epoch": 0.0015276928428044985,
+      "grad_norm": 0.14327965676784515,
+      "learning_rate": 0.0002,
+      "loss": 0.0866,
+      "step": 840
+    },
+    {
+      "epoch": 0.001545879662361695,
+      "grad_norm": 0.001585015095770359,
+      "learning_rate": 0.0002,
+      "loss": 0.0042,
+      "step": 850
+    },
+    {
+      "epoch": 0.0015640664819188913,
+      "grad_norm": 0.1842886209487915,
+      "learning_rate": 0.0002,
+      "loss": 0.3273,
+      "step": 860
+    },
+    {
+      "epoch": 0.0015822533014760878,
+      "grad_norm": 0.09671049565076828,
+      "learning_rate": 0.0002,
+      "loss": 0.1079,
+      "step": 870
+    },
+    {
+      "epoch": 0.0016004401210332842,
+      "grad_norm": 0.2730088233947754,
+      "learning_rate": 0.0002,
+      "loss": 0.1018,
+      "step": 880
+    },
+    {
+      "epoch": 0.0016186269405904805,
+      "grad_norm": 0.11702803522348404,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 890
+    },
+    {
+      "epoch": 0.001636813760147677,
+      "grad_norm": 0.004438066389411688,
+      "learning_rate": 0.0002,
+      "loss": 0.0033,
+      "step": 900
+    },
+    {
+      "epoch": 0.0016550005797048733,
+      "grad_norm": 0.18424616754055023,
+      "learning_rate": 0.0002,
+      "loss": 0.4028,
+      "step": 910
+    },
+    {
+      "epoch": 0.0016731873992620698,
+      "grad_norm": 0.12502820789813995,
+      "learning_rate": 0.0002,
+      "loss": 0.0979,
+      "step": 920
+    },
+    {
+      "epoch": 0.0016913742188192661,
+      "grad_norm": 0.05109328031539917,
+      "learning_rate": 0.0002,
+      "loss": 0.0889,
+      "step": 930
+    },
+    {
+      "epoch": 0.0017095610383764627,
+      "grad_norm": 0.18566183745861053,
+      "learning_rate": 0.0002,
+      "loss": 0.0833,
+      "step": 940
+    },
+    {
+      "epoch": 0.001727747857933659,
+      "grad_norm": 0.0012954511912539601,
+      "learning_rate": 0.0002,
+      "loss": 0.0029,
+      "step": 950
+    },
+    {
+      "epoch": 0.0017459346774908555,
+      "grad_norm": 0.06683014333248138,
+      "learning_rate": 0.0002,
+      "loss": 0.4614,
+      "step": 960
+    },
+    {
+      "epoch": 0.0017641214970480518,
+      "grad_norm": 0.27773013710975647,
+      "learning_rate": 0.0002,
+      "loss": 0.1131,
+      "step": 970
+    },
+    {
+      "epoch": 0.0017823083166052483,
+      "grad_norm": 0.1999790072441101,
+      "learning_rate": 0.0002,
+      "loss": 0.089,
+      "step": 980
+    },
+    {
+      "epoch": 0.0018004951361624446,
+      "grad_norm": 0.09625103324651718,
+      "learning_rate": 0.0002,
+      "loss": 0.0739,
+      "step": 990
+    },
+    {
+      "epoch": 0.0018186819557196412,
+      "grad_norm": 0.005470380187034607,
+      "learning_rate": 0.0002,
+      "loss": 0.0012,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0018368687752768375,
+      "grad_norm": 0.038832616060972214,
+      "learning_rate": 0.0002,
+      "loss": 0.5521,
+      "step": 1010
+    },
+    {
+      "epoch": 0.001855055594834034,
+      "grad_norm": 0.1903093159198761,
+      "learning_rate": 0.0002,
+      "loss": 0.1237,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0018732424143912303,
+      "grad_norm": 0.031102774664759636,
+      "learning_rate": 0.0002,
+      "loss": 0.0866,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0018914292339484266,
+      "grad_norm": 0.043983202427625656,
+      "learning_rate": 0.0002,
+      "loss": 0.0611,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0019096160535056232,
+      "grad_norm": 0.0002974902163259685,
+      "learning_rate": 0.0002,
+      "loss": 0.0035,
+      "step": 1050
+    },
+    {
+      "epoch": 0.0019278028730628195,
+      "grad_norm": 0.1936149299144745,
+      "learning_rate": 0.0002,
+      "loss": 0.3019,
+      "step": 1060
+    },
+    {
+      "epoch": 0.001945989692620016,
+      "grad_norm": 0.15767355263233185,
+      "learning_rate": 0.0002,
+      "loss": 0.108,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0019641765121772123,
+      "grad_norm": 0.08244495838880539,
+      "learning_rate": 0.0002,
+      "loss": 0.091,
+      "step": 1080
+    },
+    {
+      "epoch": 0.0019823633317344086,
+      "grad_norm": 0.15848897397518158,
+      "learning_rate": 0.0002,
+      "loss": 0.0655,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0020005501512916054,
+      "grad_norm": 0.0011951205087825656,
+      "learning_rate": 0.0002,
+      "loss": 0.0052,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0020187369708488017,
+      "grad_norm": 0.13027112185955048,
+      "learning_rate": 0.0002,
+      "loss": 0.2943,
+      "step": 1110
+    },
+    {
+      "epoch": 0.002036923790405998,
+      "grad_norm": 0.19413979351520538,
+      "learning_rate": 0.0002,
+      "loss": 0.1329,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0020551106099631943,
+      "grad_norm": 0.08515465259552002,
+      "learning_rate": 0.0002,
+      "loss": 0.0921,
+      "step": 1130
+    },
+    {
+      "epoch": 0.002073297429520391,
+      "grad_norm": 0.1244177296757698,
+      "learning_rate": 0.0002,
+      "loss": 0.0678,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0020914842490775873,
+      "grad_norm": 0.0016714326338842511,
+      "learning_rate": 0.0002,
+      "loss": 0.0035,
+      "step": 1150
+    },
+    {
+      "epoch": 0.0021096710686347836,
+      "grad_norm": 0.24979737401008606,
+      "learning_rate": 0.0002,
+      "loss": 0.2643,
+      "step": 1160
+    },
+    {
+      "epoch": 0.00212785788819198,
+      "grad_norm": 0.14143353700637817,
+      "learning_rate": 0.0002,
+      "loss": 0.1037,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0021460447077491763,
+      "grad_norm": 0.033794257789850235,
+      "learning_rate": 0.0002,
+      "loss": 0.087,
+      "step": 1180
+    },
+    {
+      "epoch": 0.002164231527306373,
+      "grad_norm": 0.11503162235021591,
+      "learning_rate": 0.0002,
+      "loss": 0.0659,
+      "step": 1190
+    },
+    {
+      "epoch": 0.0021824183468635693,
+      "grad_norm": 0.0014654065016657114,
+      "learning_rate": 0.0002,
+      "loss": 0.0056,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0022006051664207656,
+      "grad_norm": 0.13292767107486725,
+      "learning_rate": 0.0002,
+      "loss": 0.2956,
+      "step": 1210
+    },
+    {
+      "epoch": 0.002218791985977962,
+      "grad_norm": 0.15238040685653687,
+      "learning_rate": 0.0002,
+      "loss": 0.1122,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0022369788055351587,
+      "grad_norm": 0.045078523457050323,
+      "learning_rate": 0.0002,
+      "loss": 0.091,
+      "step": 1230
+    },
+    {
+      "epoch": 0.002255165625092355,
+      "grad_norm": 0.11438468098640442,
+      "learning_rate": 0.0002,
+      "loss": 0.0754,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0022733524446495513,
+      "grad_norm": 0.001236733514815569,
+      "learning_rate": 0.0002,
+      "loss": 0.004,
+      "step": 1250
+    },
+    {
+      "epoch": 0.0022915392642067476,
+      "grad_norm": 0.23386552929878235,
+      "learning_rate": 0.0002,
+      "loss": 0.351,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0023097260837639444,
+      "grad_norm": 0.030786139890551567,
+      "learning_rate": 0.0002,
+      "loss": 0.1074,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0023279129033211407,
+      "grad_norm": 0.150347501039505,
+      "learning_rate": 0.0002,
+      "loss": 0.1064,
+      "step": 1280
+    },
+    {
+      "epoch": 0.002346099722878337,
+      "grad_norm": 0.1402382105588913,
+      "learning_rate": 0.0002,
+      "loss": 0.0675,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0023642865424355333,
+      "grad_norm": 0.0006117303855717182,
+      "learning_rate": 0.0002,
+      "loss": 0.0031,
+      "step": 1300
+    },
+    {
+      "epoch": 0.00238247336199273,
+      "grad_norm": 0.16031372547149658,
+      "learning_rate": 0.0002,
+      "loss": 0.4344,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0024006601815499263,
+      "grad_norm": 0.11017303168773651,
+      "learning_rate": 0.0002,
+      "loss": 0.1147,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0024188470011071227,
+      "grad_norm": 0.055746905505657196,
+      "learning_rate": 0.0002,
+      "loss": 0.093,
+      "step": 1330
+    },
+    {
+      "epoch": 0.002437033820664319,
+      "grad_norm": 0.09806664288043976,
+      "learning_rate": 0.0002,
+      "loss": 0.0682,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0024552206402215153,
+      "grad_norm": 0.000555588339921087,
+      "learning_rate": 0.0002,
+      "loss": 0.0045,
+      "step": 1350
+    },
+    {
+      "epoch": 0.002473407459778712,
+      "grad_norm": 0.04899182915687561,
+      "learning_rate": 0.0002,
+      "loss": 0.3454,
+      "step": 1360
+    },
+    {
+      "epoch": 0.0024915942793359083,
+      "grad_norm": 0.02870030514895916,
+      "learning_rate": 0.0002,
+      "loss": 0.1036,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0025097810988931046,
+      "grad_norm": 0.08591730147600174,
+      "learning_rate": 0.0002,
+      "loss": 0.0962,
+      "step": 1380
+    },
+    {
+      "epoch": 0.002527967918450301,
+      "grad_norm": 0.1169242337346077,
+      "learning_rate": 0.0002,
+      "loss": 0.0627,
+      "step": 1390
+    },
+    {
+      "epoch": 0.0025461547380074977,
+      "grad_norm": 0.0008637752034701407,
+      "learning_rate": 0.0002,
+      "loss": 0.0025,
+      "step": 1400
+    },
+    {
+      "epoch": 0.002564341557564694,
+      "grad_norm": 0.11741841584444046,
+      "learning_rate": 0.0002,
+      "loss": 0.3703,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0025825283771218903,
+      "grad_norm": 0.05232485383749008,
+      "learning_rate": 0.0002,
+      "loss": 0.1072,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0026007151966790866,
+      "grad_norm": 0.025201110169291496,
+      "learning_rate": 0.0002,
+      "loss": 0.0893,
+      "step": 1430
+    },
+    {
+      "epoch": 0.0026189020162362834,
+      "grad_norm": 0.11462239921092987,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0026370888357934797,
+      "grad_norm": 0.002194227883592248,
+      "learning_rate": 0.0002,
+      "loss": 0.0049,
+      "step": 1450
+    },
+    {
+      "epoch": 0.002655275655350676,
+      "grad_norm": 0.05786404758691788,
+      "learning_rate": 0.0002,
+      "loss": 0.3187,
+      "step": 1460
+    },
+    {
+      "epoch": 0.0026734624749078723,
+      "grad_norm": 0.03776915743947029,
+      "learning_rate": 0.0002,
+      "loss": 0.1002,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0026916492944650686,
+      "grad_norm": 0.08628734946250916,
+      "learning_rate": 0.0002,
+      "loss": 0.0933,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0027098361140222653,
+      "grad_norm": 0.0933455228805542,
+      "learning_rate": 0.0002,
+      "loss": 0.0712,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0027280229335794617,
+      "grad_norm": 0.0007446192903444171,
+      "learning_rate": 0.0002,
+      "loss": 0.003,
+      "step": 1500
+    },
+    {
+      "epoch": 0.002746209753136658,
+      "grad_norm": 0.04412281885743141,
+      "learning_rate": 0.0002,
+      "loss": 0.3738,
+      "step": 1510
+    },
+    {
+      "epoch": 0.0027643965726938543,
+      "grad_norm": 0.04729326814413071,
+      "learning_rate": 0.0002,
+      "loss": 0.1015,
+      "step": 1520
+    },
+    {
+      "epoch": 0.002782583392251051,
+      "grad_norm": 0.04822024703025818,
+      "learning_rate": 0.0002,
+      "loss": 0.0913,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0028007702118082473,
+      "grad_norm": 0.15468090772628784,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 1540
+    },
+    {
+      "epoch": 0.0028189570313654436,
+      "grad_norm": 0.0011828596470877528,
+      "learning_rate": 0.0002,
+      "loss": 0.0089,
+      "step": 1550
+    },
+    {
+      "epoch": 0.00283714385092264,
+      "grad_norm": 0.030639037489891052,
+      "learning_rate": 0.0002,
+      "loss": 0.3382,
+      "step": 1560
+    },
+    {
+      "epoch": 0.0028553306704798367,
+      "grad_norm": 0.08429472148418427,
+      "learning_rate": 0.0002,
+      "loss": 0.1075,
+      "step": 1570
+    },
+    {
+      "epoch": 0.002873517490037033,
+      "grad_norm": 0.056431323289871216,
+      "learning_rate": 0.0002,
+      "loss": 0.0946,
+      "step": 1580
+    },
+    {
+      "epoch": 0.0028917043095942293,
+      "grad_norm": 0.1799512803554535,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 1590
+    },
+    {
+      "epoch": 0.0029098911291514256,
+      "grad_norm": 0.0018818675307556987,
+      "learning_rate": 0.0002,
+      "loss": 0.0082,
+      "step": 1600
+    },
+    {
+      "epoch": 0.002928077948708622,
+      "grad_norm": 0.061398155987262726,
+      "learning_rate": 0.0002,
+      "loss": 0.3414,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0029462647682658187,
+      "grad_norm": 0.0657019093632698,
+      "learning_rate": 0.0002,
+      "loss": 0.1082,
+      "step": 1620
+    },
+    {
+      "epoch": 0.002964451587823015,
+      "grad_norm": 0.04701487720012665,
+      "learning_rate": 0.0002,
+      "loss": 0.0918,
+      "step": 1630
+    },
+    {
+      "epoch": 0.0029826384073802113,
+      "grad_norm": 0.1834430694580078,
+      "learning_rate": 0.0002,
+      "loss": 0.081,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0030008252269374076,
+      "grad_norm": 0.004841644782572985,
+      "learning_rate": 0.0002,
+      "loss": 0.0138,
+      "step": 1650
+    },
+    {
+      "epoch": 0.0030190120464946043,
+      "grad_norm": 0.05793444439768791,
+      "learning_rate": 0.0002,
+      "loss": 0.2981,
+      "step": 1660
+    },
+    {
+      "epoch": 0.0030371988660518007,
+      "grad_norm": 0.049123138189315796,
+      "learning_rate": 0.0002,
+      "loss": 0.1072,
+      "step": 1670
+    },
+    {
+      "epoch": 0.003055385685608997,
+      "grad_norm": 0.033852141350507736,
+      "learning_rate": 0.0002,
+      "loss": 0.093,
+      "step": 1680
+    },
+    {
+      "epoch": 0.0030735725051661933,
+      "grad_norm": 0.16161279380321503,
+      "learning_rate": 0.0002,
+      "loss": 0.084,
+      "step": 1690
+    },
+    {
+      "epoch": 0.00309175932472339,
+      "grad_norm": 0.0011225020280107856,
+      "learning_rate": 0.0002,
+      "loss": 0.0059,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0031099461442805863,
+      "grad_norm": 0.05849582701921463,
+      "learning_rate": 0.0002,
+      "loss": 0.3878,
+      "step": 1710
+    },
+    {
+      "epoch": 0.0031281329638377826,
+      "grad_norm": 0.033466637134552,
+      "learning_rate": 0.0002,
+      "loss": 0.1096,
+      "step": 1720
+    },
+    {
+      "epoch": 0.003146319783394979,
+      "grad_norm": 0.03488466143608093,
+      "learning_rate": 0.0002,
+      "loss": 0.0895,
+      "step": 1730
+    },
+    {
+      "epoch": 0.0031645066029521757,
+      "grad_norm": 0.15636079013347626,
+      "learning_rate": 0.0002,
+      "loss": 0.0716,
+      "step": 1740
+    },
+    {
+      "epoch": 0.003182693422509372,
+      "grad_norm": 0.001519509358331561,
+      "learning_rate": 0.0002,
+      "loss": 0.0062,
+      "step": 1750
+    },
+    {
+      "epoch": 0.0032008802420665683,
+      "grad_norm": 0.04979783296585083,
+      "learning_rate": 0.0002,
+      "loss": 0.3409,
+      "step": 1760
+    },
+    {
+      "epoch": 0.0032190670616237646,
+      "grad_norm": 0.09706272929906845,
+      "learning_rate": 0.0002,
+      "loss": 0.1052,
+      "step": 1770
+    },
+    {
+      "epoch": 0.003237253881180961,
+      "grad_norm": 0.08768483251333237,
+      "learning_rate": 0.0002,
+      "loss": 0.0938,
+      "step": 1780
+    },
+    {
+      "epoch": 0.0032554407007381577,
+      "grad_norm": 0.20421457290649414,
+      "learning_rate": 0.0002,
+      "loss": 0.085,
+      "step": 1790
+    },
+    {
+      "epoch": 0.003273627520295354,
+      "grad_norm": 0.0024727964773774147,
+      "learning_rate": 0.0002,
+      "loss": 0.0147,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0032918143398525503,
+      "grad_norm": 0.04270516335964203,
+      "learning_rate": 0.0002,
+      "loss": 0.2872,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0033100011594097466,
+      "grad_norm": 0.08055799454450607,
+      "learning_rate": 0.0002,
+      "loss": 0.0992,
+      "step": 1820
+    },
+    {
+      "epoch": 0.0033281879789669433,
+      "grad_norm": 0.02607434056699276,
+      "learning_rate": 0.0002,
+      "loss": 0.0803,
+      "step": 1830
+    },
+    {
+      "epoch": 0.0033463747985241397,
+      "grad_norm": 0.16260816156864166,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 1840
+    },
+    {
+      "epoch": 0.003364561618081336,
+      "grad_norm": 0.004690333269536495,
+      "learning_rate": 0.0002,
+      "loss": 0.012,
+      "step": 1850
+    },
+    {
+      "epoch": 0.0033827484376385323,
+      "grad_norm": 0.041513338685035706,
+      "learning_rate": 0.0002,
+      "loss": 0.2491,
+      "step": 1860
+    },
+    {
+      "epoch": 0.003400935257195729,
+      "grad_norm": 0.08935420960187912,
+      "learning_rate": 0.0002,
+      "loss": 0.1001,
+      "step": 1870
+    },
+    {
+      "epoch": 0.0034191220767529253,
+      "grad_norm": 0.03826737776398659,
+      "learning_rate": 0.0002,
+      "loss": 0.0877,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0034373088963101216,
+      "grad_norm": 0.19423778355121613,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 1890
+    },
+    {
+      "epoch": 0.003455495715867318,
+      "grad_norm": 0.003520288970321417,
+      "learning_rate": 0.0002,
+      "loss": 0.013,
+      "step": 1900
+    },
+    {
+      "epoch": 0.0034736825354245143,
+      "grad_norm": 0.14648132026195526,
+      "learning_rate": 0.0002,
+      "loss": 0.3209,
+      "step": 1910
+    },
+    {
+      "epoch": 0.003491869354981711,
+      "grad_norm": 0.03780071437358856,
+      "learning_rate": 0.0002,
+      "loss": 0.0934,
+      "step": 1920
+    },
+    {
+      "epoch": 0.0035100561745389073,
+      "grad_norm": 0.05014612153172493,
+      "learning_rate": 0.0002,
+      "loss": 0.082,
+      "step": 1930
+    },
+    {
+      "epoch": 0.0035282429940961036,
+      "grad_norm": 0.12917590141296387,
+      "learning_rate": 0.0002,
+      "loss": 0.0733,
+      "step": 1940
+    },
+    {
+      "epoch": 0.0035464298136533,
+      "grad_norm": 0.0030132795218378305,
+      "learning_rate": 0.0002,
+      "loss": 0.0111,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0035646166332104967,
+      "grad_norm": 0.03008626028895378,
+      "learning_rate": 0.0002,
+      "loss": 0.2126,
+      "step": 1960
+    },
+    {
+      "epoch": 0.003582803452767693,
+      "grad_norm": 0.0915503203868866,
+      "learning_rate": 0.0002,
+      "loss": 0.1097,
+      "step": 1970
+    },
+    {
+      "epoch": 0.0036009902723248893,
+      "grad_norm": 0.06607015430927277,
+      "learning_rate": 0.0002,
+      "loss": 0.0932,
+      "step": 1980
+    },
+    {
+      "epoch": 0.0036191770918820856,
+      "grad_norm": 0.18796613812446594,
+      "learning_rate": 0.0002,
+      "loss": 0.083,
+      "step": 1990
+    },
+    {
+      "epoch": 0.0036373639114392823,
+      "grad_norm": 0.0022257096134126186,
+      "learning_rate": 0.0002,
+      "loss": 0.0147,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0036555507309964787,
+      "grad_norm": 0.0687415823340416,
+      "learning_rate": 0.0002,
+      "loss": 0.2604,
+      "step": 2010
+    },
+    {
+      "epoch": 0.003673737550553675,
+      "grad_norm": 0.025175679475069046,
+      "learning_rate": 0.0002,
+      "loss": 0.0998,
+      "step": 2020
+    },
+    {
+      "epoch": 0.0036919243701108713,
+      "grad_norm": 0.04275168478488922,
+      "learning_rate": 0.0002,
+      "loss": 0.0898,
+      "step": 2030
+    },
+    {
+      "epoch": 0.003710111189668068,
+      "grad_norm": 0.17306455969810486,
+      "learning_rate": 0.0002,
+      "loss": 0.0793,
+      "step": 2040
+    },
+    {
+      "epoch": 0.0037282980092252643,
+      "grad_norm": 0.007826454006135464,
+      "learning_rate": 0.0002,
+      "loss": 0.011,
+      "step": 2050
+    },
+    {
+      "epoch": 0.0037464848287824606,
+      "grad_norm": 0.06461178511381149,
+      "learning_rate": 0.0002,
+      "loss": 0.2597,
+      "step": 2060
+    },
+    {
+      "epoch": 0.003764671648339657,
+      "grad_norm": 0.061357177793979645,
+      "learning_rate": 0.0002,
+      "loss": 0.1001,
+      "step": 2070
+    },
+    {
+      "epoch": 0.0037828584678968533,
+      "grad_norm": 0.029154235497117043,
+      "learning_rate": 0.0002,
+      "loss": 0.0859,
+      "step": 2080
+    },
+    {
+      "epoch": 0.00380104528745405,
+      "grad_norm": 0.1350340098142624,
+      "learning_rate": 0.0002,
+      "loss": 0.0756,
+      "step": 2090
+    },
+    {
+      "epoch": 0.0038192321070112463,
+      "grad_norm": 0.0017614173702895641,
+      "learning_rate": 0.0002,
+      "loss": 0.0058,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0038374189265684426,
+      "grad_norm": 0.024254316464066505,
+      "learning_rate": 0.0002,
+      "loss": 0.3349,
+      "step": 2110
+    },
+    {
+      "epoch": 0.003855605746125639,
+      "grad_norm": 0.07142530381679535,
+      "learning_rate": 0.0002,
+      "loss": 0.0953,
+      "step": 2120
+    },
+    {
+      "epoch": 0.0038737925656828357,
+      "grad_norm": 0.05570175498723984,
+      "learning_rate": 0.0002,
+      "loss": 0.0796,
+      "step": 2130
+    },
+    {
+      "epoch": 0.003891979385240032,
+      "grad_norm": 0.16996875405311584,
+      "learning_rate": 0.0002,
+      "loss": 0.0782,
+      "step": 2140
+    },
+    {
+      "epoch": 0.003910166204797228,
+      "grad_norm": 0.0058751595206558704,
+      "learning_rate": 0.0002,
+      "loss": 0.0206,
+      "step": 2150
+    },
+    {
+      "epoch": 0.003928353024354425,
+      "grad_norm": 0.029807811602950096,
+      "learning_rate": 0.0002,
+      "loss": 0.1926,
+      "step": 2160
+    },
+    {
+      "epoch": 0.003946539843911621,
+      "grad_norm": 0.11123469471931458,
+      "learning_rate": 0.0002,
+      "loss": 0.1082,
+      "step": 2170
+    },
+    {
+      "epoch": 0.003964726663468817,
+      "grad_norm": 0.074626125395298,
+      "learning_rate": 0.0002,
+      "loss": 0.081,
+      "step": 2180
+    },
+    {
+      "epoch": 0.003982913483026014,
+      "grad_norm": 0.17397737503051758,
+      "learning_rate": 0.0002,
+      "loss": 0.0729,
+      "step": 2190
+    },
+    {
+      "epoch": 0.004001100302583211,
+      "grad_norm": 0.007995887659490108,
+      "learning_rate": 0.0002,
+      "loss": 0.022,
+      "step": 2200
+    },
+    {
+      "epoch": 0.004019287122140407,
+      "grad_norm": 0.039921898394823074,
+      "learning_rate": 0.0002,
+      "loss": 0.1883,
+      "step": 2210
+    },
+    {
+      "epoch": 0.004037473941697603,
+      "grad_norm": 0.07736324518918991,
+      "learning_rate": 0.0002,
+      "loss": 0.0941,
+      "step": 2220
+    },
+    {
+      "epoch": 0.0040556607612548,
+      "grad_norm": 0.0867881178855896,
+      "learning_rate": 0.0002,
+      "loss": 0.0873,
+      "step": 2230
+    },
+    {
+      "epoch": 0.004073847580811996,
+      "grad_norm": 0.1497400403022766,
+      "learning_rate": 0.0002,
+      "loss": 0.0829,
+      "step": 2240
+    },
+    {
+      "epoch": 0.004092034400369192,
+      "grad_norm": 0.007458314299583435,
+      "learning_rate": 0.0002,
+      "loss": 0.02,
+      "step": 2250
+    },
+    {
+      "epoch": 0.004110221219926389,
+      "grad_norm": 0.04168029874563217,
+      "learning_rate": 0.0002,
+      "loss": 0.2176,
+      "step": 2260
+    },
+    {
+      "epoch": 0.004128408039483585,
+      "grad_norm": 0.10017130523920059,
+      "learning_rate": 0.0002,
+      "loss": 0.0958,
+      "step": 2270
+    },
+    {
+      "epoch": 0.004146594859040782,
+      "grad_norm": 0.02727416157722473,
+      "learning_rate": 0.0002,
+      "loss": 0.088,
+      "step": 2280
+    },
+    {
+      "epoch": 0.004164781678597978,
+      "grad_norm": 0.15034393966197968,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 2290
+    },
+    {
+      "epoch": 0.004182968498155175,
+      "grad_norm": 0.0023451410233974457,
+      "learning_rate": 0.0002,
+      "loss": 0.0102,
+      "step": 2300
+    },
+    {
+      "epoch": 0.004201155317712371,
+      "grad_norm": 0.03462455794215202,
+      "learning_rate": 0.0002,
+      "loss": 0.3404,
+      "step": 2310
+    },
+    {
+      "epoch": 0.004219342137269567,
+      "grad_norm": 0.02866148017346859,
+      "learning_rate": 0.0002,
+      "loss": 0.0932,
+      "step": 2320
+    },
+    {
+      "epoch": 0.004237528956826764,
+      "grad_norm": 0.0685456171631813,
+      "learning_rate": 0.0002,
+      "loss": 0.0806,
+      "step": 2330
+    },
+    {
+      "epoch": 0.00425571577638396,
+      "grad_norm": 0.17208056151866913,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 2340
+    },
+    {
+      "epoch": 0.004273902595941156,
+      "grad_norm": 0.008708455599844456,
+      "learning_rate": 0.0002,
+      "loss": 0.0171,
+      "step": 2350
+    },
+    {
+      "epoch": 0.0042920894154983525,
+      "grad_norm": 0.044025715440511703,
+      "learning_rate": 0.0002,
+      "loss": 0.212,
+      "step": 2360
+    },
+    {
+      "epoch": 0.00431027623505555,
+      "grad_norm": 0.050246164202690125,
+      "learning_rate": 0.0002,
+      "loss": 0.107,
+      "step": 2370
+    },
+    {
+      "epoch": 0.004328463054612746,
+      "grad_norm": 0.05257886275649071,
+      "learning_rate": 0.0002,
+      "loss": 0.0868,
+      "step": 2380
+    },
+    {
+      "epoch": 0.004346649874169942,
+      "grad_norm": 0.16567641496658325,
+      "learning_rate": 0.0002,
+      "loss": 0.0819,
+      "step": 2390
+    },
+    {
+      "epoch": 0.004364836693727139,
+      "grad_norm": 0.0062621901743113995,
+      "learning_rate": 0.0002,
+      "loss": 0.0171,
+      "step": 2400
+    },
+    {
+      "epoch": 0.004383023513284335,
+      "grad_norm": 0.03025338612496853,
+      "learning_rate": 0.0002,
+      "loss": 0.2141,
+      "step": 2410
+    },
+    {
+      "epoch": 0.004401210332841531,
+      "grad_norm": 0.06401577591896057,
+      "learning_rate": 0.0002,
+      "loss": 0.0982,
+      "step": 2420
+    },
+    {
+      "epoch": 0.004419397152398728,
+      "grad_norm": 0.12474781274795532,
+      "learning_rate": 0.0002,
+      "loss": 0.0834,
+      "step": 2430
+    },
+    {
+      "epoch": 0.004437583971955924,
+      "grad_norm": 0.18607665598392487,
+      "learning_rate": 0.0002,
+      "loss": 0.0801,
+      "step": 2440
+    },
+    {
+      "epoch": 0.004455770791513121,
+      "grad_norm": 0.0017643098253756762,
+      "learning_rate": 0.0002,
+      "loss": 0.0129,
+      "step": 2450
+    },
+    {
+      "epoch": 0.004473957611070317,
+      "grad_norm": 0.03936386480927467,
+      "learning_rate": 0.0002,
+      "loss": 0.2541,
+      "step": 2460
+    },
+    {
+      "epoch": 0.004492144430627514,
+      "grad_norm": 0.08961635082960129,
+      "learning_rate": 0.0002,
+      "loss": 0.0961,
+      "step": 2470
+    },
+    {
+      "epoch": 0.00451033125018471,
+      "grad_norm": 0.07525113970041275,
+      "learning_rate": 0.0002,
+      "loss": 0.0844,
+      "step": 2480
+    },
+    {
+      "epoch": 0.004528518069741906,
+      "grad_norm": 0.16746751964092255,
+      "learning_rate": 0.0002,
+      "loss": 0.071,
+      "step": 2490
+    },
+    {
+      "epoch": 0.004546704889299103,
+      "grad_norm": 0.0027625334914773703,
+      "learning_rate": 0.0002,
+      "loss": 0.0151,
+      "step": 2500
+    },
+    {
+      "epoch": 0.004564891708856299,
+      "grad_norm": 0.049662694334983826,
+      "learning_rate": 0.0002,
+      "loss": 0.253,
+      "step": 2510
+    },
+    {
+      "epoch": 0.004583078528413495,
+      "grad_norm": 0.08312079310417175,
+      "learning_rate": 0.0002,
+      "loss": 0.0922,
+      "step": 2520
+    },
+    {
+      "epoch": 0.0046012653479706915,
+      "grad_norm": 0.0646345317363739,
+      "learning_rate": 0.0002,
+      "loss": 0.0889,
+      "step": 2530
+    },
+    {
+      "epoch": 0.004619452167527889,
+      "grad_norm": 0.20036271214485168,
+      "learning_rate": 0.0002,
+      "loss": 0.081,
+      "step": 2540
+    },
+    {
+      "epoch": 0.004637638987085085,
+      "grad_norm": 0.010091719217598438,
+      "learning_rate": 0.0002,
+      "loss": 0.024,
+      "step": 2550
+    },
+    {
+      "epoch": 0.004655825806642281,
+      "grad_norm": 0.048885516822338104,
+      "learning_rate": 0.0002,
+      "loss": 0.184,
+      "step": 2560
+    },
+    {
+      "epoch": 0.004674012626199478,
+      "grad_norm": 0.09142889827489853,
+      "learning_rate": 0.0002,
+      "loss": 0.0935,
+      "step": 2570
+    },
+    {
+      "epoch": 0.004692199445756674,
+      "grad_norm": 0.049207963049411774,
+      "learning_rate": 0.0002,
+      "loss": 0.0816,
+      "step": 2580
+    },
+    {
+      "epoch": 0.00471038626531387,
+      "grad_norm": 0.1498396098613739,
+      "learning_rate": 0.0002,
+      "loss": 0.0698,
+      "step": 2590
+    },
+    {
+      "epoch": 0.004728573084871067,
+      "grad_norm": 0.00522881094366312,
+      "learning_rate": 0.0002,
+      "loss": 0.0189,
+      "step": 2600
+    },
+    {
+      "epoch": 0.004746759904428263,
+      "grad_norm": 0.07461311668157578,
+      "learning_rate": 0.0002,
+      "loss": 0.1944,
+      "step": 2610
+    },
+    {
+      "epoch": 0.00476494672398546,
+      "grad_norm": 0.048005711287260056,
+      "learning_rate": 0.0002,
+      "loss": 0.0883,
+      "step": 2620
+    },
+    {
+      "epoch": 0.004783133543542656,
+      "grad_norm": 0.10151612013578415,
+      "learning_rate": 0.0002,
+      "loss": 0.0827,
+      "step": 2630
+    },
+    {
+      "epoch": 0.004801320363099853,
+      "grad_norm": 0.1504422426223755,
+      "learning_rate": 0.0002,
+      "loss": 0.078,
+      "step": 2640
+    },
+    {
+      "epoch": 0.004819507182657049,
+      "grad_norm": 0.004988422151654959,
+      "learning_rate": 0.0002,
+      "loss": 0.0229,
+      "step": 2650
+    },
+    {
+      "epoch": 0.004837694002214245,
+      "grad_norm": 0.025008924305438995,
+      "learning_rate": 0.0002,
+      "loss": 0.1818,
+      "step": 2660
+    },
+    {
+      "epoch": 0.004855880821771442,
+      "grad_norm": 0.027460169047117233,
+      "learning_rate": 0.0002,
+      "loss": 0.0966,
+      "step": 2670
+    },
+    {
+      "epoch": 0.004874067641328638,
+      "grad_norm": 0.09704197943210602,
+      "learning_rate": 0.0002,
+      "loss": 0.0824,
+      "step": 2680
+    },
+    {
+      "epoch": 0.004892254460885834,
+      "grad_norm": 0.138654425740242,
+      "learning_rate": 0.0002,
+      "loss": 0.0746,
+      "step": 2690
+    },
+    {
+      "epoch": 0.0049104412804430305,
+      "grad_norm": 0.00859556533396244,
+      "learning_rate": 0.0002,
+      "loss": 0.0187,
+      "step": 2700
+    },
+    {
+      "epoch": 0.004928628100000228,
+      "grad_norm": 0.05207522585988045,
+      "learning_rate": 0.0002,
+      "loss": 0.1985,
+      "step": 2710
+    },
+    {
+      "epoch": 0.004946814919557424,
+      "grad_norm": 0.07787417620420456,
+      "learning_rate": 0.0002,
+      "loss": 0.101,
+      "step": 2720
+    },
+    {
+      "epoch": 0.00496500173911462,
+      "grad_norm": 0.02819981426000595,
+      "learning_rate": 0.0002,
+      "loss": 0.0845,
+      "step": 2730
+    },
+    {
+      "epoch": 0.004983188558671817,
+      "grad_norm": 0.13569314777851105,
+      "learning_rate": 0.0002,
+      "loss": 0.0756,
+      "step": 2740
+    },
+    {
+      "epoch": 0.005001375378229013,
+      "grad_norm": 0.05175986513495445,
+      "learning_rate": 0.0002,
+      "loss": 0.024,
+      "step": 2750
+    },
+    {
+      "epoch": 0.005019562197786209,
+      "grad_norm": 0.037230249494314194,
+      "learning_rate": 0.0002,
+      "loss": 0.2056,
+      "step": 2760
+    },
+    {
+      "epoch": 0.005037749017343406,
+      "grad_norm": 0.05532974749803543,
+      "learning_rate": 0.0002,
+      "loss": 0.0939,
+      "step": 2770
+    },
+    {
+      "epoch": 0.005055935836900602,
+      "grad_norm": 0.06930708140134811,
+      "learning_rate": 0.0002,
+      "loss": 0.0853,
+      "step": 2780
+    },
+    {
+      "epoch": 0.005074122656457798,
+      "grad_norm": 0.16405801475048065,
+      "learning_rate": 0.0002,
+      "loss": 0.0766,
+      "step": 2790
+    },
+    {
+      "epoch": 0.005092309476014995,
+      "grad_norm": 0.006398684345185757,
+      "learning_rate": 0.0002,
+      "loss": 0.0124,
+      "step": 2800
+    },
+    {
+      "epoch": 0.005110496295572192,
+      "grad_norm": 0.06269315630197525,
+      "learning_rate": 0.0002,
+      "loss": 0.2703,
+      "step": 2810
+    },
+    {
+      "epoch": 0.005128683115129388,
+      "grad_norm": 0.049293261021375656,
+      "learning_rate": 0.0002,
+      "loss": 0.0943,
+      "step": 2820
+    },
+    {
+      "epoch": 0.005146869934686584,
+      "grad_norm": 0.08814405649900436,
+      "learning_rate": 0.0002,
+      "loss": 0.0855,
+      "step": 2830
+    },
+    {
+      "epoch": 0.005165056754243781,
+      "grad_norm": 0.17452259361743927,
+      "learning_rate": 0.0002,
+      "loss": 0.0822,
+      "step": 2840
+    },
+    {
+      "epoch": 0.005183243573800977,
+      "grad_norm": 0.005008229520171881,
+      "learning_rate": 0.0002,
+      "loss": 0.0136,
+      "step": 2850
+    },
+    {
+      "epoch": 0.005201430393358173,
+      "grad_norm": 0.04459540545940399,
+      "learning_rate": 0.0002,
+      "loss": 0.2623,
+      "step": 2860
+    },
+    {
+      "epoch": 0.0052196172129153695,
+      "grad_norm": 0.042845603078603745,
+      "learning_rate": 0.0002,
+      "loss": 0.0929,
+      "step": 2870
+    },
+    {
+      "epoch": 0.005237804032472567,
+      "grad_norm": 0.03079635463654995,
+      "learning_rate": 0.0002,
+      "loss": 0.0844,
+      "step": 2880
+    },
+    {
+      "epoch": 0.005255990852029763,
+      "grad_norm": 0.14457851648330688,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 2890
+    },
+    {
+      "epoch": 0.005274177671586959,
+      "grad_norm": 0.0009016963304020464,
+      "learning_rate": 0.0002,
+      "loss": 0.0037,
+      "step": 2900
+    },
+    {
+      "epoch": 0.005292364491144156,
+      "grad_norm": 0.0983906164765358,
+      "learning_rate": 0.0002,
+      "loss": 0.3661,
+      "step": 2910
+    },
+    {
+      "epoch": 0.005310551310701352,
+      "grad_norm": 0.08794154971837997,
+      "learning_rate": 0.0002,
+      "loss": 0.0894,
+      "step": 2920
+    },
+    {
+      "epoch": 0.005328738130258548,
+      "grad_norm": 0.026981573551893234,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 2930
+    },
+    {
+      "epoch": 0.005346924949815745,
+      "grad_norm": 0.15572553873062134,
+      "learning_rate": 0.0002,
+      "loss": 0.077,
+      "step": 2940
+    },
+    {
+      "epoch": 0.005365111769372941,
+      "grad_norm": 0.005491070915013552,
+      "learning_rate": 0.0002,
+      "loss": 0.0092,
+      "step": 2950
+    },
+    {
+      "epoch": 0.005383298588930137,
+      "grad_norm": 0.07383686304092407,
+      "learning_rate": 0.0002,
+      "loss": 0.2574,
+      "step": 2960
+    },
+    {
+      "epoch": 0.005401485408487334,
+      "grad_norm": 0.05919960141181946,
+      "learning_rate": 0.0002,
+      "loss": 0.1045,
+      "step": 2970
+    },
+    {
+      "epoch": 0.005419672228044531,
+      "grad_norm": 0.06027739867568016,
+      "learning_rate": 0.0002,
+      "loss": 0.0822,
+      "step": 2980
+    },
+    {
+      "epoch": 0.005437859047601727,
+      "grad_norm": 0.1288602501153946,
+      "learning_rate": 0.0002,
+      "loss": 0.0688,
+      "step": 2990
+    },
+    {
+      "epoch": 0.005456045867158923,
+      "grad_norm": 0.007565880194306374,
+      "learning_rate": 0.0002,
+      "loss": 0.0192,
+      "step": 3000
+    },
+    {
+      "epoch": 0.00547423268671612,
+      "grad_norm": 0.024412864819169044,
+      "learning_rate": 0.0002,
+      "loss": 0.1782,
+      "step": 3010
+    },
+    {
+      "epoch": 0.005492419506273316,
+      "grad_norm": 0.05559355765581131,
+      "learning_rate": 0.0002,
+      "loss": 0.1072,
+      "step": 3020
+    },
+    {
+      "epoch": 0.005510606325830512,
+      "grad_norm": 0.07073906064033508,
+      "learning_rate": 0.0002,
+      "loss": 0.0863,
+      "step": 3030
+    },
+    {
+      "epoch": 0.0055287931453877085,
+      "grad_norm": 0.14979414641857147,
+      "learning_rate": 0.0002,
+      "loss": 0.0788,
+      "step": 3040
+    },
+    {
+      "epoch": 0.005546979964944906,
+      "grad_norm": 0.0057297456078231335,
+      "learning_rate": 0.0002,
+      "loss": 0.0192,
+      "step": 3050
+    },
+    {
+      "epoch": 0.005565166784502102,
+      "grad_norm": 0.03195042535662651,
+      "learning_rate": 0.0002,
+      "loss": 0.1879,
+      "step": 3060
+    },
+    {
+      "epoch": 0.005583353604059298,
+      "grad_norm": 0.05925082787871361,
+      "learning_rate": 0.0002,
+      "loss": 0.0992,
+      "step": 3070
+    },
+    {
+      "epoch": 0.005601540423616495,
+      "grad_norm": 0.052063606679439545,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 3080
+    },
+    {
+      "epoch": 0.005619727243173691,
+      "grad_norm": 0.16005952656269073,
+      "learning_rate": 0.0002,
+      "loss": 0.0743,
+      "step": 3090
+    },
+    {
+      "epoch": 0.005637914062730887,
+      "grad_norm": 0.005742133595049381,
+      "learning_rate": 0.0002,
+      "loss": 0.0137,
+      "step": 3100
+    },
+    {
+      "epoch": 0.005656100882288084,
+      "grad_norm": 0.07523638010025024,
+      "learning_rate": 0.0002,
+      "loss": 0.2072,
+      "step": 3110
+    },
+    {
+      "epoch": 0.00567428770184528,
+      "grad_norm": 0.23799611628055573,
+      "learning_rate": 0.0002,
+      "loss": 0.0906,
+      "step": 3120
+    },
+    {
+      "epoch": 0.005692474521402476,
+      "grad_norm": 0.06176261603832245,
+      "learning_rate": 0.0002,
+      "loss": 0.088,
+      "step": 3130
+    },
+    {
+      "epoch": 0.005710661340959673,
+      "grad_norm": 0.13692723214626312,
+      "learning_rate": 0.0002,
+      "loss": 0.0733,
+      "step": 3140
+    },
+    {
+      "epoch": 0.00572884816051687,
+      "grad_norm": 0.007059803698211908,
+      "learning_rate": 0.0002,
+      "loss": 0.0194,
+      "step": 3150
+    },
+    {
+      "epoch": 0.005747034980074066,
+      "grad_norm": 0.08868405222892761,
+      "learning_rate": 0.0002,
+      "loss": 0.1745,
+      "step": 3160
+    },
+    {
+      "epoch": 0.005765221799631262,
+      "grad_norm": 0.05126733332872391,
+      "learning_rate": 0.0002,
+      "loss": 0.1024,
+      "step": 3170
+    },
+    {
+      "epoch": 0.005783408619188459,
+      "grad_norm": 0.06377821415662766,
+      "learning_rate": 0.0002,
+      "loss": 0.0846,
+      "step": 3180
+    },
+    {
+      "epoch": 0.005801595438745655,
+      "grad_norm": 0.10748566687107086,
+      "learning_rate": 0.0002,
+      "loss": 0.0769,
+      "step": 3190
+    },
+    {
+      "epoch": 0.005819782258302851,
+      "grad_norm": 0.004992443602532148,
+      "learning_rate": 0.0002,
+      "loss": 0.0114,
+      "step": 3200
+    },
+    {
+      "epoch": 0.0058379690778600475,
+      "grad_norm": 0.0420277863740921,
+      "learning_rate": 0.0002,
+      "loss": 0.2159,
+      "step": 3210
+    },
+    {
+      "epoch": 0.005856155897417244,
+      "grad_norm": 0.02828531712293625,
+      "learning_rate": 0.0002,
+      "loss": 0.0923,
+      "step": 3220
+    },
+    {
+      "epoch": 0.005874342716974441,
+      "grad_norm": 0.028216248378157616,
+      "learning_rate": 0.0002,
+      "loss": 0.0789,
+      "step": 3230
+    },
+    {
+      "epoch": 0.005892529536531637,
+      "grad_norm": 0.11420746147632599,
+      "learning_rate": 0.0002,
+      "loss": 0.0696,
+      "step": 3240
+    },
+    {
+      "epoch": 0.005910716356088834,
+      "grad_norm": 0.0019631448667496443,
+      "learning_rate": 0.0002,
+      "loss": 0.0128,
+      "step": 3250
+    },
+    {
+      "epoch": 0.00592890317564603,
+      "grad_norm": 0.05514012649655342,
+      "learning_rate": 0.0002,
+      "loss": 0.2609,
+      "step": 3260
+    },
+    {
+      "epoch": 0.005947089995203226,
+      "grad_norm": 0.0917636826634407,
+      "learning_rate": 0.0002,
+      "loss": 0.0996,
+      "step": 3270
+    },
+    {
+      "epoch": 0.005965276814760423,
+      "grad_norm": 0.03648284077644348,
+      "learning_rate": 0.0002,
+      "loss": 0.084,
+      "step": 3280
+    },
+    {
+      "epoch": 0.005983463634317619,
+      "grad_norm": 0.13859149813652039,
+      "learning_rate": 0.0002,
+      "loss": 0.0807,
+      "step": 3290
+    },
+    {
+      "epoch": 0.006001650453874815,
+      "grad_norm": 0.013779910281300545,
+      "learning_rate": 0.0002,
+      "loss": 0.0181,
+      "step": 3300
+    },
+    {
+      "epoch": 0.006019837273432012,
+      "grad_norm": 0.02654041163623333,
+      "learning_rate": 0.0002,
+      "loss": 0.1636,
+      "step": 3310
+    },
+    {
+      "epoch": 0.006038024092989209,
+      "grad_norm": 0.062298137694597244,
+      "learning_rate": 0.0002,
+      "loss": 0.0872,
+      "step": 3320
+    },
+    {
+      "epoch": 0.006056210912546405,
+      "grad_norm": 0.0351388119161129,
+      "learning_rate": 0.0002,
+      "loss": 0.0802,
+      "step": 3330
+    },
+    {
+      "epoch": 0.006074397732103601,
+      "grad_norm": 0.16063807904720306,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 3340
+    },
+    {
+      "epoch": 0.006092584551660798,
+      "grad_norm": 0.009991235099732876,
+      "learning_rate": 0.0002,
+      "loss": 0.016,
+      "step": 3350
+    },
+    {
+      "epoch": 0.006110771371217994,
+      "grad_norm": 0.052919622510671616,
+      "learning_rate": 0.0002,
+      "loss": 0.2027,
+      "step": 3360
+    },
+    {
+      "epoch": 0.00612895819077519,
+      "grad_norm": 0.03228602185845375,
+      "learning_rate": 0.0002,
+      "loss": 0.0985,
+      "step": 3370
+    },
+    {
+      "epoch": 0.0061471450103323865,
+      "grad_norm": 0.11311203986406326,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 3380
+    },
+    {
+      "epoch": 0.006165331829889583,
+      "grad_norm": 0.1674620360136032,
+      "learning_rate": 0.0002,
+      "loss": 0.072,
+      "step": 3390
+    },
+    {
+      "epoch": 0.00618351864944678,
+      "grad_norm": 0.015154430642724037,
+      "learning_rate": 0.0002,
+      "loss": 0.0186,
+      "step": 3400
+    },
+    {
+      "epoch": 0.006201705469003976,
+      "grad_norm": 0.043151434510946274,
+      "learning_rate": 0.0002,
+      "loss": 0.1892,
+      "step": 3410
+    },
+    {
+      "epoch": 0.006219892288561173,
+      "grad_norm": 0.12342707067728043,
+      "learning_rate": 0.0002,
+      "loss": 0.0907,
+      "step": 3420
+    },
+    {
+      "epoch": 0.006238079108118369,
+      "grad_norm": 0.08350827544927597,
+      "learning_rate": 0.0002,
+      "loss": 0.0783,
+      "step": 3430
+    },
+    {
+      "epoch": 0.006256265927675565,
+      "grad_norm": 0.11938697844743729,
+      "learning_rate": 0.0002,
+      "loss": 0.0666,
+      "step": 3440
+    },
+    {
+      "epoch": 0.006274452747232762,
+      "grad_norm": 0.015424132347106934,
+      "learning_rate": 0.0002,
+      "loss": 0.0173,
+      "step": 3450
+    },
+    {
+      "epoch": 0.006292639566789958,
+      "grad_norm": 0.04220043867826462,
+      "learning_rate": 0.0002,
+      "loss": 0.1805,
+      "step": 3460
+    },
+    {
+      "epoch": 0.006310826386347154,
+      "grad_norm": 0.08813903480768204,
+      "learning_rate": 0.0002,
+      "loss": 0.096,
+      "step": 3470
+    },
+    {
+      "epoch": 0.006329013205904351,
+      "grad_norm": 0.07647278904914856,
+      "learning_rate": 0.0002,
+      "loss": 0.0821,
+      "step": 3480
+    },
+    {
+      "epoch": 0.006347200025461548,
+      "grad_norm": 0.14242641627788544,
+      "learning_rate": 0.0002,
+      "loss": 0.0769,
+      "step": 3490
+    },
+    {
+      "epoch": 0.006365386845018744,
+      "grad_norm": 0.011115231551229954,
+      "learning_rate": 0.0002,
+      "loss": 0.0221,
+      "step": 3500
+    },
+    {
+      "epoch": 0.00638357366457594,
+      "grad_norm": 0.036351826041936874,
+      "learning_rate": 0.0002,
+      "loss": 0.1557,
+      "step": 3510
+    },
+    {
+      "epoch": 0.006401760484133137,
+      "grad_norm": 0.08549819141626358,
+      "learning_rate": 0.0002,
+      "loss": 0.0864,
+      "step": 3520
+    },
+    {
+      "epoch": 0.006419947303690333,
+      "grad_norm": 0.047141823917627335,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 3530
+    },
+    {
+      "epoch": 0.006438134123247529,
+      "grad_norm": 0.13143447041511536,
+      "learning_rate": 0.0002,
+      "loss": 0.0717,
+      "step": 3540
+    },
+    {
+      "epoch": 0.0064563209428047256,
+      "grad_norm": 0.013524871319532394,
+      "learning_rate": 0.0002,
+      "loss": 0.0149,
+      "step": 3550
+    },
+    {
+      "epoch": 0.006474507762361922,
+      "grad_norm": 0.03367459774017334,
+      "learning_rate": 0.0002,
+      "loss": 0.1715,
+      "step": 3560
+    },
+    {
+      "epoch": 0.006492694581919119,
+      "grad_norm": 0.045889757573604584,
+      "learning_rate": 0.0002,
+      "loss": 0.0949,
+      "step": 3570
+    },
+    {
+      "epoch": 0.006510881401476315,
+      "grad_norm": 0.04099202901124954,
+      "learning_rate": 0.0002,
+      "loss": 0.0813,
+      "step": 3580
+    },
+    {
+      "epoch": 0.006529068221033512,
+      "grad_norm": 0.133371040225029,
+      "learning_rate": 0.0002,
+      "loss": 0.0767,
+      "step": 3590
+    },
+    {
+      "epoch": 0.006547255040590708,
+      "grad_norm": 0.00645647756755352,
+      "learning_rate": 0.0002,
+      "loss": 0.0186,
+      "step": 3600
+    },
+    {
+      "epoch": 0.006565441860147904,
+      "grad_norm": 0.050674330443143845,
+      "learning_rate": 0.0002,
+      "loss": 0.2179,
+      "step": 3610
+    },
+    {
+      "epoch": 0.006583628679705101,
+      "grad_norm": 0.07087302207946777,
+      "learning_rate": 0.0002,
+      "loss": 0.0882,
+      "step": 3620
+    },
+    {
+      "epoch": 0.006601815499262297,
+      "grad_norm": 0.02759486250579357,
+      "learning_rate": 0.0002,
+      "loss": 0.0789,
+      "step": 3630
+    },
+    {
+      "epoch": 0.006620002318819493,
+      "grad_norm": 0.12163479626178741,
+      "learning_rate": 0.0002,
+      "loss": 0.0689,
+      "step": 3640
+    },
+    {
+      "epoch": 0.00663818913837669,
+      "grad_norm": 0.00969718024134636,
+      "learning_rate": 0.0002,
+      "loss": 0.0112,
+      "step": 3650
+    },
+    {
+      "epoch": 0.006656375957933887,
+      "grad_norm": 0.07106204330921173,
+      "learning_rate": 0.0002,
+      "loss": 0.199,
+      "step": 3660
+    },
+    {
+      "epoch": 0.006674562777491083,
+      "grad_norm": 0.08954132348299026,
+      "learning_rate": 0.0002,
+      "loss": 0.0985,
+      "step": 3670
+    },
+    {
+      "epoch": 0.006692749597048279,
+      "grad_norm": 0.09899396449327469,
+      "learning_rate": 0.0002,
+      "loss": 0.0811,
+      "step": 3680
+    },
+    {
+      "epoch": 0.006710936416605476,
+      "grad_norm": 0.12119311839342117,
+      "learning_rate": 0.0002,
+      "loss": 0.0698,
+      "step": 3690
+    },
+    {
+      "epoch": 0.006729123236162672,
+      "grad_norm": 0.013957214541733265,
+      "learning_rate": 0.0002,
+      "loss": 0.018,
+      "step": 3700
+    },
+    {
+      "epoch": 0.006747310055719868,
+      "grad_norm": 0.03089285083115101,
+      "learning_rate": 0.0002,
+      "loss": 0.1434,
+      "step": 3710
+    },
+    {
+      "epoch": 0.0067654968752770646,
+      "grad_norm": 0.025650829076766968,
+      "learning_rate": 0.0002,
+      "loss": 0.0886,
+      "step": 3720
+    },
+    {
+      "epoch": 0.006783683694834261,
+      "grad_norm": 0.044103365391492844,
+      "learning_rate": 0.0002,
+      "loss": 0.0788,
+      "step": 3730
+    },
+    {
+      "epoch": 0.006801870514391458,
+      "grad_norm": 0.09726370871067047,
+      "learning_rate": 0.0002,
+      "loss": 0.0674,
+      "step": 3740
+    },
+    {
+      "epoch": 0.006820057333948654,
+      "grad_norm": 0.018105274066329002,
+      "learning_rate": 0.0002,
+      "loss": 0.0232,
+      "step": 3750
+    },
+    {
+      "epoch": 0.006838244153505851,
+      "grad_norm": 0.021543240174651146,
+      "learning_rate": 0.0002,
+      "loss": 0.1406,
+      "step": 3760
+    },
+    {
+      "epoch": 0.006856430973063047,
+      "grad_norm": 0.09367050975561142,
+      "learning_rate": 0.0002,
+      "loss": 0.0973,
+      "step": 3770
+    },
+    {
+      "epoch": 0.006874617792620243,
+      "grad_norm": 0.06836032122373581,
+      "learning_rate": 0.0002,
+      "loss": 0.0848,
+      "step": 3780
+    },
+    {
+      "epoch": 0.00689280461217744,
+      "grad_norm": 0.11758081614971161,
+      "learning_rate": 0.0002,
+      "loss": 0.0693,
+      "step": 3790
+    },
+    {
+      "epoch": 0.006910991431734636,
+      "grad_norm": 0.008669364266097546,
+      "learning_rate": 0.0002,
+      "loss": 0.0223,
+      "step": 3800
+    },
+    {
+      "epoch": 0.006929178251291832,
+      "grad_norm": 0.03903719782829285,
+      "learning_rate": 0.0002,
+      "loss": 0.1519,
+      "step": 3810
+    },
+    {
+      "epoch": 0.0069473650708490285,
+      "grad_norm": 0.030682874843478203,
+      "learning_rate": 0.0002,
+      "loss": 0.0931,
+      "step": 3820
+    },
+    {
+      "epoch": 0.006965551890406226,
+      "grad_norm": 0.02693006955087185,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 3830
+    },
+    {
+      "epoch": 0.006983738709963422,
+      "grad_norm": 0.09535166621208191,
+      "learning_rate": 0.0002,
+      "loss": 0.0696,
+      "step": 3840
+    },
+    {
+      "epoch": 0.007001925529520618,
+      "grad_norm": 0.014680403284728527,
+      "learning_rate": 0.0002,
+      "loss": 0.0176,
+      "step": 3850
+    },
+    {
+      "epoch": 0.007020112349077815,
+      "grad_norm": 0.031090212985873222,
+      "learning_rate": 0.0002,
+      "loss": 0.1544,
+      "step": 3860
+    },
+    {
+      "epoch": 0.007038299168635011,
+      "grad_norm": 0.05870644003152847,
+      "learning_rate": 0.0002,
+      "loss": 0.0898,
+      "step": 3870
+    },
+    {
+      "epoch": 0.007056485988192207,
+      "grad_norm": 0.03480982780456543,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 3880
+    },
+    {
+      "epoch": 0.0070746728077494036,
+      "grad_norm": 0.09751418977975845,
+      "learning_rate": 0.0002,
+      "loss": 0.0724,
+      "step": 3890
+    },
+    {
+      "epoch": 0.0070928596273066,
+      "grad_norm": 0.022084850817918777,
+      "learning_rate": 0.0002,
+      "loss": 0.019,
+      "step": 3900
+    },
+    {
+      "epoch": 0.007111046446863797,
+      "grad_norm": 0.06994971632957458,
+      "learning_rate": 0.0002,
+      "loss": 0.1478,
+      "step": 3910
+    },
+    {
+      "epoch": 0.007129233266420993,
+      "grad_norm": 0.05761263892054558,
+      "learning_rate": 0.0002,
+      "loss": 0.0932,
+      "step": 3920
+    },
+    {
+      "epoch": 0.00714742008597819,
+      "grad_norm": 0.029772033914923668,
+      "learning_rate": 0.0002,
+      "loss": 0.0855,
+      "step": 3930
+    },
+    {
+      "epoch": 0.007165606905535386,
+      "grad_norm": 0.11868726462125778,
+      "learning_rate": 0.0002,
+      "loss": 0.0727,
+      "step": 3940
+    },
+    {
+      "epoch": 0.007183793725092582,
+      "grad_norm": 0.0065403408370912075,
+      "learning_rate": 0.0002,
+      "loss": 0.0174,
+      "step": 3950
+    },
+    {
+      "epoch": 0.007201980544649779,
+      "grad_norm": 0.031544361263513565,
+      "learning_rate": 0.0002,
+      "loss": 0.1827,
+      "step": 3960
+    },
+    {
+      "epoch": 0.007220167364206975,
+      "grad_norm": 0.031641531735658646,
+      "learning_rate": 0.0002,
+      "loss": 0.0867,
+      "step": 3970
+    },
+    {
+      "epoch": 0.007238354183764171,
+      "grad_norm": 0.028574040159583092,
+      "learning_rate": 0.0002,
+      "loss": 0.0798,
+      "step": 3980
+    },
+    {
+      "epoch": 0.0072565410033213675,
+      "grad_norm": 0.12866555154323578,
+      "learning_rate": 0.0002,
+      "loss": 0.0708,
+      "step": 3990
+    },
+    {
+      "epoch": 0.007274727822878565,
+      "grad_norm": 0.00843430683016777,
+      "learning_rate": 0.0002,
+      "loss": 0.0127,
+      "step": 4000
+    },
+    {
+      "epoch": 0.007292914642435761,
+      "grad_norm": 0.03737691789865494,
+      "learning_rate": 0.0002,
+      "loss": 0.2201,
+      "step": 4010
+    },
+    {
+      "epoch": 0.007311101461992957,
+      "grad_norm": 0.05326579511165619,
+      "learning_rate": 0.0002,
+      "loss": 0.0838,
+      "step": 4020
+    },
+    {
+      "epoch": 0.007329288281550154,
+      "grad_norm": 0.031934209167957306,
+      "learning_rate": 0.0002,
+      "loss": 0.0778,
+      "step": 4030
+    },
+    {
+      "epoch": 0.00734747510110735,
+      "grad_norm": 0.17401957511901855,
+      "learning_rate": 0.0002,
+      "loss": 0.073,
+      "step": 4040
+    },
+    {
+      "epoch": 0.007365661920664546,
+      "grad_norm": 0.005256639327853918,
+      "learning_rate": 0.0002,
+      "loss": 0.0122,
+      "step": 4050
+    },
+    {
+      "epoch": 0.0073838487402217426,
+      "grad_norm": 0.05043623968958855,
+      "learning_rate": 0.0002,
+      "loss": 0.2524,
+      "step": 4060
+    },
+    {
+      "epoch": 0.007402035559778939,
+      "grad_norm": 0.06662425398826599,
+      "learning_rate": 0.0002,
+      "loss": 0.0976,
+      "step": 4070
+    },
+    {
+      "epoch": 0.007420222379336136,
+      "grad_norm": 0.13419686257839203,
+      "learning_rate": 0.0002,
+      "loss": 0.0833,
+      "step": 4080
+    },
+    {
+      "epoch": 0.007438409198893332,
+      "grad_norm": 0.176285520195961,
+      "learning_rate": 0.0002,
+      "loss": 0.0761,
+      "step": 4090
+    },
+    {
+      "epoch": 0.007456596018450529,
+      "grad_norm": 0.008489354513585567,
+      "learning_rate": 0.0002,
+      "loss": 0.0182,
+      "step": 4100
+    },
+    {
+      "epoch": 0.007474782838007725,
+      "grad_norm": 0.06247509643435478,
+      "learning_rate": 0.0002,
+      "loss": 0.2232,
+      "step": 4110
+    },
+    {
+      "epoch": 0.007492969657564921,
+      "grad_norm": 0.05744702368974686,
+      "learning_rate": 0.0002,
+      "loss": 0.0875,
+      "step": 4120
+    },
+    {
+      "epoch": 0.007511156477122118,
+      "grad_norm": 0.053026407957077026,
+      "learning_rate": 0.0002,
+      "loss": 0.0807,
+      "step": 4130
+    },
+    {
+      "epoch": 0.007529343296679314,
+      "grad_norm": 0.11734003573656082,
+      "learning_rate": 0.0002,
+      "loss": 0.0724,
+      "step": 4140
+    },
+    {
+      "epoch": 0.00754753011623651,
+      "grad_norm": 0.005216363817453384,
+      "learning_rate": 0.0002,
+      "loss": 0.0129,
+      "step": 4150
+    },
+    {
+      "epoch": 0.0075657169357937065,
+      "grad_norm": 0.08154789358377457,
+      "learning_rate": 0.0002,
+      "loss": 0.2221,
+      "step": 4160
+    },
+    {
+      "epoch": 0.007583903755350904,
+      "grad_norm": 0.03619784861803055,
+      "learning_rate": 0.0002,
+      "loss": 0.0993,
+      "step": 4170
+    },
+    {
+      "epoch": 0.0076020905749081,
+      "grad_norm": 0.08239256590604782,
+      "learning_rate": 0.0002,
+      "loss": 0.0811,
+      "step": 4180
+    },
+    {
+      "epoch": 0.007620277394465296,
+      "grad_norm": 0.11934535950422287,
+      "learning_rate": 0.0002,
+      "loss": 0.0726,
+      "step": 4190
+    },
+    {
+      "epoch": 0.007638464214022493,
+      "grad_norm": 0.006965799257159233,
+      "learning_rate": 0.0002,
+      "loss": 0.0181,
+      "step": 4200
+    },
+    {
+      "epoch": 0.007656651033579689,
+      "grad_norm": 0.04328077286481857,
+      "learning_rate": 0.0002,
+      "loss": 0.1983,
+      "step": 4210
+    },
+    {
+      "epoch": 0.007674837853136885,
+      "grad_norm": 0.08253510296344757,
+      "learning_rate": 0.0002,
+      "loss": 0.0954,
+      "step": 4220
+    },
+    {
+      "epoch": 0.0076930246726940816,
+      "grad_norm": 0.06146657094359398,
+      "learning_rate": 0.0002,
+      "loss": 0.0843,
+      "step": 4230
+    },
+    {
+      "epoch": 0.007711211492251278,
+      "grad_norm": 0.13579218089580536,
+      "learning_rate": 0.0002,
+      "loss": 0.0672,
+      "step": 4240
+    },
+    {
+      "epoch": 0.007729398311808474,
+      "grad_norm": 0.0038396338932216167,
+      "learning_rate": 0.0002,
+      "loss": 0.0131,
+      "step": 4250
+    },
+    {
+      "epoch": 0.007747585131365671,
+      "grad_norm": 0.03109130822122097,
+      "learning_rate": 0.0002,
+      "loss": 0.2102,
+      "step": 4260
+    },
+    {
+      "epoch": 0.007765771950922868,
+      "grad_norm": 0.04971664398908615,
+      "learning_rate": 0.0002,
+      "loss": 0.0903,
+      "step": 4270
+    },
+    {
+      "epoch": 0.007783958770480064,
+      "grad_norm": 0.06476306915283203,
+      "learning_rate": 0.0002,
+      "loss": 0.0859,
+      "step": 4280
+    },
+    {
+      "epoch": 0.00780214559003726,
+      "grad_norm": 0.15377041697502136,
+      "learning_rate": 0.0002,
+      "loss": 0.0828,
+      "step": 4290
+    },
+    {
+      "epoch": 0.007820332409594457,
+      "grad_norm": 0.005592274013906717,
+      "learning_rate": 0.0002,
+      "loss": 0.014,
+      "step": 4300
+    },
+    {
+      "epoch": 0.007838519229151653,
+      "grad_norm": 0.04387212172150612,
+      "learning_rate": 0.0002,
+      "loss": 0.1907,
+      "step": 4310
+    },
+    {
+      "epoch": 0.00785670604870885,
+      "grad_norm": 0.06001356989145279,
+      "learning_rate": 0.0002,
+      "loss": 0.0864,
+      "step": 4320
+    },
+    {
+      "epoch": 0.007874892868266046,
+      "grad_norm": 0.030866140499711037,
+      "learning_rate": 0.0002,
+      "loss": 0.0748,
+      "step": 4330
+    },
+    {
+      "epoch": 0.007893079687823242,
+      "grad_norm": 0.13280808925628662,
+      "learning_rate": 0.0002,
+      "loss": 0.0686,
+      "step": 4340
+    },
+    {
+      "epoch": 0.007911266507380438,
+      "grad_norm": 0.015559020452201366,
+      "learning_rate": 0.0002,
+      "loss": 0.016,
+      "step": 4350
+    },
+    {
+      "epoch": 0.007929453326937634,
+      "grad_norm": 0.0669974684715271,
+      "learning_rate": 0.0002,
+      "loss": 0.1916,
+      "step": 4360
+    },
+    {
+      "epoch": 0.00794764014649483,
+      "grad_norm": 0.0759076252579689,
+      "learning_rate": 0.0002,
+      "loss": 0.0925,
+      "step": 4370
+    },
+    {
+      "epoch": 0.007965826966052029,
+      "grad_norm": 0.029388410970568657,
+      "learning_rate": 0.0002,
+      "loss": 0.086,
+      "step": 4380
+    },
+    {
+      "epoch": 0.007984013785609225,
+      "grad_norm": 0.17637981474399567,
+      "learning_rate": 0.0002,
+      "loss": 0.0697,
+      "step": 4390
+    },
+    {
+      "epoch": 0.008002200605166421,
+      "grad_norm": 0.008022189140319824,
+      "learning_rate": 0.0002,
+      "loss": 0.0196,
+      "step": 4400
+    },
+    {
+      "epoch": 0.008020387424723618,
+      "grad_norm": 0.04126167669892311,
+      "learning_rate": 0.0002,
+      "loss": 0.192,
+      "step": 4410
+    },
+    {
+      "epoch": 0.008038574244280814,
+      "grad_norm": 0.08132971078157425,
+      "learning_rate": 0.0002,
+      "loss": 0.093,
+      "step": 4420
+    },
+    {
+      "epoch": 0.00805676106383801,
+      "grad_norm": 0.07568484544754028,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 4430
+    },
+    {
+      "epoch": 0.008074947883395207,
+      "grad_norm": 0.1259222775697708,
+      "learning_rate": 0.0002,
+      "loss": 0.0696,
+      "step": 4440
+    },
+    {
+      "epoch": 0.008093134702952403,
+      "grad_norm": 0.009711826220154762,
+      "learning_rate": 0.0002,
+      "loss": 0.0232,
+      "step": 4450
+    },
+    {
+      "epoch": 0.0081113215225096,
+      "grad_norm": 0.029734279960393906,
+      "learning_rate": 0.0002,
+      "loss": 0.1595,
+      "step": 4460
+    },
+    {
+      "epoch": 0.008129508342066796,
+      "grad_norm": 0.04886960610747337,
+      "learning_rate": 0.0002,
+      "loss": 0.0919,
+      "step": 4470
+    },
+    {
+      "epoch": 0.008147695161623992,
+      "grad_norm": 0.07031470537185669,
+      "learning_rate": 0.0002,
+      "loss": 0.0813,
+      "step": 4480
+    },
+    {
+      "epoch": 0.008165881981181188,
+      "grad_norm": 0.12099859863519669,
+      "learning_rate": 0.0002,
+      "loss": 0.0731,
+      "step": 4490
+    },
+    {
+      "epoch": 0.008184068800738385,
+      "grad_norm": 0.02181529812514782,
+      "learning_rate": 0.0002,
+      "loss": 0.021,
+      "step": 4500
+    },
+    {
+      "epoch": 0.00820225562029558,
+      "grad_norm": 0.035477787256240845,
+      "learning_rate": 0.0002,
+      "loss": 0.1429,
+      "step": 4510
+    },
+    {
+      "epoch": 0.008220442439852777,
+      "grad_norm": 0.07788772135972977,
+      "learning_rate": 0.0002,
+      "loss": 0.0842,
+      "step": 4520
+    },
+    {
+      "epoch": 0.008238629259409973,
+      "grad_norm": 0.045833125710487366,
+      "learning_rate": 0.0002,
+      "loss": 0.0829,
+      "step": 4530
+    },
+    {
+      "epoch": 0.00825681607896717,
+      "grad_norm": 0.12271951884031296,
+      "learning_rate": 0.0002,
+      "loss": 0.0707,
+      "step": 4540
+    },
+    {
+      "epoch": 0.008275002898524366,
+      "grad_norm": 0.01919553242623806,
+      "learning_rate": 0.0002,
+      "loss": 0.0213,
+      "step": 4550
+    },
+    {
+      "epoch": 0.008293189718081564,
+      "grad_norm": 0.032527096569538116,
+      "learning_rate": 0.0002,
+      "loss": 0.1397,
+      "step": 4560
+    },
+    {
+      "epoch": 0.00831137653763876,
+      "grad_norm": 0.045243579894304276,
+      "learning_rate": 0.0002,
+      "loss": 0.0854,
+      "step": 4570
+    },
+    {
+      "epoch": 0.008329563357195957,
+      "grad_norm": 0.04226524010300636,
+      "learning_rate": 0.0002,
+      "loss": 0.0728,
+      "step": 4580
+    },
+    {
+      "epoch": 0.008347750176753153,
+      "grad_norm": 0.09887039661407471,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 4590
+    },
+    {
+      "epoch": 0.00836593699631035,
+      "grad_norm": 0.01822318509221077,
+      "learning_rate": 0.0002,
+      "loss": 0.0169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.008384123815867546,
+      "grad_norm": 0.05729951336979866,
+      "learning_rate": 0.0002,
+      "loss": 0.137,
+      "step": 4610
+    },
+    {
+      "epoch": 0.008402310635424742,
+      "grad_norm": 0.041520439088344574,
+      "learning_rate": 0.0002,
+      "loss": 0.0825,
+      "step": 4620
+    },
+    {
+      "epoch": 0.008420497454981938,
+      "grad_norm": 0.051164623349905014,
+      "learning_rate": 0.0002,
+      "loss": 0.0818,
+      "step": 4630
+    },
+    {
+      "epoch": 0.008438684274539135,
+      "grad_norm": 0.1289409101009369,
+      "learning_rate": 0.0002,
+      "loss": 0.0664,
+      "step": 4640
+    },
+    {
+      "epoch": 0.008456871094096331,
+      "grad_norm": 0.0085114361718297,
+      "learning_rate": 0.0002,
+      "loss": 0.0229,
+      "step": 4650
+    },
+    {
+      "epoch": 0.008475057913653527,
+      "grad_norm": 0.03594676032662392,
+      "learning_rate": 0.0002,
+      "loss": 0.1401,
+      "step": 4660
+    },
+    {
+      "epoch": 0.008493244733210724,
+      "grad_norm": 0.0316978394985199,
+      "learning_rate": 0.0002,
+      "loss": 0.0877,
+      "step": 4670
+    },
+    {
+      "epoch": 0.00851143155276792,
+      "grad_norm": 0.023302162066102028,
+      "learning_rate": 0.0002,
+      "loss": 0.0764,
+      "step": 4680
+    },
+    {
+      "epoch": 0.008529618372325116,
+      "grad_norm": 0.1329929083585739,
+      "learning_rate": 0.0002,
+      "loss": 0.0788,
+      "step": 4690
+    },
+    {
+      "epoch": 0.008547805191882312,
+      "grad_norm": 0.01048013661056757,
+      "learning_rate": 0.0002,
+      "loss": 0.0234,
+      "step": 4700
+    },
+    {
+      "epoch": 0.008565992011439509,
+      "grad_norm": 0.03505022078752518,
+      "learning_rate": 0.0002,
+      "loss": 0.1509,
+      "step": 4710
+    },
+    {
+      "epoch": 0.008584178830996705,
+      "grad_norm": 0.03877585008740425,
+      "learning_rate": 0.0002,
+      "loss": 0.0802,
+      "step": 4720
+    },
+    {
+      "epoch": 0.008602365650553903,
+      "grad_norm": 0.041193027049303055,
+      "learning_rate": 0.0002,
+      "loss": 0.0695,
+      "step": 4730
+    },
+    {
+      "epoch": 0.0086205524701111,
+      "grad_norm": 0.17310455441474915,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 4740
+    },
+    {
+      "epoch": 0.008638739289668296,
+      "grad_norm": 0.0061012376099824905,
+      "learning_rate": 0.0002,
+      "loss": 0.0158,
+      "step": 4750
+    },
+    {
+      "epoch": 0.008656926109225492,
+      "grad_norm": 0.04843207076191902,
+      "learning_rate": 0.0002,
+      "loss": 0.2103,
+      "step": 4760
+    },
+    {
+      "epoch": 0.008675112928782688,
+      "grad_norm": 0.04483436048030853,
+      "learning_rate": 0.0002,
+      "loss": 0.0878,
+      "step": 4770
+    },
+    {
+      "epoch": 0.008693299748339885,
+      "grad_norm": 0.056655965745449066,
+      "learning_rate": 0.0002,
+      "loss": 0.0752,
+      "step": 4780
+    },
+    {
+      "epoch": 0.008711486567897081,
+      "grad_norm": 0.11626063287258148,
+      "learning_rate": 0.0002,
+      "loss": 0.0685,
+      "step": 4790
+    },
+    {
+      "epoch": 0.008729673387454277,
+      "grad_norm": 0.013872025534510612,
+      "learning_rate": 0.0002,
+      "loss": 0.0198,
+      "step": 4800
+    },
+    {
+      "epoch": 0.008747860207011474,
+      "grad_norm": 0.06217370182275772,
+      "learning_rate": 0.0002,
+      "loss": 0.1371,
+      "step": 4810
+    },
+    {
+      "epoch": 0.00876604702656867,
+      "grad_norm": 0.027149083092808723,
+      "learning_rate": 0.0002,
+      "loss": 0.0849,
+      "step": 4820
+    },
+    {
+      "epoch": 0.008784233846125866,
+      "grad_norm": 0.043290987610816956,
+      "learning_rate": 0.0002,
+      "loss": 0.0739,
+      "step": 4830
+    },
+    {
+      "epoch": 0.008802420665683063,
+      "grad_norm": 0.10664638131856918,
+      "learning_rate": 0.0002,
+      "loss": 0.0722,
+      "step": 4840
+    },
+    {
+      "epoch": 0.008820607485240259,
+      "grad_norm": 0.033459801226854324,
+      "learning_rate": 0.0002,
+      "loss": 0.0234,
+      "step": 4850
+    },
+    {
+      "epoch": 0.008838794304797455,
+      "grad_norm": 0.049193184822797775,
+      "learning_rate": 0.0002,
+      "loss": 0.1173,
+      "step": 4860
+    },
+    {
+      "epoch": 0.008856981124354651,
+      "grad_norm": 0.05060647428035736,
+      "learning_rate": 0.0002,
+      "loss": 0.0883,
+      "step": 4870
+    },
+    {
+      "epoch": 0.008875167943911848,
+      "grad_norm": 0.028496885672211647,
+      "learning_rate": 0.0002,
+      "loss": 0.0747,
+      "step": 4880
+    },
+    {
+      "epoch": 0.008893354763469044,
+      "grad_norm": 0.10652820765972137,
+      "learning_rate": 0.0002,
+      "loss": 0.0707,
+      "step": 4890
+    },
+    {
+      "epoch": 0.008911541583026242,
+      "grad_norm": 0.007879966869950294,
+      "learning_rate": 0.0002,
+      "loss": 0.0178,
+      "step": 4900
+    },
+    {
+      "epoch": 0.008929728402583438,
+      "grad_norm": 0.05227983742952347,
+      "learning_rate": 0.0002,
+      "loss": 0.1379,
+      "step": 4910
+    },
+    {
+      "epoch": 0.008947915222140635,
+      "grad_norm": 0.06054231896996498,
+      "learning_rate": 0.0002,
+      "loss": 0.0934,
+      "step": 4920
+    },
+    {
+      "epoch": 0.008966102041697831,
+      "grad_norm": 0.029085835441946983,
+      "learning_rate": 0.0002,
+      "loss": 0.0816,
+      "step": 4930
+    },
+    {
+      "epoch": 0.008984288861255027,
+      "grad_norm": 0.09829402714967728,
+      "learning_rate": 0.0002,
+      "loss": 0.0672,
+      "step": 4940
+    },
+    {
+      "epoch": 0.009002475680812224,
+      "grad_norm": 0.005579107441008091,
+      "learning_rate": 0.0002,
+      "loss": 0.0172,
+      "step": 4950
+    },
+    {
+      "epoch": 0.00902066250036942,
+      "grad_norm": 0.027280857786536217,
+      "learning_rate": 0.0002,
+      "loss": 0.1659,
+      "step": 4960
+    },
+    {
+      "epoch": 0.009038849319926616,
+      "grad_norm": 0.10321583598852158,
+      "learning_rate": 0.0002,
+      "loss": 0.0947,
+      "step": 4970
+    },
+    {
+      "epoch": 0.009057036139483813,
+      "grad_norm": 0.03381946310400963,
+      "learning_rate": 0.0002,
+      "loss": 0.0837,
+      "step": 4980
+    },
+    {
+      "epoch": 0.009075222959041009,
+      "grad_norm": 0.14493779838085175,
+      "learning_rate": 0.0002,
+      "loss": 0.0736,
+      "step": 4990
+    },
+    {
+      "epoch": 0.009093409778598205,
+      "grad_norm": 0.009917684830725193,
+      "learning_rate": 0.0002,
+      "loss": 0.0188,
+      "step": 5000
+    },
+    {
+      "epoch": 0.009111596598155402,
+      "grad_norm": 1.003450632095337,
+      "learning_rate": 0.0002,
+      "loss": 0.218,
+      "step": 5010
+    },
+    {
+      "epoch": 0.009129783417712598,
+      "grad_norm": 0.09081514924764633,
+      "learning_rate": 0.0002,
+      "loss": 0.1714,
+      "step": 5020
+    },
+    {
+      "epoch": 0.009147970237269794,
+      "grad_norm": 0.042343392968177795,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 5030
+    },
+    {
+      "epoch": 0.00916615705682699,
+      "grad_norm": 0.09944835305213928,
+      "learning_rate": 0.0002,
+      "loss": 0.0667,
+      "step": 5040
+    },
+    {
+      "epoch": 0.009184343876384187,
+      "grad_norm": 0.008264658972620964,
+      "learning_rate": 0.0002,
+      "loss": 0.0122,
+      "step": 5050
+    },
+    {
+      "epoch": 0.009202530695941383,
+      "grad_norm": 0.08990125358104706,
+      "learning_rate": 0.0002,
+      "loss": 0.1685,
+      "step": 5060
+    },
+    {
+      "epoch": 0.009220717515498581,
+      "grad_norm": 0.0331488698720932,
+      "learning_rate": 0.0002,
+      "loss": 0.0885,
+      "step": 5070
+    },
+    {
+      "epoch": 0.009238904335055777,
+      "grad_norm": 0.029458707198500633,
+      "learning_rate": 0.0002,
+      "loss": 0.078,
+      "step": 5080
+    },
+    {
+      "epoch": 0.009257091154612974,
+      "grad_norm": 0.10468839108943939,
+      "learning_rate": 0.0002,
+      "loss": 0.0683,
+      "step": 5090
+    },
+    {
+      "epoch": 0.00927527797417017,
+      "grad_norm": 0.002719841431826353,
+      "learning_rate": 0.0002,
+      "loss": 0.0117,
+      "step": 5100
+    },
+    {
+      "epoch": 0.009293464793727366,
+      "grad_norm": 0.0411439947783947,
+      "learning_rate": 0.0002,
+      "loss": 0.2025,
+      "step": 5110
+    },
+    {
+      "epoch": 0.009311651613284563,
+      "grad_norm": 0.03695548698306084,
+      "learning_rate": 0.0002,
+      "loss": 0.0831,
+      "step": 5120
+    },
+    {
+      "epoch": 0.009329838432841759,
+      "grad_norm": 0.06067590415477753,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 5130
+    },
+    {
+      "epoch": 0.009348025252398955,
+      "grad_norm": 0.11754634976387024,
+      "learning_rate": 0.0002,
+      "loss": 0.0667,
+      "step": 5140
+    },
+    {
+      "epoch": 0.009366212071956152,
+      "grad_norm": 0.004248317331075668,
+      "learning_rate": 0.0002,
+      "loss": 0.0113,
+      "step": 5150
+    },
+    {
+      "epoch": 0.009384398891513348,
+      "grad_norm": 0.03073648177087307,
+      "learning_rate": 0.0002,
+      "loss": 0.2289,
+      "step": 5160
+    },
+    {
+      "epoch": 0.009402585711070544,
+      "grad_norm": 0.10287592560052872,
+      "learning_rate": 0.0002,
+      "loss": 0.0977,
+      "step": 5170
+    },
+    {
+      "epoch": 0.00942077253062774,
+      "grad_norm": 0.06832946836948395,
+      "learning_rate": 0.0002,
+      "loss": 0.0764,
+      "step": 5180
+    },
+    {
+      "epoch": 0.009438959350184937,
+      "grad_norm": 0.1760883778333664,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 5190
+    },
+    {
+      "epoch": 0.009457146169742133,
+      "grad_norm": 0.02968805655837059,
+      "learning_rate": 0.0002,
+      "loss": 0.0253,
+      "step": 5200
+    },
+    {
+      "epoch": 0.00947533298929933,
+      "grad_norm": 0.046602651476860046,
+      "learning_rate": 0.0002,
+      "loss": 0.1432,
+      "step": 5210
+    },
+    {
+      "epoch": 0.009493519808856526,
+      "grad_norm": 0.051989324390888214,
+      "learning_rate": 0.0002,
+      "loss": 0.0807,
+      "step": 5220
+    },
+    {
+      "epoch": 0.009511706628413722,
+      "grad_norm": 0.04583961144089699,
+      "learning_rate": 0.0002,
+      "loss": 0.0782,
+      "step": 5230
+    },
+    {
+      "epoch": 0.00952989344797092,
+      "grad_norm": 0.13195525109767914,
+      "learning_rate": 0.0002,
+      "loss": 0.0688,
+      "step": 5240
+    },
+    {
+      "epoch": 0.009548080267528116,
+      "grad_norm": 0.011369351297616959,
+      "learning_rate": 0.0002,
+      "loss": 0.0232,
+      "step": 5250
+    },
+    {
+      "epoch": 0.009566267087085313,
+      "grad_norm": 0.05092083290219307,
+      "learning_rate": 0.0002,
+      "loss": 0.145,
+      "step": 5260
+    },
+    {
+      "epoch": 0.009584453906642509,
+      "grad_norm": 0.05051489174365997,
+      "learning_rate": 0.0002,
+      "loss": 0.0803,
+      "step": 5270
+    },
+    {
+      "epoch": 0.009602640726199705,
+      "grad_norm": 0.05730990320444107,
+      "learning_rate": 0.0002,
+      "loss": 0.0717,
+      "step": 5280
+    },
+    {
+      "epoch": 0.009620827545756902,
+      "grad_norm": 0.11170202493667603,
+      "learning_rate": 0.0002,
+      "loss": 0.0711,
+      "step": 5290
+    },
+    {
+      "epoch": 0.009639014365314098,
+      "grad_norm": 0.011571788229048252,
+      "learning_rate": 0.0002,
+      "loss": 0.0204,
+      "step": 5300
+    },
+    {
+      "epoch": 0.009657201184871294,
+      "grad_norm": 0.04396244138479233,
+      "learning_rate": 0.0002,
+      "loss": 0.1764,
+      "step": 5310
+    },
+    {
+      "epoch": 0.00967538800442849,
+      "grad_norm": 0.047808658331632614,
+      "learning_rate": 0.0002,
+      "loss": 0.0855,
+      "step": 5320
+    },
+    {
+      "epoch": 0.009693574823985687,
+      "grad_norm": 0.09201673418283463,
+      "learning_rate": 0.0002,
+      "loss": 0.0737,
+      "step": 5330
+    },
+    {
+      "epoch": 0.009711761643542883,
+      "grad_norm": 0.12273146212100983,
+      "learning_rate": 0.0002,
+      "loss": 0.0658,
+      "step": 5340
+    },
+    {
+      "epoch": 0.00972994846310008,
+      "grad_norm": 0.014599839225411415,
+      "learning_rate": 0.0002,
+      "loss": 0.0254,
+      "step": 5350
+    },
+    {
+      "epoch": 0.009748135282657276,
+      "grad_norm": 0.049732692539691925,
+      "learning_rate": 0.0002,
+      "loss": 0.1432,
+      "step": 5360
+    },
+    {
+      "epoch": 0.009766322102214472,
+      "grad_norm": 0.07791377604007721,
+      "learning_rate": 0.0002,
+      "loss": 0.0865,
+      "step": 5370
+    },
+    {
+      "epoch": 0.009784508921771668,
+      "grad_norm": 0.06298892199993134,
+      "learning_rate": 0.0002,
+      "loss": 0.0816,
+      "step": 5380
+    },
+    {
+      "epoch": 0.009802695741328865,
+      "grad_norm": 0.08924435079097748,
+      "learning_rate": 0.0002,
+      "loss": 0.0709,
+      "step": 5390
+    },
+    {
+      "epoch": 0.009820882560886061,
+      "grad_norm": 0.02383723482489586,
+      "learning_rate": 0.0002,
+      "loss": 0.0208,
+      "step": 5400
+    },
+    {
+      "epoch": 0.009839069380443257,
+      "grad_norm": 0.042910825461149216,
+      "learning_rate": 0.0002,
+      "loss": 0.1383,
+      "step": 5410
+    },
+    {
+      "epoch": 0.009857256200000455,
+      "grad_norm": 0.05560186505317688,
+      "learning_rate": 0.0002,
+      "loss": 0.0827,
+      "step": 5420
+    },
+    {
+      "epoch": 0.009875443019557652,
+      "grad_norm": 0.08179624378681183,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 5430
+    },
+    {
+      "epoch": 0.009893629839114848,
+      "grad_norm": 0.17111806571483612,
+      "learning_rate": 0.0002,
+      "loss": 0.0688,
+      "step": 5440
+    },
+    {
+      "epoch": 0.009911816658672044,
+      "grad_norm": 0.008684845641255379,
+      "learning_rate": 0.0002,
+      "loss": 0.0177,
+      "step": 5450
+    },
+    {
+      "epoch": 0.00993000347822924,
+      "grad_norm": 0.044370412826538086,
+      "learning_rate": 0.0002,
+      "loss": 0.2036,
+      "step": 5460
+    },
+    {
+      "epoch": 0.009948190297786437,
+      "grad_norm": 0.08403154462575912,
+      "learning_rate": 0.0002,
+      "loss": 0.0878,
+      "step": 5470
+    },
+    {
+      "epoch": 0.009966377117343633,
+      "grad_norm": 0.10712645202875137,
+      "learning_rate": 0.0002,
+      "loss": 0.0787,
+      "step": 5480
+    },
+    {
+      "epoch": 0.00998456393690083,
+      "grad_norm": 0.12575705349445343,
+      "learning_rate": 0.0002,
+      "loss": 0.0637,
+      "step": 5490
+    },
+    {
+      "epoch": 0.010002750756458026,
+      "grad_norm": 0.018583891913294792,
+      "learning_rate": 0.0002,
+      "loss": 0.0179,
+      "step": 5500
+    },
+    {
+      "epoch": 0.010020937576015222,
+      "grad_norm": 0.040852561593055725,
+      "learning_rate": 0.0002,
+      "loss": 0.1545,
+      "step": 5510
+    },
+    {
+      "epoch": 0.010039124395572419,
+      "grad_norm": 0.09006325900554657,
+      "learning_rate": 0.0002,
+      "loss": 0.0888,
+      "step": 5520
+    },
+    {
+      "epoch": 0.010057311215129615,
+      "grad_norm": 0.06323093175888062,
+      "learning_rate": 0.0002,
+      "loss": 0.0778,
+      "step": 5530
+    },
+    {
+      "epoch": 0.010075498034686811,
+      "grad_norm": 0.10159824043512344,
+      "learning_rate": 0.0002,
+      "loss": 0.0662,
+      "step": 5540
+    },
+    {
+      "epoch": 0.010093684854244007,
+      "grad_norm": 0.012086872011423111,
+      "learning_rate": 0.0002,
+      "loss": 0.0237,
+      "step": 5550
+    },
+    {
+      "epoch": 0.010111871673801204,
+      "grad_norm": 0.02518664114177227,
+      "learning_rate": 0.0002,
+      "loss": 0.1246,
+      "step": 5560
+    },
+    {
+      "epoch": 0.0101300584933584,
+      "grad_norm": 0.056161828339099884,
+      "learning_rate": 0.0002,
+      "loss": 0.086,
+      "step": 5570
+    },
+    {
+      "epoch": 0.010148245312915596,
+      "grad_norm": 0.03376586362719536,
+      "learning_rate": 0.0002,
+      "loss": 0.0842,
+      "step": 5580
+    },
+    {
+      "epoch": 0.010166432132472794,
+      "grad_norm": 0.09921032190322876,
+      "learning_rate": 0.0002,
+      "loss": 0.0667,
+      "step": 5590
+    },
+    {
+      "epoch": 0.01018461895202999,
+      "grad_norm": 0.009120604954659939,
+      "learning_rate": 0.0002,
+      "loss": 0.0209,
+      "step": 5600
+    },
+    {
+      "epoch": 0.010202805771587187,
+      "grad_norm": 0.037767425179481506,
+      "learning_rate": 0.0002,
+      "loss": 0.1248,
+      "step": 5610
+    },
+    {
+      "epoch": 0.010220992591144383,
+      "grad_norm": 0.05255524069070816,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 5620
+    },
+    {
+      "epoch": 0.01023917941070158,
+      "grad_norm": 0.038734354078769684,
+      "learning_rate": 0.0002,
+      "loss": 0.0791,
+      "step": 5630
+    },
+    {
+      "epoch": 0.010257366230258776,
+      "grad_norm": 0.09293238073587418,
+      "learning_rate": 0.0002,
+      "loss": 0.064,
+      "step": 5640
+    },
+    {
+      "epoch": 0.010275553049815972,
+      "grad_norm": 0.013020232319831848,
+      "learning_rate": 0.0002,
+      "loss": 0.0174,
+      "step": 5650
+    },
+    {
+      "epoch": 0.010293739869373169,
+      "grad_norm": 0.030535893514752388,
+      "learning_rate": 0.0002,
+      "loss": 0.1615,
+      "step": 5660
+    },
+    {
+      "epoch": 0.010311926688930365,
+      "grad_norm": 0.08644227683544159,
+      "learning_rate": 0.0002,
+      "loss": 0.0856,
+      "step": 5670
+    },
+    {
+      "epoch": 0.010330113508487561,
+      "grad_norm": 0.04769067466259003,
+      "learning_rate": 0.0002,
+      "loss": 0.0777,
+      "step": 5680
+    },
+    {
+      "epoch": 0.010348300328044758,
+      "grad_norm": 0.1528550088405609,
+      "learning_rate": 0.0002,
+      "loss": 0.0757,
+      "step": 5690
+    },
+    {
+      "epoch": 0.010366487147601954,
+      "grad_norm": 0.012257793918251991,
+      "learning_rate": 0.0002,
+      "loss": 0.0163,
+      "step": 5700
+    },
+    {
+      "epoch": 0.01038467396715915,
+      "grad_norm": 0.5761304497718811,
+      "learning_rate": 0.0002,
+      "loss": 0.1787,
+      "step": 5710
+    },
+    {
+      "epoch": 0.010402860786716346,
+      "grad_norm": 0.07034485787153244,
+      "learning_rate": 0.0002,
+      "loss": 0.0964,
+      "step": 5720
+    },
+    {
+      "epoch": 0.010421047606273543,
+      "grad_norm": 0.04541708156466484,
+      "learning_rate": 0.0002,
+      "loss": 0.0793,
+      "step": 5730
+    },
+    {
+      "epoch": 0.010439234425830739,
+      "grad_norm": 0.12013612687587738,
+      "learning_rate": 0.0002,
+      "loss": 0.07,
+      "step": 5740
+    },
+    {
+      "epoch": 0.010457421245387935,
+      "grad_norm": 0.014152747578918934,
+      "learning_rate": 0.0002,
+      "loss": 0.0208,
+      "step": 5750
+    },
+    {
+      "epoch": 0.010475608064945133,
+      "grad_norm": 0.029470542445778847,
+      "learning_rate": 0.0002,
+      "loss": 0.1352,
+      "step": 5760
+    },
+    {
+      "epoch": 0.01049379488450233,
+      "grad_norm": 0.04889104515314102,
+      "learning_rate": 0.0002,
+      "loss": 0.0748,
+      "step": 5770
+    },
+    {
+      "epoch": 0.010511981704059526,
+      "grad_norm": 0.0311355609446764,
+      "learning_rate": 0.0002,
+      "loss": 0.0764,
+      "step": 5780
+    },
+    {
+      "epoch": 0.010530168523616722,
+      "grad_norm": 0.16830098628997803,
+      "learning_rate": 0.0002,
+      "loss": 0.0734,
+      "step": 5790
+    },
+    {
+      "epoch": 0.010548355343173919,
+      "grad_norm": 0.013224232010543346,
+      "learning_rate": 0.0002,
+      "loss": 0.0218,
+      "step": 5800
+    },
+    {
+      "epoch": 0.010566542162731115,
+      "grad_norm": 0.03710555657744408,
+      "learning_rate": 0.0002,
+      "loss": 0.1403,
+      "step": 5810
+    },
+    {
+      "epoch": 0.010584728982288311,
+      "grad_norm": 0.05788695067167282,
+      "learning_rate": 0.0002,
+      "loss": 0.0863,
+      "step": 5820
+    },
+    {
+      "epoch": 0.010602915801845508,
+      "grad_norm": 0.03398163616657257,
+      "learning_rate": 0.0002,
+      "loss": 0.0751,
+      "step": 5830
+    },
+    {
+      "epoch": 0.010621102621402704,
+      "grad_norm": 0.13862720131874084,
+      "learning_rate": 0.0002,
+      "loss": 0.07,
+      "step": 5840
+    },
+    {
+      "epoch": 0.0106392894409599,
+      "grad_norm": 0.016240287572145462,
+      "learning_rate": 0.0002,
+      "loss": 0.0209,
+      "step": 5850
+    },
+    {
+      "epoch": 0.010657476260517097,
+      "grad_norm": 0.030351752415299416,
+      "learning_rate": 0.0002,
+      "loss": 0.157,
+      "step": 5860
+    },
+    {
+      "epoch": 0.010675663080074293,
+      "grad_norm": 0.038465555757284164,
+      "learning_rate": 0.0002,
+      "loss": 0.072,
+      "step": 5870
+    },
+    {
+      "epoch": 0.01069384989963149,
+      "grad_norm": 0.07298482209444046,
+      "learning_rate": 0.0002,
+      "loss": 0.0796,
+      "step": 5880
+    },
+    {
+      "epoch": 0.010712036719188685,
+      "grad_norm": 0.13822157680988312,
+      "learning_rate": 0.0002,
+      "loss": 0.0687,
+      "step": 5890
+    },
+    {
+      "epoch": 0.010730223538745882,
+      "grad_norm": 0.014381729066371918,
+      "learning_rate": 0.0002,
+      "loss": 0.0192,
+      "step": 5900
+    },
+    {
+      "epoch": 0.010748410358303078,
+      "grad_norm": 0.040448348969221115,
+      "learning_rate": 0.0002,
+      "loss": 0.1714,
+      "step": 5910
+    },
+    {
+      "epoch": 0.010766597177860274,
+      "grad_norm": 0.06950225681066513,
+      "learning_rate": 0.0002,
+      "loss": 0.098,
+      "step": 5920
+    },
+    {
+      "epoch": 0.010784783997417472,
+      "grad_norm": 0.04581855982542038,
+      "learning_rate": 0.0002,
+      "loss": 0.0752,
+      "step": 5930
+    },
+    {
+      "epoch": 0.010802970816974669,
+      "grad_norm": 0.10498905926942825,
+      "learning_rate": 0.0002,
+      "loss": 0.0627,
+      "step": 5940
+    },
+    {
+      "epoch": 0.010821157636531865,
+      "grad_norm": 0.009345698170363903,
+      "learning_rate": 0.0002,
+      "loss": 0.0183,
+      "step": 5950
+    },
+    {
+      "epoch": 0.010839344456089061,
+      "grad_norm": 0.02440352365374565,
+      "learning_rate": 0.0002,
+      "loss": 0.1289,
+      "step": 5960
+    },
+    {
+      "epoch": 0.010857531275646258,
+      "grad_norm": 0.051523737609386444,
+      "learning_rate": 0.0002,
+      "loss": 0.0813,
+      "step": 5970
+    },
+    {
+      "epoch": 0.010875718095203454,
+      "grad_norm": 0.031664300709962845,
+      "learning_rate": 0.0002,
+      "loss": 0.0736,
+      "step": 5980
+    },
+    {
+      "epoch": 0.01089390491476065,
+      "grad_norm": 0.10166060924530029,
+      "learning_rate": 0.0002,
+      "loss": 0.0631,
+      "step": 5990
+    },
+    {
+      "epoch": 0.010912091734317847,
+      "grad_norm": 0.01642071269452572,
+      "learning_rate": 0.0002,
+      "loss": 0.0198,
+      "step": 6000
+    },
+    {
+      "epoch": 0.010930278553875043,
+      "grad_norm": 0.04028782621026039,
+      "learning_rate": 0.0002,
+      "loss": 0.1355,
+      "step": 6010
+    },
+    {
+      "epoch": 0.01094846537343224,
+      "grad_norm": 0.04289260134100914,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 6020
+    },
+    {
+      "epoch": 0.010966652192989436,
+      "grad_norm": 0.03854202851653099,
+      "learning_rate": 0.0002,
+      "loss": 0.0765,
+      "step": 6030
+    },
+    {
+      "epoch": 0.010984839012546632,
+      "grad_norm": 0.07910823822021484,
+      "learning_rate": 0.0002,
+      "loss": 0.0618,
+      "step": 6040
+    },
+    {
+      "epoch": 0.011003025832103828,
+      "grad_norm": 0.009719946421682835,
+      "learning_rate": 0.0002,
+      "loss": 0.0145,
+      "step": 6050
+    },
+    {
+      "epoch": 0.011021212651661024,
+      "grad_norm": 0.06853003799915314,
+      "learning_rate": 0.0002,
+      "loss": 0.1563,
+      "step": 6060
+    },
+    {
+      "epoch": 0.01103939947121822,
+      "grad_norm": 0.02887076325714588,
+      "learning_rate": 0.0002,
+      "loss": 0.0803,
+      "step": 6070
+    },
+    {
+      "epoch": 0.011057586290775417,
+      "grad_norm": 0.060147739946842194,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 6080
+    },
+    {
+      "epoch": 0.011075773110332613,
+      "grad_norm": 0.10197418928146362,
+      "learning_rate": 0.0002,
+      "loss": 0.0627,
+      "step": 6090
+    },
+    {
+      "epoch": 0.011093959929889811,
+      "grad_norm": 0.015125100500881672,
+      "learning_rate": 0.0002,
+      "loss": 0.0164,
+      "step": 6100
+    },
+    {
+      "epoch": 0.011112146749447008,
+      "grad_norm": 0.029526161029934883,
+      "learning_rate": 0.0002,
+      "loss": 0.1526,
+      "step": 6110
+    },
+    {
+      "epoch": 0.011130333569004204,
+      "grad_norm": 0.05942453444004059,
+      "learning_rate": 0.0002,
+      "loss": 0.0891,
+      "step": 6120
+    },
+    {
+      "epoch": 0.0111485203885614,
+      "grad_norm": 0.07344426214694977,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 6130
+    },
+    {
+      "epoch": 0.011166707208118597,
+      "grad_norm": 0.1394059658050537,
+      "learning_rate": 0.0002,
+      "loss": 0.0776,
+      "step": 6140
+    },
+    {
+      "epoch": 0.011184894027675793,
+      "grad_norm": 0.00965851079672575,
+      "learning_rate": 0.0002,
+      "loss": 0.019,
+      "step": 6150
+    },
+    {
+      "epoch": 0.01120308084723299,
+      "grad_norm": 0.041846372187137604,
+      "learning_rate": 0.0002,
+      "loss": 0.1776,
+      "step": 6160
+    },
+    {
+      "epoch": 0.011221267666790186,
+      "grad_norm": 0.04657486826181412,
+      "learning_rate": 0.0002,
+      "loss": 0.0878,
+      "step": 6170
+    },
+    {
+      "epoch": 0.011239454486347382,
+      "grad_norm": 0.026520246639847755,
+      "learning_rate": 0.0002,
+      "loss": 0.0768,
+      "step": 6180
+    },
+    {
+      "epoch": 0.011257641305904578,
+      "grad_norm": 0.10318096727132797,
+      "learning_rate": 0.0002,
+      "loss": 0.0617,
+      "step": 6190
+    },
+    {
+      "epoch": 0.011275828125461775,
+      "grad_norm": 0.019912905991077423,
+      "learning_rate": 0.0002,
+      "loss": 0.0202,
+      "step": 6200
+    },
+    {
+      "epoch": 0.01129401494501897,
+      "grad_norm": 0.05316480994224548,
+      "learning_rate": 0.0002,
+      "loss": 0.1412,
+      "step": 6210
+    },
+    {
+      "epoch": 0.011312201764576167,
+      "grad_norm": 0.02944323979318142,
+      "learning_rate": 0.0002,
+      "loss": 0.0829,
+      "step": 6220
+    },
+    {
+      "epoch": 0.011330388584133363,
+      "grad_norm": 0.0285831056535244,
+      "learning_rate": 0.0002,
+      "loss": 0.074,
+      "step": 6230
+    },
+    {
+      "epoch": 0.01134857540369056,
+      "grad_norm": 0.0975700169801712,
+      "learning_rate": 0.0002,
+      "loss": 0.0681,
+      "step": 6240
+    },
+    {
+      "epoch": 0.011366762223247756,
+      "grad_norm": 0.025717545300722122,
+      "learning_rate": 0.0002,
+      "loss": 0.0221,
+      "step": 6250
+    },
+    {
+      "epoch": 0.011384949042804952,
+      "grad_norm": 0.02859714813530445,
+      "learning_rate": 0.0002,
+      "loss": 0.1142,
+      "step": 6260
+    },
+    {
+      "epoch": 0.01140313586236215,
+      "grad_norm": 0.04395005479454994,
+      "learning_rate": 0.0002,
+      "loss": 0.0777,
+      "step": 6270
+    },
+    {
+      "epoch": 0.011421322681919347,
+      "grad_norm": 0.05116860568523407,
+      "learning_rate": 0.0002,
+      "loss": 0.0763,
+      "step": 6280
+    },
+    {
+      "epoch": 0.011439509501476543,
+      "grad_norm": 0.06850302964448929,
+      "learning_rate": 0.0002,
+      "loss": 0.0632,
+      "step": 6290
+    },
+    {
+      "epoch": 0.01145769632103374,
+      "grad_norm": 0.016113542020320892,
+      "learning_rate": 0.0002,
+      "loss": 0.0178,
+      "step": 6300
+    },
+    {
+      "epoch": 0.011475883140590936,
+      "grad_norm": 0.032306116074323654,
+      "learning_rate": 0.0002,
+      "loss": 0.1306,
+      "step": 6310
+    },
+    {
+      "epoch": 0.011494069960148132,
+      "grad_norm": 0.055701326578855515,
+      "learning_rate": 0.0002,
+      "loss": 0.0834,
+      "step": 6320
+    },
+    {
+      "epoch": 0.011512256779705328,
+      "grad_norm": 0.022934190928936005,
+      "learning_rate": 0.0002,
+      "loss": 0.0737,
+      "step": 6330
+    },
+    {
+      "epoch": 0.011530443599262525,
+      "grad_norm": 0.08375566452741623,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 6340
+    },
+    {
+      "epoch": 0.011548630418819721,
+      "grad_norm": 0.013614729046821594,
+      "learning_rate": 0.0002,
+      "loss": 0.0187,
+      "step": 6350
+    },
+    {
+      "epoch": 0.011566817238376917,
+      "grad_norm": 0.028269700706005096,
+      "learning_rate": 0.0002,
+      "loss": 0.1245,
+      "step": 6360
+    },
+    {
+      "epoch": 0.011585004057934114,
+      "grad_norm": 0.03646335378289223,
+      "learning_rate": 0.0002,
+      "loss": 0.0866,
+      "step": 6370
+    },
+    {
+      "epoch": 0.01160319087749131,
+      "grad_norm": 0.0371277742087841,
+      "learning_rate": 0.0002,
+      "loss": 0.0737,
+      "step": 6380
+    },
+    {
+      "epoch": 0.011621377697048506,
+      "grad_norm": 0.13698458671569824,
+      "learning_rate": 0.0002,
+      "loss": 0.0679,
+      "step": 6390
+    },
+    {
+      "epoch": 0.011639564516605702,
+      "grad_norm": 0.009350700303912163,
+      "learning_rate": 0.0002,
+      "loss": 0.024,
+      "step": 6400
+    },
+    {
+      "epoch": 0.011657751336162899,
+      "grad_norm": 0.03187236189842224,
+      "learning_rate": 0.0002,
+      "loss": 0.1555,
+      "step": 6410
+    },
+    {
+      "epoch": 0.011675938155720095,
+      "grad_norm": 0.06672242283821106,
+      "learning_rate": 0.0002,
+      "loss": 0.0835,
+      "step": 6420
+    },
+    {
+      "epoch": 0.011694124975277291,
+      "grad_norm": 0.07821471244096756,
+      "learning_rate": 0.0002,
+      "loss": 0.0746,
+      "step": 6430
+    },
+    {
+      "epoch": 0.011712311794834488,
+      "grad_norm": 0.14781107008457184,
+      "learning_rate": 0.0002,
+      "loss": 0.0662,
+      "step": 6440
+    },
+    {
+      "epoch": 0.011730498614391686,
+      "grad_norm": 0.0057207453064620495,
+      "learning_rate": 0.0002,
+      "loss": 0.0169,
+      "step": 6450
+    },
+    {
+      "epoch": 0.011748685433948882,
+      "grad_norm": 0.04252105578780174,
+      "learning_rate": 0.0002,
+      "loss": 0.1868,
+      "step": 6460
+    },
+    {
+      "epoch": 0.011766872253506078,
+      "grad_norm": 0.05041474476456642,
+      "learning_rate": 0.0002,
+      "loss": 0.0842,
+      "step": 6470
+    },
+    {
+      "epoch": 0.011785059073063275,
+      "grad_norm": 0.06584125757217407,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 6480
+    },
+    {
+      "epoch": 0.011803245892620471,
+      "grad_norm": 0.14610575139522552,
+      "learning_rate": 0.0002,
+      "loss": 0.063,
+      "step": 6490
+    },
+    {
+      "epoch": 0.011821432712177667,
+      "grad_norm": 0.01419675163924694,
+      "learning_rate": 0.0002,
+      "loss": 0.0152,
+      "step": 6500
+    },
+    {
+      "epoch": 0.011839619531734864,
+      "grad_norm": 0.03371060639619827,
+      "learning_rate": 0.0002,
+      "loss": 0.1725,
+      "step": 6510
+    },
+    {
+      "epoch": 0.01185780635129206,
+      "grad_norm": 0.028900766745209694,
+      "learning_rate": 0.0002,
+      "loss": 0.0815,
+      "step": 6520
+    },
+    {
+      "epoch": 0.011875993170849256,
+      "grad_norm": 0.059519629925489426,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 6530
+    },
+    {
+      "epoch": 0.011894179990406453,
+      "grad_norm": 0.12085167318582535,
+      "learning_rate": 0.0002,
+      "loss": 0.0615,
+      "step": 6540
+    },
+    {
+      "epoch": 0.011912366809963649,
+      "grad_norm": 0.028604619204998016,
+      "learning_rate": 0.0002,
+      "loss": 0.0196,
+      "step": 6550
+    },
+    {
+      "epoch": 0.011930553629520845,
+      "grad_norm": 0.03659407049417496,
+      "learning_rate": 0.0002,
+      "loss": 0.1403,
+      "step": 6560
+    },
+    {
+      "epoch": 0.011948740449078041,
+      "grad_norm": 0.034444138407707214,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 6570
+    },
+    {
+      "epoch": 0.011966927268635238,
+      "grad_norm": 0.029788263142108917,
+      "learning_rate": 0.0002,
+      "loss": 0.0713,
+      "step": 6580
+    },
+    {
+      "epoch": 0.011985114088192434,
+      "grad_norm": 0.1271272599697113,
+      "learning_rate": 0.0002,
+      "loss": 0.0672,
+      "step": 6590
+    },
+    {
+      "epoch": 0.01200330090774963,
+      "grad_norm": 0.018705012276768684,
+      "learning_rate": 0.0002,
+      "loss": 0.0212,
+      "step": 6600
+    },
+    {
+      "epoch": 0.012021487727306827,
+      "grad_norm": 0.02982541173696518,
+      "learning_rate": 0.0002,
+      "loss": 0.1152,
+      "step": 6610
+    },
+    {
+      "epoch": 0.012039674546864025,
+      "grad_norm": 0.06942040473222733,
+      "learning_rate": 0.0002,
+      "loss": 0.0963,
+      "step": 6620
+    },
+    {
+      "epoch": 0.012057861366421221,
+      "grad_norm": 0.06102292984724045,
+      "learning_rate": 0.0002,
+      "loss": 0.0775,
+      "step": 6630
+    },
+    {
+      "epoch": 0.012076048185978417,
+      "grad_norm": 0.10115987807512283,
+      "learning_rate": 0.0002,
+      "loss": 0.0729,
+      "step": 6640
+    },
+    {
+      "epoch": 0.012094235005535614,
+      "grad_norm": 0.011439867317676544,
+      "learning_rate": 0.0002,
+      "loss": 0.0253,
+      "step": 6650
+    },
+    {
+      "epoch": 0.01211242182509281,
+      "grad_norm": 0.062434904277324677,
+      "learning_rate": 0.0002,
+      "loss": 0.1166,
+      "step": 6660
+    },
+    {
+      "epoch": 0.012130608644650006,
+      "grad_norm": 0.055352553725242615,
+      "learning_rate": 0.0002,
+      "loss": 0.0802,
+      "step": 6670
+    },
+    {
+      "epoch": 0.012148795464207203,
+      "grad_norm": 0.031538888812065125,
+      "learning_rate": 0.0002,
+      "loss": 0.0786,
+      "step": 6680
+    },
+    {
+      "epoch": 0.012166982283764399,
+      "grad_norm": 0.10964162647724152,
+      "learning_rate": 0.0002,
+      "loss": 0.0626,
+      "step": 6690
+    },
+    {
+      "epoch": 0.012185169103321595,
+      "grad_norm": 0.011173764243721962,
+      "learning_rate": 0.0002,
+      "loss": 0.0205,
+      "step": 6700
+    },
+    {
+      "epoch": 0.012203355922878792,
+      "grad_norm": 0.035984206944704056,
+      "learning_rate": 0.0002,
+      "loss": 0.1412,
+      "step": 6710
+    },
+    {
+      "epoch": 0.012221542742435988,
+      "grad_norm": 0.07189827412366867,
+      "learning_rate": 0.0002,
+      "loss": 0.0818,
+      "step": 6720
+    },
+    {
+      "epoch": 0.012239729561993184,
+      "grad_norm": 0.0400136299431324,
+      "learning_rate": 0.0002,
+      "loss": 0.0676,
+      "step": 6730
+    },
+    {
+      "epoch": 0.01225791638155038,
+      "grad_norm": 0.14700625836849213,
+      "learning_rate": 0.0002,
+      "loss": 0.0663,
+      "step": 6740
+    },
+    {
+      "epoch": 0.012276103201107577,
+      "grad_norm": 0.007156179752200842,
+      "learning_rate": 0.0002,
+      "loss": 0.0171,
+      "step": 6750
+    },
+    {
+      "epoch": 0.012294290020664773,
+      "grad_norm": 0.04911777004599571,
+      "learning_rate": 0.0002,
+      "loss": 0.1657,
+      "step": 6760
+    },
+    {
+      "epoch": 0.01231247684022197,
+      "grad_norm": 0.03729144483804703,
+      "learning_rate": 0.0002,
+      "loss": 0.0806,
+      "step": 6770
+    },
+    {
+      "epoch": 0.012330663659779166,
+      "grad_norm": 0.037231944501399994,
+      "learning_rate": 0.0002,
+      "loss": 0.0736,
+      "step": 6780
+    },
+    {
+      "epoch": 0.012348850479336364,
+      "grad_norm": 0.09694401919841766,
+      "learning_rate": 0.0002,
+      "loss": 0.0642,
+      "step": 6790
+    },
+    {
+      "epoch": 0.01236703729889356,
+      "grad_norm": 0.025534989312291145,
+      "learning_rate": 0.0002,
+      "loss": 0.0208,
+      "step": 6800
+    },
+    {
+      "epoch": 0.012385224118450756,
+      "grad_norm": 0.033654361963272095,
+      "learning_rate": 0.0002,
+      "loss": 0.1295,
+      "step": 6810
+    },
+    {
+      "epoch": 0.012403410938007953,
+      "grad_norm": 0.04499521851539612,
+      "learning_rate": 0.0002,
+      "loss": 0.0902,
+      "step": 6820
+    },
+    {
+      "epoch": 0.012421597757565149,
+      "grad_norm": 0.0335836224257946,
+      "learning_rate": 0.0002,
+      "loss": 0.0781,
+      "step": 6830
+    },
+    {
+      "epoch": 0.012439784577122345,
+      "grad_norm": 0.1040850430727005,
+      "learning_rate": 0.0002,
+      "loss": 0.0679,
+      "step": 6840
+    },
+    {
+      "epoch": 0.012457971396679542,
+      "grad_norm": 0.015963764861226082,
+      "learning_rate": 0.0002,
+      "loss": 0.0226,
+      "step": 6850
+    },
+    {
+      "epoch": 0.012476158216236738,
+      "grad_norm": 0.05578307807445526,
+      "learning_rate": 0.0002,
+      "loss": 0.1119,
+      "step": 6860
+    },
+    {
+      "epoch": 0.012494345035793934,
+      "grad_norm": 0.0364505760371685,
+      "learning_rate": 0.0002,
+      "loss": 0.0805,
+      "step": 6870
+    },
+    {
+      "epoch": 0.01251253185535113,
+      "grad_norm": 0.027990469709038734,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 6880
+    },
+    {
+      "epoch": 0.012530718674908327,
+      "grad_norm": 0.08282670378684998,
+      "learning_rate": 0.0002,
+      "loss": 0.0685,
+      "step": 6890
+    },
+    {
+      "epoch": 0.012548905494465523,
+      "grad_norm": 0.02172144502401352,
+      "learning_rate": 0.0002,
+      "loss": 0.0259,
+      "step": 6900
+    },
+    {
+      "epoch": 0.01256709231402272,
+      "grad_norm": 0.04074740409851074,
+      "learning_rate": 0.0002,
+      "loss": 0.1211,
+      "step": 6910
+    },
+    {
+      "epoch": 0.012585279133579916,
+      "grad_norm": 0.05433020740747452,
+      "learning_rate": 0.0002,
+      "loss": 0.0776,
+      "step": 6920
+    },
+    {
+      "epoch": 0.012603465953137112,
+      "grad_norm": 0.05479983240365982,
+      "learning_rate": 0.0002,
+      "loss": 0.0778,
+      "step": 6930
+    },
+    {
+      "epoch": 0.012621652772694308,
+      "grad_norm": 1.6031180620193481,
+      "learning_rate": 0.0002,
+      "loss": 0.2265,
+      "step": 6940
+    },
+    {
+      "epoch": 0.012639839592251505,
+      "grad_norm": 1.0940366983413696,
+      "learning_rate": 0.0002,
+      "loss": 0.4586,
+      "step": 6950
+    },
+    {
+      "epoch": 0.012658026411808703,
+      "grad_norm": 0.0412282720208168,
+      "learning_rate": 0.0002,
+      "loss": 0.1072,
+      "step": 6960
+    },
+    {
+      "epoch": 0.012676213231365899,
+      "grad_norm": 0.03705910965800285,
+      "learning_rate": 0.0002,
+      "loss": 0.1014,
+      "step": 6970
+    },
+    {
+      "epoch": 0.012694400050923095,
+      "grad_norm": 0.07444313168525696,
+      "learning_rate": 0.0002,
+      "loss": 0.0881,
+      "step": 6980
+    },
+    {
+      "epoch": 0.012712586870480292,
+      "grad_norm": 0.08558017760515213,
+      "learning_rate": 0.0002,
+      "loss": 0.0619,
+      "step": 6990
+    },
+    {
+      "epoch": 0.012730773690037488,
+      "grad_norm": 0.0004157133516855538,
+      "learning_rate": 0.0002,
+      "loss": 0.0045,
+      "step": 7000
+    },
+    {
+      "epoch": 0.012748960509594684,
+      "grad_norm": 0.07950109243392944,
+      "learning_rate": 0.0002,
+      "loss": 0.1801,
+      "step": 7010
+    },
+    {
+      "epoch": 0.01276714732915188,
+      "grad_norm": 0.08424151688814163,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 7020
+    },
+    {
+      "epoch": 0.012785334148709077,
+      "grad_norm": 0.47635558247566223,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 7030
+    },
+    {
+      "epoch": 0.012803520968266273,
+      "grad_norm": 0.0452958345413208,
+      "learning_rate": 0.0002,
+      "loss": 0.0731,
+      "step": 7040
+    },
+    {
+      "epoch": 0.01282170778782347,
+      "grad_norm": 0.007719043176621199,
+      "learning_rate": 0.0002,
+      "loss": 0.0193,
+      "step": 7050
+    },
+    {
+      "epoch": 0.012839894607380666,
+      "grad_norm": 0.2408572882413864,
+      "learning_rate": 0.0002,
+      "loss": 0.4117,
+      "step": 7060
+    },
+    {
+      "epoch": 0.012858081426937862,
+      "grad_norm": 0.7272363305091858,
+      "learning_rate": 0.0002,
+      "loss": 0.0852,
+      "step": 7070
+    },
+    {
+      "epoch": 0.012876268246495058,
+      "grad_norm": 0.5539261698722839,
+      "learning_rate": 0.0002,
+      "loss": 0.075,
+      "step": 7080
+    },
+    {
+      "epoch": 0.012894455066052255,
+      "grad_norm": 4.608922481536865,
+      "learning_rate": 0.0002,
+      "loss": 0.2301,
+      "step": 7090
+    },
+    {
+      "epoch": 0.012912641885609451,
+      "grad_norm": 0.0012216357281431556,
+      "learning_rate": 0.0002,
+      "loss": 0.0034,
+      "step": 7100
+    },
+    {
+      "epoch": 0.012930828705166647,
+      "grad_norm": 0.15025563538074493,
+      "learning_rate": 0.0002,
+      "loss": 0.2717,
+      "step": 7110
+    },
+    {
+      "epoch": 0.012949015524723844,
+      "grad_norm": 0.06209970638155937,
+      "learning_rate": 0.0002,
+      "loss": 0.0852,
+      "step": 7120
+    },
+    {
+      "epoch": 0.012967202344281042,
+      "grad_norm": 0.6127016544342041,
+      "learning_rate": 0.0002,
+      "loss": 0.1271,
+      "step": 7130
+    },
+    {
+      "epoch": 0.012985389163838238,
+      "grad_norm": 0.047152891755104065,
+      "learning_rate": 0.0002,
+      "loss": 0.0626,
+      "step": 7140
+    },
+    {
+      "epoch": 0.013003575983395434,
+      "grad_norm": 0.0005132685182616115,
+      "learning_rate": 0.0002,
+      "loss": 0.0029,
+      "step": 7150
+    },
+    {
+      "epoch": 0.01302176280295263,
+      "grad_norm": 0.08946029096841812,
+      "learning_rate": 0.0002,
+      "loss": 0.309,
+      "step": 7160
+    },
+    {
+      "epoch": 0.013039949622509827,
+      "grad_norm": 0.18610751628875732,
+      "learning_rate": 0.0002,
+      "loss": 0.0867,
+      "step": 7170
+    },
+    {
+      "epoch": 0.013058136442067023,
+      "grad_norm": 0.07280854880809784,
+      "learning_rate": 0.0002,
+      "loss": 0.0832,
+      "step": 7180
+    },
+    {
+      "epoch": 0.01307632326162422,
+      "grad_norm": 0.11997990310192108,
+      "learning_rate": 0.0002,
+      "loss": 0.0746,
+      "step": 7190
+    },
+    {
+      "epoch": 0.013094510081181416,
+      "grad_norm": 0.00019475500448606908,
+      "learning_rate": 0.0002,
+      "loss": 0.01,
+      "step": 7200
+    },
+    {
+      "epoch": 0.013112696900738612,
+      "grad_norm": 0.07719916105270386,
+      "learning_rate": 0.0002,
+      "loss": 0.3035,
+      "step": 7210
+    },
+    {
+      "epoch": 0.013130883720295809,
+      "grad_norm": 0.0990060344338417,
+      "learning_rate": 0.0002,
+      "loss": 0.0902,
+      "step": 7220
+    },
+    {
+      "epoch": 0.013149070539853005,
+      "grad_norm": 0.22215688228607178,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 7230
+    },
+    {
+      "epoch": 0.013167257359410201,
+      "grad_norm": 0.08412040770053864,
+      "learning_rate": 0.0002,
+      "loss": 0.0646,
+      "step": 7240
+    },
+    {
+      "epoch": 0.013185444178967397,
+      "grad_norm": 0.0017518314998596907,
+      "learning_rate": 0.0002,
+      "loss": 0.007,
+      "step": 7250
+    },
+    {
+      "epoch": 0.013203630998524594,
+      "grad_norm": 0.1554754078388214,
+      "learning_rate": 0.0002,
+      "loss": 0.2319,
+      "step": 7260
+    },
+    {
+      "epoch": 0.01322181781808179,
+      "grad_norm": 0.052371326833963394,
+      "learning_rate": 0.0002,
+      "loss": 0.0832,
+      "step": 7270
+    },
+    {
+      "epoch": 0.013240004637638986,
+      "grad_norm": 0.9168817400932312,
+      "learning_rate": 0.0002,
+      "loss": 0.0791,
+      "step": 7280
+    },
+    {
+      "epoch": 0.013258191457196183,
+      "grad_norm": 0.07169363647699356,
+      "learning_rate": 0.0002,
+      "loss": 0.0602,
+      "step": 7290
+    },
+    {
+      "epoch": 0.01327637827675338,
+      "grad_norm": 0.0009911650558933616,
+      "learning_rate": 0.0002,
+      "loss": 0.0041,
+      "step": 7300
+    },
+    {
+      "epoch": 0.013294565096310577,
+      "grad_norm": 0.2644541263580322,
+      "learning_rate": 0.0002,
+      "loss": 0.2193,
+      "step": 7310
+    },
+    {
+      "epoch": 0.013312751915867773,
+      "grad_norm": 0.12140689790248871,
+      "learning_rate": 0.0002,
+      "loss": 0.0944,
+      "step": 7320
+    },
+    {
+      "epoch": 0.01333093873542497,
+      "grad_norm": 0.03627191483974457,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 7330
+    },
+    {
+      "epoch": 0.013349125554982166,
+      "grad_norm": 0.06252894550561905,
+      "learning_rate": 0.0002,
+      "loss": 0.0596,
+      "step": 7340
+    },
+    {
+      "epoch": 0.013367312374539362,
+      "grad_norm": 0.20318441092967987,
+      "learning_rate": 0.0002,
+      "loss": 0.0064,
+      "step": 7350
+    },
+    {
+      "epoch": 0.013385499194096559,
+      "grad_norm": 0.4231732189655304,
+      "learning_rate": 0.0002,
+      "loss": 0.4329,
+      "step": 7360
+    },
+    {
+      "epoch": 0.013403686013653755,
+      "grad_norm": 0.07567082345485687,
+      "learning_rate": 0.0002,
+      "loss": 0.089,
+      "step": 7370
+    },
+    {
+      "epoch": 0.013421872833210951,
+      "grad_norm": 0.23021474480628967,
+      "learning_rate": 0.0002,
+      "loss": 0.0833,
+      "step": 7380
+    },
+    {
+      "epoch": 0.013440059652768148,
+      "grad_norm": 0.09458985179662704,
+      "learning_rate": 0.0002,
+      "loss": 0.1391,
+      "step": 7390
+    },
+    {
+      "epoch": 0.013458246472325344,
+      "grad_norm": 0.010052111931145191,
+      "learning_rate": 0.0002,
+      "loss": 0.0073,
+      "step": 7400
+    },
+    {
+      "epoch": 0.01347643329188254,
+      "grad_norm": 0.2159787267446518,
+      "learning_rate": 0.0002,
+      "loss": 0.2249,
+      "step": 7410
+    },
+    {
+      "epoch": 0.013494620111439736,
+      "grad_norm": 0.11222853511571884,
+      "learning_rate": 0.0002,
+      "loss": 0.093,
+      "step": 7420
+    },
+    {
+      "epoch": 0.013512806930996933,
+      "grad_norm": 0.08586139976978302,
+      "learning_rate": 0.0002,
+      "loss": 0.0821,
+      "step": 7430
+    },
+    {
+      "epoch": 0.013530993750554129,
+      "grad_norm": 0.12232748419046402,
+      "learning_rate": 0.0002,
+      "loss": 0.0736,
+      "step": 7440
+    },
+    {
+      "epoch": 0.013549180570111325,
+      "grad_norm": 0.006977527402341366,
+      "learning_rate": 0.0002,
+      "loss": 0.0115,
+      "step": 7450
+    },
+    {
+      "epoch": 0.013567367389668522,
+      "grad_norm": 0.051690369844436646,
+      "learning_rate": 0.0002,
+      "loss": 0.2247,
+      "step": 7460
+    },
+    {
+      "epoch": 0.013585554209225718,
+      "grad_norm": 0.06542158871889114,
+      "learning_rate": 0.0002,
+      "loss": 0.1056,
+      "step": 7470
+    },
+    {
+      "epoch": 0.013603741028782916,
+      "grad_norm": 0.18546995520591736,
+      "learning_rate": 0.0002,
+      "loss": 0.1102,
+      "step": 7480
+    },
+    {
+      "epoch": 0.013621927848340112,
+      "grad_norm": 13.399182319641113,
+      "learning_rate": 0.0002,
+      "loss": 2.0806,
+      "step": 7490
+    },
+    {
+      "epoch": 0.013640114667897309,
+      "grad_norm": 0.0982588455080986,
+      "learning_rate": 0.0002,
+      "loss": 0.2158,
+      "step": 7500
+    },
+    {
+      "epoch": 0.013658301487454505,
+      "grad_norm": 0.07860754430294037,
+      "learning_rate": 0.0002,
+      "loss": 0.125,
+      "step": 7510
+    },
+    {
+      "epoch": 0.013676488307011701,
+      "grad_norm": 0.1165497750043869,
+      "learning_rate": 0.0002,
+      "loss": 0.0899,
+      "step": 7520
+    },
+    {
+      "epoch": 0.013694675126568898,
+      "grad_norm": 0.2813965380191803,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 7530
+    },
+    {
+      "epoch": 0.013712861946126094,
+      "grad_norm": 0.33458462357521057,
+      "learning_rate": 0.0002,
+      "loss": 0.0683,
+      "step": 7540
+    },
+    {
+      "epoch": 0.01373104876568329,
+      "grad_norm": 0.012062279507517815,
+      "learning_rate": 0.0002,
+      "loss": 0.0135,
+      "step": 7550
+    },
+    {
+      "epoch": 0.013749235585240487,
+      "grad_norm": 0.1787721961736679,
+      "learning_rate": 0.0002,
+      "loss": 0.1763,
+      "step": 7560
+    },
+    {
+      "epoch": 0.013767422404797683,
+      "grad_norm": 0.05922751501202583,
+      "learning_rate": 0.0002,
+      "loss": 0.1223,
+      "step": 7570
+    },
+    {
+      "epoch": 0.01378560922435488,
+      "grad_norm": 0.11594684422016144,
+      "learning_rate": 0.0002,
+      "loss": 0.0934,
+      "step": 7580
+    },
+    {
+      "epoch": 0.013803796043912075,
+      "grad_norm": 0.2290794998407364,
+      "learning_rate": 0.0002,
+      "loss": 0.0799,
+      "step": 7590
+    },
+    {
+      "epoch": 0.013821982863469272,
+      "grad_norm": 0.04903063178062439,
+      "learning_rate": 0.0002,
+      "loss": 0.0195,
+      "step": 7600
+    },
+    {
+      "epoch": 0.013840169683026468,
+      "grad_norm": 123.61300659179688,
+      "learning_rate": 0.0002,
+      "loss": 0.8359,
+      "step": 7610
+    },
+    {
+      "epoch": 0.013858356502583664,
+      "grad_norm": 0.25403347611427307,
+      "learning_rate": 0.0002,
+      "loss": 1.0464,
+      "step": 7620
+    },
+    {
+      "epoch": 0.01387654332214086,
+      "grad_norm": 0.08144152164459229,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 7630
+    },
+    {
+      "epoch": 0.013894730141698057,
+      "grad_norm": 0.11679713428020477,
+      "learning_rate": 0.0002,
+      "loss": 0.0706,
+      "step": 7640
+    },
+    {
+      "epoch": 0.013912916961255255,
+      "grad_norm": 0.00391317019239068,
+      "learning_rate": 0.0002,
+      "loss": 0.0075,
+      "step": 7650
+    },
+    {
+      "epoch": 0.013931103780812451,
+      "grad_norm": 0.13209663331508636,
+      "learning_rate": 0.0002,
+      "loss": 0.2228,
+      "step": 7660
+    },
+    {
+      "epoch": 0.013949290600369648,
+      "grad_norm": 0.06067880615592003,
+      "learning_rate": 0.0002,
+      "loss": 0.089,
+      "step": 7670
+    },
+    {
+      "epoch": 0.013967477419926844,
+      "grad_norm": 0.04806550592184067,
+      "learning_rate": 0.0002,
+      "loss": 0.0769,
+      "step": 7680
+    },
+    {
+      "epoch": 0.01398566423948404,
+      "grad_norm": 0.09506970643997192,
+      "learning_rate": 0.0002,
+      "loss": 0.0689,
+      "step": 7690
+    },
+    {
+      "epoch": 0.014003851059041237,
+      "grad_norm": 0.002536884741857648,
+      "learning_rate": 0.0002,
+      "loss": 0.0172,
+      "step": 7700
+    },
+    {
+      "epoch": 0.014022037878598433,
+      "grad_norm": 0.13837113976478577,
+      "learning_rate": 0.0002,
+      "loss": 0.2243,
+      "step": 7710
+    },
+    {
+      "epoch": 0.01404022469815563,
+      "grad_norm": 0.08101535588502884,
+      "learning_rate": 0.0002,
+      "loss": 0.0878,
+      "step": 7720
+    },
+    {
+      "epoch": 0.014058411517712826,
+      "grad_norm": 0.04018868878483772,
+      "learning_rate": 0.0002,
+      "loss": 0.0776,
+      "step": 7730
+    },
+    {
+      "epoch": 0.014076598337270022,
+      "grad_norm": 0.1377197653055191,
+      "learning_rate": 0.0002,
+      "loss": 0.0681,
+      "step": 7740
+    },
+    {
+      "epoch": 0.014094785156827218,
+      "grad_norm": 0.0006735012284480035,
+      "learning_rate": 0.0002,
+      "loss": 0.0041,
+      "step": 7750
+    },
+    {
+      "epoch": 0.014112971976384414,
+      "grad_norm": 0.17503094673156738,
+      "learning_rate": 0.0002,
+      "loss": 0.3114,
+      "step": 7760
+    },
+    {
+      "epoch": 0.01413115879594161,
+      "grad_norm": 0.07190551608800888,
+      "learning_rate": 0.0002,
+      "loss": 0.1018,
+      "step": 7770
+    },
+    {
+      "epoch": 0.014149345615498807,
+      "grad_norm": 0.036945659667253494,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 7780
+    },
+    {
+      "epoch": 0.014167532435056003,
+      "grad_norm": 0.13999724388122559,
+      "learning_rate": 0.0002,
+      "loss": 0.0725,
+      "step": 7790
+    },
+    {
+      "epoch": 0.0141857192546132,
+      "grad_norm": 0.0031171294394880533,
+      "learning_rate": 0.0002,
+      "loss": 0.0144,
+      "step": 7800
+    },
+    {
+      "epoch": 0.014203906074170396,
+      "grad_norm": 0.059554051607847214,
+      "learning_rate": 0.0002,
+      "loss": 0.2442,
+      "step": 7810
+    },
+    {
+      "epoch": 0.014222092893727594,
+      "grad_norm": 0.06873622536659241,
+      "learning_rate": 0.0002,
+      "loss": 0.0904,
+      "step": 7820
+    },
+    {
+      "epoch": 0.01424027971328479,
+      "grad_norm": 0.11261582374572754,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 7830
+    },
+    {
+      "epoch": 0.014258466532841987,
+      "grad_norm": 1.497631311416626,
+      "learning_rate": 0.0002,
+      "loss": 0.0689,
+      "step": 7840
+    },
+    {
+      "epoch": 0.014276653352399183,
+      "grad_norm": 0.004822546616196632,
+      "learning_rate": 0.0002,
+      "loss": 0.0156,
+      "step": 7850
+    },
+    {
+      "epoch": 0.01429484017195638,
+      "grad_norm": 0.0575052835047245,
+      "learning_rate": 0.0002,
+      "loss": 0.1895,
+      "step": 7860
+    },
+    {
+      "epoch": 0.014313026991513576,
+      "grad_norm": 0.10657750070095062,
+      "learning_rate": 0.0002,
+      "loss": 0.0855,
+      "step": 7870
+    },
+    {
+      "epoch": 0.014331213811070772,
+      "grad_norm": 0.07080844044685364,
+      "learning_rate": 0.0002,
+      "loss": 0.0716,
+      "step": 7880
+    },
+    {
+      "epoch": 0.014349400630627968,
+      "grad_norm": 0.1628514677286148,
+      "learning_rate": 0.0002,
+      "loss": 0.071,
+      "step": 7890
+    },
+    {
+      "epoch": 0.014367587450185165,
+      "grad_norm": 0.013860347680747509,
+      "learning_rate": 0.0002,
+      "loss": 0.0227,
+      "step": 7900
+    },
+    {
+      "epoch": 0.014385774269742361,
+      "grad_norm": 0.5240967869758606,
+      "learning_rate": 0.0002,
+      "loss": 0.1854,
+      "step": 7910
+    },
+    {
+      "epoch": 0.014403961089299557,
+      "grad_norm": 1.0027457475662231,
+      "learning_rate": 0.0002,
+      "loss": 0.0942,
+      "step": 7920
+    },
+    {
+      "epoch": 0.014422147908856753,
+      "grad_norm": 0.05730056390166283,
+      "learning_rate": 0.0002,
+      "loss": 0.0778,
+      "step": 7930
+    },
+    {
+      "epoch": 0.01444033472841395,
+      "grad_norm": 0.1485404521226883,
+      "learning_rate": 0.0002,
+      "loss": 0.0719,
+      "step": 7940
+    },
+    {
+      "epoch": 0.014458521547971146,
+      "grad_norm": 0.009702637791633606,
+      "learning_rate": 0.0002,
+      "loss": 0.0136,
+      "step": 7950
+    },
+    {
+      "epoch": 0.014476708367528342,
+      "grad_norm": 0.046543315052986145,
+      "learning_rate": 0.0002,
+      "loss": 0.1697,
+      "step": 7960
+    },
+    {
+      "epoch": 0.014494895187085539,
+      "grad_norm": 0.05248842388391495,
+      "learning_rate": 0.0002,
+      "loss": 0.0888,
+      "step": 7970
+    },
+    {
+      "epoch": 0.014513082006642735,
+      "grad_norm": 0.047813788056373596,
+      "learning_rate": 0.0002,
+      "loss": 0.0806,
+      "step": 7980
+    },
+    {
+      "epoch": 0.014531268826199933,
+      "grad_norm": 0.19744129478931427,
+      "learning_rate": 0.0002,
+      "loss": 0.0788,
+      "step": 7990
+    },
+    {
+      "epoch": 0.01454945564575713,
+      "grad_norm": 0.005265017040073872,
+      "learning_rate": 0.0002,
+      "loss": 0.014,
+      "step": 8000
+    },
+    {
+      "epoch": 0.014567642465314326,
+      "grad_norm": 0.0564056858420372,
+      "learning_rate": 0.0002,
+      "loss": 0.2681,
+      "step": 8010
+    },
+    {
+      "epoch": 0.014585829284871522,
+      "grad_norm": 0.0958496481180191,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 8020
+    },
+    {
+      "epoch": 0.014604016104428718,
+      "grad_norm": 0.12000919133424759,
+      "learning_rate": 0.0002,
+      "loss": 0.073,
+      "step": 8030
+    },
+    {
+      "epoch": 0.014622202923985915,
+      "grad_norm": 0.15912771224975586,
+      "learning_rate": 0.0002,
+      "loss": 0.075,
+      "step": 8040
+    },
+    {
+      "epoch": 0.014640389743543111,
+      "grad_norm": 0.004025776404887438,
+      "learning_rate": 0.0002,
+      "loss": 0.012,
+      "step": 8050
+    },
+    {
+      "epoch": 0.014658576563100307,
+      "grad_norm": 0.1682930886745453,
+      "learning_rate": 0.0002,
+      "loss": 0.2926,
+      "step": 8060
+    },
+    {
+      "epoch": 0.014676763382657504,
+      "grad_norm": 0.057362254709005356,
+      "learning_rate": 0.0002,
+      "loss": 0.0869,
+      "step": 8070
+    },
+    {
+      "epoch": 0.0146949502022147,
+      "grad_norm": 0.0814078077673912,
+      "learning_rate": 0.0002,
+      "loss": 0.0825,
+      "step": 8080
+    },
+    {
+      "epoch": 0.014713137021771896,
+      "grad_norm": 0.18205074965953827,
+      "learning_rate": 0.0002,
+      "loss": 0.0699,
+      "step": 8090
+    },
+    {
+      "epoch": 0.014731323841329092,
+      "grad_norm": 0.013200881890952587,
+      "learning_rate": 0.0002,
+      "loss": 0.0228,
+      "step": 8100
+    },
+    {
+      "epoch": 0.014749510660886289,
+      "grad_norm": 0.21043474972248077,
+      "learning_rate": 0.0002,
+      "loss": 0.2138,
+      "step": 8110
+    },
+    {
+      "epoch": 0.014767697480443485,
+      "grad_norm": 0.1000015065073967,
+      "learning_rate": 0.0002,
+      "loss": 0.0906,
+      "step": 8120
+    },
+    {
+      "epoch": 0.014785884300000681,
+      "grad_norm": 0.045657768845558167,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 8130
+    },
+    {
+      "epoch": 0.014804071119557878,
+      "grad_norm": 0.13545630872249603,
+      "learning_rate": 0.0002,
+      "loss": 0.066,
+      "step": 8140
+    },
+    {
+      "epoch": 0.014822257939115074,
+      "grad_norm": 0.01422254927456379,
+      "learning_rate": 0.0002,
+      "loss": 0.0179,
+      "step": 8150
+    },
+    {
+      "epoch": 0.014840444758672272,
+      "grad_norm": 0.12108676135540009,
+      "learning_rate": 0.0002,
+      "loss": 0.1717,
+      "step": 8160
+    },
+    {
+      "epoch": 0.014858631578229468,
+      "grad_norm": 0.10441934317350388,
+      "learning_rate": 0.0002,
+      "loss": 0.106,
+      "step": 8170
+    },
+    {
+      "epoch": 0.014876818397786665,
+      "grad_norm": 0.08105968683958054,
+      "learning_rate": 0.0002,
+      "loss": 0.0829,
+      "step": 8180
+    },
+    {
+      "epoch": 0.014895005217343861,
+      "grad_norm": 0.12230301648378372,
+      "learning_rate": 0.0002,
+      "loss": 0.0731,
+      "step": 8190
+    },
+    {
+      "epoch": 0.014913192036901057,
+      "grad_norm": 0.033857300877571106,
+      "learning_rate": 0.0002,
+      "loss": 0.029,
+      "step": 8200
+    },
+    {
+      "epoch": 0.014931378856458254,
+      "grad_norm": 0.04827893525362015,
+      "learning_rate": 0.0002,
+      "loss": 0.1369,
+      "step": 8210
+    },
+    {
+      "epoch": 0.01494956567601545,
+      "grad_norm": 0.056212421506643295,
+      "learning_rate": 0.0002,
+      "loss": 0.0879,
+      "step": 8220
+    },
+    {
+      "epoch": 0.014967752495572646,
+      "grad_norm": 0.03163846209645271,
+      "learning_rate": 0.0002,
+      "loss": 0.0698,
+      "step": 8230
+    },
+    {
+      "epoch": 0.014985939315129843,
+      "grad_norm": 0.09394920617341995,
+      "learning_rate": 0.0002,
+      "loss": 0.0688,
+      "step": 8240
+    },
+    {
+      "epoch": 0.015004126134687039,
+      "grad_norm": 0.024936649948358536,
+      "learning_rate": 0.0002,
+      "loss": 0.0211,
+      "step": 8250
+    },
+    {
+      "epoch": 0.015022312954244235,
+      "grad_norm": 4.499615669250488,
+      "learning_rate": 0.0002,
+      "loss": 2.7596,
+      "step": 8260
+    },
+    {
+      "epoch": 0.015040499773801431,
+      "grad_norm": 9.221298217773438,
+      "learning_rate": 0.0002,
+      "loss": 0.9135,
+      "step": 8270
+    },
+    {
+      "epoch": 0.015058686593358628,
+      "grad_norm": 0.5199778079986572,
+      "learning_rate": 0.0002,
+      "loss": 0.1441,
+      "step": 8280
+    },
+    {
+      "epoch": 0.015076873412915824,
+      "grad_norm": 0.07028087228536606,
+      "learning_rate": 0.0002,
+      "loss": 0.0793,
+      "step": 8290
+    },
+    {
+      "epoch": 0.01509506023247302,
+      "grad_norm": 0.0003307730657979846,
+      "learning_rate": 0.0002,
+      "loss": 0.0006,
+      "step": 8300
+    },
+    {
+      "epoch": 0.015113247052030217,
+      "grad_norm": 0.7940683960914612,
+      "learning_rate": 0.0002,
+      "loss": 0.7233,
+      "step": 8310
+    },
+    {
+      "epoch": 0.015131433871587413,
+      "grad_norm": 0.09774448722600937,
+      "learning_rate": 0.0002,
+      "loss": 0.1451,
+      "step": 8320
+    },
+    {
+      "epoch": 0.015149620691144611,
+      "grad_norm": 0.3088306188583374,
+      "learning_rate": 0.0002,
+      "loss": 0.0986,
+      "step": 8330
+    },
+    {
+      "epoch": 0.015167807510701807,
+      "grad_norm": 0.08629265427589417,
+      "learning_rate": 0.0002,
+      "loss": 0.0581,
+      "step": 8340
+    },
+    {
+      "epoch": 0.015185994330259004,
+      "grad_norm": 0.0011582528240978718,
+      "learning_rate": 0.0002,
+      "loss": 0.0008,
+      "step": 8350
+    },
+    {
+      "epoch": 0.0152041811498162,
+      "grad_norm": 0.48978063464164734,
+      "learning_rate": 0.0002,
+      "loss": 0.5111,
+      "step": 8360
+    },
+    {
+      "epoch": 0.015222367969373396,
+      "grad_norm": 0.2633112668991089,
+      "learning_rate": 0.0002,
+      "loss": 0.1354,
+      "step": 8370
+    },
+    {
+      "epoch": 0.015240554788930593,
+      "grad_norm": 0.058184925466775894,
+      "learning_rate": 0.0002,
+      "loss": 0.0963,
+      "step": 8380
+    },
+    {
+      "epoch": 0.015258741608487789,
+      "grad_norm": 0.397290974855423,
+      "learning_rate": 0.0002,
+      "loss": 0.0915,
+      "step": 8390
+    },
+    {
+      "epoch": 0.015276928428044985,
+      "grad_norm": 0.0013334077084437013,
+      "learning_rate": 0.0002,
+      "loss": 0.0117,
+      "step": 8400
+    },
+    {
+      "epoch": 0.015295115247602182,
+      "grad_norm": 3.2027626037597656,
+      "learning_rate": 0.0002,
+      "loss": 0.3642,
+      "step": 8410
+    },
+    {
+      "epoch": 0.015313302067159378,
+      "grad_norm": 0.4110456705093384,
+      "learning_rate": 0.0002,
+      "loss": 0.1347,
+      "step": 8420
+    },
+    {
+      "epoch": 0.015331488886716574,
+      "grad_norm": 0.19789688289165497,
+      "learning_rate": 0.0002,
+      "loss": 0.0946,
+      "step": 8430
+    },
+    {
+      "epoch": 0.01534967570627377,
+      "grad_norm": 0.15914630889892578,
+      "learning_rate": 0.0002,
+      "loss": 0.0619,
+      "step": 8440
+    },
+    {
+      "epoch": 0.015367862525830967,
+      "grad_norm": 0.004021051339805126,
+      "learning_rate": 0.0002,
+      "loss": 0.0081,
+      "step": 8450
+    },
+    {
+      "epoch": 0.015386049345388163,
+      "grad_norm": 0.25250542163848877,
+      "learning_rate": 0.0002,
+      "loss": 0.2409,
+      "step": 8460
+    },
+    {
+      "epoch": 0.01540423616494536,
+      "grad_norm": 0.16660314798355103,
+      "learning_rate": 0.0002,
+      "loss": 0.1041,
+      "step": 8470
+    },
+    {
+      "epoch": 0.015422422984502556,
+      "grad_norm": 0.09435573220252991,
+      "learning_rate": 0.0002,
+      "loss": 0.0838,
+      "step": 8480
+    },
+    {
+      "epoch": 0.015440609804059752,
+      "grad_norm": 0.1622086614370346,
+      "learning_rate": 0.0002,
+      "loss": 0.0648,
+      "step": 8490
+    },
+    {
+      "epoch": 0.015458796623616948,
+      "grad_norm": 0.002267120871692896,
+      "learning_rate": 0.0002,
+      "loss": 0.0057,
+      "step": 8500
+    },
+    {
+      "epoch": 0.015476983443174146,
+      "grad_norm": 0.11559420824050903,
+      "learning_rate": 0.0002,
+      "loss": 0.2994,
+      "step": 8510
+    },
+    {
+      "epoch": 0.015495170262731343,
+      "grad_norm": 0.18291179835796356,
+      "learning_rate": 0.0002,
+      "loss": 0.0908,
+      "step": 8520
+    },
+    {
+      "epoch": 0.015513357082288539,
+      "grad_norm": 0.14989323914051056,
+      "learning_rate": 0.0002,
+      "loss": 0.0912,
+      "step": 8530
+    },
+    {
+      "epoch": 0.015531543901845735,
+      "grad_norm": 0.09752708673477173,
+      "learning_rate": 0.0002,
+      "loss": 0.0586,
+      "step": 8540
+    },
+    {
+      "epoch": 0.015549730721402932,
+      "grad_norm": 0.0005314307054504752,
+      "learning_rate": 0.0002,
+      "loss": 0.0026,
+      "step": 8550
+    },
+    {
+      "epoch": 0.015567917540960128,
+      "grad_norm": 0.18309178948402405,
+      "learning_rate": 0.0002,
+      "loss": 0.3059,
+      "step": 8560
+    },
+    {
+      "epoch": 0.015586104360517324,
+      "grad_norm": 0.8144251108169556,
+      "learning_rate": 0.0002,
+      "loss": 0.1103,
+      "step": 8570
+    },
+    {
+      "epoch": 0.01560429118007452,
+      "grad_norm": 0.0331404022872448,
+      "learning_rate": 0.0002,
+      "loss": 0.0898,
+      "step": 8580
+    },
+    {
+      "epoch": 0.015622477999631717,
+      "grad_norm": 0.1460132598876953,
+      "learning_rate": 0.0002,
+      "loss": 0.0656,
+      "step": 8590
+    },
+    {
+      "epoch": 0.015640664819188913,
+      "grad_norm": 0.013606027700006962,
+      "learning_rate": 0.0002,
+      "loss": 0.0076,
+      "step": 8600
+    },
+    {
+      "epoch": 0.01565885163874611,
+      "grad_norm": 0.22224061191082,
+      "learning_rate": 0.0002,
+      "loss": 0.2609,
+      "step": 8610
+    },
+    {
+      "epoch": 0.015677038458303306,
+      "grad_norm": 0.22729800641536713,
+      "learning_rate": 0.0002,
+      "loss": 0.1028,
+      "step": 8620
+    },
+    {
+      "epoch": 0.015695225277860502,
+      "grad_norm": 0.0848810002207756,
+      "learning_rate": 0.0002,
+      "loss": 0.0871,
+      "step": 8630
+    },
+    {
+      "epoch": 0.0157134120974177,
+      "grad_norm": 0.17896370589733124,
+      "learning_rate": 0.0002,
+      "loss": 0.0636,
+      "step": 8640
+    },
+    {
+      "epoch": 0.015731598916974895,
+      "grad_norm": 0.006263076793402433,
+      "learning_rate": 0.0002,
+      "loss": 0.0068,
+      "step": 8650
+    },
+    {
+      "epoch": 0.01574978573653209,
+      "grad_norm": 0.29927679896354675,
+      "learning_rate": 0.0002,
+      "loss": 0.2761,
+      "step": 8660
+    },
+    {
+      "epoch": 0.015767972556089287,
+      "grad_norm": 0.05662700906395912,
+      "learning_rate": 0.0002,
+      "loss": 0.1029,
+      "step": 8670
+    },
+    {
+      "epoch": 0.015786159375646484,
+      "grad_norm": 0.09140895307064056,
+      "learning_rate": 0.0002,
+      "loss": 0.0854,
+      "step": 8680
+    },
+    {
+      "epoch": 0.01580434619520368,
+      "grad_norm": 0.21034927666187286,
+      "learning_rate": 0.0002,
+      "loss": 0.0691,
+      "step": 8690
+    },
+    {
+      "epoch": 0.015822533014760876,
+      "grad_norm": 0.0010229075560346246,
+      "learning_rate": 0.0002,
+      "loss": 0.0033,
+      "step": 8700
+    },
+    {
+      "epoch": 0.015840719834318073,
+      "grad_norm": 0.0626237690448761,
+      "learning_rate": 0.0002,
+      "loss": 0.3583,
+      "step": 8710
+    },
+    {
+      "epoch": 0.01585890665387527,
+      "grad_norm": 0.10027278959751129,
+      "learning_rate": 0.0002,
+      "loss": 0.0959,
+      "step": 8720
+    },
+    {
+      "epoch": 0.015877093473432465,
+      "grad_norm": 0.0870286151766777,
+      "learning_rate": 0.0002,
+      "loss": 0.0767,
+      "step": 8730
+    },
+    {
+      "epoch": 0.01589528029298966,
+      "grad_norm": 0.16106969118118286,
+      "learning_rate": 0.0002,
+      "loss": 0.0763,
+      "step": 8740
+    },
+    {
+      "epoch": 0.015913467112546858,
+      "grad_norm": 0.0022529088892042637,
+      "learning_rate": 0.0002,
+      "loss": 0.0108,
+      "step": 8750
+    },
+    {
+      "epoch": 0.015931653932104058,
+      "grad_norm": 0.06070050224661827,
+      "learning_rate": 0.0002,
+      "loss": 0.2606,
+      "step": 8760
+    },
+    {
+      "epoch": 0.015949840751661254,
+      "grad_norm": 0.09406338632106781,
+      "learning_rate": 0.0002,
+      "loss": 0.1062,
+      "step": 8770
+    },
+    {
+      "epoch": 0.01596802757121845,
+      "grad_norm": 0.1367248147726059,
+      "learning_rate": 0.0002,
+      "loss": 0.0757,
+      "step": 8780
+    },
+    {
+      "epoch": 0.015986214390775647,
+      "grad_norm": 0.26938319206237793,
+      "learning_rate": 0.0002,
+      "loss": 0.0733,
+      "step": 8790
+    },
+    {
+      "epoch": 0.016004401210332843,
+      "grad_norm": 0.011559409089386463,
+      "learning_rate": 0.0002,
+      "loss": 0.0176,
+      "step": 8800
+    },
+    {
+      "epoch": 0.01602258802989004,
+      "grad_norm": 0.12351766228675842,
+      "learning_rate": 0.0002,
+      "loss": 0.24,
+      "step": 8810
+    },
+    {
+      "epoch": 0.016040774849447235,
+      "grad_norm": 0.08965809643268585,
+      "learning_rate": 0.0002,
+      "loss": 0.0947,
+      "step": 8820
+    },
+    {
+      "epoch": 0.016058961669004432,
+      "grad_norm": 0.027005961164832115,
+      "learning_rate": 0.0002,
+      "loss": 0.0738,
+      "step": 8830
+    },
+    {
+      "epoch": 0.016077148488561628,
+      "grad_norm": 0.18656685948371887,
+      "learning_rate": 0.0002,
+      "loss": 0.0667,
+      "step": 8840
+    },
+    {
+      "epoch": 0.016095335308118824,
+      "grad_norm": 0.003148626768961549,
+      "learning_rate": 0.0002,
+      "loss": 0.0119,
+      "step": 8850
+    },
+    {
+      "epoch": 0.01611352212767602,
+      "grad_norm": 0.07959452271461487,
+      "learning_rate": 0.0002,
+      "loss": 0.275,
+      "step": 8860
+    },
+    {
+      "epoch": 0.016131708947233217,
+      "grad_norm": 0.19433775544166565,
+      "learning_rate": 0.0002,
+      "loss": 0.0872,
+      "step": 8870
+    },
+    {
+      "epoch": 0.016149895766790413,
+      "grad_norm": 0.1376393735408783,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 8880
+    },
+    {
+      "epoch": 0.01616808258634761,
+      "grad_norm": 0.18282419443130493,
+      "learning_rate": 0.0002,
+      "loss": 0.068,
+      "step": 8890
+    },
+    {
+      "epoch": 0.016186269405904806,
+      "grad_norm": 0.0112565653398633,
+      "learning_rate": 0.0002,
+      "loss": 0.0137,
+      "step": 8900
+    },
+    {
+      "epoch": 0.016204456225462002,
+      "grad_norm": 0.08975637704133987,
+      "learning_rate": 0.0002,
+      "loss": 0.23,
+      "step": 8910
+    },
+    {
+      "epoch": 0.0162226430450192,
+      "grad_norm": 0.19316238164901733,
+      "learning_rate": 0.0002,
+      "loss": 0.0975,
+      "step": 8920
+    },
+    {
+      "epoch": 0.016240829864576395,
+      "grad_norm": 0.1870724856853485,
+      "learning_rate": 0.0002,
+      "loss": 0.0862,
+      "step": 8930
+    },
+    {
+      "epoch": 0.01625901668413359,
+      "grad_norm": 0.19031721353530884,
+      "learning_rate": 0.0002,
+      "loss": 0.0735,
+      "step": 8940
+    },
+    {
+      "epoch": 0.016277203503690788,
+      "grad_norm": 0.015979783609509468,
+      "learning_rate": 0.0002,
+      "loss": 0.0249,
+      "step": 8950
+    },
+    {
+      "epoch": 0.016295390323247984,
+      "grad_norm": 0.09105712175369263,
+      "learning_rate": 0.0002,
+      "loss": 0.1573,
+      "step": 8960
+    },
+    {
+      "epoch": 0.01631357714280518,
+      "grad_norm": 0.13035650551319122,
+      "learning_rate": 0.0002,
+      "loss": 0.0958,
+      "step": 8970
+    },
+    {
+      "epoch": 0.016331763962362376,
+      "grad_norm": 0.18613573908805847,
+      "learning_rate": 0.0002,
+      "loss": 0.088,
+      "step": 8980
+    },
+    {
+      "epoch": 0.016349950781919573,
+      "grad_norm": 0.2518664300441742,
+      "learning_rate": 0.0002,
+      "loss": 0.0725,
+      "step": 8990
+    },
+    {
+      "epoch": 0.01636813760147677,
+      "grad_norm": 0.03324449062347412,
+      "learning_rate": 0.0002,
+      "loss": 0.0256,
+      "step": 9000
+    },
+    {
+      "epoch": 0.016386324421033965,
+      "grad_norm": 0.08766523003578186,
+      "learning_rate": 0.0002,
+      "loss": 0.1531,
+      "step": 9010
+    },
+    {
+      "epoch": 0.01640451124059116,
+      "grad_norm": 0.14177583158016205,
+      "learning_rate": 0.0002,
+      "loss": 0.0861,
+      "step": 9020
+    },
+    {
+      "epoch": 0.016422698060148358,
+      "grad_norm": 0.1354762315750122,
+      "learning_rate": 0.0002,
+      "loss": 0.0782,
+      "step": 9030
+    },
+    {
+      "epoch": 0.016440884879705554,
+      "grad_norm": 0.15894347429275513,
+      "learning_rate": 0.0002,
+      "loss": 0.073,
+      "step": 9040
+    },
+    {
+      "epoch": 0.01645907169926275,
+      "grad_norm": 0.02154761180281639,
+      "learning_rate": 0.0002,
+      "loss": 0.0156,
+      "step": 9050
+    },
+    {
+      "epoch": 0.016477258518819947,
+      "grad_norm": 0.06432317197322845,
+      "learning_rate": 0.0002,
+      "loss": 0.1384,
+      "step": 9060
+    },
+    {
+      "epoch": 0.016495445338377143,
+      "grad_norm": 0.12112505733966827,
+      "learning_rate": 0.0002,
+      "loss": 0.093,
+      "step": 9070
+    },
+    {
+      "epoch": 0.01651363215793434,
+      "grad_norm": 0.10628003627061844,
+      "learning_rate": 0.0002,
+      "loss": 0.0738,
+      "step": 9080
+    },
+    {
+      "epoch": 0.016531818977491536,
+      "grad_norm": 0.1930958330631256,
+      "learning_rate": 0.0002,
+      "loss": 0.0678,
+      "step": 9090
+    },
+    {
+      "epoch": 0.016550005797048732,
+      "grad_norm": 0.03878525644540787,
+      "learning_rate": 0.0002,
+      "loss": 0.0235,
+      "step": 9100
+    },
+    {
+      "epoch": 0.016568192616605932,
+      "grad_norm": 0.0920896977186203,
+      "learning_rate": 0.0002,
+      "loss": 0.1661,
+      "step": 9110
+    },
+    {
+      "epoch": 0.016586379436163128,
+      "grad_norm": 0.11687818914651871,
+      "learning_rate": 0.0002,
+      "loss": 0.0847,
+      "step": 9120
+    },
+    {
+      "epoch": 0.016604566255720325,
+      "grad_norm": 0.10511167347431183,
+      "learning_rate": 0.0002,
+      "loss": 0.0832,
+      "step": 9130
+    },
+    {
+      "epoch": 0.01662275307527752,
+      "grad_norm": 0.26365017890930176,
+      "learning_rate": 0.0002,
+      "loss": 0.0717,
+      "step": 9140
+    },
+    {
+      "epoch": 0.016640939894834717,
+      "grad_norm": 0.02445841394364834,
+      "learning_rate": 0.0002,
+      "loss": 0.0233,
+      "step": 9150
+    },
+    {
+      "epoch": 0.016659126714391913,
+      "grad_norm": 0.08213133364915848,
+      "learning_rate": 0.0002,
+      "loss": 0.1439,
+      "step": 9160
+    },
+    {
+      "epoch": 0.01667731353394911,
+      "grad_norm": 0.17025598883628845,
+      "learning_rate": 0.0002,
+      "loss": 0.0852,
+      "step": 9170
+    },
+    {
+      "epoch": 0.016695500353506306,
+      "grad_norm": 0.098059743642807,
+      "learning_rate": 0.0002,
+      "loss": 0.0761,
+      "step": 9180
+    },
+    {
+      "epoch": 0.016713687173063502,
+      "grad_norm": 0.18436011672019958,
+      "learning_rate": 0.0002,
+      "loss": 0.0674,
+      "step": 9190
+    },
+    {
+      "epoch": 0.0167318739926207,
+      "grad_norm": 0.011012010276317596,
+      "learning_rate": 0.0002,
+      "loss": 0.0221,
+      "step": 9200
+    },
+    {
+      "epoch": 0.016750060812177895,
+      "grad_norm": 0.07544030994176865,
+      "learning_rate": 0.0002,
+      "loss": 0.161,
+      "step": 9210
+    },
+    {
+      "epoch": 0.01676824763173509,
+      "grad_norm": 0.16041946411132812,
+      "learning_rate": 0.0002,
+      "loss": 0.0824,
+      "step": 9220
+    },
+    {
+      "epoch": 0.016786434451292288,
+      "grad_norm": 0.17295844852924347,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 9230
+    },
+    {
+      "epoch": 0.016804621270849484,
+      "grad_norm": 0.1818791776895523,
+      "learning_rate": 0.0002,
+      "loss": 0.0683,
+      "step": 9240
+    },
+    {
+      "epoch": 0.01682280809040668,
+      "grad_norm": 0.019515013322234154,
+      "learning_rate": 0.0002,
+      "loss": 0.0188,
+      "step": 9250
+    },
+    {
+      "epoch": 0.016840994909963877,
+      "grad_norm": 0.15059705078601837,
+      "learning_rate": 0.0002,
+      "loss": 0.1743,
+      "step": 9260
+    },
+    {
+      "epoch": 0.016859181729521073,
+      "grad_norm": 0.1481601595878601,
+      "learning_rate": 0.0002,
+      "loss": 0.0906,
+      "step": 9270
+    },
+    {
+      "epoch": 0.01687736854907827,
+      "grad_norm": 0.07433108985424042,
+      "learning_rate": 0.0002,
+      "loss": 0.08,
+      "step": 9280
+    },
+    {
+      "epoch": 0.016895555368635466,
+      "grad_norm": 0.1752692312002182,
+      "learning_rate": 0.0002,
+      "loss": 0.06,
+      "step": 9290
+    },
+    {
+      "epoch": 0.016913742188192662,
+      "grad_norm": 0.027612384408712387,
+      "learning_rate": 0.0002,
+      "loss": 0.0157,
+      "step": 9300
+    },
+    {
+      "epoch": 0.016931929007749858,
+      "grad_norm": 0.08575212955474854,
+      "learning_rate": 0.0002,
+      "loss": 0.1679,
+      "step": 9310
+    },
+    {
+      "epoch": 0.016950115827307054,
+      "grad_norm": 0.11127147823572159,
+      "learning_rate": 0.0002,
+      "loss": 0.0848,
+      "step": 9320
+    },
+    {
+      "epoch": 0.01696830264686425,
+      "grad_norm": 0.08989393711090088,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 9330
+    },
+    {
+      "epoch": 0.016986489466421447,
+      "grad_norm": 0.18898548185825348,
+      "learning_rate": 0.0002,
+      "loss": 0.0687,
+      "step": 9340
+    },
+    {
+      "epoch": 0.017004676285978643,
+      "grad_norm": 0.023646721616387367,
+      "learning_rate": 0.0002,
+      "loss": 0.0244,
+      "step": 9350
+    },
+    {
+      "epoch": 0.01702286310553584,
+      "grad_norm": 0.11511775106191635,
+      "learning_rate": 0.0002,
+      "loss": 0.1642,
+      "step": 9360
+    },
+    {
+      "epoch": 0.017041049925093036,
+      "grad_norm": 0.1458021104335785,
+      "learning_rate": 0.0002,
+      "loss": 0.084,
+      "step": 9370
+    },
+    {
+      "epoch": 0.017059236744650232,
+      "grad_norm": 0.060528095811605453,
+      "learning_rate": 0.0002,
+      "loss": 0.0809,
+      "step": 9380
+    },
+    {
+      "epoch": 0.01707742356420743,
+      "grad_norm": 0.16314280033111572,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 9390
+    },
+    {
+      "epoch": 0.017095610383764625,
+      "grad_norm": 0.03078557923436165,
+      "learning_rate": 0.0002,
+      "loss": 0.015,
+      "step": 9400
+    },
+    {
+      "epoch": 0.01711379720332182,
+      "grad_norm": 0.11488370597362518,
+      "learning_rate": 0.0002,
+      "loss": 0.1712,
+      "step": 9410
+    },
+    {
+      "epoch": 0.017131984022879018,
+      "grad_norm": 0.0972781702876091,
+      "learning_rate": 0.0002,
+      "loss": 0.0856,
+      "step": 9420
+    },
+    {
+      "epoch": 0.017150170842436214,
+      "grad_norm": 0.08523645251989365,
+      "learning_rate": 0.0002,
+      "loss": 0.0744,
+      "step": 9430
+    },
+    {
+      "epoch": 0.01716835766199341,
+      "grad_norm": 0.18629521131515503,
+      "learning_rate": 0.0002,
+      "loss": 0.0659,
+      "step": 9440
+    },
+    {
+      "epoch": 0.01718654448155061,
+      "grad_norm": 0.00908618327230215,
+      "learning_rate": 0.0002,
+      "loss": 0.0219,
+      "step": 9450
+    },
+    {
+      "epoch": 0.017204731301107806,
+      "grad_norm": 0.05552325397729874,
+      "learning_rate": 0.0002,
+      "loss": 0.1377,
+      "step": 9460
+    },
+    {
+      "epoch": 0.017222918120665003,
+      "grad_norm": 0.16133128106594086,
+      "learning_rate": 0.0002,
+      "loss": 0.0885,
+      "step": 9470
+    },
+    {
+      "epoch": 0.0172411049402222,
+      "grad_norm": 0.0965205654501915,
+      "learning_rate": 0.0002,
+      "loss": 0.0713,
+      "step": 9480
+    },
+    {
+      "epoch": 0.017259291759779395,
+      "grad_norm": 0.21675604581832886,
+      "learning_rate": 0.0002,
+      "loss": 0.0658,
+      "step": 9490
+    },
+    {
+      "epoch": 0.01727747857933659,
+      "grad_norm": 0.043898243457078934,
+      "learning_rate": 0.0002,
+      "loss": 0.0213,
+      "step": 9500
+    },
+    {
+      "epoch": 0.017295665398893788,
+      "grad_norm": 0.0968618243932724,
+      "learning_rate": 0.0002,
+      "loss": 0.1391,
+      "step": 9510
+    },
+    {
+      "epoch": 0.017313852218450984,
+      "grad_norm": 0.15061378479003906,
+      "learning_rate": 0.0002,
+      "loss": 0.0879,
+      "step": 9520
+    },
+    {
+      "epoch": 0.01733203903800818,
+      "grad_norm": 0.08481590449810028,
+      "learning_rate": 0.0002,
+      "loss": 0.0771,
+      "step": 9530
+    },
+    {
+      "epoch": 0.017350225857565377,
+      "grad_norm": 0.20935995876789093,
+      "learning_rate": 0.0002,
+      "loss": 0.0705,
+      "step": 9540
+    },
+    {
+      "epoch": 0.017368412677122573,
+      "grad_norm": 0.04010302573442459,
+      "learning_rate": 0.0002,
+      "loss": 0.0257,
+      "step": 9550
+    },
+    {
+      "epoch": 0.01738659949667977,
+      "grad_norm": 0.10532956570386887,
+      "learning_rate": 0.0002,
+      "loss": 0.1528,
+      "step": 9560
+    },
+    {
+      "epoch": 0.017404786316236966,
+      "grad_norm": 0.1484638750553131,
+      "learning_rate": 0.0002,
+      "loss": 0.0847,
+      "step": 9570
+    },
+    {
+      "epoch": 0.017422973135794162,
+      "grad_norm": 0.05873465910553932,
+      "learning_rate": 0.0002,
+      "loss": 0.0765,
+      "step": 9580
+    },
+    {
+      "epoch": 0.01744115995535136,
+      "grad_norm": 0.1689092516899109,
+      "learning_rate": 0.0002,
+      "loss": 0.0673,
+      "step": 9590
+    },
+    {
+      "epoch": 0.017459346774908555,
+      "grad_norm": 0.014237391762435436,
+      "learning_rate": 0.0002,
+      "loss": 0.0165,
+      "step": 9600
+    },
+    {
+      "epoch": 0.01747753359446575,
+      "grad_norm": 0.06250491738319397,
+      "learning_rate": 0.0002,
+      "loss": 0.1635,
+      "step": 9610
+    },
+    {
+      "epoch": 0.017495720414022947,
+      "grad_norm": 0.08895017951726913,
+      "learning_rate": 0.0002,
+      "loss": 0.0765,
+      "step": 9620
+    },
+    {
+      "epoch": 0.017513907233580144,
+      "grad_norm": 0.08614445477724075,
+      "learning_rate": 0.0002,
+      "loss": 0.0852,
+      "step": 9630
+    },
+    {
+      "epoch": 0.01753209405313734,
+      "grad_norm": 0.25440698862075806,
+      "learning_rate": 0.0002,
+      "loss": 0.0735,
+      "step": 9640
+    },
+    {
+      "epoch": 0.017550280872694536,
+      "grad_norm": 0.015447271056473255,
+      "learning_rate": 0.0002,
+      "loss": 0.0199,
+      "step": 9650
+    },
+    {
+      "epoch": 0.017568467692251732,
+      "grad_norm": 0.08685171604156494,
+      "learning_rate": 0.0002,
+      "loss": 0.1721,
+      "step": 9660
+    },
+    {
+      "epoch": 0.01758665451180893,
+      "grad_norm": 0.1007658839225769,
+      "learning_rate": 0.0002,
+      "loss": 0.0858,
+      "step": 9670
+    },
+    {
+      "epoch": 0.017604841331366125,
+      "grad_norm": 0.1291055977344513,
+      "learning_rate": 0.0002,
+      "loss": 0.0817,
+      "step": 9680
+    },
+    {
+      "epoch": 0.01762302815092332,
+      "grad_norm": 0.21103522181510925,
+      "learning_rate": 0.0002,
+      "loss": 0.0707,
+      "step": 9690
+    },
+    {
+      "epoch": 0.017641214970480518,
+      "grad_norm": 0.027955593541264534,
+      "learning_rate": 0.0002,
+      "loss": 0.0199,
+      "step": 9700
+    },
+    {
+      "epoch": 0.017659401790037714,
+      "grad_norm": 0.06710019707679749,
+      "learning_rate": 0.0002,
+      "loss": 0.1623,
+      "step": 9710
+    },
+    {
+      "epoch": 0.01767758860959491,
+      "grad_norm": 0.09083720296621323,
+      "learning_rate": 0.0002,
+      "loss": 0.0845,
+      "step": 9720
+    },
+    {
+      "epoch": 0.017695775429152107,
+      "grad_norm": 0.07230041921138763,
+      "learning_rate": 0.0002,
+      "loss": 0.0767,
+      "step": 9730
+    },
+    {
+      "epoch": 0.017713962248709303,
+      "grad_norm": 0.19016912579536438,
+      "learning_rate": 0.0002,
+      "loss": 0.0648,
+      "step": 9740
+    },
+    {
+      "epoch": 0.0177321490682665,
+      "grad_norm": 0.03999534249305725,
+      "learning_rate": 0.0002,
+      "loss": 0.0216,
+      "step": 9750
+    },
+    {
+      "epoch": 0.017750335887823696,
+      "grad_norm": 0.08057496696710587,
+      "learning_rate": 0.0002,
+      "loss": 0.1251,
+      "step": 9760
+    },
+    {
+      "epoch": 0.017768522707380892,
+      "grad_norm": 0.16494789719581604,
+      "learning_rate": 0.0002,
+      "loss": 0.0896,
+      "step": 9770
+    },
+    {
+      "epoch": 0.017786709526938088,
+      "grad_norm": 0.07119818776845932,
+      "learning_rate": 0.0002,
+      "loss": 0.0749,
+      "step": 9780
+    },
+    {
+      "epoch": 0.017804896346495288,
+      "grad_norm": 0.1790028065443039,
+      "learning_rate": 0.0002,
+      "loss": 0.0718,
+      "step": 9790
+    },
+    {
+      "epoch": 0.017823083166052484,
+      "grad_norm": 0.055643875151872635,
+      "learning_rate": 0.0002,
+      "loss": 0.0294,
+      "step": 9800
+    },
+    {
+      "epoch": 0.01784126998560968,
+      "grad_norm": 0.15530900657176971,
+      "learning_rate": 0.0002,
+      "loss": 0.1343,
+      "step": 9810
+    },
+    {
+      "epoch": 0.017859456805166877,
+      "grad_norm": 0.08989892899990082,
+      "learning_rate": 0.0002,
+      "loss": 0.0778,
+      "step": 9820
+    },
+    {
+      "epoch": 0.017877643624724073,
+      "grad_norm": 0.038054581731557846,
+      "learning_rate": 0.0002,
+      "loss": 0.0842,
+      "step": 9830
+    },
+    {
+      "epoch": 0.01789583044428127,
+      "grad_norm": 0.12264154851436615,
+      "learning_rate": 0.0002,
+      "loss": 0.065,
+      "step": 9840
+    },
+    {
+      "epoch": 0.017914017263838466,
+      "grad_norm": 0.03432893753051758,
+      "learning_rate": 0.0002,
+      "loss": 0.02,
+      "step": 9850
+    },
+    {
+      "epoch": 0.017932204083395662,
+      "grad_norm": 0.0516468770802021,
+      "learning_rate": 0.0002,
+      "loss": 0.1339,
+      "step": 9860
+    },
+    {
+      "epoch": 0.01795039090295286,
+      "grad_norm": 0.11306226998567581,
+      "learning_rate": 0.0002,
+      "loss": 0.0842,
+      "step": 9870
+    },
+    {
+      "epoch": 0.017968577722510055,
+      "grad_norm": 0.051579318940639496,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 9880
+    },
+    {
+      "epoch": 0.01798676454206725,
+      "grad_norm": 0.19050930440425873,
+      "learning_rate": 0.0002,
+      "loss": 0.0673,
+      "step": 9890
+    },
+    {
+      "epoch": 0.018004951361624447,
+      "grad_norm": 0.015286738984286785,
+      "learning_rate": 0.0002,
+      "loss": 0.0169,
+      "step": 9900
+    },
+    {
+      "epoch": 0.018023138181181644,
+      "grad_norm": 0.16055025160312653,
+      "learning_rate": 0.0002,
+      "loss": 0.1655,
+      "step": 9910
+    },
+    {
+      "epoch": 0.01804132500073884,
+      "grad_norm": 0.05445674806833267,
+      "learning_rate": 0.0002,
+      "loss": 0.0786,
+      "step": 9920
+    },
+    {
+      "epoch": 0.018059511820296036,
+      "grad_norm": 0.07221481204032898,
+      "learning_rate": 0.0002,
+      "loss": 0.0726,
+      "step": 9930
+    },
+    {
+      "epoch": 0.018077698639853233,
+      "grad_norm": 0.15800146758556366,
+      "learning_rate": 0.0002,
+      "loss": 0.0607,
+      "step": 9940
+    },
+    {
+      "epoch": 0.01809588545941043,
+      "grad_norm": 0.007713336031883955,
+      "learning_rate": 0.0002,
+      "loss": 0.0148,
+      "step": 9950
+    },
+    {
+      "epoch": 0.018114072278967625,
+      "grad_norm": 0.04677269607782364,
+      "learning_rate": 0.0002,
+      "loss": 0.1718,
+      "step": 9960
+    },
+    {
+      "epoch": 0.01813225909852482,
+      "grad_norm": 0.1699189841747284,
+      "learning_rate": 0.0002,
+      "loss": 0.0865,
+      "step": 9970
+    },
+    {
+      "epoch": 0.018150445918082018,
+      "grad_norm": 0.04046279937028885,
+      "learning_rate": 0.0002,
+      "loss": 0.0781,
+      "step": 9980
+    },
+    {
+      "epoch": 0.018168632737639214,
+      "grad_norm": 0.164504736661911,
+      "learning_rate": 0.0002,
+      "loss": 0.0645,
+      "step": 9990
+    },
+    {
+      "epoch": 0.01818681955719641,
+      "grad_norm": 0.014479747042059898,
+      "learning_rate": 0.0002,
+      "loss": 0.0186,
+      "step": 10000
+    },
+    {
+      "epoch": 0.018205006376753607,
+      "grad_norm": 0.051388438791036606,
+      "learning_rate": 0.0002,
+      "loss": 0.1414,
+      "step": 10010
+    },
+    {
+      "epoch": 0.018223193196310803,
+      "grad_norm": 0.11734543740749359,
+      "learning_rate": 0.0002,
+      "loss": 0.0894,
+      "step": 10020
+    },
+    {
+      "epoch": 0.018241380015868,
+      "grad_norm": 0.022312749177217484,
+      "learning_rate": 0.0002,
+      "loss": 0.0775,
+      "step": 10030
+    },
+    {
+      "epoch": 0.018259566835425196,
+      "grad_norm": 0.1579144448041916,
+      "learning_rate": 0.0002,
+      "loss": 0.0668,
+      "step": 10040
+    },
+    {
+      "epoch": 0.018277753654982392,
+      "grad_norm": 0.02757895737886429,
+      "learning_rate": 0.0002,
+      "loss": 0.0197,
+      "step": 10050
+    },
+    {
+      "epoch": 0.01829594047453959,
+      "grad_norm": 0.07557844370603561,
+      "learning_rate": 0.0002,
+      "loss": 0.1526,
+      "step": 10060
+    },
+    {
+      "epoch": 0.018314127294096785,
+      "grad_norm": 0.1216227188706398,
+      "learning_rate": 0.0002,
+      "loss": 0.0871,
+      "step": 10070
+    },
+    {
+      "epoch": 0.01833231411365398,
+      "grad_norm": 0.04201141744852066,
+      "learning_rate": 0.0002,
+      "loss": 0.0723,
+      "step": 10080
+    },
+    {
+      "epoch": 0.018350500933211177,
+      "grad_norm": 0.151902437210083,
+      "learning_rate": 0.0002,
+      "loss": 0.063,
+      "step": 10090
+    },
+    {
+      "epoch": 0.018368687752768374,
+      "grad_norm": 0.028730260208249092,
+      "learning_rate": 0.0002,
+      "loss": 0.0154,
+      "step": 10100
+    },
+    {
+      "epoch": 0.01838687457232557,
+      "grad_norm": 0.0815989300608635,
+      "learning_rate": 0.0002,
+      "loss": 0.1439,
+      "step": 10110
+    },
+    {
+      "epoch": 0.018405061391882766,
+      "grad_norm": 0.16359028220176697,
+      "learning_rate": 0.0002,
+      "loss": 0.0901,
+      "step": 10120
+    },
+    {
+      "epoch": 0.018423248211439962,
+      "grad_norm": 0.055030226707458496,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 10130
+    },
+    {
+      "epoch": 0.018441435030997162,
+      "grad_norm": 0.17064853012561798,
+      "learning_rate": 0.0002,
+      "loss": 0.0713,
+      "step": 10140
+    },
+    {
+      "epoch": 0.01845962185055436,
+      "grad_norm": 0.024902408942580223,
+      "learning_rate": 0.0002,
+      "loss": 0.0201,
+      "step": 10150
+    },
+    {
+      "epoch": 0.018477808670111555,
+      "grad_norm": 0.037377748638391495,
+      "learning_rate": 0.0002,
+      "loss": 0.1394,
+      "step": 10160
+    },
+    {
+      "epoch": 0.01849599548966875,
+      "grad_norm": 0.14072410762310028,
+      "learning_rate": 0.0002,
+      "loss": 0.088,
+      "step": 10170
+    },
+    {
+      "epoch": 0.018514182309225947,
+      "grad_norm": 0.07339414954185486,
+      "learning_rate": 0.0002,
+      "loss": 0.0739,
+      "step": 10180
+    },
+    {
+      "epoch": 0.018532369128783144,
+      "grad_norm": 0.166766956448555,
+      "learning_rate": 0.0002,
+      "loss": 0.0648,
+      "step": 10190
+    },
+    {
+      "epoch": 0.01855055594834034,
+      "grad_norm": 0.009605699218809605,
+      "learning_rate": 0.0002,
+      "loss": 0.0148,
+      "step": 10200
+    },
+    {
+      "epoch": 0.018568742767897536,
+      "grad_norm": 0.045747216790914536,
+      "learning_rate": 0.0002,
+      "loss": 0.1426,
+      "step": 10210
+    },
+    {
+      "epoch": 0.018586929587454733,
+      "grad_norm": 0.09927495568990707,
+      "learning_rate": 0.0002,
+      "loss": 0.0757,
+      "step": 10220
+    },
+    {
+      "epoch": 0.01860511640701193,
+      "grad_norm": 0.032050736248493195,
+      "learning_rate": 0.0002,
+      "loss": 0.0732,
+      "step": 10230
+    },
+    {
+      "epoch": 0.018623303226569125,
+      "grad_norm": 0.14915086328983307,
+      "learning_rate": 0.0002,
+      "loss": 0.0619,
+      "step": 10240
+    },
+    {
+      "epoch": 0.01864149004612632,
+      "grad_norm": 0.019674960523843765,
+      "learning_rate": 0.0002,
+      "loss": 0.0176,
+      "step": 10250
+    },
+    {
+      "epoch": 0.018659676865683518,
+      "grad_norm": 0.0990150198340416,
+      "learning_rate": 0.0002,
+      "loss": 0.156,
+      "step": 10260
+    },
+    {
+      "epoch": 0.018677863685240714,
+      "grad_norm": 0.1409665048122406,
+      "learning_rate": 0.0002,
+      "loss": 0.0843,
+      "step": 10270
+    },
+    {
+      "epoch": 0.01869605050479791,
+      "grad_norm": 0.0232121329754591,
+      "learning_rate": 0.0002,
+      "loss": 0.0712,
+      "step": 10280
+    },
+    {
+      "epoch": 0.018714237324355107,
+      "grad_norm": 0.14811532199382782,
+      "learning_rate": 0.0002,
+      "loss": 0.0649,
+      "step": 10290
+    },
+    {
+      "epoch": 0.018732424143912303,
+      "grad_norm": 0.025812385603785515,
+      "learning_rate": 0.0002,
+      "loss": 0.0191,
+      "step": 10300
+    },
+    {
+      "epoch": 0.0187506109634695,
+      "grad_norm": 0.03710811957716942,
+      "learning_rate": 0.0002,
+      "loss": 0.1323,
+      "step": 10310
+    },
+    {
+      "epoch": 0.018768797783026696,
+      "grad_norm": 0.16586032509803772,
+      "learning_rate": 0.0002,
+      "loss": 0.0868,
+      "step": 10320
+    },
+    {
+      "epoch": 0.018786984602583892,
+      "grad_norm": 0.09154761582612991,
+      "learning_rate": 0.0002,
+      "loss": 0.0762,
+      "step": 10330
+    },
+    {
+      "epoch": 0.01880517142214109,
+      "grad_norm": 0.20400644838809967,
+      "learning_rate": 0.0002,
+      "loss": 0.072,
+      "step": 10340
+    },
+    {
+      "epoch": 0.018823358241698285,
+      "grad_norm": 0.04426256939768791,
+      "learning_rate": 0.0002,
+      "loss": 0.0281,
+      "step": 10350
+    },
+    {
+      "epoch": 0.01884154506125548,
+      "grad_norm": 0.10118848830461502,
+      "learning_rate": 0.0002,
+      "loss": 0.116,
+      "step": 10360
+    },
+    {
+      "epoch": 0.018859731880812677,
+      "grad_norm": 0.11934473365545273,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 10370
+    },
+    {
+      "epoch": 0.018877918700369874,
+      "grad_norm": 0.04116957262158394,
+      "learning_rate": 0.0002,
+      "loss": 0.0811,
+      "step": 10380
+    },
+    {
+      "epoch": 0.01889610551992707,
+      "grad_norm": 0.16668827831745148,
+      "learning_rate": 0.0002,
+      "loss": 0.064,
+      "step": 10390
+    },
+    {
+      "epoch": 0.018914292339484266,
+      "grad_norm": 0.04703928530216217,
+      "learning_rate": 0.0002,
+      "loss": 0.0189,
+      "step": 10400
+    },
+    {
+      "epoch": 0.018932479159041463,
+      "grad_norm": 0.10670439153909683,
+      "learning_rate": 0.0002,
+      "loss": 0.1329,
+      "step": 10410
+    },
+    {
+      "epoch": 0.01895066597859866,
+      "grad_norm": 0.033486492931842804,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 10420
+    },
+    {
+      "epoch": 0.018968852798155855,
+      "grad_norm": 0.03778929263353348,
+      "learning_rate": 0.0002,
+      "loss": 0.0739,
+      "step": 10430
+    },
+    {
+      "epoch": 0.01898703961771305,
+      "grad_norm": 0.1499231606721878,
+      "learning_rate": 0.0002,
+      "loss": 0.0691,
+      "step": 10440
+    },
+    {
+      "epoch": 0.019005226437270248,
+      "grad_norm": 0.020496509969234467,
+      "learning_rate": 0.0002,
+      "loss": 0.0166,
+      "step": 10450
+    },
+    {
+      "epoch": 0.019023413256827444,
+      "grad_norm": 0.07973606884479523,
+      "learning_rate": 0.0002,
+      "loss": 0.1647,
+      "step": 10460
+    },
+    {
+      "epoch": 0.01904160007638464,
+      "grad_norm": 0.2187214344739914,
+      "learning_rate": 0.0002,
+      "loss": 0.0851,
+      "step": 10470
+    },
+    {
+      "epoch": 0.01905978689594184,
+      "grad_norm": 0.05374719575047493,
+      "learning_rate": 0.0002,
+      "loss": 0.0763,
+      "step": 10480
+    },
+    {
+      "epoch": 0.019077973715499037,
+      "grad_norm": 0.20388802886009216,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 10490
+    },
+    {
+      "epoch": 0.019096160535056233,
+      "grad_norm": 0.023114027455449104,
+      "learning_rate": 0.0002,
+      "loss": 0.0196,
+      "step": 10500
+    },
+    {
+      "epoch": 0.01911434735461343,
+      "grad_norm": 0.07263924926519394,
+      "learning_rate": 0.0002,
+      "loss": 0.1397,
+      "step": 10510
+    },
+    {
+      "epoch": 0.019132534174170625,
+      "grad_norm": 0.13590484857559204,
+      "learning_rate": 0.0002,
+      "loss": 0.0866,
+      "step": 10520
+    },
+    {
+      "epoch": 0.019150720993727822,
+      "grad_norm": 0.03279007971286774,
+      "learning_rate": 0.0002,
+      "loss": 0.0845,
+      "step": 10530
+    },
+    {
+      "epoch": 0.019168907813285018,
+      "grad_norm": 0.16929341852664948,
+      "learning_rate": 0.0002,
+      "loss": 0.0638,
+      "step": 10540
+    },
+    {
+      "epoch": 0.019187094632842214,
+      "grad_norm": 0.043504901230335236,
+      "learning_rate": 0.0002,
+      "loss": 0.0217,
+      "step": 10550
+    },
+    {
+      "epoch": 0.01920528145239941,
+      "grad_norm": 0.05582214519381523,
+      "learning_rate": 0.0002,
+      "loss": 0.1454,
+      "step": 10560
+    },
+    {
+      "epoch": 0.019223468271956607,
+      "grad_norm": 0.12112174928188324,
+      "learning_rate": 0.0002,
+      "loss": 0.0773,
+      "step": 10570
+    },
+    {
+      "epoch": 0.019241655091513803,
+      "grad_norm": 0.028584860265254974,
+      "learning_rate": 0.0002,
+      "loss": 0.0766,
+      "step": 10580
+    },
+    {
+      "epoch": 0.019259841911071,
+      "grad_norm": 0.14817841351032257,
+      "learning_rate": 0.0002,
+      "loss": 0.07,
+      "step": 10590
+    },
+    {
+      "epoch": 0.019278028730628196,
+      "grad_norm": 0.0354049950838089,
+      "learning_rate": 0.0002,
+      "loss": 0.0205,
+      "step": 10600
+    },
+    {
+      "epoch": 0.019296215550185392,
+      "grad_norm": 0.0580359622836113,
+      "learning_rate": 0.0002,
+      "loss": 0.126,
+      "step": 10610
+    },
+    {
+      "epoch": 0.01931440236974259,
+      "grad_norm": 0.1495518982410431,
+      "learning_rate": 0.0002,
+      "loss": 0.0759,
+      "step": 10620
+    },
+    {
+      "epoch": 0.019332589189299785,
+      "grad_norm": 0.029057197272777557,
+      "learning_rate": 0.0002,
+      "loss": 0.0751,
+      "step": 10630
+    },
+    {
+      "epoch": 0.01935077600885698,
+      "grad_norm": 0.17057828605175018,
+      "learning_rate": 0.0002,
+      "loss": 0.0675,
+      "step": 10640
+    },
+    {
+      "epoch": 0.019368962828414178,
+      "grad_norm": 0.029123524203896523,
+      "learning_rate": 0.0002,
+      "loss": 0.0207,
+      "step": 10650
+    },
+    {
+      "epoch": 0.019387149647971374,
+      "grad_norm": 0.06929099559783936,
+      "learning_rate": 0.0002,
+      "loss": 0.1272,
+      "step": 10660
+    },
+    {
+      "epoch": 0.01940533646752857,
+      "grad_norm": 0.0806749165058136,
+      "learning_rate": 0.0002,
+      "loss": 0.0825,
+      "step": 10670
+    },
+    {
+      "epoch": 0.019423523287085766,
+      "grad_norm": 0.025454839691519737,
+      "learning_rate": 0.0002,
+      "loss": 0.0786,
+      "step": 10680
+    },
+    {
+      "epoch": 0.019441710106642963,
+      "grad_norm": 0.1879327893257141,
+      "learning_rate": 0.0002,
+      "loss": 0.0664,
+      "step": 10690
+    },
+    {
+      "epoch": 0.01945989692620016,
+      "grad_norm": 0.03334587439894676,
+      "learning_rate": 0.0002,
+      "loss": 0.0217,
+      "step": 10700
+    },
+    {
+      "epoch": 0.019478083745757355,
+      "grad_norm": 0.05760979652404785,
+      "learning_rate": 0.0002,
+      "loss": 0.141,
+      "step": 10710
+    },
+    {
+      "epoch": 0.01949627056531455,
+      "grad_norm": 0.03565089777112007,
+      "learning_rate": 0.0002,
+      "loss": 0.0849,
+      "step": 10720
+    },
+    {
+      "epoch": 0.019514457384871748,
+      "grad_norm": 0.1484966278076172,
+      "learning_rate": 0.0002,
+      "loss": 0.0839,
+      "step": 10730
+    },
+    {
+      "epoch": 0.019532644204428944,
+      "grad_norm": 0.22200991213321686,
+      "learning_rate": 0.0002,
+      "loss": 0.0673,
+      "step": 10740
+    },
+    {
+      "epoch": 0.01955083102398614,
+      "grad_norm": 0.017915472388267517,
+      "learning_rate": 0.0002,
+      "loss": 0.0203,
+      "step": 10750
+    },
+    {
+      "epoch": 0.019569017843543337,
+      "grad_norm": 0.11213338375091553,
+      "learning_rate": 0.0002,
+      "loss": 0.126,
+      "step": 10760
+    },
+    {
+      "epoch": 0.019587204663100533,
+      "grad_norm": 0.1563912183046341,
+      "learning_rate": 0.0002,
+      "loss": 0.0803,
+      "step": 10770
+    },
+    {
+      "epoch": 0.01960539148265773,
+      "grad_norm": 0.02315036952495575,
+      "learning_rate": 0.0002,
+      "loss": 0.0801,
+      "step": 10780
+    },
+    {
+      "epoch": 0.019623578302214926,
+      "grad_norm": 0.14482071995735168,
+      "learning_rate": 0.0002,
+      "loss": 0.0701,
+      "step": 10790
+    },
+    {
+      "epoch": 0.019641765121772122,
+      "grad_norm": 0.0369495190680027,
+      "learning_rate": 0.0002,
+      "loss": 0.022,
+      "step": 10800
+    },
+    {
+      "epoch": 0.01965995194132932,
+      "grad_norm": 0.0659516304731369,
+      "learning_rate": 0.0002,
+      "loss": 0.1282,
+      "step": 10810
+    },
+    {
+      "epoch": 0.019678138760886515,
+      "grad_norm": 0.09046377241611481,
+      "learning_rate": 0.0002,
+      "loss": 0.0748,
+      "step": 10820
+    },
+    {
+      "epoch": 0.019696325580443715,
+      "grad_norm": 0.05669049918651581,
+      "learning_rate": 0.0002,
+      "loss": 0.0808,
+      "step": 10830
+    },
+    {
+      "epoch": 0.01971451240000091,
+      "grad_norm": 0.16696439683437347,
+      "learning_rate": 0.0002,
+      "loss": 0.0696,
+      "step": 10840
+    },
+    {
+      "epoch": 0.019732699219558107,
+      "grad_norm": 0.02596648782491684,
+      "learning_rate": 0.0002,
+      "loss": 0.0189,
+      "step": 10850
+    },
+    {
+      "epoch": 0.019750886039115303,
+      "grad_norm": 0.030568787828087807,
+      "learning_rate": 0.0002,
+      "loss": 0.1431,
+      "step": 10860
+    },
+    {
+      "epoch": 0.0197690728586725,
+      "grad_norm": 0.11519906669855118,
+      "learning_rate": 0.0002,
+      "loss": 0.0832,
+      "step": 10870
+    },
+    {
+      "epoch": 0.019787259678229696,
+      "grad_norm": 0.12018325924873352,
+      "learning_rate": 0.0002,
+      "loss": 0.0769,
+      "step": 10880
+    },
+    {
+      "epoch": 0.019805446497786892,
+      "grad_norm": 0.15875691175460815,
+      "learning_rate": 0.0002,
+      "loss": 0.0679,
+      "step": 10890
+    },
+    {
+      "epoch": 0.01982363331734409,
+      "grad_norm": 0.02812560275197029,
+      "learning_rate": 0.0002,
+      "loss": 0.0236,
+      "step": 10900
+    },
+    {
+      "epoch": 0.019841820136901285,
+      "grad_norm": 0.039342913776636124,
+      "learning_rate": 0.0002,
+      "loss": 0.1433,
+      "step": 10910
+    },
+    {
+      "epoch": 0.01986000695645848,
+      "grad_norm": 0.1218978762626648,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 10920
+    },
+    {
+      "epoch": 0.019878193776015678,
+      "grad_norm": 0.02437124028801918,
+      "learning_rate": 0.0002,
+      "loss": 0.081,
+      "step": 10930
+    },
+    {
+      "epoch": 0.019896380595572874,
+      "grad_norm": 0.16295987367630005,
+      "learning_rate": 0.0002,
+      "loss": 0.0615,
+      "step": 10940
+    },
+    {
+      "epoch": 0.01991456741513007,
+      "grad_norm": 0.03147517144680023,
+      "learning_rate": 0.0002,
+      "loss": 0.0167,
+      "step": 10950
+    },
+    {
+      "epoch": 0.019932754234687267,
+      "grad_norm": 0.051139310002326965,
+      "learning_rate": 0.0002,
+      "loss": 0.1486,
+      "step": 10960
+    },
+    {
+      "epoch": 0.019950941054244463,
+      "grad_norm": 0.10385333746671677,
+      "learning_rate": 0.0002,
+      "loss": 0.0835,
+      "step": 10970
+    },
+    {
+      "epoch": 0.01996912787380166,
+      "grad_norm": 0.029570510610938072,
+      "learning_rate": 0.0002,
+      "loss": 0.0808,
+      "step": 10980
+    },
+    {
+      "epoch": 0.019987314693358856,
+      "grad_norm": 0.1457994282245636,
+      "learning_rate": 0.0002,
+      "loss": 0.062,
+      "step": 10990
+    },
+    {
+      "epoch": 0.020005501512916052,
+      "grad_norm": 0.013582763262093067,
+      "learning_rate": 0.0002,
+      "loss": 0.0149,
+      "step": 11000
+    },
+    {
+      "epoch": 0.020023688332473248,
+      "grad_norm": 0.13736847043037415,
+      "learning_rate": 0.0002,
+      "loss": 0.164,
+      "step": 11010
+    },
+    {
+      "epoch": 0.020041875152030444,
+      "grad_norm": 0.146778866648674,
+      "learning_rate": 0.0002,
+      "loss": 0.0865,
+      "step": 11020
+    },
+    {
+      "epoch": 0.02006006197158764,
+      "grad_norm": 0.09848106652498245,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 11030
+    },
+    {
+      "epoch": 0.020078248791144837,
+      "grad_norm": 0.19981160759925842,
+      "learning_rate": 0.0002,
+      "loss": 0.0738,
+      "step": 11040
+    },
+    {
+      "epoch": 0.020096435610702033,
+      "grad_norm": 0.0248726736754179,
+      "learning_rate": 0.0002,
+      "loss": 0.0232,
+      "step": 11050
+    },
+    {
+      "epoch": 0.02011462243025923,
+      "grad_norm": 0.09688897430896759,
+      "learning_rate": 0.0002,
+      "loss": 0.1361,
+      "step": 11060
+    },
+    {
+      "epoch": 0.020132809249816426,
+      "grad_norm": 0.09953918308019638,
+      "learning_rate": 0.0002,
+      "loss": 0.0815,
+      "step": 11070
+    },
+    {
+      "epoch": 0.020150996069373622,
+      "grad_norm": 0.05801590532064438,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 11080
+    },
+    {
+      "epoch": 0.02016918288893082,
+      "grad_norm": 0.2029600441455841,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 11090
+    },
+    {
+      "epoch": 0.020187369708488015,
+      "grad_norm": 0.026677627116441727,
+      "learning_rate": 0.0002,
+      "loss": 0.022,
+      "step": 11100
+    },
+    {
+      "epoch": 0.02020555652804521,
+      "grad_norm": 0.054907217621803284,
+      "learning_rate": 0.0002,
+      "loss": 0.1356,
+      "step": 11110
+    },
+    {
+      "epoch": 0.020223743347602408,
+      "grad_norm": 0.16302120685577393,
+      "learning_rate": 0.0002,
+      "loss": 0.0721,
+      "step": 11120
+    },
+    {
+      "epoch": 0.020241930167159604,
+      "grad_norm": 0.03393812105059624,
+      "learning_rate": 0.0002,
+      "loss": 0.0733,
+      "step": 11130
+    },
+    {
+      "epoch": 0.0202601169867168,
+      "grad_norm": 0.16455304622650146,
+      "learning_rate": 0.0002,
+      "loss": 0.062,
+      "step": 11140
+    },
+    {
+      "epoch": 0.020278303806273996,
+      "grad_norm": 0.026239484548568726,
+      "learning_rate": 0.0002,
+      "loss": 0.0148,
+      "step": 11150
+    },
+    {
+      "epoch": 0.020296490625831193,
+      "grad_norm": 0.10048040002584457,
+      "learning_rate": 0.0002,
+      "loss": 0.1398,
+      "step": 11160
+    },
+    {
+      "epoch": 0.020314677445388393,
+      "grad_norm": 0.14221400022506714,
+      "learning_rate": 0.0002,
+      "loss": 0.0782,
+      "step": 11170
+    },
+    {
+      "epoch": 0.02033286426494559,
+      "grad_norm": 0.08432412147521973,
+      "learning_rate": 0.0002,
+      "loss": 0.0807,
+      "step": 11180
+    },
+    {
+      "epoch": 0.020351051084502785,
+      "grad_norm": 0.172295480966568,
+      "learning_rate": 0.0002,
+      "loss": 0.0655,
+      "step": 11190
+    },
+    {
+      "epoch": 0.02036923790405998,
+      "grad_norm": 0.023976756259799004,
+      "learning_rate": 0.0002,
+      "loss": 0.0218,
+      "step": 11200
+    },
+    {
+      "epoch": 0.020387424723617178,
+      "grad_norm": 0.03286349028348923,
+      "learning_rate": 0.0002,
+      "loss": 0.1441,
+      "step": 11210
+    },
+    {
+      "epoch": 0.020405611543174374,
+      "grad_norm": 0.04403531551361084,
+      "learning_rate": 0.0002,
+      "loss": 0.0825,
+      "step": 11220
+    },
+    {
+      "epoch": 0.02042379836273157,
+      "grad_norm": 0.0398452989757061,
+      "learning_rate": 0.0002,
+      "loss": 0.0755,
+      "step": 11230
+    },
+    {
+      "epoch": 0.020441985182288767,
+      "grad_norm": 0.15185104310512543,
+      "learning_rate": 0.0002,
+      "loss": 0.0591,
+      "step": 11240
+    },
+    {
+      "epoch": 0.020460172001845963,
+      "grad_norm": 0.005839187186211348,
+      "learning_rate": 0.0002,
+      "loss": 0.0172,
+      "step": 11250
+    },
+    {
+      "epoch": 0.02047835882140316,
+      "grad_norm": 0.031195368617773056,
+      "learning_rate": 0.0002,
+      "loss": 0.1594,
+      "step": 11260
+    },
+    {
+      "epoch": 0.020496545640960356,
+      "grad_norm": 0.1997426599264145,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 11270
+    },
+    {
+      "epoch": 0.020514732460517552,
+      "grad_norm": 0.03075752593576908,
+      "learning_rate": 0.0002,
+      "loss": 0.0798,
+      "step": 11280
+    },
+    {
+      "epoch": 0.02053291928007475,
+      "grad_norm": 0.17717675864696503,
+      "learning_rate": 0.0002,
+      "loss": 0.0654,
+      "step": 11290
+    },
+    {
+      "epoch": 0.020551106099631945,
+      "grad_norm": 0.036260057240724564,
+      "learning_rate": 0.0002,
+      "loss": 0.0196,
+      "step": 11300
+    },
+    {
+      "epoch": 0.02056929291918914,
+      "grad_norm": 0.11961262673139572,
+      "learning_rate": 0.0002,
+      "loss": 0.1313,
+      "step": 11310
+    },
+    {
+      "epoch": 0.020587479738746337,
+      "grad_norm": 0.12344212830066681,
+      "learning_rate": 0.0002,
+      "loss": 0.0798,
+      "step": 11320
+    },
+    {
+      "epoch": 0.020605666558303534,
+      "grad_norm": 0.12796273827552795,
+      "learning_rate": 0.0002,
+      "loss": 0.0693,
+      "step": 11330
+    },
+    {
+      "epoch": 0.02062385337786073,
+      "grad_norm": 0.12038332223892212,
+      "learning_rate": 0.0002,
+      "loss": 0.0637,
+      "step": 11340
+    },
+    {
+      "epoch": 0.020642040197417926,
+      "grad_norm": 0.013724497519433498,
+      "learning_rate": 0.0002,
+      "loss": 0.0134,
+      "step": 11350
+    },
+    {
+      "epoch": 0.020660227016975122,
+      "grad_norm": 0.030014917254447937,
+      "learning_rate": 0.0002,
+      "loss": 0.1355,
+      "step": 11360
+    },
+    {
+      "epoch": 0.02067841383653232,
+      "grad_norm": 0.05455614998936653,
+      "learning_rate": 0.0002,
+      "loss": 0.0824,
+      "step": 11370
+    },
+    {
+      "epoch": 0.020696600656089515,
+      "grad_norm": 0.09036605060100555,
+      "learning_rate": 0.0002,
+      "loss": 0.0712,
+      "step": 11380
+    },
+    {
+      "epoch": 0.02071478747564671,
+      "grad_norm": 0.15607796609401703,
+      "learning_rate": 0.0002,
+      "loss": 0.0613,
+      "step": 11390
+    },
+    {
+      "epoch": 0.020732974295203908,
+      "grad_norm": 0.029900453984737396,
+      "learning_rate": 0.0002,
+      "loss": 0.0216,
+      "step": 11400
+    },
+    {
+      "epoch": 0.020751161114761104,
+      "grad_norm": 0.06108042970299721,
+      "learning_rate": 0.0002,
+      "loss": 0.1223,
+      "step": 11410
+    },
+    {
+      "epoch": 0.0207693479343183,
+      "grad_norm": 0.052377600222826004,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 11420
+    },
+    {
+      "epoch": 0.020787534753875497,
+      "grad_norm": 0.063735231757164,
+      "learning_rate": 0.0002,
+      "loss": 0.0746,
+      "step": 11430
+    },
+    {
+      "epoch": 0.020805721573432693,
+      "grad_norm": 0.16977328062057495,
+      "learning_rate": 0.0002,
+      "loss": 0.0634,
+      "step": 11440
+    },
+    {
+      "epoch": 0.02082390839298989,
+      "grad_norm": 0.04451785981655121,
+      "learning_rate": 0.0002,
+      "loss": 0.0298,
+      "step": 11450
+    },
+    {
+      "epoch": 0.020842095212547086,
+      "grad_norm": 1.1584863662719727,
+      "learning_rate": 0.0002,
+      "loss": 0.1133,
+      "step": 11460
+    },
+    {
+      "epoch": 0.020860282032104282,
+      "grad_norm": 0.09867832064628601,
+      "learning_rate": 0.0002,
+      "loss": 0.0774,
+      "step": 11470
+    },
+    {
+      "epoch": 0.020878468851661478,
+      "grad_norm": 0.05493566766381264,
+      "learning_rate": 0.0002,
+      "loss": 0.0752,
+      "step": 11480
+    },
+    {
+      "epoch": 0.020896655671218674,
+      "grad_norm": 0.2149093896150589,
+      "learning_rate": 0.0002,
+      "loss": 0.0682,
+      "step": 11490
+    },
+    {
+      "epoch": 0.02091484249077587,
+      "grad_norm": 0.02243107184767723,
+      "learning_rate": 0.0002,
+      "loss": 0.0191,
+      "step": 11500
+    },
+    {
+      "epoch": 0.02093302931033307,
+      "grad_norm": 0.27817150950431824,
+      "learning_rate": 0.0002,
+      "loss": 0.1658,
+      "step": 11510
+    },
+    {
+      "epoch": 0.020951216129890267,
+      "grad_norm": 0.14467410743236542,
+      "learning_rate": 0.0002,
+      "loss": 0.083,
+      "step": 11520
+    },
+    {
+      "epoch": 0.020969402949447463,
+      "grad_norm": 0.1027064323425293,
+      "learning_rate": 0.0002,
+      "loss": 0.0825,
+      "step": 11530
+    },
+    {
+      "epoch": 0.02098758976900466,
+      "grad_norm": 0.2156657725572586,
+      "learning_rate": 0.0002,
+      "loss": 0.0694,
+      "step": 11540
+    },
+    {
+      "epoch": 0.021005776588561856,
+      "grad_norm": 0.023746902123093605,
+      "learning_rate": 0.0002,
+      "loss": 0.024,
+      "step": 11550
+    },
+    {
+      "epoch": 0.021023963408119052,
+      "grad_norm": 0.19738778471946716,
+      "learning_rate": 0.0002,
+      "loss": 0.1473,
+      "step": 11560
+    },
+    {
+      "epoch": 0.02104215022767625,
+      "grad_norm": 0.19759760797023773,
+      "learning_rate": 0.0002,
+      "loss": 0.078,
+      "step": 11570
+    },
+    {
+      "epoch": 0.021060337047233445,
+      "grad_norm": 9.88092041015625,
+      "learning_rate": 0.0002,
+      "loss": 0.0866,
+      "step": 11580
+    },
+    {
+      "epoch": 0.02107852386679064,
+      "grad_norm": 0.22301238775253296,
+      "learning_rate": 0.0002,
+      "loss": 0.0685,
+      "step": 11590
+    },
+    {
+      "epoch": 0.021096710686347837,
+      "grad_norm": 0.023191403597593307,
+      "learning_rate": 0.0002,
+      "loss": 0.0468,
+      "step": 11600
+    },
+    {
+      "epoch": 0.021114897505905034,
+      "grad_norm": 0.10442623496055603,
+      "learning_rate": 0.0002,
+      "loss": 0.2046,
+      "step": 11610
+    },
+    {
+      "epoch": 0.02113308432546223,
+      "grad_norm": 0.18771864473819733,
+      "learning_rate": 0.0002,
+      "loss": 0.0805,
+      "step": 11620
+    },
+    {
+      "epoch": 0.021151271145019426,
+      "grad_norm": 0.05516243353486061,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 11630
+    },
+    {
+      "epoch": 0.021169457964576623,
+      "grad_norm": 0.21308554708957672,
+      "learning_rate": 0.0002,
+      "loss": 0.0725,
+      "step": 11640
+    },
+    {
+      "epoch": 0.02118764478413382,
+      "grad_norm": 0.010607315227389336,
+      "learning_rate": 0.0002,
+      "loss": 0.0241,
+      "step": 11650
+    },
+    {
+      "epoch": 0.021205831603691015,
+      "grad_norm": 0.0542677640914917,
+      "learning_rate": 0.0002,
+      "loss": 0.1648,
+      "step": 11660
+    },
+    {
+      "epoch": 0.02122401842324821,
+      "grad_norm": 0.11239166557788849,
+      "learning_rate": 0.0002,
+      "loss": 0.0825,
+      "step": 11670
+    },
+    {
+      "epoch": 0.021242205242805408,
+      "grad_norm": 0.032700493931770325,
+      "learning_rate": 0.0002,
+      "loss": 0.0727,
+      "step": 11680
+    },
+    {
+      "epoch": 0.021260392062362604,
+      "grad_norm": 0.2005159705877304,
+      "learning_rate": 0.0002,
+      "loss": 0.0708,
+      "step": 11690
+    },
+    {
+      "epoch": 0.0212785788819198,
+      "grad_norm": 0.01741277053952217,
+      "learning_rate": 0.0002,
+      "loss": 0.0232,
+      "step": 11700
+    },
+    {
+      "epoch": 0.021296765701476997,
+      "grad_norm": 0.04048267379403114,
+      "learning_rate": 0.0002,
+      "loss": 0.1403,
+      "step": 11710
+    },
+    {
+      "epoch": 0.021314952521034193,
+      "grad_norm": 0.18796616792678833,
+      "learning_rate": 0.0002,
+      "loss": 0.0886,
+      "step": 11720
+    },
+    {
+      "epoch": 0.02133313934059139,
+      "grad_norm": 0.06360754370689392,
+      "learning_rate": 0.0002,
+      "loss": 0.0731,
+      "step": 11730
+    },
+    {
+      "epoch": 0.021351326160148586,
+      "grad_norm": 0.14168913662433624,
+      "learning_rate": 0.0002,
+      "loss": 0.0622,
+      "step": 11740
+    },
+    {
+      "epoch": 0.021369512979705782,
+      "grad_norm": 0.012988853268325329,
+      "learning_rate": 0.0002,
+      "loss": 0.0144,
+      "step": 11750
+    },
+    {
+      "epoch": 0.02138769979926298,
+      "grad_norm": 0.09176674485206604,
+      "learning_rate": 0.0002,
+      "loss": 0.1574,
+      "step": 11760
+    },
+    {
+      "epoch": 0.021405886618820175,
+      "grad_norm": 0.11934395134449005,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 11770
+    },
+    {
+      "epoch": 0.02142407343837737,
+      "grad_norm": 0.11853605508804321,
+      "learning_rate": 0.0002,
+      "loss": 0.076,
+      "step": 11780
+    },
+    {
+      "epoch": 0.021442260257934567,
+      "grad_norm": 0.1625816971063614,
+      "learning_rate": 0.0002,
+      "loss": 0.0649,
+      "step": 11790
+    },
+    {
+      "epoch": 0.021460447077491764,
+      "grad_norm": 0.023221928626298904,
+      "learning_rate": 0.0002,
+      "loss": 0.0228,
+      "step": 11800
+    },
+    {
+      "epoch": 0.02147863389704896,
+      "grad_norm": 0.0494253933429718,
+      "learning_rate": 0.0002,
+      "loss": 0.1418,
+      "step": 11810
+    },
+    {
+      "epoch": 0.021496820716606156,
+      "grad_norm": 0.18250688910484314,
+      "learning_rate": 0.0002,
+      "loss": 0.0827,
+      "step": 11820
+    },
+    {
+      "epoch": 0.021515007536163352,
+      "grad_norm": 0.13340160250663757,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 11830
+    },
+    {
+      "epoch": 0.02153319435572055,
+      "grad_norm": 0.15497778356075287,
+      "learning_rate": 0.0002,
+      "loss": 0.0613,
+      "step": 11840
+    },
+    {
+      "epoch": 0.021551381175277745,
+      "grad_norm": 0.03259354829788208,
+      "learning_rate": 0.0002,
+      "loss": 0.023,
+      "step": 11850
+    },
+    {
+      "epoch": 0.021569567994834945,
+      "grad_norm": 0.09126435220241547,
+      "learning_rate": 0.0002,
+      "loss": 0.1235,
+      "step": 11860
+    },
+    {
+      "epoch": 0.02158775481439214,
+      "grad_norm": 0.13455496728420258,
+      "learning_rate": 0.0002,
+      "loss": 0.0806,
+      "step": 11870
+    },
+    {
+      "epoch": 0.021605941633949338,
+      "grad_norm": 0.10817539691925049,
+      "learning_rate": 0.0002,
+      "loss": 0.0829,
+      "step": 11880
+    },
+    {
+      "epoch": 0.021624128453506534,
+      "grad_norm": 0.1913878321647644,
+      "learning_rate": 0.0002,
+      "loss": 0.0636,
+      "step": 11890
+    },
+    {
+      "epoch": 0.02164231527306373,
+      "grad_norm": 0.025634530931711197,
+      "learning_rate": 0.0002,
+      "loss": 0.0216,
+      "step": 11900
+    },
+    {
+      "epoch": 0.021660502092620926,
+      "grad_norm": 0.10507725924253464,
+      "learning_rate": 0.0002,
+      "loss": 0.1326,
+      "step": 11910
+    },
+    {
+      "epoch": 0.021678688912178123,
+      "grad_norm": 0.09721452742815018,
+      "learning_rate": 0.0002,
+      "loss": 0.0857,
+      "step": 11920
+    },
+    {
+      "epoch": 0.02169687573173532,
+      "grad_norm": 0.028759269043803215,
+      "learning_rate": 0.0002,
+      "loss": 0.0751,
+      "step": 11930
+    },
+    {
+      "epoch": 0.021715062551292515,
+      "grad_norm": 0.17618104815483093,
+      "learning_rate": 0.0002,
+      "loss": 0.062,
+      "step": 11940
+    },
+    {
+      "epoch": 0.02173324937084971,
+      "grad_norm": 0.02503124624490738,
+      "learning_rate": 0.0002,
+      "loss": 0.0182,
+      "step": 11950
+    },
+    {
+      "epoch": 0.021751436190406908,
+      "grad_norm": 0.10976126044988632,
+      "learning_rate": 0.0002,
+      "loss": 0.1564,
+      "step": 11960
+    },
+    {
+      "epoch": 0.021769623009964104,
+      "grad_norm": 0.0833989605307579,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 11970
+    },
+    {
+      "epoch": 0.0217878098295213,
+      "grad_norm": 0.06359647959470749,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 11980
+    },
+    {
+      "epoch": 0.021805996649078497,
+      "grad_norm": 0.1677824705839157,
+      "learning_rate": 0.0002,
+      "loss": 0.0699,
+      "step": 11990
+    },
+    {
+      "epoch": 0.021824183468635693,
+      "grad_norm": 0.018009621649980545,
+      "learning_rate": 0.0002,
+      "loss": 0.0185,
+      "step": 12000
+    },
+    {
+      "epoch": 0.02184237028819289,
+      "grad_norm": 0.12256644666194916,
+      "learning_rate": 0.0002,
+      "loss": 0.1839,
+      "step": 12010
+    },
+    {
+      "epoch": 0.021860557107750086,
+      "grad_norm": 0.11677028983831406,
+      "learning_rate": 0.0002,
+      "loss": 0.0829,
+      "step": 12020
+    },
+    {
+      "epoch": 0.021878743927307282,
+      "grad_norm": 0.12885046005249023,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 12030
+    },
+    {
+      "epoch": 0.02189693074686448,
+      "grad_norm": 0.1394425481557846,
+      "learning_rate": 0.0002,
+      "loss": 0.0668,
+      "step": 12040
+    },
+    {
+      "epoch": 0.021915117566421675,
+      "grad_norm": 0.024974076077342033,
+      "learning_rate": 0.0002,
+      "loss": 0.0192,
+      "step": 12050
+    },
+    {
+      "epoch": 0.02193330438597887,
+      "grad_norm": 0.11284986138343811,
+      "learning_rate": 0.0002,
+      "loss": 0.1492,
+      "step": 12060
+    },
+    {
+      "epoch": 0.021951491205536067,
+      "grad_norm": 0.0605492927134037,
+      "learning_rate": 0.0002,
+      "loss": 0.0787,
+      "step": 12070
+    },
+    {
+      "epoch": 0.021969678025093264,
+      "grad_norm": 0.040298718959093094,
+      "learning_rate": 0.0002,
+      "loss": 0.0778,
+      "step": 12080
+    },
+    {
+      "epoch": 0.02198786484465046,
+      "grad_norm": 0.1555332988500595,
+      "learning_rate": 0.0002,
+      "loss": 0.0683,
+      "step": 12090
+    },
+    {
+      "epoch": 0.022006051664207656,
+      "grad_norm": 0.022474724799394608,
+      "learning_rate": 0.0002,
+      "loss": 0.0139,
+      "step": 12100
+    },
+    {
+      "epoch": 0.022024238483764853,
+      "grad_norm": 0.08212363719940186,
+      "learning_rate": 0.0002,
+      "loss": 0.1513,
+      "step": 12110
+    },
+    {
+      "epoch": 0.02204242530332205,
+      "grad_norm": 0.16297335922718048,
+      "learning_rate": 0.0002,
+      "loss": 0.087,
+      "step": 12120
+    },
+    {
+      "epoch": 0.022060612122879245,
+      "grad_norm": 0.026817265897989273,
+      "learning_rate": 0.0002,
+      "loss": 0.0763,
+      "step": 12130
+    },
+    {
+      "epoch": 0.02207879894243644,
+      "grad_norm": 0.15199647843837738,
+      "learning_rate": 0.0002,
+      "loss": 0.0632,
+      "step": 12140
+    },
+    {
+      "epoch": 0.022096985761993638,
+      "grad_norm": 0.021619049832224846,
+      "learning_rate": 0.0002,
+      "loss": 0.0221,
+      "step": 12150
+    },
+    {
+      "epoch": 0.022115172581550834,
+      "grad_norm": 0.071327805519104,
+      "learning_rate": 0.0002,
+      "loss": 0.138,
+      "step": 12160
+    },
+    {
+      "epoch": 0.02213335940110803,
+      "grad_norm": 0.07506705075502396,
+      "learning_rate": 0.0002,
+      "loss": 0.0802,
+      "step": 12170
+    },
+    {
+      "epoch": 0.022151546220665227,
+      "grad_norm": 0.05193526670336723,
+      "learning_rate": 0.0002,
+      "loss": 0.0726,
+      "step": 12180
+    },
+    {
+      "epoch": 0.022169733040222423,
+      "grad_norm": 0.125730961561203,
+      "learning_rate": 0.0002,
+      "loss": 0.0658,
+      "step": 12190
+    },
+    {
+      "epoch": 0.022187919859779623,
+      "grad_norm": 0.01939002424478531,
+      "learning_rate": 0.0002,
+      "loss": 0.0174,
+      "step": 12200
+    },
+    {
+      "epoch": 0.02220610667933682,
+      "grad_norm": 0.05645585432648659,
+      "learning_rate": 0.0002,
+      "loss": 0.1447,
+      "step": 12210
+    },
+    {
+      "epoch": 0.022224293498894016,
+      "grad_norm": 0.12416274845600128,
+      "learning_rate": 0.0002,
+      "loss": 0.0727,
+      "step": 12220
+    },
+    {
+      "epoch": 0.022242480318451212,
+      "grad_norm": 0.05618545040488243,
+      "learning_rate": 0.0002,
+      "loss": 0.0801,
+      "step": 12230
+    },
+    {
+      "epoch": 0.022260667138008408,
+      "grad_norm": 0.12334968894720078,
+      "learning_rate": 0.0002,
+      "loss": 0.0598,
+      "step": 12240
+    },
+    {
+      "epoch": 0.022278853957565604,
+      "grad_norm": 0.024331970140337944,
+      "learning_rate": 0.0002,
+      "loss": 0.0179,
+      "step": 12250
+    },
+    {
+      "epoch": 0.0222970407771228,
+      "grad_norm": 0.05856281518936157,
+      "learning_rate": 0.0002,
+      "loss": 0.126,
+      "step": 12260
+    },
+    {
+      "epoch": 0.022315227596679997,
+      "grad_norm": 0.07432300597429276,
+      "learning_rate": 0.0002,
+      "loss": 0.0839,
+      "step": 12270
+    },
+    {
+      "epoch": 0.022333414416237193,
+      "grad_norm": 0.07249715179204941,
+      "learning_rate": 0.0002,
+      "loss": 0.0815,
+      "step": 12280
+    },
+    {
+      "epoch": 0.02235160123579439,
+      "grad_norm": 0.14335612952709198,
+      "learning_rate": 0.0002,
+      "loss": 0.0605,
+      "step": 12290
+    },
+    {
+      "epoch": 0.022369788055351586,
+      "grad_norm": 0.03603110462427139,
+      "learning_rate": 0.0002,
+      "loss": 0.0185,
+      "step": 12300
+    },
+    {
+      "epoch": 0.022387974874908782,
+      "grad_norm": 0.08532091230154037,
+      "learning_rate": 0.0002,
+      "loss": 0.1339,
+      "step": 12310
+    },
+    {
+      "epoch": 0.02240616169446598,
+      "grad_norm": 0.13663236796855927,
+      "learning_rate": 0.0002,
+      "loss": 0.0761,
+      "step": 12320
+    },
+    {
+      "epoch": 0.022424348514023175,
+      "grad_norm": 0.10088011622428894,
+      "learning_rate": 0.0002,
+      "loss": 0.0737,
+      "step": 12330
+    },
+    {
+      "epoch": 0.02244253533358037,
+      "grad_norm": 0.17186152935028076,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 12340
+    },
+    {
+      "epoch": 0.022460722153137568,
+      "grad_norm": 0.01941334828734398,
+      "learning_rate": 0.0002,
+      "loss": 0.0135,
+      "step": 12350
+    },
+    {
+      "epoch": 0.022478908972694764,
+      "grad_norm": 0.12438862770795822,
+      "learning_rate": 0.0002,
+      "loss": 0.1474,
+      "step": 12360
+    },
+    {
+      "epoch": 0.02249709579225196,
+      "grad_norm": 0.08050791174173355,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 12370
+    },
+    {
+      "epoch": 0.022515282611809156,
+      "grad_norm": 0.04660952091217041,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 12380
+    },
+    {
+      "epoch": 0.022533469431366353,
+      "grad_norm": 0.16433311998844147,
+      "learning_rate": 0.0002,
+      "loss": 0.0691,
+      "step": 12390
+    },
+    {
+      "epoch": 0.02255165625092355,
+      "grad_norm": 0.04376552626490593,
+      "learning_rate": 0.0002,
+      "loss": 0.0219,
+      "step": 12400
+    },
+    {
+      "epoch": 0.022569843070480745,
+      "grad_norm": 0.06648654490709305,
+      "learning_rate": 0.0002,
+      "loss": 0.1346,
+      "step": 12410
+    },
+    {
+      "epoch": 0.02258802989003794,
+      "grad_norm": 0.11318199336528778,
+      "learning_rate": 0.0002,
+      "loss": 0.0736,
+      "step": 12420
+    },
+    {
+      "epoch": 0.022606216709595138,
+      "grad_norm": 0.0922408252954483,
+      "learning_rate": 0.0002,
+      "loss": 0.0819,
+      "step": 12430
+    },
+    {
+      "epoch": 0.022624403529152334,
+      "grad_norm": 0.1696896106004715,
+      "learning_rate": 0.0002,
+      "loss": 0.0642,
+      "step": 12440
+    },
+    {
+      "epoch": 0.02264259034870953,
+      "grad_norm": 0.03212421387434006,
+      "learning_rate": 0.0002,
+      "loss": 0.0247,
+      "step": 12450
+    },
+    {
+      "epoch": 0.022660777168266727,
+      "grad_norm": 0.12295889109373093,
+      "learning_rate": 0.0002,
+      "loss": 0.1504,
+      "step": 12460
+    },
+    {
+      "epoch": 0.022678963987823923,
+      "grad_norm": 0.10351194441318512,
+      "learning_rate": 0.0002,
+      "loss": 0.0757,
+      "step": 12470
+    },
+    {
+      "epoch": 0.02269715080738112,
+      "grad_norm": 0.022580118849873543,
+      "learning_rate": 0.0002,
+      "loss": 0.0756,
+      "step": 12480
+    },
+    {
+      "epoch": 0.022715337626938316,
+      "grad_norm": 0.16330066323280334,
+      "learning_rate": 0.0002,
+      "loss": 0.0645,
+      "step": 12490
+    },
+    {
+      "epoch": 0.022733524446495512,
+      "grad_norm": 0.021431026980280876,
+      "learning_rate": 0.0002,
+      "loss": 0.0224,
+      "step": 12500
+    },
+    {
+      "epoch": 0.02275171126605271,
+      "grad_norm": 0.053853604942560196,
+      "learning_rate": 0.0002,
+      "loss": 0.1304,
+      "step": 12510
+    },
+    {
+      "epoch": 0.022769898085609905,
+      "grad_norm": 0.129705548286438,
+      "learning_rate": 0.0002,
+      "loss": 0.0799,
+      "step": 12520
+    },
+    {
+      "epoch": 0.0227880849051671,
+      "grad_norm": 0.027473529800772667,
+      "learning_rate": 0.0002,
+      "loss": 0.0771,
+      "step": 12530
+    },
+    {
+      "epoch": 0.0228062717247243,
+      "grad_norm": 0.2045305222272873,
+      "learning_rate": 0.0002,
+      "loss": 0.0615,
+      "step": 12540
+    },
+    {
+      "epoch": 0.022824458544281497,
+      "grad_norm": 0.041042860597372055,
+      "learning_rate": 0.0002,
+      "loss": 0.026,
+      "step": 12550
+    },
+    {
+      "epoch": 0.022842645363838694,
+      "grad_norm": 0.05624527484178543,
+      "learning_rate": 0.0002,
+      "loss": 0.1327,
+      "step": 12560
+    },
+    {
+      "epoch": 0.02286083218339589,
+      "grad_norm": 0.09647081047296524,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 12570
+    },
+    {
+      "epoch": 0.022879019002953086,
+      "grad_norm": 0.03362264856696129,
+      "learning_rate": 0.0002,
+      "loss": 0.076,
+      "step": 12580
+    },
+    {
+      "epoch": 0.022897205822510282,
+      "grad_norm": 0.1459503322839737,
+      "learning_rate": 0.0002,
+      "loss": 0.0603,
+      "step": 12590
+    },
+    {
+      "epoch": 0.02291539264206748,
+      "grad_norm": 0.025729481130838394,
+      "learning_rate": 0.0002,
+      "loss": 0.0196,
+      "step": 12600
+    },
+    {
+      "epoch": 0.022933579461624675,
+      "grad_norm": 0.19940927624702454,
+      "learning_rate": 0.0002,
+      "loss": 0.1298,
+      "step": 12610
+    },
+    {
+      "epoch": 0.02295176628118187,
+      "grad_norm": 0.13796600699424744,
+      "learning_rate": 0.0002,
+      "loss": 0.086,
+      "step": 12620
+    },
+    {
+      "epoch": 0.022969953100739068,
+      "grad_norm": 0.08884158730506897,
+      "learning_rate": 0.0002,
+      "loss": 0.0808,
+      "step": 12630
+    },
+    {
+      "epoch": 0.022988139920296264,
+      "grad_norm": 0.15814751386642456,
+      "learning_rate": 0.0002,
+      "loss": 0.0658,
+      "step": 12640
+    },
+    {
+      "epoch": 0.02300632673985346,
+      "grad_norm": 0.03503837063908577,
+      "learning_rate": 0.0002,
+      "loss": 0.0232,
+      "step": 12650
+    },
+    {
+      "epoch": 0.023024513559410657,
+      "grad_norm": 0.09701854735612869,
+      "learning_rate": 0.0002,
+      "loss": 0.136,
+      "step": 12660
+    },
+    {
+      "epoch": 0.023042700378967853,
+      "grad_norm": 0.13909977674484253,
+      "learning_rate": 0.0002,
+      "loss": 0.0839,
+      "step": 12670
+    },
+    {
+      "epoch": 0.02306088719852505,
+      "grad_norm": 0.03152406960725784,
+      "learning_rate": 0.0002,
+      "loss": 0.0733,
+      "step": 12680
+    },
+    {
+      "epoch": 0.023079074018082246,
+      "grad_norm": 0.13872750103473663,
+      "learning_rate": 0.0002,
+      "loss": 0.0604,
+      "step": 12690
+    },
+    {
+      "epoch": 0.023097260837639442,
+      "grad_norm": 0.03626656159758568,
+      "learning_rate": 0.0002,
+      "loss": 0.0234,
+      "step": 12700
+    },
+    {
+      "epoch": 0.023115447657196638,
+      "grad_norm": 0.10111619532108307,
+      "learning_rate": 0.0002,
+      "loss": 0.1507,
+      "step": 12710
+    },
+    {
+      "epoch": 0.023133634476753834,
+      "grad_norm": 0.09038366377353668,
+      "learning_rate": 0.0002,
+      "loss": 0.0839,
+      "step": 12720
+    },
+    {
+      "epoch": 0.02315182129631103,
+      "grad_norm": 0.026116544380784035,
+      "learning_rate": 0.0002,
+      "loss": 0.0777,
+      "step": 12730
+    },
+    {
+      "epoch": 0.023170008115868227,
+      "grad_norm": 0.2067679613828659,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 12740
+    },
+    {
+      "epoch": 0.023188194935425423,
+      "grad_norm": 0.02005072310566902,
+      "learning_rate": 0.0002,
+      "loss": 0.0165,
+      "step": 12750
+    },
+    {
+      "epoch": 0.02320638175498262,
+      "grad_norm": 0.03261101245880127,
+      "learning_rate": 0.0002,
+      "loss": 0.159,
+      "step": 12760
+    },
+    {
+      "epoch": 0.023224568574539816,
+      "grad_norm": 0.1416555494070053,
+      "learning_rate": 0.0002,
+      "loss": 0.0856,
+      "step": 12770
+    },
+    {
+      "epoch": 0.023242755394097012,
+      "grad_norm": 0.09400717914104462,
+      "learning_rate": 0.0002,
+      "loss": 0.0745,
+      "step": 12780
+    },
+    {
+      "epoch": 0.02326094221365421,
+      "grad_norm": 0.17093195021152496,
+      "learning_rate": 0.0002,
+      "loss": 0.0691,
+      "step": 12790
+    },
+    {
+      "epoch": 0.023279129033211405,
+      "grad_norm": 0.0209200382232666,
+      "learning_rate": 0.0002,
+      "loss": 0.0168,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0232973158527686,
+      "grad_norm": 0.10523302853107452,
+      "learning_rate": 0.0002,
+      "loss": 0.1628,
+      "step": 12810
+    },
+    {
+      "epoch": 0.023315502672325798,
+      "grad_norm": 0.06932856142520905,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 12820
+    },
+    {
+      "epoch": 0.023333689491882994,
+      "grad_norm": 0.03244032710790634,
+      "learning_rate": 0.0002,
+      "loss": 0.0699,
+      "step": 12830
+    },
+    {
+      "epoch": 0.02335187631144019,
+      "grad_norm": 0.13403338193893433,
+      "learning_rate": 0.0002,
+      "loss": 0.0619,
+      "step": 12840
+    },
+    {
+      "epoch": 0.023370063130997386,
+      "grad_norm": 0.034033093601465225,
+      "learning_rate": 0.0002,
+      "loss": 0.0166,
+      "step": 12850
+    },
+    {
+      "epoch": 0.023388249950554583,
+      "grad_norm": 0.07277385890483856,
+      "learning_rate": 0.0002,
+      "loss": 0.1377,
+      "step": 12860
+    },
+    {
+      "epoch": 0.02340643677011178,
+      "grad_norm": 0.10873163491487503,
+      "learning_rate": 0.0002,
+      "loss": 0.0895,
+      "step": 12870
+    },
+    {
+      "epoch": 0.023424623589668975,
+      "grad_norm": 0.06244732066988945,
+      "learning_rate": 0.0002,
+      "loss": 0.0745,
+      "step": 12880
+    },
+    {
+      "epoch": 0.023442810409226175,
+      "grad_norm": 0.1937248259782791,
+      "learning_rate": 0.0002,
+      "loss": 0.0633,
+      "step": 12890
+    },
+    {
+      "epoch": 0.02346099722878337,
+      "grad_norm": 0.03432930260896683,
+      "learning_rate": 0.0002,
+      "loss": 0.0246,
+      "step": 12900
+    },
+    {
+      "epoch": 0.023479184048340568,
+      "grad_norm": 0.33358234167099,
+      "learning_rate": 0.0002,
+      "loss": 0.1249,
+      "step": 12910
+    },
+    {
+      "epoch": 0.023497370867897764,
+      "grad_norm": 0.12039615213871002,
+      "learning_rate": 0.0002,
+      "loss": 0.0734,
+      "step": 12920
+    },
+    {
+      "epoch": 0.02351555768745496,
+      "grad_norm": 0.02666555717587471,
+      "learning_rate": 0.0002,
+      "loss": 0.0849,
+      "step": 12930
+    },
+    {
+      "epoch": 0.023533744507012157,
+      "grad_norm": 0.128091961145401,
+      "learning_rate": 0.0002,
+      "loss": 0.0647,
+      "step": 12940
+    },
+    {
+      "epoch": 0.023551931326569353,
+      "grad_norm": 0.030916422605514526,
+      "learning_rate": 0.0002,
+      "loss": 0.0217,
+      "step": 12950
+    },
+    {
+      "epoch": 0.02357011814612655,
+      "grad_norm": 0.09280567616224289,
+      "learning_rate": 0.0002,
+      "loss": 0.1281,
+      "step": 12960
+    },
+    {
+      "epoch": 0.023588304965683746,
+      "grad_norm": 0.09032955765724182,
+      "learning_rate": 0.0002,
+      "loss": 0.0834,
+      "step": 12970
+    },
+    {
+      "epoch": 0.023606491785240942,
+      "grad_norm": 0.3660918176174164,
+      "learning_rate": 0.0002,
+      "loss": 0.0776,
+      "step": 12980
+    },
+    {
+      "epoch": 0.02362467860479814,
+      "grad_norm": 0.15715408325195312,
+      "learning_rate": 0.0002,
+      "loss": 0.0611,
+      "step": 12990
+    },
+    {
+      "epoch": 0.023642865424355335,
+      "grad_norm": 0.03867153823375702,
+      "learning_rate": 0.0002,
+      "loss": 0.0214,
+      "step": 13000
+    },
+    {
+      "epoch": 0.02366105224391253,
+      "grad_norm": 0.37568527460098267,
+      "learning_rate": 0.0002,
+      "loss": 0.2529,
+      "step": 13010
+    },
+    {
+      "epoch": 0.023679239063469727,
+      "grad_norm": 0.14888867735862732,
+      "learning_rate": 0.0002,
+      "loss": 0.0801,
+      "step": 13020
+    },
+    {
+      "epoch": 0.023697425883026924,
+      "grad_norm": 0.04271422699093819,
+      "learning_rate": 0.0002,
+      "loss": 0.0863,
+      "step": 13030
+    },
+    {
+      "epoch": 0.02371561270258412,
+      "grad_norm": 0.190608948469162,
+      "learning_rate": 0.0002,
+      "loss": 0.0773,
+      "step": 13040
+    },
+    {
+      "epoch": 0.023733799522141316,
+      "grad_norm": 0.020333535969257355,
+      "learning_rate": 0.0002,
+      "loss": 0.0201,
+      "step": 13050
+    },
+    {
+      "epoch": 0.023751986341698512,
+      "grad_norm": 0.143577441573143,
+      "learning_rate": 0.0002,
+      "loss": 0.1709,
+      "step": 13060
+    },
+    {
+      "epoch": 0.02377017316125571,
+      "grad_norm": 0.09225071966648102,
+      "learning_rate": 0.0002,
+      "loss": 0.0854,
+      "step": 13070
+    },
+    {
+      "epoch": 0.023788359980812905,
+      "grad_norm": 0.08655473589897156,
+      "learning_rate": 0.0002,
+      "loss": 0.0727,
+      "step": 13080
+    },
+    {
+      "epoch": 0.0238065468003701,
+      "grad_norm": 0.14465250074863434,
+      "learning_rate": 0.0002,
+      "loss": 0.0632,
+      "step": 13090
+    },
+    {
+      "epoch": 0.023824733619927298,
+      "grad_norm": 0.019399341195821762,
+      "learning_rate": 0.0002,
+      "loss": 0.0204,
+      "step": 13100
+    },
+    {
+      "epoch": 0.023842920439484494,
+      "grad_norm": 0.09221036732196808,
+      "learning_rate": 0.0002,
+      "loss": 0.1646,
+      "step": 13110
+    },
+    {
+      "epoch": 0.02386110725904169,
+      "grad_norm": 0.1308157742023468,
+      "learning_rate": 0.0002,
+      "loss": 0.089,
+      "step": 13120
+    },
+    {
+      "epoch": 0.023879294078598887,
+      "grad_norm": 0.04212506487965584,
+      "learning_rate": 0.0002,
+      "loss": 0.0776,
+      "step": 13130
+    },
+    {
+      "epoch": 0.023897480898156083,
+      "grad_norm": 0.13541243970394135,
+      "learning_rate": 0.0002,
+      "loss": 0.0694,
+      "step": 13140
+    },
+    {
+      "epoch": 0.02391566771771328,
+      "grad_norm": 0.016859933733940125,
+      "learning_rate": 0.0002,
+      "loss": 0.0191,
+      "step": 13150
+    },
+    {
+      "epoch": 0.023933854537270476,
+      "grad_norm": 0.1553143709897995,
+      "learning_rate": 0.0002,
+      "loss": 0.1653,
+      "step": 13160
+    },
+    {
+      "epoch": 0.023952041356827672,
+      "grad_norm": 0.07960142940282822,
+      "learning_rate": 0.0002,
+      "loss": 0.0938,
+      "step": 13170
+    },
+    {
+      "epoch": 0.023970228176384868,
+      "grad_norm": 0.0719163790345192,
+      "learning_rate": 0.0002,
+      "loss": 0.0767,
+      "step": 13180
+    },
+    {
+      "epoch": 0.023988414995942065,
+      "grad_norm": 0.14845407009124756,
+      "learning_rate": 0.0002,
+      "loss": 0.0642,
+      "step": 13190
+    },
+    {
+      "epoch": 0.02400660181549926,
+      "grad_norm": 0.01817360520362854,
+      "learning_rate": 0.0002,
+      "loss": 0.0229,
+      "step": 13200
+    },
+    {
+      "epoch": 0.024024788635056457,
+      "grad_norm": 0.03876543045043945,
+      "learning_rate": 0.0002,
+      "loss": 0.1377,
+      "step": 13210
+    },
+    {
+      "epoch": 0.024042975454613653,
+      "grad_norm": 0.05972164496779442,
+      "learning_rate": 0.0002,
+      "loss": 0.0802,
+      "step": 13220
+    },
+    {
+      "epoch": 0.024061162274170853,
+      "grad_norm": 0.09239703416824341,
+      "learning_rate": 0.0002,
+      "loss": 0.0816,
+      "step": 13230
+    },
+    {
+      "epoch": 0.02407934909372805,
+      "grad_norm": 0.15912885963916779,
+      "learning_rate": 0.0002,
+      "loss": 0.0598,
+      "step": 13240
+    },
+    {
+      "epoch": 0.024097535913285246,
+      "grad_norm": 0.024279551580548286,
+      "learning_rate": 0.0002,
+      "loss": 0.0235,
+      "step": 13250
+    },
+    {
+      "epoch": 0.024115722732842442,
+      "grad_norm": 0.06568270921707153,
+      "learning_rate": 0.0002,
+      "loss": 0.1255,
+      "step": 13260
+    },
+    {
+      "epoch": 0.02413390955239964,
+      "grad_norm": 0.04041383042931557,
+      "learning_rate": 0.0002,
+      "loss": 0.0718,
+      "step": 13270
+    },
+    {
+      "epoch": 0.024152096371956835,
+      "grad_norm": 0.046768829226493835,
+      "learning_rate": 0.0002,
+      "loss": 0.0741,
+      "step": 13280
+    },
+    {
+      "epoch": 0.02417028319151403,
+      "grad_norm": 0.21418194472789764,
+      "learning_rate": 0.0002,
+      "loss": 0.0683,
+      "step": 13290
+    },
+    {
+      "epoch": 0.024188470011071227,
+      "grad_norm": 0.04398053511977196,
+      "learning_rate": 0.0002,
+      "loss": 0.0262,
+      "step": 13300
+    },
+    {
+      "epoch": 0.024206656830628424,
+      "grad_norm": 0.1672079861164093,
+      "learning_rate": 0.0002,
+      "loss": 0.1408,
+      "step": 13310
+    },
+    {
+      "epoch": 0.02422484365018562,
+      "grad_norm": 0.05705881491303444,
+      "learning_rate": 0.0002,
+      "loss": 0.0773,
+      "step": 13320
+    },
+    {
+      "epoch": 0.024243030469742816,
+      "grad_norm": 0.0667627677321434,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 13330
+    },
+    {
+      "epoch": 0.024261217289300013,
+      "grad_norm": 0.16610710322856903,
+      "learning_rate": 0.0002,
+      "loss": 0.0682,
+      "step": 13340
+    },
+    {
+      "epoch": 0.02427940410885721,
+      "grad_norm": 0.028300171718001366,
+      "learning_rate": 0.0002,
+      "loss": 0.0185,
+      "step": 13350
+    },
+    {
+      "epoch": 0.024297590928414405,
+      "grad_norm": 0.10226302593946457,
+      "learning_rate": 0.0002,
+      "loss": 0.1406,
+      "step": 13360
+    },
+    {
+      "epoch": 0.0243157777479716,
+      "grad_norm": 0.0939667820930481,
+      "learning_rate": 0.0002,
+      "loss": 0.0755,
+      "step": 13370
+    },
+    {
+      "epoch": 0.024333964567528798,
+      "grad_norm": 0.029998745769262314,
+      "learning_rate": 0.0002,
+      "loss": 0.0748,
+      "step": 13380
+    },
+    {
+      "epoch": 0.024352151387085994,
+      "grad_norm": 0.1240144744515419,
+      "learning_rate": 0.0002,
+      "loss": 0.0639,
+      "step": 13390
+    },
+    {
+      "epoch": 0.02437033820664319,
+      "grad_norm": 0.017499787732958794,
+      "learning_rate": 0.0002,
+      "loss": 0.0156,
+      "step": 13400
+    },
+    {
+      "epoch": 0.024388525026200387,
+      "grad_norm": 0.11781036853790283,
+      "learning_rate": 0.0002,
+      "loss": 0.1385,
+      "step": 13410
+    },
+    {
+      "epoch": 0.024406711845757583,
+      "grad_norm": 0.09330960363149643,
+      "learning_rate": 0.0002,
+      "loss": 0.0789,
+      "step": 13420
+    },
+    {
+      "epoch": 0.02442489866531478,
+      "grad_norm": 0.03347505256533623,
+      "learning_rate": 0.0002,
+      "loss": 0.0742,
+      "step": 13430
+    },
+    {
+      "epoch": 0.024443085484871976,
+      "grad_norm": 0.18877847492694855,
+      "learning_rate": 0.0002,
+      "loss": 0.0701,
+      "step": 13440
+    },
+    {
+      "epoch": 0.024461272304429172,
+      "grad_norm": 0.03831986337900162,
+      "learning_rate": 0.0002,
+      "loss": 0.0243,
+      "step": 13450
+    },
+    {
+      "epoch": 0.02447945912398637,
+      "grad_norm": 0.07360157370567322,
+      "learning_rate": 0.0002,
+      "loss": 0.1237,
+      "step": 13460
+    },
+    {
+      "epoch": 0.024497645943543565,
+      "grad_norm": 0.0442088283598423,
+      "learning_rate": 0.0002,
+      "loss": 0.0742,
+      "step": 13470
+    },
+    {
+      "epoch": 0.02451583276310076,
+      "grad_norm": 0.07053640484809875,
+      "learning_rate": 0.0002,
+      "loss": 0.0793,
+      "step": 13480
+    },
+    {
+      "epoch": 0.024534019582657957,
+      "grad_norm": 0.20134539902210236,
+      "learning_rate": 0.0002,
+      "loss": 0.0621,
+      "step": 13490
+    },
+    {
+      "epoch": 0.024552206402215154,
+      "grad_norm": 0.016353536397218704,
+      "learning_rate": 0.0002,
+      "loss": 0.0204,
+      "step": 13500
+    },
+    {
+      "epoch": 0.02457039322177235,
+      "grad_norm": 0.15373657643795013,
+      "learning_rate": 0.0002,
+      "loss": 0.1446,
+      "step": 13510
+    },
+    {
+      "epoch": 0.024588580041329546,
+      "grad_norm": 2.457998037338257,
+      "learning_rate": 0.0002,
+      "loss": 0.0959,
+      "step": 13520
+    },
+    {
+      "epoch": 0.024606766860886743,
+      "grad_norm": 0.11631426215171814,
+      "learning_rate": 0.0002,
+      "loss": 0.0718,
+      "step": 13530
+    },
+    {
+      "epoch": 0.02462495368044394,
+      "grad_norm": 0.15928395092487335,
+      "learning_rate": 0.0002,
+      "loss": 0.0638,
+      "step": 13540
+    },
+    {
+      "epoch": 0.024643140500001135,
+      "grad_norm": 0.01724998839199543,
+      "learning_rate": 0.0002,
+      "loss": 0.0127,
+      "step": 13550
+    },
+    {
+      "epoch": 0.02466132731955833,
+      "grad_norm": 0.10434440523386002,
+      "learning_rate": 0.0002,
+      "loss": 0.1676,
+      "step": 13560
+    },
+    {
+      "epoch": 0.02467951413911553,
+      "grad_norm": 0.09029936045408249,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 13570
+    },
+    {
+      "epoch": 0.024697700958672728,
+      "grad_norm": 0.07413540780544281,
+      "learning_rate": 0.0002,
+      "loss": 0.074,
+      "step": 13580
+    },
+    {
+      "epoch": 0.024715887778229924,
+      "grad_norm": 0.15171368420124054,
+      "learning_rate": 0.0002,
+      "loss": 0.0646,
+      "step": 13590
+    },
+    {
+      "epoch": 0.02473407459778712,
+      "grad_norm": 0.03615165874361992,
+      "learning_rate": 0.0002,
+      "loss": 0.0253,
+      "step": 13600
+    },
+    {
+      "epoch": 0.024752261417344316,
+      "grad_norm": 0.08074207603931427,
+      "learning_rate": 0.0002,
+      "loss": 0.1251,
+      "step": 13610
+    },
+    {
+      "epoch": 0.024770448236901513,
+      "grad_norm": 0.12725302577018738,
+      "learning_rate": 0.0002,
+      "loss": 0.0868,
+      "step": 13620
+    },
+    {
+      "epoch": 0.02478863505645871,
+      "grad_norm": 0.02872832864522934,
+      "learning_rate": 0.0002,
+      "loss": 0.072,
+      "step": 13630
+    },
+    {
+      "epoch": 0.024806821876015905,
+      "grad_norm": 0.14573116600513458,
+      "learning_rate": 0.0002,
+      "loss": 0.06,
+      "step": 13640
+    },
+    {
+      "epoch": 0.0248250086955731,
+      "grad_norm": 0.039421938359737396,
+      "learning_rate": 0.0002,
+      "loss": 0.0259,
+      "step": 13650
+    },
+    {
+      "epoch": 0.024843195515130298,
+      "grad_norm": 0.08786037564277649,
+      "learning_rate": 0.0002,
+      "loss": 0.1255,
+      "step": 13660
+    },
+    {
+      "epoch": 0.024861382334687494,
+      "grad_norm": 0.7118334174156189,
+      "learning_rate": 0.0002,
+      "loss": 0.1096,
+      "step": 13670
+    },
+    {
+      "epoch": 0.02487956915424469,
+      "grad_norm": 0.05718977376818657,
+      "learning_rate": 0.0002,
+      "loss": 0.1057,
+      "step": 13680
+    },
+    {
+      "epoch": 0.024897755973801887,
+      "grad_norm": 0.19388055801391602,
+      "learning_rate": 0.0002,
+      "loss": 0.0668,
+      "step": 13690
+    },
+    {
+      "epoch": 0.024915942793359083,
+      "grad_norm": 0.02519839070737362,
+      "learning_rate": 0.0002,
+      "loss": 0.0182,
+      "step": 13700
+    },
+    {
+      "epoch": 0.02493412961291628,
+      "grad_norm": 0.15939857065677643,
+      "learning_rate": 0.0002,
+      "loss": 0.1685,
+      "step": 13710
+    },
+    {
+      "epoch": 0.024952316432473476,
+      "grad_norm": 0.07893367856740952,
+      "learning_rate": 0.0002,
+      "loss": 0.0781,
+      "step": 13720
+    },
+    {
+      "epoch": 0.024970503252030672,
+      "grad_norm": 0.0573757067322731,
+      "learning_rate": 0.0002,
+      "loss": 0.0819,
+      "step": 13730
+    },
+    {
+      "epoch": 0.02498869007158787,
+      "grad_norm": 0.1089317575097084,
+      "learning_rate": 0.0002,
+      "loss": 0.0645,
+      "step": 13740
+    },
+    {
+      "epoch": 0.025006876891145065,
+      "grad_norm": 0.03239568695425987,
+      "learning_rate": 0.0002,
+      "loss": 0.0199,
+      "step": 13750
+    },
+    {
+      "epoch": 0.02502506371070226,
+      "grad_norm": 0.04015114903450012,
+      "learning_rate": 0.0002,
+      "loss": 0.146,
+      "step": 13760
+    },
+    {
+      "epoch": 0.025043250530259457,
+      "grad_norm": 0.15218386054039001,
+      "learning_rate": 0.0002,
+      "loss": 0.0854,
+      "step": 13770
+    },
+    {
+      "epoch": 0.025061437349816654,
+      "grad_norm": 0.04461386427283287,
+      "learning_rate": 0.0002,
+      "loss": 0.0734,
+      "step": 13780
+    },
+    {
+      "epoch": 0.02507962416937385,
+      "grad_norm": 0.17443357408046722,
+      "learning_rate": 0.0002,
+      "loss": 0.0677,
+      "step": 13790
+    },
+    {
+      "epoch": 0.025097810988931046,
+      "grad_norm": 1.0899302959442139,
+      "learning_rate": 0.0002,
+      "loss": 0.0312,
+      "step": 13800
+    },
+    {
+      "epoch": 0.025115997808488243,
+      "grad_norm": 0.04115718603134155,
+      "learning_rate": 0.0002,
+      "loss": 0.1392,
+      "step": 13810
+    },
+    {
+      "epoch": 0.02513418462804544,
+      "grad_norm": 0.06605038046836853,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 13820
+    },
+    {
+      "epoch": 0.025152371447602635,
+      "grad_norm": 0.115416020154953,
+      "learning_rate": 0.0002,
+      "loss": 0.0709,
+      "step": 13830
+    },
+    {
+      "epoch": 0.02517055826715983,
+      "grad_norm": 0.1582881212234497,
+      "learning_rate": 0.0002,
+      "loss": 0.066,
+      "step": 13840
+    },
+    {
+      "epoch": 0.025188745086717028,
+      "grad_norm": 0.037643156945705414,
+      "learning_rate": 0.0002,
+      "loss": 0.0226,
+      "step": 13850
+    },
+    {
+      "epoch": 0.025206931906274224,
+      "grad_norm": 0.08343279361724854,
+      "learning_rate": 0.0002,
+      "loss": 0.1197,
+      "step": 13860
+    },
+    {
+      "epoch": 0.02522511872583142,
+      "grad_norm": 0.13482169806957245,
+      "learning_rate": 0.0002,
+      "loss": 0.0799,
+      "step": 13870
+    },
+    {
+      "epoch": 0.025243305545388617,
+      "grad_norm": 0.10373103618621826,
+      "learning_rate": 0.0002,
+      "loss": 0.075,
+      "step": 13880
+    },
+    {
+      "epoch": 0.025261492364945813,
+      "grad_norm": 0.1348303109407425,
+      "learning_rate": 0.0002,
+      "loss": 0.0603,
+      "step": 13890
+    },
+    {
+      "epoch": 0.02527967918450301,
+      "grad_norm": 0.058479245752096176,
+      "learning_rate": 0.0002,
+      "loss": 0.0252,
+      "step": 13900
+    },
+    {
+      "epoch": 0.025297866004060206,
+      "grad_norm": 0.19177350401878357,
+      "learning_rate": 0.0002,
+      "loss": 0.122,
+      "step": 13910
+    },
+    {
+      "epoch": 0.025316052823617406,
+      "grad_norm": 0.11044300347566605,
+      "learning_rate": 0.0002,
+      "loss": 0.0724,
+      "step": 13920
+    },
+    {
+      "epoch": 0.025334239643174602,
+      "grad_norm": 0.05279375612735748,
+      "learning_rate": 0.0002,
+      "loss": 0.0836,
+      "step": 13930
+    },
+    {
+      "epoch": 0.025352426462731798,
+      "grad_norm": 0.12162257730960846,
+      "learning_rate": 0.0002,
+      "loss": 0.0615,
+      "step": 13940
+    },
+    {
+      "epoch": 0.025370613282288994,
+      "grad_norm": 0.026728983968496323,
+      "learning_rate": 0.0002,
+      "loss": 0.0207,
+      "step": 13950
+    },
+    {
+      "epoch": 0.02538880010184619,
+      "grad_norm": 0.08440329879522324,
+      "learning_rate": 0.0002,
+      "loss": 0.1171,
+      "step": 13960
+    },
+    {
+      "epoch": 0.025406986921403387,
+      "grad_norm": 0.10090481489896774,
+      "learning_rate": 0.0002,
+      "loss": 0.0851,
+      "step": 13970
+    },
+    {
+      "epoch": 0.025425173740960583,
+      "grad_norm": 0.03063822351396084,
+      "learning_rate": 0.0002,
+      "loss": 0.0783,
+      "step": 13980
+    },
+    {
+      "epoch": 0.02544336056051778,
+      "grad_norm": 0.14754973351955414,
+      "learning_rate": 0.0002,
+      "loss": 0.0662,
+      "step": 13990
+    },
+    {
+      "epoch": 0.025461547380074976,
+      "grad_norm": 0.04844941198825836,
+      "learning_rate": 0.0002,
+      "loss": 0.0204,
+      "step": 14000
+    },
+    {
+      "epoch": 0.025479734199632172,
+      "grad_norm": 0.08291894942522049,
+      "learning_rate": 0.0002,
+      "loss": 0.13,
+      "step": 14010
+    },
+    {
+      "epoch": 0.02549792101918937,
+      "grad_norm": 0.05875542387366295,
+      "learning_rate": 0.0002,
+      "loss": 0.0732,
+      "step": 14020
+    },
+    {
+      "epoch": 0.025516107838746565,
+      "grad_norm": 0.04103298857808113,
+      "learning_rate": 0.0002,
+      "loss": 0.0796,
+      "step": 14030
+    },
+    {
+      "epoch": 0.02553429465830376,
+      "grad_norm": 0.20349934697151184,
+      "learning_rate": 0.0002,
+      "loss": 0.0672,
+      "step": 14040
+    },
+    {
+      "epoch": 0.025552481477860958,
+      "grad_norm": 0.05419473722577095,
+      "learning_rate": 0.0002,
+      "loss": 0.0231,
+      "step": 14050
+    },
+    {
+      "epoch": 0.025570668297418154,
+      "grad_norm": 0.05501960590481758,
+      "learning_rate": 0.0002,
+      "loss": 0.1281,
+      "step": 14060
+    },
+    {
+      "epoch": 0.02558885511697535,
+      "grad_norm": 0.07140739262104034,
+      "learning_rate": 0.0002,
+      "loss": 0.0746,
+      "step": 14070
+    },
+    {
+      "epoch": 0.025607041936532546,
+      "grad_norm": 0.04564960300922394,
+      "learning_rate": 0.0002,
+      "loss": 0.0746,
+      "step": 14080
+    },
+    {
+      "epoch": 0.025625228756089743,
+      "grad_norm": 0.16987308859825134,
+      "learning_rate": 0.0002,
+      "loss": 0.0642,
+      "step": 14090
+    },
+    {
+      "epoch": 0.02564341557564694,
+      "grad_norm": 0.017460890114307404,
+      "learning_rate": 0.0002,
+      "loss": 0.0218,
+      "step": 14100
+    },
+    {
+      "epoch": 0.025661602395204135,
+      "grad_norm": 0.15666340291500092,
+      "learning_rate": 0.0002,
+      "loss": 0.1572,
+      "step": 14110
+    },
+    {
+      "epoch": 0.02567978921476133,
+      "grad_norm": 0.06847309321165085,
+      "learning_rate": 0.0002,
+      "loss": 0.0744,
+      "step": 14120
+    },
+    {
+      "epoch": 0.025697976034318528,
+      "grad_norm": 0.03678276389837265,
+      "learning_rate": 0.0002,
+      "loss": 0.0718,
+      "step": 14130
+    },
+    {
+      "epoch": 0.025716162853875724,
+      "grad_norm": 0.1861123889684677,
+      "learning_rate": 0.0002,
+      "loss": 0.06,
+      "step": 14140
+    },
+    {
+      "epoch": 0.02573434967343292,
+      "grad_norm": 0.010294788517057896,
+      "learning_rate": 0.0002,
+      "loss": 0.0183,
+      "step": 14150
+    },
+    {
+      "epoch": 0.025752536492990117,
+      "grad_norm": 0.0643458440899849,
+      "learning_rate": 0.0002,
+      "loss": 0.1594,
+      "step": 14160
+    },
+    {
+      "epoch": 0.025770723312547313,
+      "grad_norm": 0.10639938712120056,
+      "learning_rate": 0.0002,
+      "loss": 0.0761,
+      "step": 14170
+    },
+    {
+      "epoch": 0.02578891013210451,
+      "grad_norm": 0.056529924273490906,
+      "learning_rate": 0.0002,
+      "loss": 0.082,
+      "step": 14180
+    },
+    {
+      "epoch": 0.025807096951661706,
+      "grad_norm": 0.18884658813476562,
+      "learning_rate": 0.0002,
+      "loss": 0.0683,
+      "step": 14190
+    },
+    {
+      "epoch": 0.025825283771218902,
+      "grad_norm": 0.035667784512043,
+      "learning_rate": 0.0002,
+      "loss": 0.0263,
+      "step": 14200
+    },
+    {
+      "epoch": 0.0258434705907761,
+      "grad_norm": 0.14650103449821472,
+      "learning_rate": 0.0002,
+      "loss": 0.1314,
+      "step": 14210
+    },
+    {
+      "epoch": 0.025861657410333295,
+      "grad_norm": 0.12219654768705368,
+      "learning_rate": 0.0002,
+      "loss": 0.0755,
+      "step": 14220
+    },
+    {
+      "epoch": 0.02587984422989049,
+      "grad_norm": 0.05271647870540619,
+      "learning_rate": 0.0002,
+      "loss": 0.0789,
+      "step": 14230
+    },
+    {
+      "epoch": 0.025898031049447687,
+      "grad_norm": 0.1669916957616806,
+      "learning_rate": 0.0002,
+      "loss": 0.0641,
+      "step": 14240
+    },
+    {
+      "epoch": 0.025916217869004884,
+      "grad_norm": 0.035175371915102005,
+      "learning_rate": 0.0002,
+      "loss": 0.0222,
+      "step": 14250
+    },
+    {
+      "epoch": 0.025934404688562084,
+      "grad_norm": 0.14658409357070923,
+      "learning_rate": 0.0002,
+      "loss": 0.1382,
+      "step": 14260
+    },
+    {
+      "epoch": 0.02595259150811928,
+      "grad_norm": 0.07525639981031418,
+      "learning_rate": 0.0002,
+      "loss": 0.0815,
+      "step": 14270
+    },
+    {
+      "epoch": 0.025970778327676476,
+      "grad_norm": 0.02428872510790825,
+      "learning_rate": 0.0002,
+      "loss": 0.076,
+      "step": 14280
+    },
+    {
+      "epoch": 0.025988965147233672,
+      "grad_norm": 0.1825665533542633,
+      "learning_rate": 0.0002,
+      "loss": 0.0652,
+      "step": 14290
+    },
+    {
+      "epoch": 0.02600715196679087,
+      "grad_norm": 0.033867619931697845,
+      "learning_rate": 0.0002,
+      "loss": 0.0206,
+      "step": 14300
+    },
+    {
+      "epoch": 0.026025338786348065,
+      "grad_norm": 0.051891107112169266,
+      "learning_rate": 0.0002,
+      "loss": 0.1576,
+      "step": 14310
+    },
+    {
+      "epoch": 0.02604352560590526,
+      "grad_norm": 0.1111353188753128,
+      "learning_rate": 0.0002,
+      "loss": 0.0889,
+      "step": 14320
+    },
+    {
+      "epoch": 0.026061712425462458,
+      "grad_norm": 0.04253942146897316,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 14330
+    },
+    {
+      "epoch": 0.026079899245019654,
+      "grad_norm": 0.17151106894016266,
+      "learning_rate": 0.0002,
+      "loss": 0.0678,
+      "step": 14340
+    },
+    {
+      "epoch": 0.02609808606457685,
+      "grad_norm": 0.03877005726099014,
+      "learning_rate": 0.0002,
+      "loss": 0.0206,
+      "step": 14350
+    },
+    {
+      "epoch": 0.026116272884134047,
+      "grad_norm": 0.03517235442996025,
+      "learning_rate": 0.0002,
+      "loss": 0.1343,
+      "step": 14360
+    },
+    {
+      "epoch": 0.026134459703691243,
+      "grad_norm": 0.08157488703727722,
+      "learning_rate": 0.0002,
+      "loss": 0.0764,
+      "step": 14370
+    },
+    {
+      "epoch": 0.02615264652324844,
+      "grad_norm": 0.03245632350444794,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 14380
+    },
+    {
+      "epoch": 0.026170833342805636,
+      "grad_norm": 0.20079655945301056,
+      "learning_rate": 0.0002,
+      "loss": 0.0691,
+      "step": 14390
+    },
+    {
+      "epoch": 0.026189020162362832,
+      "grad_norm": 0.03477077558636665,
+      "learning_rate": 0.0002,
+      "loss": 0.0232,
+      "step": 14400
+    },
+    {
+      "epoch": 0.026207206981920028,
+      "grad_norm": 0.14853888750076294,
+      "learning_rate": 0.0002,
+      "loss": 0.1436,
+      "step": 14410
+    },
+    {
+      "epoch": 0.026225393801477224,
+      "grad_norm": 0.12416905164718628,
+      "learning_rate": 0.0002,
+      "loss": 0.0755,
+      "step": 14420
+    },
+    {
+      "epoch": 0.02624358062103442,
+      "grad_norm": 0.03126871958374977,
+      "learning_rate": 0.0002,
+      "loss": 0.0762,
+      "step": 14430
+    },
+    {
+      "epoch": 0.026261767440591617,
+      "grad_norm": 0.20726743340492249,
+      "learning_rate": 0.0002,
+      "loss": 0.0614,
+      "step": 14440
+    },
+    {
+      "epoch": 0.026279954260148813,
+      "grad_norm": 0.039617493748664856,
+      "learning_rate": 0.0002,
+      "loss": 0.0181,
+      "step": 14450
+    },
+    {
+      "epoch": 0.02629814107970601,
+      "grad_norm": 0.08146277070045471,
+      "learning_rate": 0.0002,
+      "loss": 0.132,
+      "step": 14460
+    },
+    {
+      "epoch": 0.026316327899263206,
+      "grad_norm": 0.07181694358587265,
+      "learning_rate": 0.0002,
+      "loss": 0.0706,
+      "step": 14470
+    },
+    {
+      "epoch": 0.026334514718820402,
+      "grad_norm": 0.04080040752887726,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 14480
+    },
+    {
+      "epoch": 0.0263527015383776,
+      "grad_norm": 0.1903056502342224,
+      "learning_rate": 0.0002,
+      "loss": 0.0647,
+      "step": 14490
+    },
+    {
+      "epoch": 0.026370888357934795,
+      "grad_norm": 0.027256207540631294,
+      "learning_rate": 0.0002,
+      "loss": 0.0202,
+      "step": 14500
+    },
+    {
+      "epoch": 0.02638907517749199,
+      "grad_norm": 0.1434287130832672,
+      "learning_rate": 0.0002,
+      "loss": 0.1262,
+      "step": 14510
+    },
+    {
+      "epoch": 0.026407261997049188,
+      "grad_norm": 0.06977452337741852,
+      "learning_rate": 0.0002,
+      "loss": 0.0722,
+      "step": 14520
+    },
+    {
+      "epoch": 0.026425448816606384,
+      "grad_norm": 0.03453589975833893,
+      "learning_rate": 0.0002,
+      "loss": 0.0778,
+      "step": 14530
+    },
+    {
+      "epoch": 0.02644363563616358,
+      "grad_norm": 0.1455768346786499,
+      "learning_rate": 0.0002,
+      "loss": 0.0678,
+      "step": 14540
+    },
+    {
+      "epoch": 0.026461822455720777,
+      "grad_norm": 0.02977900207042694,
+      "learning_rate": 0.0002,
+      "loss": 0.0227,
+      "step": 14550
+    },
+    {
+      "epoch": 0.026480009275277973,
+      "grad_norm": 0.06667467951774597,
+      "learning_rate": 0.0002,
+      "loss": 0.1345,
+      "step": 14560
+    },
+    {
+      "epoch": 0.02649819609483517,
+      "grad_norm": 0.05125528201460838,
+      "learning_rate": 0.0002,
+      "loss": 0.078,
+      "step": 14570
+    },
+    {
+      "epoch": 0.026516382914392365,
+      "grad_norm": 0.02796974405646324,
+      "learning_rate": 0.0002,
+      "loss": 0.0782,
+      "step": 14580
+    },
+    {
+      "epoch": 0.026534569733949562,
+      "grad_norm": 0.18518763780593872,
+      "learning_rate": 0.0002,
+      "loss": 0.0722,
+      "step": 14590
+    },
+    {
+      "epoch": 0.02655275655350676,
+      "grad_norm": 0.01827179454267025,
+      "learning_rate": 0.0002,
+      "loss": 0.0193,
+      "step": 14600
+    },
+    {
+      "epoch": 0.026570943373063958,
+      "grad_norm": 0.1146678775548935,
+      "learning_rate": 0.0002,
+      "loss": 0.1651,
+      "step": 14610
+    },
+    {
+      "epoch": 0.026589130192621154,
+      "grad_norm": 3.385193109512329,
+      "learning_rate": 0.0002,
+      "loss": 0.2165,
+      "step": 14620
+    },
+    {
+      "epoch": 0.02660731701217835,
+      "grad_norm": 0.3052279055118561,
+      "learning_rate": 0.0002,
+      "loss": 0.1489,
+      "step": 14630
+    },
+    {
+      "epoch": 0.026625503831735547,
+      "grad_norm": 0.12762853503227234,
+      "learning_rate": 0.0002,
+      "loss": 0.0693,
+      "step": 14640
+    },
+    {
+      "epoch": 0.026643690651292743,
+      "grad_norm": 0.003925936296582222,
+      "learning_rate": 0.0002,
+      "loss": 0.0078,
+      "step": 14650
+    },
+    {
+      "epoch": 0.02666187747084994,
+      "grad_norm": 0.28632932901382446,
+      "learning_rate": 0.0002,
+      "loss": 0.2533,
+      "step": 14660
+    },
+    {
+      "epoch": 0.026680064290407136,
+      "grad_norm": 0.037552788853645325,
+      "learning_rate": 0.0002,
+      "loss": 0.0852,
+      "step": 14670
+    },
+    {
+      "epoch": 0.026698251109964332,
+      "grad_norm": 0.0911126434803009,
+      "learning_rate": 0.0002,
+      "loss": 0.0751,
+      "step": 14680
+    },
+    {
+      "epoch": 0.02671643792952153,
+      "grad_norm": 0.18434865772724152,
+      "learning_rate": 0.0002,
+      "loss": 0.084,
+      "step": 14690
+    },
+    {
+      "epoch": 0.026734624749078725,
+      "grad_norm": 0.03813793510198593,
+      "learning_rate": 0.0002,
+      "loss": 0.0165,
+      "step": 14700
+    },
+    {
+      "epoch": 0.02675281156863592,
+      "grad_norm": 0.04764392226934433,
+      "learning_rate": 0.0002,
+      "loss": 0.1642,
+      "step": 14710
+    },
+    {
+      "epoch": 0.026770998388193117,
+      "grad_norm": 0.04611713066697121,
+      "learning_rate": 0.0002,
+      "loss": 0.088,
+      "step": 14720
+    },
+    {
+      "epoch": 0.026789185207750314,
+      "grad_norm": 0.07171179354190826,
+      "learning_rate": 0.0002,
+      "loss": 0.1417,
+      "step": 14730
+    },
+    {
+      "epoch": 0.02680737202730751,
+      "grad_norm": 0.14135649800300598,
+      "learning_rate": 0.0002,
+      "loss": 0.0692,
+      "step": 14740
+    },
+    {
+      "epoch": 0.026825558846864706,
+      "grad_norm": 0.004508219193667173,
+      "learning_rate": 0.0002,
+      "loss": 0.016,
+      "step": 14750
+    },
+    {
+      "epoch": 0.026843745666421902,
+      "grad_norm": 0.09732682257890701,
+      "learning_rate": 0.0002,
+      "loss": 0.2089,
+      "step": 14760
+    },
+    {
+      "epoch": 0.0268619324859791,
+      "grad_norm": 0.12676575779914856,
+      "learning_rate": 0.0002,
+      "loss": 0.0849,
+      "step": 14770
+    },
+    {
+      "epoch": 0.026880119305536295,
+      "grad_norm": 0.0696650817990303,
+      "learning_rate": 0.0002,
+      "loss": 0.08,
+      "step": 14780
+    },
+    {
+      "epoch": 0.02689830612509349,
+      "grad_norm": 0.17883484065532684,
+      "learning_rate": 0.0002,
+      "loss": 0.0682,
+      "step": 14790
+    },
+    {
+      "epoch": 0.026916492944650688,
+      "grad_norm": 0.0567975677549839,
+      "learning_rate": 0.0002,
+      "loss": 0.0149,
+      "step": 14800
+    },
+    {
+      "epoch": 0.026934679764207884,
+      "grad_norm": 0.4884565472602844,
+      "learning_rate": 0.0002,
+      "loss": 0.6381,
+      "step": 14810
+    },
+    {
+      "epoch": 0.02695286658376508,
+      "grad_norm": 0.0742981806397438,
+      "learning_rate": 0.0002,
+      "loss": 0.0845,
+      "step": 14820
+    },
+    {
+      "epoch": 0.026971053403322277,
+      "grad_norm": 0.030466781929135323,
+      "learning_rate": 0.0002,
+      "loss": 0.0818,
+      "step": 14830
+    },
+    {
+      "epoch": 0.026989240222879473,
+      "grad_norm": 0.13108357787132263,
+      "learning_rate": 0.0002,
+      "loss": 0.065,
+      "step": 14840
+    },
+    {
+      "epoch": 0.02700742704243667,
+      "grad_norm": 0.019065184518694878,
+      "learning_rate": 0.0002,
+      "loss": 0.0168,
+      "step": 14850
+    },
+    {
+      "epoch": 0.027025613861993866,
+      "grad_norm": 0.21891777217388153,
+      "learning_rate": 0.0002,
+      "loss": 0.1456,
+      "step": 14860
+    },
+    {
+      "epoch": 0.027043800681551062,
+      "grad_norm": 0.0836934968829155,
+      "learning_rate": 0.0002,
+      "loss": 0.0781,
+      "step": 14870
+    },
+    {
+      "epoch": 0.027061987501108258,
+      "grad_norm": 0.0643845945596695,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 14880
+    },
+    {
+      "epoch": 0.027080174320665455,
+      "grad_norm": 0.27108556032180786,
+      "learning_rate": 0.0002,
+      "loss": 0.0722,
+      "step": 14890
+    },
+    {
+      "epoch": 0.02709836114022265,
+      "grad_norm": 0.008289041928946972,
+      "learning_rate": 0.0002,
+      "loss": 0.0201,
+      "step": 14900
+    },
+    {
+      "epoch": 0.027116547959779847,
+      "grad_norm": 0.03284185752272606,
+      "learning_rate": 0.0002,
+      "loss": 0.1509,
+      "step": 14910
+    },
+    {
+      "epoch": 0.027134734779337043,
+      "grad_norm": 0.051129039376974106,
+      "learning_rate": 0.0002,
+      "loss": 0.0831,
+      "step": 14920
+    },
+    {
+      "epoch": 0.02715292159889424,
+      "grad_norm": 0.046401191502809525,
+      "learning_rate": 0.0002,
+      "loss": 0.0694,
+      "step": 14930
+    },
+    {
+      "epoch": 0.027171108418451436,
+      "grad_norm": 0.19945313036441803,
+      "learning_rate": 0.0002,
+      "loss": 0.0734,
+      "step": 14940
+    },
+    {
+      "epoch": 0.027189295238008636,
+      "grad_norm": 0.03877973556518555,
+      "learning_rate": 0.0002,
+      "loss": 0.026,
+      "step": 14950
+    },
+    {
+      "epoch": 0.027207482057565832,
+      "grad_norm": 0.19090695679187775,
+      "learning_rate": 0.0002,
+      "loss": 0.136,
+      "step": 14960
+    },
+    {
+      "epoch": 0.02722566887712303,
+      "grad_norm": 0.11352288722991943,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 14970
+    },
+    {
+      "epoch": 0.027243855696680225,
+      "grad_norm": 0.055218834429979324,
+      "learning_rate": 0.0002,
+      "loss": 0.0763,
+      "step": 14980
+    },
+    {
+      "epoch": 0.02726204251623742,
+      "grad_norm": 0.1060803234577179,
+      "learning_rate": 0.0002,
+      "loss": 0.059,
+      "step": 14990
+    },
+    {
+      "epoch": 0.027280229335794617,
+      "grad_norm": 0.03370797634124756,
+      "learning_rate": 0.0002,
+      "loss": 0.0172,
+      "step": 15000
+    },
+    {
+      "epoch": 0.027298416155351814,
+      "grad_norm": 0.19884982705116272,
+      "learning_rate": 0.0002,
+      "loss": 0.1408,
+      "step": 15010
+    },
+    {
+      "epoch": 0.02731660297490901,
+      "grad_norm": 0.1186273992061615,
+      "learning_rate": 0.0002,
+      "loss": 0.0769,
+      "step": 15020
+    },
+    {
+      "epoch": 0.027334789794466206,
+      "grad_norm": 0.0494297556579113,
+      "learning_rate": 0.0002,
+      "loss": 0.0818,
+      "step": 15030
+    },
+    {
+      "epoch": 0.027352976614023403,
+      "grad_norm": 0.17990480363368988,
+      "learning_rate": 0.0002,
+      "loss": 0.06,
+      "step": 15040
+    },
+    {
+      "epoch": 0.0273711634335806,
+      "grad_norm": 0.015269913710653782,
+      "learning_rate": 0.0002,
+      "loss": 0.0143,
+      "step": 15050
+    },
+    {
+      "epoch": 0.027389350253137795,
+      "grad_norm": 0.1387794464826584,
+      "learning_rate": 0.0002,
+      "loss": 0.171,
+      "step": 15060
+    },
+    {
+      "epoch": 0.02740753707269499,
+      "grad_norm": 0.11648393422365189,
+      "learning_rate": 0.0002,
+      "loss": 0.0769,
+      "step": 15070
+    },
+    {
+      "epoch": 0.027425723892252188,
+      "grad_norm": 0.04039733111858368,
+      "learning_rate": 0.0002,
+      "loss": 0.0707,
+      "step": 15080
+    },
+    {
+      "epoch": 0.027443910711809384,
+      "grad_norm": 0.19274230301380157,
+      "learning_rate": 0.0002,
+      "loss": 0.0657,
+      "step": 15090
+    },
+    {
+      "epoch": 0.02746209753136658,
+      "grad_norm": 0.03266929090023041,
+      "learning_rate": 0.0002,
+      "loss": 0.0155,
+      "step": 15100
+    },
+    {
+      "epoch": 0.027480284350923777,
+      "grad_norm": 0.44524702429771423,
+      "learning_rate": 0.0002,
+      "loss": 0.3075,
+      "step": 15110
+    },
+    {
+      "epoch": 0.027498471170480973,
+      "grad_norm": 0.15604422986507416,
+      "learning_rate": 0.0002,
+      "loss": 0.0874,
+      "step": 15120
+    },
+    {
+      "epoch": 0.02751665799003817,
+      "grad_norm": 0.043061114847660065,
+      "learning_rate": 0.0002,
+      "loss": 0.0814,
+      "step": 15130
+    },
+    {
+      "epoch": 0.027534844809595366,
+      "grad_norm": 0.2331482172012329,
+      "learning_rate": 0.0002,
+      "loss": 0.0638,
+      "step": 15140
+    },
+    {
+      "epoch": 0.027553031629152562,
+      "grad_norm": 0.011037157848477364,
+      "learning_rate": 0.0002,
+      "loss": 0.0197,
+      "step": 15150
+    },
+    {
+      "epoch": 0.02757121844870976,
+      "grad_norm": 0.0758776143193245,
+      "learning_rate": 0.0002,
+      "loss": 0.1481,
+      "step": 15160
+    },
+    {
+      "epoch": 0.027589405268266955,
+      "grad_norm": 0.18878699839115143,
+      "learning_rate": 0.0002,
+      "loss": 0.083,
+      "step": 15170
+    },
+    {
+      "epoch": 0.02760759208782415,
+      "grad_norm": 0.042469121515750885,
+      "learning_rate": 0.0002,
+      "loss": 0.0799,
+      "step": 15180
+    },
+    {
+      "epoch": 0.027625778907381347,
+      "grad_norm": 0.1603335440158844,
+      "learning_rate": 0.0002,
+      "loss": 0.0579,
+      "step": 15190
+    },
+    {
+      "epoch": 0.027643965726938544,
+      "grad_norm": 0.03533349186182022,
+      "learning_rate": 0.0002,
+      "loss": 0.0195,
+      "step": 15200
+    },
+    {
+      "epoch": 0.02766215254649574,
+      "grad_norm": 0.2014724314212799,
+      "learning_rate": 0.0002,
+      "loss": 0.1443,
+      "step": 15210
+    },
+    {
+      "epoch": 0.027680339366052936,
+      "grad_norm": 0.04604899883270264,
+      "learning_rate": 0.0002,
+      "loss": 0.0701,
+      "step": 15220
+    },
+    {
+      "epoch": 0.027698526185610133,
+      "grad_norm": 0.04726789519190788,
+      "learning_rate": 0.0002,
+      "loss": 0.078,
+      "step": 15230
+    },
+    {
+      "epoch": 0.02771671300516733,
+      "grad_norm": 0.16189764440059662,
+      "learning_rate": 0.0002,
+      "loss": 0.0686,
+      "step": 15240
+    },
+    {
+      "epoch": 0.027734899824724525,
+      "grad_norm": 0.018077973276376724,
+      "learning_rate": 0.0002,
+      "loss": 0.0155,
+      "step": 15250
+    },
+    {
+      "epoch": 0.02775308664428172,
+      "grad_norm": 0.09486963599920273,
+      "learning_rate": 0.0002,
+      "loss": 0.1695,
+      "step": 15260
+    },
+    {
+      "epoch": 0.027771273463838918,
+      "grad_norm": 0.19950449466705322,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 15270
+    },
+    {
+      "epoch": 0.027789460283396114,
+      "grad_norm": 0.03350493311882019,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 15280
+    },
+    {
+      "epoch": 0.027807647102953314,
+      "grad_norm": 0.14408868551254272,
+      "learning_rate": 0.0002,
+      "loss": 0.0624,
+      "step": 15290
+    },
+    {
+      "epoch": 0.02782583392251051,
+      "grad_norm": 0.03824521601200104,
+      "learning_rate": 0.0002,
+      "loss": 0.0182,
+      "step": 15300
+    },
+    {
+      "epoch": 0.027844020742067706,
+      "grad_norm": 0.051167964935302734,
+      "learning_rate": 0.0002,
+      "loss": 0.1342,
+      "step": 15310
+    },
+    {
+      "epoch": 0.027862207561624903,
+      "grad_norm": 0.08440420031547546,
+      "learning_rate": 0.0002,
+      "loss": 0.0775,
+      "step": 15320
+    },
+    {
+      "epoch": 0.0278803943811821,
+      "grad_norm": 0.05162487551569939,
+      "learning_rate": 0.0002,
+      "loss": 0.0824,
+      "step": 15330
+    },
+    {
+      "epoch": 0.027898581200739295,
+      "grad_norm": 0.1576220989227295,
+      "learning_rate": 0.0002,
+      "loss": 0.0607,
+      "step": 15340
+    },
+    {
+      "epoch": 0.02791676802029649,
+      "grad_norm": 0.03840797394514084,
+      "learning_rate": 0.0002,
+      "loss": 0.0197,
+      "step": 15350
+    },
+    {
+      "epoch": 0.027934954839853688,
+      "grad_norm": 0.1418246179819107,
+      "learning_rate": 0.0002,
+      "loss": 0.151,
+      "step": 15360
+    },
+    {
+      "epoch": 0.027953141659410884,
+      "grad_norm": 0.07326096296310425,
+      "learning_rate": 0.0002,
+      "loss": 0.0764,
+      "step": 15370
+    },
+    {
+      "epoch": 0.02797132847896808,
+      "grad_norm": 0.0582844614982605,
+      "learning_rate": 0.0002,
+      "loss": 0.0745,
+      "step": 15380
+    },
+    {
+      "epoch": 0.027989515298525277,
+      "grad_norm": 0.2234935164451599,
+      "learning_rate": 0.0002,
+      "loss": 0.0687,
+      "step": 15390
+    },
+    {
+      "epoch": 0.028007702118082473,
+      "grad_norm": 0.04384669288992882,
+      "learning_rate": 0.0002,
+      "loss": 0.023,
+      "step": 15400
+    },
+    {
+      "epoch": 0.02802588893763967,
+      "grad_norm": 0.14306089282035828,
+      "learning_rate": 0.0002,
+      "loss": 0.1477,
+      "step": 15410
+    },
+    {
+      "epoch": 0.028044075757196866,
+      "grad_norm": 0.1326105296611786,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 15420
+    },
+    {
+      "epoch": 0.028062262576754062,
+      "grad_norm": 0.05531894043087959,
+      "learning_rate": 0.0002,
+      "loss": 0.0813,
+      "step": 15430
+    },
+    {
+      "epoch": 0.02808044939631126,
+      "grad_norm": 0.14875297248363495,
+      "learning_rate": 0.0002,
+      "loss": 0.0622,
+      "step": 15440
+    },
+    {
+      "epoch": 0.028098636215868455,
+      "grad_norm": 0.03749268501996994,
+      "learning_rate": 0.0002,
+      "loss": 0.0181,
+      "step": 15450
+    },
+    {
+      "epoch": 0.02811682303542565,
+      "grad_norm": 0.05747106671333313,
+      "learning_rate": 0.0002,
+      "loss": 0.1157,
+      "step": 15460
+    },
+    {
+      "epoch": 0.028135009854982847,
+      "grad_norm": 0.06197863444685936,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 15470
+    },
+    {
+      "epoch": 0.028153196674540044,
+      "grad_norm": 0.09997677057981491,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 15480
+    },
+    {
+      "epoch": 0.02817138349409724,
+      "grad_norm": 0.18067684769630432,
+      "learning_rate": 0.0002,
+      "loss": 0.0728,
+      "step": 15490
+    },
+    {
+      "epoch": 0.028189570313654436,
+      "grad_norm": 0.03378088399767876,
+      "learning_rate": 0.0002,
+      "loss": 0.0252,
+      "step": 15500
+    },
+    {
+      "epoch": 0.028207757133211633,
+      "grad_norm": 0.14048723876476288,
+      "learning_rate": 0.0002,
+      "loss": 0.1392,
+      "step": 15510
+    },
+    {
+      "epoch": 0.02822594395276883,
+      "grad_norm": 0.09573493152856827,
+      "learning_rate": 0.0002,
+      "loss": 0.0751,
+      "step": 15520
+    },
+    {
+      "epoch": 0.028244130772326025,
+      "grad_norm": 0.11000777781009674,
+      "learning_rate": 0.0002,
+      "loss": 0.08,
+      "step": 15530
+    },
+    {
+      "epoch": 0.02826231759188322,
+      "grad_norm": 0.17712855339050293,
+      "learning_rate": 0.0002,
+      "loss": 0.0658,
+      "step": 15540
+    },
+    {
+      "epoch": 0.028280504411440418,
+      "grad_norm": 0.0183733981102705,
+      "learning_rate": 0.0002,
+      "loss": 0.0188,
+      "step": 15550
+    },
+    {
+      "epoch": 0.028298691230997614,
+      "grad_norm": 0.15027762949466705,
+      "learning_rate": 0.0002,
+      "loss": 0.1235,
+      "step": 15560
+    },
+    {
+      "epoch": 0.02831687805055481,
+      "grad_norm": 0.10586661100387573,
+      "learning_rate": 0.0002,
+      "loss": 0.0791,
+      "step": 15570
+    },
+    {
+      "epoch": 0.028335064870112007,
+      "grad_norm": 0.031083540990948677,
+      "learning_rate": 0.0002,
+      "loss": 0.0765,
+      "step": 15580
+    },
+    {
+      "epoch": 0.028353251689669203,
+      "grad_norm": 0.12294827401638031,
+      "learning_rate": 0.0002,
+      "loss": 0.0615,
+      "step": 15590
+    },
+    {
+      "epoch": 0.0283714385092264,
+      "grad_norm": 0.03652534633874893,
+      "learning_rate": 0.0002,
+      "loss": 0.0203,
+      "step": 15600
+    },
+    {
+      "epoch": 0.028389625328783596,
+      "grad_norm": 0.046638645231723785,
+      "learning_rate": 0.0002,
+      "loss": 0.1327,
+      "step": 15610
+    },
+    {
+      "epoch": 0.028407812148340792,
+      "grad_norm": 0.07200415432453156,
+      "learning_rate": 0.0002,
+      "loss": 0.0765,
+      "step": 15620
+    },
+    {
+      "epoch": 0.028425998967897992,
+      "grad_norm": 0.040679559111595154,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 15630
+    },
+    {
+      "epoch": 0.028444185787455188,
+      "grad_norm": 0.1572960615158081,
+      "learning_rate": 0.0002,
+      "loss": 0.0637,
+      "step": 15640
+    },
+    {
+      "epoch": 0.028462372607012384,
+      "grad_norm": 0.036091506481170654,
+      "learning_rate": 0.0002,
+      "loss": 0.0266,
+      "step": 15650
+    },
+    {
+      "epoch": 0.02848055942656958,
+      "grad_norm": 0.10555437207221985,
+      "learning_rate": 0.0002,
+      "loss": 0.1093,
+      "step": 15660
+    },
+    {
+      "epoch": 0.028498746246126777,
+      "grad_norm": 0.08854329586029053,
+      "learning_rate": 0.0002,
+      "loss": 0.0741,
+      "step": 15670
+    },
+    {
+      "epoch": 0.028516933065683973,
+      "grad_norm": 0.02908560261130333,
+      "learning_rate": 0.0002,
+      "loss": 0.0732,
+      "step": 15680
+    },
+    {
+      "epoch": 0.02853511988524117,
+      "grad_norm": 0.1568380743265152,
+      "learning_rate": 0.0002,
+      "loss": 0.0586,
+      "step": 15690
+    },
+    {
+      "epoch": 0.028553306704798366,
+      "grad_norm": 0.04985487833619118,
+      "learning_rate": 0.0002,
+      "loss": 0.0247,
+      "step": 15700
+    },
+    {
+      "epoch": 0.028571493524355562,
+      "grad_norm": 0.07582605630159378,
+      "learning_rate": 0.0002,
+      "loss": 0.1196,
+      "step": 15710
+    },
+    {
+      "epoch": 0.02858968034391276,
+      "grad_norm": 0.02401849813759327,
+      "learning_rate": 0.0002,
+      "loss": 0.075,
+      "step": 15720
+    },
+    {
+      "epoch": 0.028607867163469955,
+      "grad_norm": 0.032545965164899826,
+      "learning_rate": 0.0002,
+      "loss": 0.0761,
+      "step": 15730
+    },
+    {
+      "epoch": 0.02862605398302715,
+      "grad_norm": 0.1098649650812149,
+      "learning_rate": 0.0002,
+      "loss": 0.0599,
+      "step": 15740
+    },
+    {
+      "epoch": 0.028644240802584348,
+      "grad_norm": 0.021166007965803146,
+      "learning_rate": 0.0002,
+      "loss": 0.0169,
+      "step": 15750
+    },
+    {
+      "epoch": 0.028662427622141544,
+      "grad_norm": 0.0823541134595871,
+      "learning_rate": 0.0002,
+      "loss": 0.1337,
+      "step": 15760
+    },
+    {
+      "epoch": 0.02868061444169874,
+      "grad_norm": 0.1009572371840477,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 15770
+    },
+    {
+      "epoch": 0.028698801261255937,
+      "grad_norm": 0.09160738438367844,
+      "learning_rate": 0.0002,
+      "loss": 0.0766,
+      "step": 15780
+    },
+    {
+      "epoch": 0.028716988080813133,
+      "grad_norm": 0.14419673383235931,
+      "learning_rate": 0.0002,
+      "loss": 0.0594,
+      "step": 15790
+    },
+    {
+      "epoch": 0.02873517490037033,
+      "grad_norm": 0.01628550887107849,
+      "learning_rate": 0.0002,
+      "loss": 0.0218,
+      "step": 15800
+    },
+    {
+      "epoch": 0.028753361719927525,
+      "grad_norm": 0.15207678079605103,
+      "learning_rate": 0.0002,
+      "loss": 0.1262,
+      "step": 15810
+    },
+    {
+      "epoch": 0.028771548539484722,
+      "grad_norm": 0.14951761066913605,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 15820
+    },
+    {
+      "epoch": 0.028789735359041918,
+      "grad_norm": 0.028078215196728706,
+      "learning_rate": 0.0002,
+      "loss": 0.0783,
+      "step": 15830
+    },
+    {
+      "epoch": 0.028807922178599114,
+      "grad_norm": 0.16079741716384888,
+      "learning_rate": 0.0002,
+      "loss": 0.0633,
+      "step": 15840
+    },
+    {
+      "epoch": 0.02882610899815631,
+      "grad_norm": 0.04218870773911476,
+      "learning_rate": 0.0002,
+      "loss": 0.0217,
+      "step": 15850
+    },
+    {
+      "epoch": 0.028844295817713507,
+      "grad_norm": 0.13758492469787598,
+      "learning_rate": 0.0002,
+      "loss": 0.1358,
+      "step": 15860
+    },
+    {
+      "epoch": 0.028862482637270703,
+      "grad_norm": 0.10366559028625488,
+      "learning_rate": 0.0002,
+      "loss": 0.0787,
+      "step": 15870
+    },
+    {
+      "epoch": 0.0288806694568279,
+      "grad_norm": 0.04433147609233856,
+      "learning_rate": 0.0002,
+      "loss": 0.0781,
+      "step": 15880
+    },
+    {
+      "epoch": 0.028898856276385096,
+      "grad_norm": 0.16709402203559875,
+      "learning_rate": 0.0002,
+      "loss": 0.0684,
+      "step": 15890
+    },
+    {
+      "epoch": 0.028917043095942292,
+      "grad_norm": 0.03370310738682747,
+      "learning_rate": 0.0002,
+      "loss": 0.0191,
+      "step": 15900
+    },
+    {
+      "epoch": 0.02893522991549949,
+      "grad_norm": 0.15469267964363098,
+      "learning_rate": 0.0002,
+      "loss": 0.1487,
+      "step": 15910
+    },
+    {
+      "epoch": 0.028953416735056685,
+      "grad_norm": 0.19974654912948608,
+      "learning_rate": 0.0002,
+      "loss": 0.0769,
+      "step": 15920
+    },
+    {
+      "epoch": 0.02897160355461388,
+      "grad_norm": 0.04307623952627182,
+      "learning_rate": 0.0002,
+      "loss": 0.075,
+      "step": 15930
+    },
+    {
+      "epoch": 0.028989790374171077,
+      "grad_norm": 0.21828149259090424,
+      "learning_rate": 0.0002,
+      "loss": 0.0691,
+      "step": 15940
+    },
+    {
+      "epoch": 0.029007977193728274,
+      "grad_norm": 0.0268656387925148,
+      "learning_rate": 0.0002,
+      "loss": 0.022,
+      "step": 15950
+    },
+    {
+      "epoch": 0.02902616401328547,
+      "grad_norm": 0.11213699728250504,
+      "learning_rate": 0.0002,
+      "loss": 0.1326,
+      "step": 15960
+    },
+    {
+      "epoch": 0.029044350832842666,
+      "grad_norm": 0.2018963098526001,
+      "learning_rate": 0.0002,
+      "loss": 0.0772,
+      "step": 15970
+    },
+    {
+      "epoch": 0.029062537652399866,
+      "grad_norm": 0.06034110113978386,
+      "learning_rate": 0.0002,
+      "loss": 0.0712,
+      "step": 15980
+    },
+    {
+      "epoch": 0.029080724471957062,
+      "grad_norm": 0.1817707121372223,
+      "learning_rate": 0.0002,
+      "loss": 0.0692,
+      "step": 15990
+    },
+    {
+      "epoch": 0.02909891129151426,
+      "grad_norm": 0.03466440737247467,
+      "learning_rate": 0.0002,
+      "loss": 0.0205,
+      "step": 16000
+    },
+    {
+      "epoch": 0.029117098111071455,
+      "grad_norm": 0.1375580132007599,
+      "learning_rate": 0.0002,
+      "loss": 0.1499,
+      "step": 16010
+    },
+    {
+      "epoch": 0.02913528493062865,
+      "grad_norm": 0.14308910071849823,
+      "learning_rate": 0.0002,
+      "loss": 0.083,
+      "step": 16020
+    },
+    {
+      "epoch": 0.029153471750185848,
+      "grad_norm": 0.041022926568984985,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 16030
+    },
+    {
+      "epoch": 0.029171658569743044,
+      "grad_norm": 0.1701498180627823,
+      "learning_rate": 0.0002,
+      "loss": 0.0656,
+      "step": 16040
+    },
+    {
+      "epoch": 0.02918984538930024,
+      "grad_norm": 0.023075805976986885,
+      "learning_rate": 0.0002,
+      "loss": 0.0225,
+      "step": 16050
+    },
+    {
+      "epoch": 0.029208032208857437,
+      "grad_norm": 0.05303549766540527,
+      "learning_rate": 0.0002,
+      "loss": 0.1369,
+      "step": 16060
+    },
+    {
+      "epoch": 0.029226219028414633,
+      "grad_norm": 0.044178470969200134,
+      "learning_rate": 0.0002,
+      "loss": 0.0754,
+      "step": 16070
+    },
+    {
+      "epoch": 0.02924440584797183,
+      "grad_norm": 0.03951259329915047,
+      "learning_rate": 0.0002,
+      "loss": 0.0759,
+      "step": 16080
+    },
+    {
+      "epoch": 0.029262592667529026,
+      "grad_norm": 0.13762067258358002,
+      "learning_rate": 0.0002,
+      "loss": 0.0605,
+      "step": 16090
+    },
+    {
+      "epoch": 0.029280779487086222,
+      "grad_norm": 0.021227868273854256,
+      "learning_rate": 0.0002,
+      "loss": 0.0173,
+      "step": 16100
+    },
+    {
+      "epoch": 0.029298966306643418,
+      "grad_norm": 0.19493195414543152,
+      "learning_rate": 0.0002,
+      "loss": 0.1307,
+      "step": 16110
+    },
+    {
+      "epoch": 0.029317153126200615,
+      "grad_norm": 0.09980791062116623,
+      "learning_rate": 0.0002,
+      "loss": 0.0724,
+      "step": 16120
+    },
+    {
+      "epoch": 0.02933533994575781,
+      "grad_norm": 0.08762095868587494,
+      "learning_rate": 0.0002,
+      "loss": 0.0734,
+      "step": 16130
+    },
+    {
+      "epoch": 0.029353526765315007,
+      "grad_norm": 0.14261308312416077,
+      "learning_rate": 0.0002,
+      "loss": 0.071,
+      "step": 16140
+    },
+    {
+      "epoch": 0.029371713584872203,
+      "grad_norm": 0.033154651522636414,
+      "learning_rate": 0.0002,
+      "loss": 0.0238,
+      "step": 16150
+    },
+    {
+      "epoch": 0.0293899004044294,
+      "grad_norm": 0.1422877162694931,
+      "learning_rate": 0.0002,
+      "loss": 0.1285,
+      "step": 16160
+    },
+    {
+      "epoch": 0.029408087223986596,
+      "grad_norm": 0.1342266947031021,
+      "learning_rate": 0.0002,
+      "loss": 0.0765,
+      "step": 16170
+    },
+    {
+      "epoch": 0.029426274043543792,
+      "grad_norm": 0.031525906175374985,
+      "learning_rate": 0.0002,
+      "loss": 0.0772,
+      "step": 16180
+    },
+    {
+      "epoch": 0.02944446086310099,
+      "grad_norm": 0.14790122210979462,
+      "learning_rate": 0.0002,
+      "loss": 0.0627,
+      "step": 16190
+    },
+    {
+      "epoch": 0.029462647682658185,
+      "grad_norm": 0.025354932993650436,
+      "learning_rate": 0.0002,
+      "loss": 0.0212,
+      "step": 16200
+    },
+    {
+      "epoch": 0.02948083450221538,
+      "grad_norm": 0.1287624090909958,
+      "learning_rate": 0.0002,
+      "loss": 0.1457,
+      "step": 16210
+    },
+    {
+      "epoch": 0.029499021321772578,
+      "grad_norm": 0.1079782247543335,
+      "learning_rate": 0.0002,
+      "loss": 0.0819,
+      "step": 16220
+    },
+    {
+      "epoch": 0.029517208141329774,
+      "grad_norm": 0.04884497448801994,
+      "learning_rate": 0.0002,
+      "loss": 0.0843,
+      "step": 16230
+    },
+    {
+      "epoch": 0.02953539496088697,
+      "grad_norm": 0.14452646672725677,
+      "learning_rate": 0.0002,
+      "loss": 0.0664,
+      "step": 16240
+    },
+    {
+      "epoch": 0.029553581780444167,
+      "grad_norm": 0.029236188158392906,
+      "learning_rate": 0.0002,
+      "loss": 0.0182,
+      "step": 16250
+    },
+    {
+      "epoch": 0.029571768600001363,
+      "grad_norm": 0.18048252165317535,
+      "learning_rate": 0.0002,
+      "loss": 0.1382,
+      "step": 16260
+    },
+    {
+      "epoch": 0.02958995541955856,
+      "grad_norm": 0.08402508497238159,
+      "learning_rate": 0.0002,
+      "loss": 0.078,
+      "step": 16270
+    },
+    {
+      "epoch": 0.029608142239115755,
+      "grad_norm": 0.07740433514118195,
+      "learning_rate": 0.0002,
+      "loss": 0.0776,
+      "step": 16280
+    },
+    {
+      "epoch": 0.029626329058672952,
+      "grad_norm": 0.1414123773574829,
+      "learning_rate": 0.0002,
+      "loss": 0.0611,
+      "step": 16290
+    },
+    {
+      "epoch": 0.029644515878230148,
+      "grad_norm": 0.03296574577689171,
+      "learning_rate": 0.0002,
+      "loss": 0.0228,
+      "step": 16300
+    },
+    {
+      "epoch": 0.029662702697787344,
+      "grad_norm": 0.09312735497951508,
+      "learning_rate": 0.0002,
+      "loss": 0.1213,
+      "step": 16310
+    },
+    {
+      "epoch": 0.029680889517344544,
+      "grad_norm": 0.07857484370470047,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 16320
+    },
+    {
+      "epoch": 0.02969907633690174,
+      "grad_norm": 0.0680379793047905,
+      "learning_rate": 0.0002,
+      "loss": 0.0774,
+      "step": 16330
+    },
+    {
+      "epoch": 0.029717263156458937,
+      "grad_norm": 0.18506748974323273,
+      "learning_rate": 0.0002,
+      "loss": 0.0675,
+      "step": 16340
+    },
+    {
+      "epoch": 0.029735449976016133,
+      "grad_norm": 0.029233543202280998,
+      "learning_rate": 0.0002,
+      "loss": 0.0187,
+      "step": 16350
+    },
+    {
+      "epoch": 0.02975363679557333,
+      "grad_norm": 0.1133171021938324,
+      "learning_rate": 0.0002,
+      "loss": 0.1217,
+      "step": 16360
+    },
+    {
+      "epoch": 0.029771823615130526,
+      "grad_norm": 0.06985988467931747,
+      "learning_rate": 0.0002,
+      "loss": 0.0761,
+      "step": 16370
+    },
+    {
+      "epoch": 0.029790010434687722,
+      "grad_norm": 0.13158757984638214,
+      "learning_rate": 0.0002,
+      "loss": 0.0764,
+      "step": 16380
+    },
+    {
+      "epoch": 0.02980819725424492,
+      "grad_norm": 0.19751304388046265,
+      "learning_rate": 0.0002,
+      "loss": 0.0652,
+      "step": 16390
+    },
+    {
+      "epoch": 0.029826384073802115,
+      "grad_norm": 0.019567493349313736,
+      "learning_rate": 0.0002,
+      "loss": 0.0166,
+      "step": 16400
+    },
+    {
+      "epoch": 0.02984457089335931,
+      "grad_norm": 0.1859702467918396,
+      "learning_rate": 0.0002,
+      "loss": 0.1482,
+      "step": 16410
+    },
+    {
+      "epoch": 0.029862757712916507,
+      "grad_norm": 0.03211350366473198,
+      "learning_rate": 0.0002,
+      "loss": 0.073,
+      "step": 16420
+    },
+    {
+      "epoch": 0.029880944532473704,
+      "grad_norm": 0.10664219409227371,
+      "learning_rate": 0.0002,
+      "loss": 0.075,
+      "step": 16430
+    },
+    {
+      "epoch": 0.0298991313520309,
+      "grad_norm": 0.18254978954792023,
+      "learning_rate": 0.0002,
+      "loss": 0.0666,
+      "step": 16440
+    },
+    {
+      "epoch": 0.029917318171588096,
+      "grad_norm": 0.03076091594994068,
+      "learning_rate": 0.0002,
+      "loss": 0.0217,
+      "step": 16450
+    },
+    {
+      "epoch": 0.029935504991145293,
+      "grad_norm": 0.11172248423099518,
+      "learning_rate": 0.0002,
+      "loss": 0.1115,
+      "step": 16460
+    },
+    {
+      "epoch": 0.02995369181070249,
+      "grad_norm": 0.1121174767613411,
+      "learning_rate": 0.0002,
+      "loss": 0.0838,
+      "step": 16470
+    },
+    {
+      "epoch": 0.029971878630259685,
+      "grad_norm": 0.05544061213731766,
+      "learning_rate": 0.0002,
+      "loss": 0.0773,
+      "step": 16480
+    },
+    {
+      "epoch": 0.02999006544981688,
+      "grad_norm": 0.13899610936641693,
+      "learning_rate": 0.0002,
+      "loss": 0.0648,
+      "step": 16490
+    },
+    {
+      "epoch": 0.030008252269374078,
+      "grad_norm": 0.031017031520605087,
+      "learning_rate": 0.0002,
+      "loss": 0.0205,
+      "step": 16500
+    },
+    {
+      "epoch": 0.030026439088931274,
+      "grad_norm": 0.5919166803359985,
+      "learning_rate": 0.0002,
+      "loss": 0.1454,
+      "step": 16510
+    },
+    {
+      "epoch": 0.03004462590848847,
+      "grad_norm": 2.5127646923065186,
+      "learning_rate": 0.0002,
+      "loss": 0.0925,
+      "step": 16520
+    },
+    {
+      "epoch": 0.030062812728045667,
+      "grad_norm": 0.12587642669677734,
+      "learning_rate": 0.0002,
+      "loss": 0.0896,
+      "step": 16530
+    },
+    {
+      "epoch": 0.030080999547602863,
+      "grad_norm": 0.29352524876594543,
+      "learning_rate": 0.0002,
+      "loss": 0.0692,
+      "step": 16540
+    },
+    {
+      "epoch": 0.03009918636716006,
+      "grad_norm": 0.012585405260324478,
+      "learning_rate": 0.0002,
+      "loss": 0.021,
+      "step": 16550
+    },
+    {
+      "epoch": 0.030117373186717256,
+      "grad_norm": 2.432018756866455,
+      "learning_rate": 0.0002,
+      "loss": 0.239,
+      "step": 16560
+    },
+    {
+      "epoch": 0.030135560006274452,
+      "grad_norm": 0.09337054193019867,
+      "learning_rate": 0.0002,
+      "loss": 0.0859,
+      "step": 16570
+    },
+    {
+      "epoch": 0.030153746825831648,
+      "grad_norm": 0.05135548114776611,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 16580
+    },
+    {
+      "epoch": 0.030171933645388845,
+      "grad_norm": 0.15056684613227844,
+      "learning_rate": 0.0002,
+      "loss": 0.0697,
+      "step": 16590
+    },
+    {
+      "epoch": 0.03019012046494604,
+      "grad_norm": 5.883757694391534e-05,
+      "learning_rate": 0.0002,
+      "loss": 0.0085,
+      "step": 16600
+    },
+    {
+      "epoch": 0.030208307284503237,
+      "grad_norm": 1.0368543863296509,
+      "learning_rate": 0.0002,
+      "loss": 0.1861,
+      "step": 16610
+    },
+    {
+      "epoch": 0.030226494104060433,
+      "grad_norm": 0.07987317442893982,
+      "learning_rate": 0.0002,
+      "loss": 0.0938,
+      "step": 16620
+    },
+    {
+      "epoch": 0.03024468092361763,
+      "grad_norm": 0.02812887355685234,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 16630
+    },
+    {
+      "epoch": 0.030262867743174826,
+      "grad_norm": 0.24061231315135956,
+      "learning_rate": 0.0002,
+      "loss": 0.0653,
+      "step": 16640
+    },
+    {
+      "epoch": 0.030281054562732022,
+      "grad_norm": 0.0402507558465004,
+      "learning_rate": 0.0002,
+      "loss": 0.0266,
+      "step": 16650
+    },
+    {
+      "epoch": 0.030299241382289222,
+      "grad_norm": 0.13552093505859375,
+      "learning_rate": 0.0002,
+      "loss": 0.1709,
+      "step": 16660
+    },
+    {
+      "epoch": 0.03031742820184642,
+      "grad_norm": 0.6093604564666748,
+      "learning_rate": 0.0002,
+      "loss": 0.0857,
+      "step": 16670
+    },
+    {
+      "epoch": 0.030335615021403615,
+      "grad_norm": 0.11608528345823288,
+      "learning_rate": 0.0002,
+      "loss": 0.0874,
+      "step": 16680
+    },
+    {
+      "epoch": 0.03035380184096081,
+      "grad_norm": 0.23376339673995972,
+      "learning_rate": 0.0002,
+      "loss": 0.0688,
+      "step": 16690
+    },
+    {
+      "epoch": 0.030371988660518007,
+      "grad_norm": 0.03484225273132324,
+      "learning_rate": 0.0002,
+      "loss": 0.0172,
+      "step": 16700
+    },
+    {
+      "epoch": 0.030390175480075204,
+      "grad_norm": 0.30532532930374146,
+      "learning_rate": 0.0002,
+      "loss": 0.1686,
+      "step": 16710
+    },
+    {
+      "epoch": 0.0304083622996324,
+      "grad_norm": 0.05142231658101082,
+      "learning_rate": 0.0002,
+      "loss": 0.0766,
+      "step": 16720
+    },
+    {
+      "epoch": 0.030426549119189596,
+      "grad_norm": 0.08218207955360413,
+      "learning_rate": 0.0002,
+      "loss": 0.0839,
+      "step": 16730
+    },
+    {
+      "epoch": 0.030444735938746793,
+      "grad_norm": 0.15296520292758942,
+      "learning_rate": 0.0002,
+      "loss": 0.0717,
+      "step": 16740
+    },
+    {
+      "epoch": 0.03046292275830399,
+      "grad_norm": 0.009951476007699966,
+      "learning_rate": 0.0002,
+      "loss": 0.0103,
+      "step": 16750
+    },
+    {
+      "epoch": 0.030481109577861185,
+      "grad_norm": 0.18752850592136383,
+      "learning_rate": 0.0002,
+      "loss": 0.2382,
+      "step": 16760
+    },
+    {
+      "epoch": 0.03049929639741838,
+      "grad_norm": 0.1473335325717926,
+      "learning_rate": 0.0002,
+      "loss": 0.0975,
+      "step": 16770
+    },
+    {
+      "epoch": 0.030517483216975578,
+      "grad_norm": 0.04578230902552605,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 16780
+    },
+    {
+      "epoch": 0.030535670036532774,
+      "grad_norm": 0.2557182312011719,
+      "learning_rate": 0.0002,
+      "loss": 0.0691,
+      "step": 16790
+    },
+    {
+      "epoch": 0.03055385685608997,
+      "grad_norm": 1.473021388053894,
+      "learning_rate": 0.0002,
+      "loss": 0.2088,
+      "step": 16800
+    },
+    {
+      "epoch": 0.030572043675647167,
+      "grad_norm": 1.0227181911468506,
+      "learning_rate": 0.0002,
+      "loss": 0.7207,
+      "step": 16810
+    },
+    {
+      "epoch": 0.030590230495204363,
+      "grad_norm": 0.11395780742168427,
+      "learning_rate": 0.0002,
+      "loss": 0.0943,
+      "step": 16820
+    },
+    {
+      "epoch": 0.03060841731476156,
+      "grad_norm": 6.501937389373779,
+      "learning_rate": 0.0002,
+      "loss": 0.0871,
+      "step": 16830
+    },
+    {
+      "epoch": 0.030626604134318756,
+      "grad_norm": 0.17187578976154327,
+      "learning_rate": 0.0002,
+      "loss": 0.0672,
+      "step": 16840
+    },
+    {
+      "epoch": 0.030644790953875952,
+      "grad_norm": 0.03396519273519516,
+      "learning_rate": 0.0002,
+      "loss": 0.0224,
+      "step": 16850
+    },
+    {
+      "epoch": 0.03066297777343315,
+      "grad_norm": 3.397012948989868,
+      "learning_rate": 0.0002,
+      "loss": 0.1641,
+      "step": 16860
+    },
+    {
+      "epoch": 0.030681164592990345,
+      "grad_norm": 0.44838130474090576,
+      "learning_rate": 0.0002,
+      "loss": 0.0868,
+      "step": 16870
+    },
+    {
+      "epoch": 0.03069935141254754,
+      "grad_norm": 0.08598771691322327,
+      "learning_rate": 0.0002,
+      "loss": 0.0766,
+      "step": 16880
+    },
+    {
+      "epoch": 0.030717538232104737,
+      "grad_norm": 0.15339739620685577,
+      "learning_rate": 0.0002,
+      "loss": 0.0609,
+      "step": 16890
+    },
+    {
+      "epoch": 0.030735725051661934,
+      "grad_norm": 0.04086040332913399,
+      "learning_rate": 0.0002,
+      "loss": 0.0218,
+      "step": 16900
+    },
+    {
+      "epoch": 0.03075391187121913,
+      "grad_norm": 0.40313076972961426,
+      "learning_rate": 0.0002,
+      "loss": 0.2017,
+      "step": 16910
+    },
+    {
+      "epoch": 0.030772098690776326,
+      "grad_norm": 0.2068721503019333,
+      "learning_rate": 0.0002,
+      "loss": 0.0906,
+      "step": 16920
+    },
+    {
+      "epoch": 0.030790285510333523,
+      "grad_norm": 0.12770770490169525,
+      "learning_rate": 0.0002,
+      "loss": 0.0801,
+      "step": 16930
+    },
+    {
+      "epoch": 0.03080847232989072,
+      "grad_norm": 17.294641494750977,
+      "learning_rate": 0.0002,
+      "loss": 0.0701,
+      "step": 16940
+    },
+    {
+      "epoch": 0.030826659149447915,
+      "grad_norm": 0.04612286388874054,
+      "learning_rate": 0.0002,
+      "loss": 0.0287,
+      "step": 16950
+    },
+    {
+      "epoch": 0.03084484596900511,
+      "grad_norm": 0.10311487317085266,
+      "learning_rate": 0.0002,
+      "loss": 0.136,
+      "step": 16960
+    },
+    {
+      "epoch": 0.030863032788562308,
+      "grad_norm": 0.20878446102142334,
+      "learning_rate": 0.0002,
+      "loss": 0.0886,
+      "step": 16970
+    },
+    {
+      "epoch": 0.030881219608119504,
+      "grad_norm": 1.412353515625,
+      "learning_rate": 0.0002,
+      "loss": 0.0843,
+      "step": 16980
+    },
+    {
+      "epoch": 0.0308994064276767,
+      "grad_norm": 0.27046918869018555,
+      "learning_rate": 0.0002,
+      "loss": 0.0755,
+      "step": 16990
+    },
+    {
+      "epoch": 0.030917593247233897,
+      "grad_norm": 0.5227788090705872,
+      "learning_rate": 0.0002,
+      "loss": 0.0234,
+      "step": 17000
+    },
+    {
+      "epoch": 0.030935780066791096,
+      "grad_norm": 0.16006655991077423,
+      "learning_rate": 0.0002,
+      "loss": 0.183,
+      "step": 17010
+    },
+    {
+      "epoch": 0.030953966886348293,
+      "grad_norm": 0.1297607421875,
+      "learning_rate": 0.0002,
+      "loss": 0.0868,
+      "step": 17020
+    },
+    {
+      "epoch": 0.03097215370590549,
+      "grad_norm": 11.198999404907227,
+      "learning_rate": 0.0002,
+      "loss": 0.0998,
+      "step": 17030
+    },
+    {
+      "epoch": 0.030990340525462685,
+      "grad_norm": 0.39887136220932007,
+      "learning_rate": 0.0002,
+      "loss": 0.0898,
+      "step": 17040
+    },
+    {
+      "epoch": 0.03100852734501988,
+      "grad_norm": 0.009262642823159695,
+      "learning_rate": 0.0002,
+      "loss": 0.0215,
+      "step": 17050
+    },
+    {
+      "epoch": 0.031026714164577078,
+      "grad_norm": 0.15820527076721191,
+      "learning_rate": 0.0002,
+      "loss": 0.2017,
+      "step": 17060
+    },
+    {
+      "epoch": 0.031044900984134274,
+      "grad_norm": 0.11645558476448059,
+      "learning_rate": 0.0002,
+      "loss": 0.085,
+      "step": 17070
+    },
+    {
+      "epoch": 0.03106308780369147,
+      "grad_norm": 0.03981775790452957,
+      "learning_rate": 0.0002,
+      "loss": 0.0803,
+      "step": 17080
+    },
+    {
+      "epoch": 0.031081274623248667,
+      "grad_norm": 0.1584177166223526,
+      "learning_rate": 0.0002,
+      "loss": 0.0635,
+      "step": 17090
+    },
+    {
+      "epoch": 0.031099461442805863,
+      "grad_norm": 0.0005907397717237473,
+      "learning_rate": 0.0002,
+      "loss": 0.006,
+      "step": 17100
+    },
+    {
+      "epoch": 0.03111764826236306,
+      "grad_norm": 0.05344061553478241,
+      "learning_rate": 0.0002,
+      "loss": 0.3098,
+      "step": 17110
+    },
+    {
+      "epoch": 0.031135835081920256,
+      "grad_norm": 0.05249408632516861,
+      "learning_rate": 0.0002,
+      "loss": 0.1002,
+      "step": 17120
+    },
+    {
+      "epoch": 0.031154021901477452,
+      "grad_norm": 0.04177263006567955,
+      "learning_rate": 0.0002,
+      "loss": 0.0969,
+      "step": 17130
+    },
+    {
+      "epoch": 0.03117220872103465,
+      "grad_norm": 0.18396486341953278,
+      "learning_rate": 0.0002,
+      "loss": 0.0727,
+      "step": 17140
+    },
+    {
+      "epoch": 0.031190395540591845,
+      "grad_norm": 0.0019848416559398174,
+      "learning_rate": 0.0002,
+      "loss": 0.0092,
+      "step": 17150
+    },
+    {
+      "epoch": 0.03120858236014904,
+      "grad_norm": 0.23747271299362183,
+      "learning_rate": 0.0002,
+      "loss": 0.3243,
+      "step": 17160
+    },
+    {
+      "epoch": 0.031226769179706237,
+      "grad_norm": 0.2365376353263855,
+      "learning_rate": 0.0002,
+      "loss": 0.094,
+      "step": 17170
+    },
+    {
+      "epoch": 0.031244955999263434,
+      "grad_norm": 0.21784919500350952,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 17180
+    },
+    {
+      "epoch": 0.03126314281882063,
+      "grad_norm": 0.27253153920173645,
+      "learning_rate": 0.0002,
+      "loss": 0.0748,
+      "step": 17190
+    },
+    {
+      "epoch": 0.031281329638377826,
+      "grad_norm": 0.004298684187233448,
+      "learning_rate": 0.0002,
+      "loss": 0.014,
+      "step": 17200
+    },
+    {
+      "epoch": 0.03129951645793502,
+      "grad_norm": 0.267871230840683,
+      "learning_rate": 0.0002,
+      "loss": 0.2938,
+      "step": 17210
+    },
+    {
+      "epoch": 0.03131770327749222,
+      "grad_norm": 0.1428530067205429,
+      "learning_rate": 0.0002,
+      "loss": 0.0901,
+      "step": 17220
+    },
+    {
+      "epoch": 0.031335890097049415,
+      "grad_norm": 0.10623782873153687,
+      "learning_rate": 0.0002,
+      "loss": 0.0752,
+      "step": 17230
+    },
+    {
+      "epoch": 0.03135407691660661,
+      "grad_norm": 0.2869247496128082,
+      "learning_rate": 0.0002,
+      "loss": 0.0707,
+      "step": 17240
+    },
+    {
+      "epoch": 0.03137226373616381,
+      "grad_norm": 0.011321209371089935,
+      "learning_rate": 0.0002,
+      "loss": 0.0168,
+      "step": 17250
+    },
+    {
+      "epoch": 0.031390450555721004,
+      "grad_norm": 0.09432020783424377,
+      "learning_rate": 0.0002,
+      "loss": 0.2046,
+      "step": 17260
+    },
+    {
+      "epoch": 0.0314086373752782,
+      "grad_norm": 0.190867081284523,
+      "learning_rate": 0.0002,
+      "loss": 0.0866,
+      "step": 17270
+    },
+    {
+      "epoch": 0.0314268241948354,
+      "grad_norm": 0.14274829626083374,
+      "learning_rate": 0.0002,
+      "loss": 0.0796,
+      "step": 17280
+    },
+    {
+      "epoch": 0.03144501101439259,
+      "grad_norm": 0.29910504817962646,
+      "learning_rate": 0.0002,
+      "loss": 0.0711,
+      "step": 17290
+    },
+    {
+      "epoch": 0.03146319783394979,
+      "grad_norm": 0.031730011105537415,
+      "learning_rate": 0.0002,
+      "loss": 0.0217,
+      "step": 17300
+    },
+    {
+      "epoch": 0.031481384653506986,
+      "grad_norm": 0.23042625188827515,
+      "learning_rate": 0.0002,
+      "loss": 0.1491,
+      "step": 17310
+    },
+    {
+      "epoch": 0.03149957147306418,
+      "grad_norm": 0.15560220181941986,
+      "learning_rate": 0.0002,
+      "loss": 0.0761,
+      "step": 17320
+    },
+    {
+      "epoch": 0.03151775829262138,
+      "grad_norm": 0.051929160952568054,
+      "learning_rate": 0.0002,
+      "loss": 0.0893,
+      "step": 17330
+    },
+    {
+      "epoch": 0.031535945112178575,
+      "grad_norm": 0.16162756085395813,
+      "learning_rate": 0.0002,
+      "loss": 0.0623,
+      "step": 17340
+    },
+    {
+      "epoch": 0.03155413193173577,
+      "grad_norm": 0.019480068236589432,
+      "learning_rate": 0.0002,
+      "loss": 0.0137,
+      "step": 17350
+    },
+    {
+      "epoch": 0.03157231875129297,
+      "grad_norm": 0.24700693786144257,
+      "learning_rate": 0.0002,
+      "loss": 0.1481,
+      "step": 17360
+    },
+    {
+      "epoch": 0.031590505570850164,
+      "grad_norm": 0.17574873566627502,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 17370
+    },
+    {
+      "epoch": 0.03160869239040736,
+      "grad_norm": 0.10368580371141434,
+      "learning_rate": 0.0002,
+      "loss": 0.0811,
+      "step": 17380
+    },
+    {
+      "epoch": 0.031626879209964556,
+      "grad_norm": 0.23330622911453247,
+      "learning_rate": 0.0002,
+      "loss": 0.0669,
+      "step": 17390
+    },
+    {
+      "epoch": 0.03164506602952175,
+      "grad_norm": 0.031393859535455704,
+      "learning_rate": 0.0002,
+      "loss": 0.0183,
+      "step": 17400
+    },
+    {
+      "epoch": 0.03166325284907895,
+      "grad_norm": 0.22080129384994507,
+      "learning_rate": 0.0002,
+      "loss": 0.1567,
+      "step": 17410
+    },
+    {
+      "epoch": 0.031681439668636145,
+      "grad_norm": 0.177025705575943,
+      "learning_rate": 0.0002,
+      "loss": 0.0798,
+      "step": 17420
+    },
+    {
+      "epoch": 0.03169962648819334,
+      "grad_norm": 0.054285600781440735,
+      "learning_rate": 0.0002,
+      "loss": 0.0709,
+      "step": 17430
+    },
+    {
+      "epoch": 0.03171781330775054,
+      "grad_norm": 0.20625421404838562,
+      "learning_rate": 0.0002,
+      "loss": 0.0592,
+      "step": 17440
+    },
+    {
+      "epoch": 0.031736000127307734,
+      "grad_norm": 0.042640089988708496,
+      "learning_rate": 0.0002,
+      "loss": 0.0199,
+      "step": 17450
+    },
+    {
+      "epoch": 0.03175418694686493,
+      "grad_norm": 0.2505437731742859,
+      "learning_rate": 0.0002,
+      "loss": 0.131,
+      "step": 17460
+    },
+    {
+      "epoch": 0.03177237376642213,
+      "grad_norm": 0.24848629534244537,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 17470
+    },
+    {
+      "epoch": 0.03179056058597932,
+      "grad_norm": 0.056854844093322754,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 17480
+    },
+    {
+      "epoch": 0.03180874740553652,
+      "grad_norm": 0.23022660613059998,
+      "learning_rate": 0.0002,
+      "loss": 0.0703,
+      "step": 17490
+    },
+    {
+      "epoch": 0.031826934225093716,
+      "grad_norm": 0.033501993864774704,
+      "learning_rate": 0.0002,
+      "loss": 0.0229,
+      "step": 17500
+    },
+    {
+      "epoch": 0.03184512104465091,
+      "grad_norm": 0.25061148405075073,
+      "learning_rate": 0.0002,
+      "loss": 0.1588,
+      "step": 17510
+    },
+    {
+      "epoch": 0.031863307864208115,
+      "grad_norm": 0.21534167230129242,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 17520
+    },
+    {
+      "epoch": 0.03188149468376531,
+      "grad_norm": 0.04823959991335869,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 17530
+    },
+    {
+      "epoch": 0.03189968150332251,
+      "grad_norm": 0.23680952191352844,
+      "learning_rate": 0.0002,
+      "loss": 0.0617,
+      "step": 17540
+    },
+    {
+      "epoch": 0.031917868322879704,
+      "grad_norm": 0.016636351123452187,
+      "learning_rate": 0.0002,
+      "loss": 0.0143,
+      "step": 17550
+    },
+    {
+      "epoch": 0.0319360551424369,
+      "grad_norm": 0.3684225082397461,
+      "learning_rate": 0.0002,
+      "loss": 0.2011,
+      "step": 17560
+    },
+    {
+      "epoch": 0.0319542419619941,
+      "grad_norm": 0.07126643508672714,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 17570
+    },
+    {
+      "epoch": 0.03197242878155129,
+      "grad_norm": 0.05354290455579758,
+      "learning_rate": 0.0002,
+      "loss": 0.0831,
+      "step": 17580
+    },
+    {
+      "epoch": 0.03199061560110849,
+      "grad_norm": 0.20318995416164398,
+      "learning_rate": 0.0002,
+      "loss": 0.0617,
+      "step": 17590
+    },
+    {
+      "epoch": 0.032008802420665686,
+      "grad_norm": 0.021502351388335228,
+      "learning_rate": 0.0002,
+      "loss": 0.0137,
+      "step": 17600
+    },
+    {
+      "epoch": 0.03202698924022288,
+      "grad_norm": 0.3471545875072479,
+      "learning_rate": 0.0002,
+      "loss": 0.1823,
+      "step": 17610
+    },
+    {
+      "epoch": 0.03204517605978008,
+      "grad_norm": 0.23191972076892853,
+      "learning_rate": 0.0002,
+      "loss": 0.0837,
+      "step": 17620
+    },
+    {
+      "epoch": 0.032063362879337275,
+      "grad_norm": 0.0479818731546402,
+      "learning_rate": 0.0002,
+      "loss": 0.0845,
+      "step": 17630
+    },
+    {
+      "epoch": 0.03208154969889447,
+      "grad_norm": 0.2193339467048645,
+      "learning_rate": 0.0002,
+      "loss": 0.068,
+      "step": 17640
+    },
+    {
+      "epoch": 0.03209973651845167,
+      "grad_norm": 0.03661821037530899,
+      "learning_rate": 0.0002,
+      "loss": 0.0234,
+      "step": 17650
+    },
+    {
+      "epoch": 0.032117923338008864,
+      "grad_norm": 0.10396943986415863,
+      "learning_rate": 0.0002,
+      "loss": 0.1295,
+      "step": 17660
+    },
+    {
+      "epoch": 0.03213611015756606,
+      "grad_norm": 0.16999179124832153,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 17670
+    },
+    {
+      "epoch": 0.032154296977123256,
+      "grad_norm": 0.09069819748401642,
+      "learning_rate": 0.0002,
+      "loss": 0.0748,
+      "step": 17680
+    },
+    {
+      "epoch": 0.03217248379668045,
+      "grad_norm": 0.24210433661937714,
+      "learning_rate": 0.0002,
+      "loss": 0.0611,
+      "step": 17690
+    },
+    {
+      "epoch": 0.03219067061623765,
+      "grad_norm": 0.028281020000576973,
+      "learning_rate": 0.0002,
+      "loss": 0.018,
+      "step": 17700
+    },
+    {
+      "epoch": 0.032208857435794845,
+      "grad_norm": 0.4133516252040863,
+      "learning_rate": 0.0002,
+      "loss": 0.1704,
+      "step": 17710
+    },
+    {
+      "epoch": 0.03222704425535204,
+      "grad_norm": 0.20207400619983673,
+      "learning_rate": 0.0002,
+      "loss": 0.0804,
+      "step": 17720
+    },
+    {
+      "epoch": 0.03224523107490924,
+      "grad_norm": 0.043604232370853424,
+      "learning_rate": 0.0002,
+      "loss": 0.0929,
+      "step": 17730
+    },
+    {
+      "epoch": 0.032263417894466434,
+      "grad_norm": 0.1995580494403839,
+      "learning_rate": 0.0002,
+      "loss": 0.062,
+      "step": 17740
+    },
+    {
+      "epoch": 0.03228160471402363,
+      "grad_norm": 0.03241848200559616,
+      "learning_rate": 0.0002,
+      "loss": 0.0137,
+      "step": 17750
+    },
+    {
+      "epoch": 0.03229979153358083,
+      "grad_norm": 0.28819000720977783,
+      "learning_rate": 0.0002,
+      "loss": 0.1696,
+      "step": 17760
+    },
+    {
+      "epoch": 0.03231797835313802,
+      "grad_norm": 0.2625056803226471,
+      "learning_rate": 0.0002,
+      "loss": 0.0704,
+      "step": 17770
+    },
+    {
+      "epoch": 0.03233616517269522,
+      "grad_norm": 0.03986202925443649,
+      "learning_rate": 0.0002,
+      "loss": 0.0848,
+      "step": 17780
+    },
+    {
+      "epoch": 0.032354351992252416,
+      "grad_norm": 0.24770867824554443,
+      "learning_rate": 0.0002,
+      "loss": 0.0608,
+      "step": 17790
+    },
+    {
+      "epoch": 0.03237253881180961,
+      "grad_norm": 0.031353630125522614,
+      "learning_rate": 0.0002,
+      "loss": 0.0145,
+      "step": 17800
+    },
+    {
+      "epoch": 0.03239072563136681,
+      "grad_norm": 0.2273588478565216,
+      "learning_rate": 0.0002,
+      "loss": 0.1765,
+      "step": 17810
+    },
+    {
+      "epoch": 0.032408912450924005,
+      "grad_norm": 0.19741755723953247,
+      "learning_rate": 0.0002,
+      "loss": 0.0818,
+      "step": 17820
+    },
+    {
+      "epoch": 0.0324270992704812,
+      "grad_norm": 0.03193483129143715,
+      "learning_rate": 0.0002,
+      "loss": 0.0737,
+      "step": 17830
+    },
+    {
+      "epoch": 0.0324452860900384,
+      "grad_norm": 0.13962946832180023,
+      "learning_rate": 0.0002,
+      "loss": 0.0575,
+      "step": 17840
+    },
+    {
+      "epoch": 0.03246347290959559,
+      "grad_norm": 0.01755092851817608,
+      "learning_rate": 0.0002,
+      "loss": 0.0159,
+      "step": 17850
+    },
+    {
+      "epoch": 0.03248165972915279,
+      "grad_norm": 0.21713244915008545,
+      "learning_rate": 0.0002,
+      "loss": 0.1476,
+      "step": 17860
+    },
+    {
+      "epoch": 0.032499846548709986,
+      "grad_norm": 0.15362155437469482,
+      "learning_rate": 0.0002,
+      "loss": 0.0747,
+      "step": 17870
+    },
+    {
+      "epoch": 0.03251803336826718,
+      "grad_norm": 0.02643916755914688,
+      "learning_rate": 0.0002,
+      "loss": 0.0793,
+      "step": 17880
+    },
+    {
+      "epoch": 0.03253622018782438,
+      "grad_norm": 0.2702760100364685,
+      "learning_rate": 0.0002,
+      "loss": 0.0641,
+      "step": 17890
+    },
+    {
+      "epoch": 0.032554407007381575,
+      "grad_norm": 0.05910428613424301,
+      "learning_rate": 0.0002,
+      "loss": 0.022,
+      "step": 17900
+    },
+    {
+      "epoch": 0.03257259382693877,
+      "grad_norm": 0.17692551016807556,
+      "learning_rate": 0.0002,
+      "loss": 0.1407,
+      "step": 17910
+    },
+    {
+      "epoch": 0.03259078064649597,
+      "grad_norm": 0.19877870380878448,
+      "learning_rate": 0.0002,
+      "loss": 0.0798,
+      "step": 17920
+    },
+    {
+      "epoch": 0.032608967466053164,
+      "grad_norm": 0.06731924414634705,
+      "learning_rate": 0.0002,
+      "loss": 0.0798,
+      "step": 17930
+    },
+    {
+      "epoch": 0.03262715428561036,
+      "grad_norm": 0.20342952013015747,
+      "learning_rate": 0.0002,
+      "loss": 0.0571,
+      "step": 17940
+    },
+    {
+      "epoch": 0.03264534110516756,
+      "grad_norm": 0.06299301236867905,
+      "learning_rate": 0.0002,
+      "loss": 0.0154,
+      "step": 17950
+    },
+    {
+      "epoch": 0.03266352792472475,
+      "grad_norm": 0.30317986011505127,
+      "learning_rate": 0.0002,
+      "loss": 0.1496,
+      "step": 17960
+    },
+    {
+      "epoch": 0.03268171474428195,
+      "grad_norm": 0.2737327218055725,
+      "learning_rate": 0.0002,
+      "loss": 0.0777,
+      "step": 17970
+    },
+    {
+      "epoch": 0.032699901563839145,
+      "grad_norm": 0.03226702660322189,
+      "learning_rate": 0.0002,
+      "loss": 0.0799,
+      "step": 17980
+    },
+    {
+      "epoch": 0.03271808838339634,
+      "grad_norm": 0.20195341110229492,
+      "learning_rate": 0.0002,
+      "loss": 0.0654,
+      "step": 17990
+    },
+    {
+      "epoch": 0.03273627520295354,
+      "grad_norm": 0.03351292014122009,
+      "learning_rate": 0.0002,
+      "loss": 0.0194,
+      "step": 18000
+    },
+    {
+      "epoch": 0.032754462022510734,
+      "grad_norm": 0.2281372845172882,
+      "learning_rate": 0.0002,
+      "loss": 0.154,
+      "step": 18010
+    },
+    {
+      "epoch": 0.03277264884206793,
+      "grad_norm": 0.19263891875743866,
+      "learning_rate": 0.0002,
+      "loss": 0.0803,
+      "step": 18020
+    },
+    {
+      "epoch": 0.03279083566162513,
+      "grad_norm": 0.04183288663625717,
+      "learning_rate": 0.0002,
+      "loss": 0.0842,
+      "step": 18030
+    },
+    {
+      "epoch": 0.03280902248118232,
+      "grad_norm": 0.284759521484375,
+      "learning_rate": 0.0002,
+      "loss": 0.067,
+      "step": 18040
+    },
+    {
+      "epoch": 0.03282720930073952,
+      "grad_norm": 0.02972390688955784,
+      "learning_rate": 0.0002,
+      "loss": 0.016,
+      "step": 18050
+    },
+    {
+      "epoch": 0.032845396120296716,
+      "grad_norm": 0.28630614280700684,
+      "learning_rate": 0.0002,
+      "loss": 0.1866,
+      "step": 18060
+    },
+    {
+      "epoch": 0.03286358293985391,
+      "grad_norm": 0.16426514089107513,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 18070
+    },
+    {
+      "epoch": 0.03288176975941111,
+      "grad_norm": 0.05643441155552864,
+      "learning_rate": 0.0002,
+      "loss": 0.0773,
+      "step": 18080
+    },
+    {
+      "epoch": 0.032899956578968305,
+      "grad_norm": 0.19082742929458618,
+      "learning_rate": 0.0002,
+      "loss": 0.0582,
+      "step": 18090
+    },
+    {
+      "epoch": 0.0329181433985255,
+      "grad_norm": 0.017512233927845955,
+      "learning_rate": 0.0002,
+      "loss": 0.0174,
+      "step": 18100
+    },
+    {
+      "epoch": 0.0329363302180827,
+      "grad_norm": 0.22619640827178955,
+      "learning_rate": 0.0002,
+      "loss": 0.166,
+      "step": 18110
+    },
+    {
+      "epoch": 0.032954517037639894,
+      "grad_norm": 0.10430974513292313,
+      "learning_rate": 0.0002,
+      "loss": 0.0716,
+      "step": 18120
+    },
+    {
+      "epoch": 0.03297270385719709,
+      "grad_norm": 0.07371710985898972,
+      "learning_rate": 0.0002,
+      "loss": 0.0733,
+      "step": 18130
+    },
+    {
+      "epoch": 0.032990890676754286,
+      "grad_norm": 0.19163483381271362,
+      "learning_rate": 0.0002,
+      "loss": 0.0609,
+      "step": 18140
+    },
+    {
+      "epoch": 0.03300907749631148,
+      "grad_norm": 0.03743975609540939,
+      "learning_rate": 0.0002,
+      "loss": 0.017,
+      "step": 18150
+    },
+    {
+      "epoch": 0.03302726431586868,
+      "grad_norm": 0.19496546685695648,
+      "learning_rate": 0.0002,
+      "loss": 0.1622,
+      "step": 18160
+    },
+    {
+      "epoch": 0.033045451135425875,
+      "grad_norm": 0.13054883480072021,
+      "learning_rate": 0.0002,
+      "loss": 0.0728,
+      "step": 18170
+    },
+    {
+      "epoch": 0.03306363795498307,
+      "grad_norm": 0.10058756172657013,
+      "learning_rate": 0.0002,
+      "loss": 0.0738,
+      "step": 18180
+    },
+    {
+      "epoch": 0.03308182477454027,
+      "grad_norm": 0.220932736992836,
+      "learning_rate": 0.0002,
+      "loss": 0.063,
+      "step": 18190
+    },
+    {
+      "epoch": 0.033100011594097464,
+      "grad_norm": 0.04396356642246246,
+      "learning_rate": 0.0002,
+      "loss": 0.0207,
+      "step": 18200
+    },
+    {
+      "epoch": 0.03311819841365467,
+      "grad_norm": 0.23554326593875885,
+      "learning_rate": 0.0002,
+      "loss": 0.1484,
+      "step": 18210
+    },
+    {
+      "epoch": 0.033136385233211864,
+      "grad_norm": 0.11277181655168533,
+      "learning_rate": 0.0002,
+      "loss": 0.0763,
+      "step": 18220
+    },
+    {
+      "epoch": 0.03315457205276906,
+      "grad_norm": 0.05176365375518799,
+      "learning_rate": 0.0002,
+      "loss": 0.076,
+      "step": 18230
+    },
+    {
+      "epoch": 0.033172758872326256,
+      "grad_norm": 0.1521395444869995,
+      "learning_rate": 0.0002,
+      "loss": 0.0605,
+      "step": 18240
+    },
+    {
+      "epoch": 0.03319094569188345,
+      "grad_norm": 0.04682580381631851,
+      "learning_rate": 0.0002,
+      "loss": 0.0149,
+      "step": 18250
+    },
+    {
+      "epoch": 0.03320913251144065,
+      "grad_norm": 0.16890883445739746,
+      "learning_rate": 0.0002,
+      "loss": 0.1402,
+      "step": 18260
+    },
+    {
+      "epoch": 0.033227319330997845,
+      "grad_norm": 0.17221559584140778,
+      "learning_rate": 0.0002,
+      "loss": 0.0819,
+      "step": 18270
+    },
+    {
+      "epoch": 0.03324550615055504,
+      "grad_norm": 0.07434559613466263,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 18280
+    },
+    {
+      "epoch": 0.03326369297011224,
+      "grad_norm": 0.1912834346294403,
+      "learning_rate": 0.0002,
+      "loss": 0.0614,
+      "step": 18290
+    },
+    {
+      "epoch": 0.033281879789669434,
+      "grad_norm": 0.04286884889006615,
+      "learning_rate": 0.0002,
+      "loss": 0.0185,
+      "step": 18300
+    },
+    {
+      "epoch": 0.03330006660922663,
+      "grad_norm": 0.29059842228889465,
+      "learning_rate": 0.0002,
+      "loss": 0.1357,
+      "step": 18310
+    },
+    {
+      "epoch": 0.03331825342878383,
+      "grad_norm": 0.2289486825466156,
+      "learning_rate": 0.0002,
+      "loss": 0.0865,
+      "step": 18320
+    },
+    {
+      "epoch": 0.03333644024834102,
+      "grad_norm": 0.027094636112451553,
+      "learning_rate": 0.0002,
+      "loss": 0.0841,
+      "step": 18330
+    },
+    {
+      "epoch": 0.03335462706789822,
+      "grad_norm": 0.21263600885868073,
+      "learning_rate": 0.0002,
+      "loss": 0.0628,
+      "step": 18340
+    },
+    {
+      "epoch": 0.033372813887455416,
+      "grad_norm": 0.03497980535030365,
+      "learning_rate": 0.0002,
+      "loss": 0.0158,
+      "step": 18350
+    },
+    {
+      "epoch": 0.03339100070701261,
+      "grad_norm": 0.20155973732471466,
+      "learning_rate": 0.0002,
+      "loss": 0.1523,
+      "step": 18360
+    },
+    {
+      "epoch": 0.03340918752656981,
+      "grad_norm": 0.03746286779642105,
+      "learning_rate": 0.0002,
+      "loss": 0.0781,
+      "step": 18370
+    },
+    {
+      "epoch": 0.033427374346127005,
+      "grad_norm": 0.06747066229581833,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 18380
+    },
+    {
+      "epoch": 0.0334455611656842,
+      "grad_norm": 0.23699060082435608,
+      "learning_rate": 0.0002,
+      "loss": 0.0651,
+      "step": 18390
+    },
+    {
+      "epoch": 0.0334637479852414,
+      "grad_norm": 0.047832150012254715,
+      "learning_rate": 0.0002,
+      "loss": 0.0181,
+      "step": 18400
+    },
+    {
+      "epoch": 0.033481934804798594,
+      "grad_norm": 0.3178698420524597,
+      "learning_rate": 0.0002,
+      "loss": 0.1537,
+      "step": 18410
+    },
+    {
+      "epoch": 0.03350012162435579,
+      "grad_norm": 0.16258081793785095,
+      "learning_rate": 0.0002,
+      "loss": 0.0722,
+      "step": 18420
+    },
+    {
+      "epoch": 0.033518308443912986,
+      "grad_norm": 0.02807716652750969,
+      "learning_rate": 0.0002,
+      "loss": 0.0844,
+      "step": 18430
+    },
+    {
+      "epoch": 0.03353649526347018,
+      "grad_norm": 0.16596710681915283,
+      "learning_rate": 0.0002,
+      "loss": 0.0607,
+      "step": 18440
+    },
+    {
+      "epoch": 0.03355468208302738,
+      "grad_norm": 0.04448723793029785,
+      "learning_rate": 0.0002,
+      "loss": 0.0183,
+      "step": 18450
+    },
+    {
+      "epoch": 0.033572868902584575,
+      "grad_norm": 0.39318934082984924,
+      "learning_rate": 0.0002,
+      "loss": 0.1497,
+      "step": 18460
+    },
+    {
+      "epoch": 0.03359105572214177,
+      "grad_norm": 0.17387263476848602,
+      "learning_rate": 0.0002,
+      "loss": 0.0787,
+      "step": 18470
+    },
+    {
+      "epoch": 0.03360924254169897,
+      "grad_norm": 0.14859163761138916,
+      "learning_rate": 0.0002,
+      "loss": 0.0837,
+      "step": 18480
+    },
+    {
+      "epoch": 0.033627429361256164,
+      "grad_norm": 0.24148601293563843,
+      "learning_rate": 0.0002,
+      "loss": 0.0655,
+      "step": 18490
+    },
+    {
+      "epoch": 0.03364561618081336,
+      "grad_norm": 0.04743284359574318,
+      "learning_rate": 0.0002,
+      "loss": 0.0174,
+      "step": 18500
+    },
+    {
+      "epoch": 0.03366380300037056,
+      "grad_norm": 0.25396591424942017,
+      "learning_rate": 0.0002,
+      "loss": 0.1438,
+      "step": 18510
+    },
+    {
+      "epoch": 0.03368198981992775,
+      "grad_norm": 0.1759178638458252,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 18520
+    },
+    {
+      "epoch": 0.03370017663948495,
+      "grad_norm": 0.06611669808626175,
+      "learning_rate": 0.0002,
+      "loss": 0.0787,
+      "step": 18530
+    },
+    {
+      "epoch": 0.033718363459042146,
+      "grad_norm": 0.22699445486068726,
+      "learning_rate": 0.0002,
+      "loss": 0.0697,
+      "step": 18540
+    },
+    {
+      "epoch": 0.03373655027859934,
+      "grad_norm": 0.02634899877011776,
+      "learning_rate": 0.0002,
+      "loss": 0.0189,
+      "step": 18550
+    },
+    {
+      "epoch": 0.03375473709815654,
+      "grad_norm": 0.3238360285758972,
+      "learning_rate": 0.0002,
+      "loss": 0.1496,
+      "step": 18560
+    },
+    {
+      "epoch": 0.033772923917713735,
+      "grad_norm": 0.16044601798057556,
+      "learning_rate": 0.0002,
+      "loss": 0.076,
+      "step": 18570
+    },
+    {
+      "epoch": 0.03379111073727093,
+      "grad_norm": 0.029841836541891098,
+      "learning_rate": 0.0002,
+      "loss": 0.0718,
+      "step": 18580
+    },
+    {
+      "epoch": 0.03380929755682813,
+      "grad_norm": 0.21851007640361786,
+      "learning_rate": 0.0002,
+      "loss": 0.0656,
+      "step": 18590
+    },
+    {
+      "epoch": 0.033827484376385324,
+      "grad_norm": 0.02096417360007763,
+      "learning_rate": 0.0002,
+      "loss": 0.0173,
+      "step": 18600
+    },
+    {
+      "epoch": 0.03384567119594252,
+      "grad_norm": 0.29625844955444336,
+      "learning_rate": 0.0002,
+      "loss": 0.1716,
+      "step": 18610
+    },
+    {
+      "epoch": 0.033863858015499716,
+      "grad_norm": 0.1510130614042282,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 18620
+    },
+    {
+      "epoch": 0.03388204483505691,
+      "grad_norm": 0.04192917421460152,
+      "learning_rate": 0.0002,
+      "loss": 0.0717,
+      "step": 18630
+    },
+    {
+      "epoch": 0.03390023165461411,
+      "grad_norm": 0.23139427602291107,
+      "learning_rate": 0.0002,
+      "loss": 0.0609,
+      "step": 18640
+    },
+    {
+      "epoch": 0.033918418474171305,
+      "grad_norm": 0.03887970373034477,
+      "learning_rate": 0.0002,
+      "loss": 0.0127,
+      "step": 18650
+    },
+    {
+      "epoch": 0.0339366052937285,
+      "grad_norm": 0.1315147578716278,
+      "learning_rate": 0.0002,
+      "loss": 0.1434,
+      "step": 18660
+    },
+    {
+      "epoch": 0.0339547921132857,
+      "grad_norm": 0.13328243792057037,
+      "learning_rate": 0.0002,
+      "loss": 0.0673,
+      "step": 18670
+    },
+    {
+      "epoch": 0.033972978932842894,
+      "grad_norm": 0.07161080092191696,
+      "learning_rate": 0.0002,
+      "loss": 0.0692,
+      "step": 18680
+    },
+    {
+      "epoch": 0.03399116575240009,
+      "grad_norm": 0.16019296646118164,
+      "learning_rate": 0.0002,
+      "loss": 0.0641,
+      "step": 18690
+    },
+    {
+      "epoch": 0.03400935257195729,
+      "grad_norm": 0.042882539331912994,
+      "learning_rate": 0.0002,
+      "loss": 0.0196,
+      "step": 18700
+    },
+    {
+      "epoch": 0.03402753939151448,
+      "grad_norm": 0.15019817650318146,
+      "learning_rate": 0.0002,
+      "loss": 0.1239,
+      "step": 18710
+    },
+    {
+      "epoch": 0.03404572621107168,
+      "grad_norm": 0.140267476439476,
+      "learning_rate": 0.0002,
+      "loss": 0.0715,
+      "step": 18720
+    },
+    {
+      "epoch": 0.034063913030628876,
+      "grad_norm": 0.060760073363780975,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 18730
+    },
+    {
+      "epoch": 0.03408209985018607,
+      "grad_norm": 0.1783122718334198,
+      "learning_rate": 0.0002,
+      "loss": 0.0616,
+      "step": 18740
+    },
+    {
+      "epoch": 0.03410028666974327,
+      "grad_norm": 0.023139121010899544,
+      "learning_rate": 0.0002,
+      "loss": 0.0171,
+      "step": 18750
+    },
+    {
+      "epoch": 0.034118473489300465,
+      "grad_norm": 0.2645978331565857,
+      "learning_rate": 0.0002,
+      "loss": 0.1355,
+      "step": 18760
+    },
+    {
+      "epoch": 0.03413666030885766,
+      "grad_norm": 0.21009914577007294,
+      "learning_rate": 0.0002,
+      "loss": 0.0757,
+      "step": 18770
+    },
+    {
+      "epoch": 0.03415484712841486,
+      "grad_norm": 0.13494494557380676,
+      "learning_rate": 0.0002,
+      "loss": 0.0774,
+      "step": 18780
+    },
+    {
+      "epoch": 0.034173033947972054,
+      "grad_norm": 0.19806784391403198,
+      "learning_rate": 0.0002,
+      "loss": 0.0636,
+      "step": 18790
+    },
+    {
+      "epoch": 0.03419122076752925,
+      "grad_norm": 0.020482519641518593,
+      "learning_rate": 0.0002,
+      "loss": 0.0194,
+      "step": 18800
+    },
+    {
+      "epoch": 0.034209407587086446,
+      "grad_norm": 0.34826937317848206,
+      "learning_rate": 0.0002,
+      "loss": 0.1521,
+      "step": 18810
+    },
+    {
+      "epoch": 0.03422759440664364,
+      "grad_norm": 0.1293957680463791,
+      "learning_rate": 0.0002,
+      "loss": 0.0742,
+      "step": 18820
+    },
+    {
+      "epoch": 0.03424578122620084,
+      "grad_norm": 0.06574539095163345,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 18830
+    },
+    {
+      "epoch": 0.034263968045758035,
+      "grad_norm": 0.2005399614572525,
+      "learning_rate": 0.0002,
+      "loss": 0.0618,
+      "step": 18840
+    },
+    {
+      "epoch": 0.03428215486531523,
+      "grad_norm": 0.04699913039803505,
+      "learning_rate": 0.0002,
+      "loss": 0.0176,
+      "step": 18850
+    },
+    {
+      "epoch": 0.03430034168487243,
+      "grad_norm": 0.2593109905719757,
+      "learning_rate": 0.0002,
+      "loss": 0.1709,
+      "step": 18860
+    },
+    {
+      "epoch": 0.034318528504429624,
+      "grad_norm": 0.587365448474884,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 18870
+    },
+    {
+      "epoch": 0.03433671532398682,
+      "grad_norm": 0.0371614433825016,
+      "learning_rate": 0.0002,
+      "loss": 0.076,
+      "step": 18880
+    },
+    {
+      "epoch": 0.03435490214354402,
+      "grad_norm": 0.2164178341627121,
+      "learning_rate": 0.0002,
+      "loss": 0.0577,
+      "step": 18890
+    },
+    {
+      "epoch": 0.03437308896310122,
+      "grad_norm": 0.028071587905287743,
+      "learning_rate": 0.0002,
+      "loss": 0.0184,
+      "step": 18900
+    },
+    {
+      "epoch": 0.034391275782658416,
+      "grad_norm": 0.25464126467704773,
+      "learning_rate": 0.0002,
+      "loss": 0.1616,
+      "step": 18910
+    },
+    {
+      "epoch": 0.03440946260221561,
+      "grad_norm": 0.2830415368080139,
+      "learning_rate": 0.0002,
+      "loss": 0.0795,
+      "step": 18920
+    },
+    {
+      "epoch": 0.03442764942177281,
+      "grad_norm": 0.07880273461341858,
+      "learning_rate": 0.0002,
+      "loss": 0.0717,
+      "step": 18930
+    },
+    {
+      "epoch": 0.034445836241330005,
+      "grad_norm": 0.19671671092510223,
+      "learning_rate": 0.0002,
+      "loss": 0.0625,
+      "step": 18940
+    },
+    {
+      "epoch": 0.0344640230608872,
+      "grad_norm": 0.038350027054548264,
+      "learning_rate": 0.0002,
+      "loss": 0.0172,
+      "step": 18950
+    },
+    {
+      "epoch": 0.0344822098804444,
+      "grad_norm": 0.196768656373024,
+      "learning_rate": 0.0002,
+      "loss": 0.1586,
+      "step": 18960
+    },
+    {
+      "epoch": 0.034500396700001594,
+      "grad_norm": 0.1861678808927536,
+      "learning_rate": 0.0002,
+      "loss": 0.0871,
+      "step": 18970
+    },
+    {
+      "epoch": 0.03451858351955879,
+      "grad_norm": 0.1074979305267334,
+      "learning_rate": 0.0002,
+      "loss": 0.0697,
+      "step": 18980
+    },
+    {
+      "epoch": 0.03453677033911599,
+      "grad_norm": 0.18214645981788635,
+      "learning_rate": 0.0002,
+      "loss": 0.0594,
+      "step": 18990
+    },
+    {
+      "epoch": 0.03455495715867318,
+      "grad_norm": 0.035948049277067184,
+      "learning_rate": 0.0002,
+      "loss": 0.0177,
+      "step": 19000
+    },
+    {
+      "epoch": 0.03457314397823038,
+      "grad_norm": 0.2434094399213791,
+      "learning_rate": 0.0002,
+      "loss": 0.1402,
+      "step": 19010
+    },
+    {
+      "epoch": 0.034591330797787576,
+      "grad_norm": 0.06897670775651932,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 19020
+    },
+    {
+      "epoch": 0.03460951761734477,
+      "grad_norm": 0.13107649981975555,
+      "learning_rate": 0.0002,
+      "loss": 0.0826,
+      "step": 19030
+    },
+    {
+      "epoch": 0.03462770443690197,
+      "grad_norm": 0.1787865310907364,
+      "learning_rate": 0.0002,
+      "loss": 0.0619,
+      "step": 19040
+    },
+    {
+      "epoch": 0.034645891256459165,
+      "grad_norm": 0.0460963137447834,
+      "learning_rate": 0.0002,
+      "loss": 0.0203,
+      "step": 19050
+    },
+    {
+      "epoch": 0.03466407807601636,
+      "grad_norm": 0.20582084357738495,
+      "learning_rate": 0.0002,
+      "loss": 0.1325,
+      "step": 19060
+    },
+    {
+      "epoch": 0.03468226489557356,
+      "grad_norm": 0.16120313107967377,
+      "learning_rate": 0.0002,
+      "loss": 0.08,
+      "step": 19070
+    },
+    {
+      "epoch": 0.03470045171513075,
+      "grad_norm": 0.04322347044944763,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 19080
+    },
+    {
+      "epoch": 0.03471863853468795,
+      "grad_norm": 0.1764109879732132,
+      "learning_rate": 0.0002,
+      "loss": 0.0618,
+      "step": 19090
+    },
+    {
+      "epoch": 0.034736825354245146,
+      "grad_norm": 0.04453815147280693,
+      "learning_rate": 0.0002,
+      "loss": 0.0172,
+      "step": 19100
+    },
+    {
+      "epoch": 0.03475501217380234,
+      "grad_norm": 0.32023972272872925,
+      "learning_rate": 0.0002,
+      "loss": 0.1394,
+      "step": 19110
+    },
+    {
+      "epoch": 0.03477319899335954,
+      "grad_norm": 0.09920009225606918,
+      "learning_rate": 0.0002,
+      "loss": 0.0788,
+      "step": 19120
+    },
+    {
+      "epoch": 0.034791385812916735,
+      "grad_norm": 0.047868456691503525,
+      "learning_rate": 0.0002,
+      "loss": 0.0745,
+      "step": 19130
+    },
+    {
+      "epoch": 0.03480957263247393,
+      "grad_norm": 0.219430074095726,
+      "learning_rate": 0.0002,
+      "loss": 0.063,
+      "step": 19140
+    },
+    {
+      "epoch": 0.03482775945203113,
+      "grad_norm": 0.04879681020975113,
+      "learning_rate": 0.0002,
+      "loss": 0.0161,
+      "step": 19150
+    },
+    {
+      "epoch": 0.034845946271588324,
+      "grad_norm": 0.21360138058662415,
+      "learning_rate": 0.0002,
+      "loss": 0.1602,
+      "step": 19160
+    },
+    {
+      "epoch": 0.03486413309114552,
+      "grad_norm": 0.1391269713640213,
+      "learning_rate": 0.0002,
+      "loss": 0.0798,
+      "step": 19170
+    },
+    {
+      "epoch": 0.03488231991070272,
+      "grad_norm": 0.06293737888336182,
+      "learning_rate": 0.0002,
+      "loss": 0.0717,
+      "step": 19180
+    },
+    {
+      "epoch": 0.03490050673025991,
+      "grad_norm": 0.20241963863372803,
+      "learning_rate": 0.0002,
+      "loss": 0.0612,
+      "step": 19190
+    },
+    {
+      "epoch": 0.03491869354981711,
+      "grad_norm": 0.06246611103415489,
+      "learning_rate": 0.0002,
+      "loss": 0.0148,
+      "step": 19200
+    },
+    {
+      "epoch": 0.034936880369374305,
+      "grad_norm": 0.16479995846748352,
+      "learning_rate": 0.0002,
+      "loss": 0.1611,
+      "step": 19210
+    },
+    {
+      "epoch": 0.0349550671889315,
+      "grad_norm": 0.12036983668804169,
+      "learning_rate": 0.0002,
+      "loss": 0.0724,
+      "step": 19220
+    },
+    {
+      "epoch": 0.0349732540084887,
+      "grad_norm": 0.03939517214894295,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 19230
+    },
+    {
+      "epoch": 0.034991440828045894,
+      "grad_norm": 0.17047277092933655,
+      "learning_rate": 0.0002,
+      "loss": 0.066,
+      "step": 19240
+    },
+    {
+      "epoch": 0.03500962764760309,
+      "grad_norm": 0.031782686710357666,
+      "learning_rate": 0.0002,
+      "loss": 0.0203,
+      "step": 19250
+    },
+    {
+      "epoch": 0.03502781446716029,
+      "grad_norm": 0.2545730471611023,
+      "learning_rate": 0.0002,
+      "loss": 0.1716,
+      "step": 19260
+    },
+    {
+      "epoch": 0.03504600128671748,
+      "grad_norm": 0.11225811392068863,
+      "learning_rate": 0.0002,
+      "loss": 0.0791,
+      "step": 19270
+    },
+    {
+      "epoch": 0.03506418810627468,
+      "grad_norm": 0.049140989780426025,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 19280
+    },
+    {
+      "epoch": 0.035082374925831876,
+      "grad_norm": 0.16942913830280304,
+      "learning_rate": 0.0002,
+      "loss": 0.0638,
+      "step": 19290
+    },
+    {
+      "epoch": 0.03510056174538907,
+      "grad_norm": 0.03836115077137947,
+      "learning_rate": 0.0002,
+      "loss": 0.0193,
+      "step": 19300
+    },
+    {
+      "epoch": 0.03511874856494627,
+      "grad_norm": 0.13004787266254425,
+      "learning_rate": 0.0002,
+      "loss": 0.1477,
+      "step": 19310
+    },
+    {
+      "epoch": 0.035136935384503465,
+      "grad_norm": 0.2054329216480255,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 19320
+    },
+    {
+      "epoch": 0.03515512220406066,
+      "grad_norm": 0.06592074781656265,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 19330
+    },
+    {
+      "epoch": 0.03517330902361786,
+      "grad_norm": 0.19228027760982513,
+      "learning_rate": 0.0002,
+      "loss": 0.067,
+      "step": 19340
+    },
+    {
+      "epoch": 0.035191495843175054,
+      "grad_norm": 0.04050719738006592,
+      "learning_rate": 0.0002,
+      "loss": 0.017,
+      "step": 19350
+    },
+    {
+      "epoch": 0.03520968266273225,
+      "grad_norm": 0.28715401887893677,
+      "learning_rate": 0.0002,
+      "loss": 0.1499,
+      "step": 19360
+    },
+    {
+      "epoch": 0.035227869482289446,
+      "grad_norm": 0.13954712450504303,
+      "learning_rate": 0.0002,
+      "loss": 0.0787,
+      "step": 19370
+    },
+    {
+      "epoch": 0.03524605630184664,
+      "grad_norm": 0.08851815015077591,
+      "learning_rate": 0.0002,
+      "loss": 0.0739,
+      "step": 19380
+    },
+    {
+      "epoch": 0.03526424312140384,
+      "grad_norm": 0.1788545697927475,
+      "learning_rate": 0.0002,
+      "loss": 0.0576,
+      "step": 19390
+    },
+    {
+      "epoch": 0.035282429940961035,
+      "grad_norm": 0.03644658252596855,
+      "learning_rate": 0.0002,
+      "loss": 0.0143,
+      "step": 19400
+    },
+    {
+      "epoch": 0.03530061676051823,
+      "grad_norm": 0.3140568137168884,
+      "learning_rate": 0.0002,
+      "loss": 0.1498,
+      "step": 19410
+    },
+    {
+      "epoch": 0.03531880358007543,
+      "grad_norm": 0.14550529420375824,
+      "learning_rate": 0.0002,
+      "loss": 0.0748,
+      "step": 19420
+    },
+    {
+      "epoch": 0.035336990399632624,
+      "grad_norm": 0.10995481163263321,
+      "learning_rate": 0.0002,
+      "loss": 0.0801,
+      "step": 19430
+    },
+    {
+      "epoch": 0.03535517721918982,
+      "grad_norm": 0.17238560318946838,
+      "learning_rate": 0.0002,
+      "loss": 0.0608,
+      "step": 19440
+    },
+    {
+      "epoch": 0.03537336403874702,
+      "grad_norm": 0.031363293528556824,
+      "learning_rate": 0.0002,
+      "loss": 0.0154,
+      "step": 19450
+    },
+    {
+      "epoch": 0.03539155085830421,
+      "grad_norm": 0.14145390689373016,
+      "learning_rate": 0.0002,
+      "loss": 0.1511,
+      "step": 19460
+    },
+    {
+      "epoch": 0.03540973767786141,
+      "grad_norm": 0.19073855876922607,
+      "learning_rate": 0.0002,
+      "loss": 0.0725,
+      "step": 19470
+    },
+    {
+      "epoch": 0.035427924497418606,
+      "grad_norm": 0.15639430284500122,
+      "learning_rate": 0.0002,
+      "loss": 0.0836,
+      "step": 19480
+    },
+    {
+      "epoch": 0.0354461113169758,
+      "grad_norm": 0.2566238045692444,
+      "learning_rate": 0.0002,
+      "loss": 0.0617,
+      "step": 19490
+    },
+    {
+      "epoch": 0.035464298136533,
+      "grad_norm": 0.055755820125341415,
+      "learning_rate": 0.0002,
+      "loss": 0.0178,
+      "step": 19500
+    },
+    {
+      "epoch": 0.035482484956090195,
+      "grad_norm": 0.2835562527179718,
+      "learning_rate": 0.0002,
+      "loss": 0.1306,
+      "step": 19510
+    },
+    {
+      "epoch": 0.03550067177564739,
+      "grad_norm": 0.2310812920331955,
+      "learning_rate": 0.0002,
+      "loss": 0.0766,
+      "step": 19520
+    },
+    {
+      "epoch": 0.03551885859520459,
+      "grad_norm": 0.1287071257829666,
+      "learning_rate": 0.0002,
+      "loss": 0.0791,
+      "step": 19530
+    },
+    {
+      "epoch": 0.035537045414761784,
+      "grad_norm": 0.21308869123458862,
+      "learning_rate": 0.0002,
+      "loss": 0.0584,
+      "step": 19540
+    },
+    {
+      "epoch": 0.03555523223431898,
+      "grad_norm": 0.0662735179066658,
+      "learning_rate": 0.0002,
+      "loss": 0.0207,
+      "step": 19550
+    },
+    {
+      "epoch": 0.035573419053876176,
+      "grad_norm": 0.21706523001194,
+      "learning_rate": 0.0002,
+      "loss": 0.1308,
+      "step": 19560
+    },
+    {
+      "epoch": 0.03559160587343337,
+      "grad_norm": 0.09376335144042969,
+      "learning_rate": 0.0002,
+      "loss": 0.0677,
+      "step": 19570
+    },
+    {
+      "epoch": 0.035609792692990576,
+      "grad_norm": 0.1093437597155571,
+      "learning_rate": 0.0002,
+      "loss": 0.0741,
+      "step": 19580
+    },
+    {
+      "epoch": 0.03562797951254777,
+      "grad_norm": 0.21057911217212677,
+      "learning_rate": 0.0002,
+      "loss": 0.0637,
+      "step": 19590
+    },
+    {
+      "epoch": 0.03564616633210497,
+      "grad_norm": 0.04383830726146698,
+      "learning_rate": 0.0002,
+      "loss": 0.019,
+      "step": 19600
+    },
+    {
+      "epoch": 0.035664353151662165,
+      "grad_norm": 0.3657427132129669,
+      "learning_rate": 0.0002,
+      "loss": 0.1421,
+      "step": 19610
+    },
+    {
+      "epoch": 0.03568253997121936,
+      "grad_norm": 0.17154265940189362,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 19620
+    },
+    {
+      "epoch": 0.03570072679077656,
+      "grad_norm": 0.041993435472249985,
+      "learning_rate": 0.0002,
+      "loss": 0.0768,
+      "step": 19630
+    },
+    {
+      "epoch": 0.035718913610333754,
+      "grad_norm": 0.1658252775669098,
+      "learning_rate": 0.0002,
+      "loss": 0.0602,
+      "step": 19640
+    },
+    {
+      "epoch": 0.03573710042989095,
+      "grad_norm": 0.028523078188300133,
+      "learning_rate": 0.0002,
+      "loss": 0.0151,
+      "step": 19650
+    },
+    {
+      "epoch": 0.035755287249448146,
+      "grad_norm": 0.2624453902244568,
+      "learning_rate": 0.0002,
+      "loss": 0.1355,
+      "step": 19660
+    },
+    {
+      "epoch": 0.03577347406900534,
+      "grad_norm": 0.12055794149637222,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 19670
+    },
+    {
+      "epoch": 0.03579166088856254,
+      "grad_norm": 0.043441224843263626,
+      "learning_rate": 0.0002,
+      "loss": 0.0722,
+      "step": 19680
+    },
+    {
+      "epoch": 0.035809847708119735,
+      "grad_norm": 0.2464340627193451,
+      "learning_rate": 0.0002,
+      "loss": 0.0673,
+      "step": 19690
+    },
+    {
+      "epoch": 0.03582803452767693,
+      "grad_norm": 0.04004153981804848,
+      "learning_rate": 0.0002,
+      "loss": 0.0212,
+      "step": 19700
+    },
+    {
+      "epoch": 0.03584622134723413,
+      "grad_norm": 0.3159453570842743,
+      "learning_rate": 0.0002,
+      "loss": 0.1806,
+      "step": 19710
+    },
+    {
+      "epoch": 0.035864408166791324,
+      "grad_norm": 0.11327318102121353,
+      "learning_rate": 0.0002,
+      "loss": 0.0748,
+      "step": 19720
+    },
+    {
+      "epoch": 0.03588259498634852,
+      "grad_norm": 0.0980909988284111,
+      "learning_rate": 0.0002,
+      "loss": 0.0807,
+      "step": 19730
+    },
+    {
+      "epoch": 0.03590078180590572,
+      "grad_norm": 0.15508098900318146,
+      "learning_rate": 0.0002,
+      "loss": 0.0576,
+      "step": 19740
+    },
+    {
+      "epoch": 0.03591896862546291,
+      "grad_norm": 0.019624806940555573,
+      "learning_rate": 0.0002,
+      "loss": 0.0135,
+      "step": 19750
+    },
+    {
+      "epoch": 0.03593715544502011,
+      "grad_norm": 0.20336109399795532,
+      "learning_rate": 0.0002,
+      "loss": 0.1702,
+      "step": 19760
+    },
+    {
+      "epoch": 0.035955342264577306,
+      "grad_norm": 0.12767620384693146,
+      "learning_rate": 0.0002,
+      "loss": 0.0776,
+      "step": 19770
+    },
+    {
+      "epoch": 0.0359735290841345,
+      "grad_norm": 0.19050805270671844,
+      "learning_rate": 0.0002,
+      "loss": 0.0838,
+      "step": 19780
+    },
+    {
+      "epoch": 0.0359917159036917,
+      "grad_norm": 0.17471866309642792,
+      "learning_rate": 0.0002,
+      "loss": 0.0561,
+      "step": 19790
+    },
+    {
+      "epoch": 0.036009902723248895,
+      "grad_norm": 0.044348277151584625,
+      "learning_rate": 0.0002,
+      "loss": 0.0159,
+      "step": 19800
+    },
+    {
+      "epoch": 0.03602808954280609,
+      "grad_norm": 0.30847081542015076,
+      "learning_rate": 0.0002,
+      "loss": 0.1686,
+      "step": 19810
+    },
+    {
+      "epoch": 0.03604627636236329,
+      "grad_norm": 0.08963622897863388,
+      "learning_rate": 0.0002,
+      "loss": 0.078,
+      "step": 19820
+    },
+    {
+      "epoch": 0.036064463181920484,
+      "grad_norm": 0.0580587275326252,
+      "learning_rate": 0.0002,
+      "loss": 0.0741,
+      "step": 19830
+    },
+    {
+      "epoch": 0.03608265000147768,
+      "grad_norm": 0.1698184460401535,
+      "learning_rate": 0.0002,
+      "loss": 0.0631,
+      "step": 19840
+    },
+    {
+      "epoch": 0.036100836821034876,
+      "grad_norm": 0.025531867519021034,
+      "learning_rate": 0.0002,
+      "loss": 0.0166,
+      "step": 19850
+    },
+    {
+      "epoch": 0.03611902364059207,
+      "grad_norm": 0.3544731140136719,
+      "learning_rate": 0.0002,
+      "loss": 0.1886,
+      "step": 19860
+    },
+    {
+      "epoch": 0.03613721046014927,
+      "grad_norm": 0.2552841901779175,
+      "learning_rate": 0.0002,
+      "loss": 0.0859,
+      "step": 19870
+    },
+    {
+      "epoch": 0.036155397279706465,
+      "grad_norm": 0.07771942019462585,
+      "learning_rate": 0.0002,
+      "loss": 0.0859,
+      "step": 19880
+    },
+    {
+      "epoch": 0.03617358409926366,
+      "grad_norm": 0.15945585072040558,
+      "learning_rate": 0.0002,
+      "loss": 0.0609,
+      "step": 19890
+    },
+    {
+      "epoch": 0.03619177091882086,
+      "grad_norm": 0.04583865404129028,
+      "learning_rate": 0.0002,
+      "loss": 0.0196,
+      "step": 19900
+    },
+    {
+      "epoch": 0.036209957738378054,
+      "grad_norm": 0.2110920548439026,
+      "learning_rate": 0.0002,
+      "loss": 0.1305,
+      "step": 19910
+    },
+    {
+      "epoch": 0.03622814455793525,
+      "grad_norm": 0.22165755927562714,
+      "learning_rate": 0.0002,
+      "loss": 0.0767,
+      "step": 19920
+    },
+    {
+      "epoch": 0.03624633137749245,
+      "grad_norm": 0.0866742879152298,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 19930
+    },
+    {
+      "epoch": 0.03626451819704964,
+      "grad_norm": 0.19838224351406097,
+      "learning_rate": 0.0002,
+      "loss": 0.0663,
+      "step": 19940
+    },
+    {
+      "epoch": 0.03628270501660684,
+      "grad_norm": 0.05543521046638489,
+      "learning_rate": 0.0002,
+      "loss": 0.023,
+      "step": 19950
+    },
+    {
+      "epoch": 0.036300891836164036,
+      "grad_norm": 0.20800183713436127,
+      "learning_rate": 0.0002,
+      "loss": 0.1468,
+      "step": 19960
+    },
+    {
+      "epoch": 0.03631907865572123,
+      "grad_norm": 0.14951092004776,
+      "learning_rate": 0.0002,
+      "loss": 0.0698,
+      "step": 19970
+    },
+    {
+      "epoch": 0.03633726547527843,
+      "grad_norm": 0.10162603855133057,
+      "learning_rate": 0.0002,
+      "loss": 0.0841,
+      "step": 19980
+    },
+    {
+      "epoch": 0.036355452294835625,
+      "grad_norm": 0.24774019420146942,
+      "learning_rate": 0.0002,
+      "loss": 0.0658,
+      "step": 19990
+    },
+    {
+      "epoch": 0.03637363911439282,
+      "grad_norm": 0.02705777995288372,
+      "learning_rate": 0.0002,
+      "loss": 0.02,
+      "step": 20000
+    },
+    {
+      "epoch": 0.03639182593395002,
+      "grad_norm": 0.2509992718696594,
+      "learning_rate": 0.0002,
+      "loss": 0.1529,
+      "step": 20010
+    },
+    {
+      "epoch": 0.036410012753507214,
+      "grad_norm": 0.2126697599887848,
+      "learning_rate": 0.0002,
+      "loss": 0.0716,
+      "step": 20020
+    },
+    {
+      "epoch": 0.03642819957306441,
+      "grad_norm": 0.1463591754436493,
+      "learning_rate": 0.0002,
+      "loss": 0.076,
+      "step": 20030
+    },
+    {
+      "epoch": 0.036446386392621606,
+      "grad_norm": 0.21879518032073975,
+      "learning_rate": 0.0002,
+      "loss": 0.0677,
+      "step": 20040
+    },
+    {
+      "epoch": 0.0364645732121788,
+      "grad_norm": 0.028337355703115463,
+      "learning_rate": 0.0002,
+      "loss": 0.0131,
+      "step": 20050
+    },
+    {
+      "epoch": 0.036482760031736,
+      "grad_norm": 0.335788756608963,
+      "learning_rate": 0.0002,
+      "loss": 0.1693,
+      "step": 20060
+    },
+    {
+      "epoch": 0.036500946851293195,
+      "grad_norm": 0.17615728080272675,
+      "learning_rate": 0.0002,
+      "loss": 0.0791,
+      "step": 20070
+    },
+    {
+      "epoch": 0.03651913367085039,
+      "grad_norm": 0.034229181706905365,
+      "learning_rate": 0.0002,
+      "loss": 0.0774,
+      "step": 20080
+    },
+    {
+      "epoch": 0.03653732049040759,
+      "grad_norm": 0.20637790858745575,
+      "learning_rate": 0.0002,
+      "loss": 0.0544,
+      "step": 20090
+    },
+    {
+      "epoch": 0.036555507309964784,
+      "grad_norm": 0.033659741282463074,
+      "learning_rate": 0.0002,
+      "loss": 0.0128,
+      "step": 20100
+    },
+    {
+      "epoch": 0.03657369412952198,
+      "grad_norm": 0.18249601125717163,
+      "learning_rate": 0.0002,
+      "loss": 0.1939,
+      "step": 20110
+    },
+    {
+      "epoch": 0.03659188094907918,
+      "grad_norm": 0.18065877258777618,
+      "learning_rate": 0.0002,
+      "loss": 0.0816,
+      "step": 20120
+    },
+    {
+      "epoch": 0.03661006776863637,
+      "grad_norm": 0.4361811876296997,
+      "learning_rate": 0.0002,
+      "loss": 0.0978,
+      "step": 20130
+    },
+    {
+      "epoch": 0.03662825458819357,
+      "grad_norm": 0.24488002061843872,
+      "learning_rate": 0.0002,
+      "loss": 0.0742,
+      "step": 20140
+    },
+    {
+      "epoch": 0.036646441407750766,
+      "grad_norm": 0.023062752559781075,
+      "learning_rate": 0.0002,
+      "loss": 0.0196,
+      "step": 20150
+    },
+    {
+      "epoch": 0.03666462822730796,
+      "grad_norm": 0.22796255350112915,
+      "learning_rate": 0.0002,
+      "loss": 0.1457,
+      "step": 20160
+    },
+    {
+      "epoch": 0.03668281504686516,
+      "grad_norm": 0.16665758192539215,
+      "learning_rate": 0.0002,
+      "loss": 0.138,
+      "step": 20170
+    },
+    {
+      "epoch": 0.036701001866422354,
+      "grad_norm": 0.0503946952521801,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 20180
+    },
+    {
+      "epoch": 0.03671918868597955,
+      "grad_norm": 0.1672963798046112,
+      "learning_rate": 0.0002,
+      "loss": 0.0621,
+      "step": 20190
+    },
+    {
+      "epoch": 0.03673737550553675,
+      "grad_norm": 0.06765859574079514,
+      "learning_rate": 0.0002,
+      "loss": 0.0171,
+      "step": 20200
+    },
+    {
+      "epoch": 0.03675556232509394,
+      "grad_norm": 0.6076682806015015,
+      "learning_rate": 0.0002,
+      "loss": 0.6804,
+      "step": 20210
+    },
+    {
+      "epoch": 0.03677374914465114,
+      "grad_norm": 0.04764563590288162,
+      "learning_rate": 0.0002,
+      "loss": 0.0965,
+      "step": 20220
+    },
+    {
+      "epoch": 0.036791935964208336,
+      "grad_norm": 0.6847806572914124,
+      "learning_rate": 0.0002,
+      "loss": 0.0784,
+      "step": 20230
+    },
+    {
+      "epoch": 0.03681012278376553,
+      "grad_norm": 0.2678837478160858,
+      "learning_rate": 0.0002,
+      "loss": 0.069,
+      "step": 20240
+    },
+    {
+      "epoch": 0.03682830960332273,
+      "grad_norm": 0.039824239909648895,
+      "learning_rate": 0.0002,
+      "loss": 0.0206,
+      "step": 20250
+    },
+    {
+      "epoch": 0.036846496422879925,
+      "grad_norm": 0.19583609700202942,
+      "learning_rate": 0.0002,
+      "loss": 0.1588,
+      "step": 20260
+    },
+    {
+      "epoch": 0.03686468324243713,
+      "grad_norm": 0.08613055944442749,
+      "learning_rate": 0.0002,
+      "loss": 0.0777,
+      "step": 20270
+    },
+    {
+      "epoch": 0.036882870061994324,
+      "grad_norm": 0.028818165883421898,
+      "learning_rate": 0.0002,
+      "loss": 0.0704,
+      "step": 20280
+    },
+    {
+      "epoch": 0.03690105688155152,
+      "grad_norm": 0.19514115154743195,
+      "learning_rate": 0.0002,
+      "loss": 0.0654,
+      "step": 20290
+    },
+    {
+      "epoch": 0.03691924370110872,
+      "grad_norm": 0.043222617357969284,
+      "learning_rate": 0.0002,
+      "loss": 0.0216,
+      "step": 20300
+    },
+    {
+      "epoch": 0.03693743052066591,
+      "grad_norm": 0.2490546703338623,
+      "learning_rate": 0.0002,
+      "loss": 0.1472,
+      "step": 20310
+    },
+    {
+      "epoch": 0.03695561734022311,
+      "grad_norm": 0.16989269852638245,
+      "learning_rate": 0.0002,
+      "loss": 0.081,
+      "step": 20320
+    },
+    {
+      "epoch": 0.036973804159780306,
+      "grad_norm": 0.09191739559173584,
+      "learning_rate": 0.0002,
+      "loss": 0.0733,
+      "step": 20330
+    },
+    {
+      "epoch": 0.0369919909793375,
+      "grad_norm": 0.18435023725032806,
+      "learning_rate": 0.0002,
+      "loss": 0.0654,
+      "step": 20340
+    },
+    {
+      "epoch": 0.0370101777988947,
+      "grad_norm": 0.031144114211201668,
+      "learning_rate": 0.0002,
+      "loss": 0.0226,
+      "step": 20350
+    },
+    {
+      "epoch": 0.037028364618451895,
+      "grad_norm": 0.3244694769382477,
+      "learning_rate": 0.0002,
+      "loss": 0.1304,
+      "step": 20360
+    },
+    {
+      "epoch": 0.03704655143800909,
+      "grad_norm": 0.13787488639354706,
+      "learning_rate": 0.0002,
+      "loss": 0.0811,
+      "step": 20370
+    },
+    {
+      "epoch": 0.03706473825756629,
+      "grad_norm": 0.058523450046777725,
+      "learning_rate": 0.0002,
+      "loss": 0.0806,
+      "step": 20380
+    },
+    {
+      "epoch": 0.037082925077123484,
+      "grad_norm": 0.3001325726509094,
+      "learning_rate": 0.0002,
+      "loss": 0.0694,
+      "step": 20390
+    },
+    {
+      "epoch": 0.03710111189668068,
+      "grad_norm": 0.04447292909026146,
+      "learning_rate": 0.0002,
+      "loss": 0.0218,
+      "step": 20400
+    },
+    {
+      "epoch": 0.037119298716237877,
+      "grad_norm": 0.25786396861076355,
+      "learning_rate": 0.0002,
+      "loss": 0.1499,
+      "step": 20410
+    },
+    {
+      "epoch": 0.03713748553579507,
+      "grad_norm": 0.11381134390830994,
+      "learning_rate": 0.0002,
+      "loss": 0.0822,
+      "step": 20420
+    },
+    {
+      "epoch": 0.03715567235535227,
+      "grad_norm": 0.022713568061590195,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 20430
+    },
+    {
+      "epoch": 0.037173859174909465,
+      "grad_norm": 0.15770909190177917,
+      "learning_rate": 0.0002,
+      "loss": 0.0625,
+      "step": 20440
+    },
+    {
+      "epoch": 0.03719204599446666,
+      "grad_norm": 0.021412041038274765,
+      "learning_rate": 0.0002,
+      "loss": 0.0126,
+      "step": 20450
+    },
+    {
+      "epoch": 0.03721023281402386,
+      "grad_norm": 0.24260753393173218,
+      "learning_rate": 0.0002,
+      "loss": 0.1777,
+      "step": 20460
+    },
+    {
+      "epoch": 0.037228419633581054,
+      "grad_norm": 0.10953031480312347,
+      "learning_rate": 0.0002,
+      "loss": 0.073,
+      "step": 20470
+    },
+    {
+      "epoch": 0.03724660645313825,
+      "grad_norm": 0.03975062072277069,
+      "learning_rate": 0.0002,
+      "loss": 0.0907,
+      "step": 20480
+    },
+    {
+      "epoch": 0.03726479327269545,
+      "grad_norm": 0.2025018036365509,
+      "learning_rate": 0.0002,
+      "loss": 0.0631,
+      "step": 20490
+    },
+    {
+      "epoch": 0.03728298009225264,
+      "grad_norm": 0.031849734485149384,
+      "learning_rate": 0.0002,
+      "loss": 0.0156,
+      "step": 20500
+    },
+    {
+      "epoch": 0.03730116691180984,
+      "grad_norm": 0.2650098502635956,
+      "learning_rate": 0.0002,
+      "loss": 0.1569,
+      "step": 20510
+    },
+    {
+      "epoch": 0.037319353731367036,
+      "grad_norm": 0.14113937318325043,
+      "learning_rate": 0.0002,
+      "loss": 0.0824,
+      "step": 20520
+    },
+    {
+      "epoch": 0.03733754055092423,
+      "grad_norm": 0.10276420414447784,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 20530
+    },
+    {
+      "epoch": 0.03735572737048143,
+      "grad_norm": 0.2258286476135254,
+      "learning_rate": 0.0002,
+      "loss": 0.0671,
+      "step": 20540
+    },
+    {
+      "epoch": 0.037373914190038625,
+      "grad_norm": 0.10343242436647415,
+      "learning_rate": 0.0002,
+      "loss": 0.0178,
+      "step": 20550
+    },
+    {
+      "epoch": 0.03739210100959582,
+      "grad_norm": 0.19423982501029968,
+      "learning_rate": 0.0002,
+      "loss": 0.1423,
+      "step": 20560
+    },
+    {
+      "epoch": 0.03741028782915302,
+      "grad_norm": 0.12046124786138535,
+      "learning_rate": 0.0002,
+      "loss": 0.0827,
+      "step": 20570
+    },
+    {
+      "epoch": 0.037428474648710214,
+      "grad_norm": 0.026751041412353516,
+      "learning_rate": 0.0002,
+      "loss": 0.0743,
+      "step": 20580
+    },
+    {
+      "epoch": 0.03744666146826741,
+      "grad_norm": 0.23576834797859192,
+      "learning_rate": 0.0002,
+      "loss": 0.0629,
+      "step": 20590
+    },
+    {
+      "epoch": 0.037464848287824606,
+      "grad_norm": 0.05146399885416031,
+      "learning_rate": 0.0002,
+      "loss": 0.0205,
+      "step": 20600
+    },
+    {
+      "epoch": 0.0374830351073818,
+      "grad_norm": 0.21750135719776154,
+      "learning_rate": 0.0002,
+      "loss": 0.1397,
+      "step": 20610
+    },
+    {
+      "epoch": 0.037501221926939,
+      "grad_norm": 0.08351115882396698,
+      "learning_rate": 0.0002,
+      "loss": 0.0801,
+      "step": 20620
+    },
+    {
+      "epoch": 0.037519408746496195,
+      "grad_norm": 0.07272092998027802,
+      "learning_rate": 0.0002,
+      "loss": 0.0881,
+      "step": 20630
+    },
+    {
+      "epoch": 0.03753759556605339,
+      "grad_norm": 0.23707769811153412,
+      "learning_rate": 0.0002,
+      "loss": 0.0706,
+      "step": 20640
+    },
+    {
+      "epoch": 0.03755578238561059,
+      "grad_norm": 0.05208323150873184,
+      "learning_rate": 0.0002,
+      "loss": 0.024,
+      "step": 20650
+    },
+    {
+      "epoch": 0.037573969205167784,
+      "grad_norm": 0.4163022041320801,
+      "learning_rate": 0.0002,
+      "loss": 0.159,
+      "step": 20660
+    },
+    {
+      "epoch": 0.03759215602472498,
+      "grad_norm": 0.1036575511097908,
+      "learning_rate": 0.0002,
+      "loss": 0.0814,
+      "step": 20670
+    },
+    {
+      "epoch": 0.03761034284428218,
+      "grad_norm": 0.09861626476049423,
+      "learning_rate": 0.0002,
+      "loss": 0.0828,
+      "step": 20680
+    },
+    {
+      "epoch": 0.03762852966383937,
+      "grad_norm": 0.1685744971036911,
+      "learning_rate": 0.0002,
+      "loss": 0.0597,
+      "step": 20690
+    },
+    {
+      "epoch": 0.03764671648339657,
+      "grad_norm": 0.02716050110757351,
+      "learning_rate": 0.0002,
+      "loss": 0.0164,
+      "step": 20700
+    },
+    {
+      "epoch": 0.037664903302953766,
+      "grad_norm": 0.46858713030815125,
+      "learning_rate": 0.0002,
+      "loss": 0.1596,
+      "step": 20710
+    },
+    {
+      "epoch": 0.03768309012251096,
+      "grad_norm": 0.15260715782642365,
+      "learning_rate": 0.0002,
+      "loss": 0.0835,
+      "step": 20720
+    },
+    {
+      "epoch": 0.03770127694206816,
+      "grad_norm": 0.2063397914171219,
+      "learning_rate": 0.0002,
+      "loss": 0.0845,
+      "step": 20730
+    },
+    {
+      "epoch": 0.037719463761625355,
+      "grad_norm": 0.16447599232196808,
+      "learning_rate": 0.0002,
+      "loss": 0.0595,
+      "step": 20740
+    },
+    {
+      "epoch": 0.03773765058118255,
+      "grad_norm": 0.020755184814333916,
+      "learning_rate": 0.0002,
+      "loss": 0.0164,
+      "step": 20750
+    },
+    {
+      "epoch": 0.03775583740073975,
+      "grad_norm": 0.23675021529197693,
+      "learning_rate": 0.0002,
+      "loss": 0.1634,
+      "step": 20760
+    },
+    {
+      "epoch": 0.037774024220296944,
+      "grad_norm": 0.08625516295433044,
+      "learning_rate": 0.0002,
+      "loss": 0.0685,
+      "step": 20770
+    },
+    {
+      "epoch": 0.03779221103985414,
+      "grad_norm": 0.043796882033348083,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 20780
+    },
+    {
+      "epoch": 0.037810397859411336,
+      "grad_norm": 0.20600435137748718,
+      "learning_rate": 0.0002,
+      "loss": 0.0651,
+      "step": 20790
+    },
+    {
+      "epoch": 0.03782858467896853,
+      "grad_norm": 0.04963940382003784,
+      "learning_rate": 0.0002,
+      "loss": 0.0202,
+      "step": 20800
+    },
+    {
+      "epoch": 0.03784677149852573,
+      "grad_norm": 0.34920167922973633,
+      "learning_rate": 0.0002,
+      "loss": 0.1494,
+      "step": 20810
+    },
+    {
+      "epoch": 0.037864958318082925,
+      "grad_norm": 0.18662041425704956,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 20820
+    },
+    {
+      "epoch": 0.03788314513764012,
+      "grad_norm": 0.12615887820720673,
+      "learning_rate": 0.0002,
+      "loss": 0.0856,
+      "step": 20830
+    },
+    {
+      "epoch": 0.03790133195719732,
+      "grad_norm": 0.1857282668352127,
+      "learning_rate": 0.0002,
+      "loss": 0.0676,
+      "step": 20840
+    },
+    {
+      "epoch": 0.037919518776754514,
+      "grad_norm": 0.05569197237491608,
+      "learning_rate": 0.0002,
+      "loss": 0.0181,
+      "step": 20850
+    },
+    {
+      "epoch": 0.03793770559631171,
+      "grad_norm": 0.29011765122413635,
+      "learning_rate": 0.0002,
+      "loss": 0.1418,
+      "step": 20860
+    },
+    {
+      "epoch": 0.03795589241586891,
+      "grad_norm": 0.14119744300842285,
+      "learning_rate": 0.0002,
+      "loss": 0.0812,
+      "step": 20870
+    },
+    {
+      "epoch": 0.0379740792354261,
+      "grad_norm": 0.039884984493255615,
+      "learning_rate": 0.0002,
+      "loss": 0.0781,
+      "step": 20880
+    },
+    {
+      "epoch": 0.0379922660549833,
+      "grad_norm": 0.23705685138702393,
+      "learning_rate": 0.0002,
+      "loss": 0.0621,
+      "step": 20890
+    },
+    {
+      "epoch": 0.038010452874540496,
+      "grad_norm": 0.07462739199399948,
+      "learning_rate": 0.0002,
+      "loss": 0.022,
+      "step": 20900
+    },
+    {
+      "epoch": 0.03802863969409769,
+      "grad_norm": 0.2610052824020386,
+      "learning_rate": 0.0002,
+      "loss": 0.1517,
+      "step": 20910
+    },
+    {
+      "epoch": 0.03804682651365489,
+      "grad_norm": 0.12775090336799622,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 20920
+    },
+    {
+      "epoch": 0.038065013333212085,
+      "grad_norm": 0.03661905974149704,
+      "learning_rate": 0.0002,
+      "loss": 0.0738,
+      "step": 20930
+    },
+    {
+      "epoch": 0.03808320015276928,
+      "grad_norm": 0.20907218754291534,
+      "learning_rate": 0.0002,
+      "loss": 0.0627,
+      "step": 20940
+    },
+    {
+      "epoch": 0.03810138697232648,
+      "grad_norm": 0.022804679349064827,
+      "learning_rate": 0.0002,
+      "loss": 0.0205,
+      "step": 20950
+    },
+    {
+      "epoch": 0.03811957379188368,
+      "grad_norm": 0.258284330368042,
+      "learning_rate": 0.0002,
+      "loss": 0.1428,
+      "step": 20960
+    },
+    {
+      "epoch": 0.03813776061144088,
+      "grad_norm": 0.1477317065000534,
+      "learning_rate": 0.0002,
+      "loss": 0.0789,
+      "step": 20970
+    },
+    {
+      "epoch": 0.03815594743099807,
+      "grad_norm": 0.0610325001180172,
+      "learning_rate": 0.0002,
+      "loss": 0.0836,
+      "step": 20980
+    },
+    {
+      "epoch": 0.03817413425055527,
+      "grad_norm": 0.18825507164001465,
+      "learning_rate": 0.0002,
+      "loss": 0.0621,
+      "step": 20990
+    },
+    {
+      "epoch": 0.038192321070112466,
+      "grad_norm": 0.03943372145295143,
+      "learning_rate": 0.0002,
+      "loss": 0.0185,
+      "step": 21000
+    },
+    {
+      "epoch": 0.03821050788966966,
+      "grad_norm": 0.34519344568252563,
+      "learning_rate": 0.0002,
+      "loss": 0.1345,
+      "step": 21010
+    },
+    {
+      "epoch": 0.03822869470922686,
+      "grad_norm": 0.09635084867477417,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 21020
+    },
+    {
+      "epoch": 0.038246881528784055,
+      "grad_norm": 0.032520972192287445,
+      "learning_rate": 0.0002,
+      "loss": 0.082,
+      "step": 21030
+    },
+    {
+      "epoch": 0.03826506834834125,
+      "grad_norm": 0.18068930506706238,
+      "learning_rate": 0.0002,
+      "loss": 0.0609,
+      "step": 21040
+    },
+    {
+      "epoch": 0.03828325516789845,
+      "grad_norm": 0.05550973862409592,
+      "learning_rate": 0.0002,
+      "loss": 0.0241,
+      "step": 21050
+    },
+    {
+      "epoch": 0.038301441987455644,
+      "grad_norm": 0.19561107456684113,
+      "learning_rate": 0.0002,
+      "loss": 0.1337,
+      "step": 21060
+    },
+    {
+      "epoch": 0.03831962880701284,
+      "grad_norm": 0.1852179914712906,
+      "learning_rate": 0.0002,
+      "loss": 0.0724,
+      "step": 21070
+    },
+    {
+      "epoch": 0.038337815626570036,
+      "grad_norm": 0.11915116757154465,
+      "learning_rate": 0.0002,
+      "loss": 0.0836,
+      "step": 21080
+    },
+    {
+      "epoch": 0.03835600244612723,
+      "grad_norm": 0.21116836369037628,
+      "learning_rate": 0.0002,
+      "loss": 0.0628,
+      "step": 21090
+    },
+    {
+      "epoch": 0.03837418926568443,
+      "grad_norm": 0.042745884507894516,
+      "learning_rate": 0.0002,
+      "loss": 0.0214,
+      "step": 21100
+    },
+    {
+      "epoch": 0.038392376085241625,
+      "grad_norm": 0.43089792132377625,
+      "learning_rate": 0.0002,
+      "loss": 0.1351,
+      "step": 21110
+    },
+    {
+      "epoch": 0.03841056290479882,
+      "grad_norm": 0.09607810527086258,
+      "learning_rate": 0.0002,
+      "loss": 0.0778,
+      "step": 21120
+    },
+    {
+      "epoch": 0.03842874972435602,
+      "grad_norm": 0.13603460788726807,
+      "learning_rate": 0.0002,
+      "loss": 0.0787,
+      "step": 21130
+    },
+    {
+      "epoch": 0.038446936543913214,
+      "grad_norm": 0.20110103487968445,
+      "learning_rate": 0.0002,
+      "loss": 0.067,
+      "step": 21140
+    },
+    {
+      "epoch": 0.03846512336347041,
+      "grad_norm": 0.042503997683525085,
+      "learning_rate": 0.0002,
+      "loss": 0.0194,
+      "step": 21150
+    },
+    {
+      "epoch": 0.03848331018302761,
+      "grad_norm": 0.2605084478855133,
+      "learning_rate": 0.0002,
+      "loss": 0.1374,
+      "step": 21160
+    },
+    {
+      "epoch": 0.0385014970025848,
+      "grad_norm": 0.09476794302463531,
+      "learning_rate": 0.0002,
+      "loss": 0.078,
+      "step": 21170
+    },
+    {
+      "epoch": 0.038519683822142,
+      "grad_norm": 0.03458428382873535,
+      "learning_rate": 0.0002,
+      "loss": 0.08,
+      "step": 21180
+    },
+    {
+      "epoch": 0.038537870641699196,
+      "grad_norm": 0.31196194887161255,
+      "learning_rate": 0.0002,
+      "loss": 0.0664,
+      "step": 21190
+    },
+    {
+      "epoch": 0.03855605746125639,
+      "grad_norm": 0.037113118916749954,
+      "learning_rate": 0.0002,
+      "loss": 0.0221,
+      "step": 21200
+    },
+    {
+      "epoch": 0.03857424428081359,
+      "grad_norm": 0.3699415922164917,
+      "learning_rate": 0.0002,
+      "loss": 0.1534,
+      "step": 21210
+    },
+    {
+      "epoch": 0.038592431100370785,
+      "grad_norm": 0.06454256922006607,
+      "learning_rate": 0.0002,
+      "loss": 0.0762,
+      "step": 21220
+    },
+    {
+      "epoch": 0.03861061791992798,
+      "grad_norm": 0.09858033806085587,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 21230
+    },
+    {
+      "epoch": 0.03862880473948518,
+      "grad_norm": 0.1482791304588318,
+      "learning_rate": 0.0002,
+      "loss": 0.062,
+      "step": 21240
+    },
+    {
+      "epoch": 0.038646991559042373,
+      "grad_norm": 0.031473588198423386,
+      "learning_rate": 0.0002,
+      "loss": 0.0163,
+      "step": 21250
+    },
+    {
+      "epoch": 0.03866517837859957,
+      "grad_norm": 0.09360513091087341,
+      "learning_rate": 0.0002,
+      "loss": 0.1397,
+      "step": 21260
+    },
+    {
+      "epoch": 0.038683365198156766,
+      "grad_norm": 0.10830901563167572,
+      "learning_rate": 0.0002,
+      "loss": 0.0789,
+      "step": 21270
+    },
+    {
+      "epoch": 0.03870155201771396,
+      "grad_norm": 0.08910014480352402,
+      "learning_rate": 0.0002,
+      "loss": 0.0758,
+      "step": 21280
+    },
+    {
+      "epoch": 0.03871973883727116,
+      "grad_norm": 0.21524523198604584,
+      "learning_rate": 0.0002,
+      "loss": 0.0628,
+      "step": 21290
+    },
+    {
+      "epoch": 0.038737925656828355,
+      "grad_norm": 0.03794678673148155,
+      "learning_rate": 0.0002,
+      "loss": 0.0229,
+      "step": 21300
+    },
+    {
+      "epoch": 0.03875611247638555,
+      "grad_norm": 0.46754345297813416,
+      "learning_rate": 0.0002,
+      "loss": 0.1291,
+      "step": 21310
+    },
+    {
+      "epoch": 0.03877429929594275,
+      "grad_norm": 0.07472983002662659,
+      "learning_rate": 0.0002,
+      "loss": 0.076,
+      "step": 21320
+    },
+    {
+      "epoch": 0.038792486115499944,
+      "grad_norm": 0.11820811778306961,
+      "learning_rate": 0.0002,
+      "loss": 0.0772,
+      "step": 21330
+    },
+    {
+      "epoch": 0.03881067293505714,
+      "grad_norm": 0.21140390634536743,
+      "learning_rate": 0.0002,
+      "loss": 0.0539,
+      "step": 21340
+    },
+    {
+      "epoch": 0.03882885975461434,
+      "grad_norm": 0.044819217175245285,
+      "learning_rate": 0.0002,
+      "loss": 0.0228,
+      "step": 21350
+    },
+    {
+      "epoch": 0.03884704657417153,
+      "grad_norm": 0.2267816811800003,
+      "learning_rate": 0.0002,
+      "loss": 0.1462,
+      "step": 21360
+    },
+    {
+      "epoch": 0.03886523339372873,
+      "grad_norm": 0.10087496787309647,
+      "learning_rate": 0.0002,
+      "loss": 0.0766,
+      "step": 21370
+    },
+    {
+      "epoch": 0.038883420213285926,
+      "grad_norm": 0.09982341527938843,
+      "learning_rate": 0.0002,
+      "loss": 0.0798,
+      "step": 21380
+    },
+    {
+      "epoch": 0.03890160703284312,
+      "grad_norm": 0.21729151904582977,
+      "learning_rate": 0.0002,
+      "loss": 0.0586,
+      "step": 21390
+    },
+    {
+      "epoch": 0.03891979385240032,
+      "grad_norm": 0.020691821351647377,
+      "learning_rate": 0.0002,
+      "loss": 0.0175,
+      "step": 21400
+    },
+    {
+      "epoch": 0.038937980671957514,
+      "grad_norm": 0.33531665802001953,
+      "learning_rate": 0.0002,
+      "loss": 0.149,
+      "step": 21410
+    },
+    {
+      "epoch": 0.03895616749151471,
+      "grad_norm": 0.11777795851230621,
+      "learning_rate": 0.0002,
+      "loss": 0.0736,
+      "step": 21420
+    },
+    {
+      "epoch": 0.03897435431107191,
+      "grad_norm": 0.07860718667507172,
+      "learning_rate": 0.0002,
+      "loss": 0.0788,
+      "step": 21430
+    },
+    {
+      "epoch": 0.0389925411306291,
+      "grad_norm": 0.16030597686767578,
+      "learning_rate": 0.0002,
+      "loss": 0.0581,
+      "step": 21440
+    },
+    {
+      "epoch": 0.0390107279501863,
+      "grad_norm": 0.01747356541454792,
+      "learning_rate": 0.0002,
+      "loss": 0.0185,
+      "step": 21450
+    },
+    {
+      "epoch": 0.039028914769743496,
+      "grad_norm": 0.2313859909772873,
+      "learning_rate": 0.0002,
+      "loss": 0.1383,
+      "step": 21460
+    },
+    {
+      "epoch": 0.03904710158930069,
+      "grad_norm": 0.14510080218315125,
+      "learning_rate": 0.0002,
+      "loss": 0.0805,
+      "step": 21470
+    },
+    {
+      "epoch": 0.03906528840885789,
+      "grad_norm": 0.04511871561408043,
+      "learning_rate": 0.0002,
+      "loss": 0.0793,
+      "step": 21480
+    },
+    {
+      "epoch": 0.039083475228415085,
+      "grad_norm": 0.24205265939235687,
+      "learning_rate": 0.0002,
+      "loss": 0.0624,
+      "step": 21490
+    },
+    {
+      "epoch": 0.03910166204797228,
+      "grad_norm": 0.08096791058778763,
+      "learning_rate": 0.0002,
+      "loss": 0.0208,
+      "step": 21500
+    },
+    {
+      "epoch": 0.03911984886752948,
+      "grad_norm": 0.14405490458011627,
+      "learning_rate": 0.0002,
+      "loss": 0.1189,
+      "step": 21510
+    },
+    {
+      "epoch": 0.039138035687086674,
+      "grad_norm": 0.06753374636173248,
+      "learning_rate": 0.0002,
+      "loss": 0.0772,
+      "step": 21520
+    },
+    {
+      "epoch": 0.03915622250664387,
+      "grad_norm": 0.029025042429566383,
+      "learning_rate": 0.0002,
+      "loss": 0.0761,
+      "step": 21530
+    },
+    {
+      "epoch": 0.039174409326201066,
+      "grad_norm": 0.2987070381641388,
+      "learning_rate": 0.0002,
+      "loss": 0.0656,
+      "step": 21540
+    },
+    {
+      "epoch": 0.03919259614575826,
+      "grad_norm": 0.04445091262459755,
+      "learning_rate": 0.0002,
+      "loss": 0.0241,
+      "step": 21550
+    },
+    {
+      "epoch": 0.03921078296531546,
+      "grad_norm": 0.34976306557655334,
+      "learning_rate": 0.0002,
+      "loss": 0.138,
+      "step": 21560
+    },
+    {
+      "epoch": 0.039228969784872655,
+      "grad_norm": 0.07521916925907135,
+      "learning_rate": 0.0002,
+      "loss": 0.0774,
+      "step": 21570
+    },
+    {
+      "epoch": 0.03924715660442985,
+      "grad_norm": 0.1445412039756775,
+      "learning_rate": 0.0002,
+      "loss": 0.087,
+      "step": 21580
+    },
+    {
+      "epoch": 0.03926534342398705,
+      "grad_norm": 0.2688128352165222,
+      "learning_rate": 0.0002,
+      "loss": 0.0712,
+      "step": 21590
+    },
+    {
+      "epoch": 0.039283530243544244,
+      "grad_norm": 0.05321233719587326,
+      "learning_rate": 0.0002,
+      "loss": 0.0245,
+      "step": 21600
+    },
+    {
+      "epoch": 0.03930171706310144,
+      "grad_norm": 0.44459134340286255,
+      "learning_rate": 0.0002,
+      "loss": 0.1524,
+      "step": 21610
+    },
+    {
+      "epoch": 0.03931990388265864,
+      "grad_norm": 0.13169553875923157,
+      "learning_rate": 0.0002,
+      "loss": 0.0726,
+      "step": 21620
+    },
+    {
+      "epoch": 0.03933809070221583,
+      "grad_norm": 0.0908237174153328,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 21630
+    },
+    {
+      "epoch": 0.03935627752177303,
+      "grad_norm": 0.18110623955726624,
+      "learning_rate": 0.0002,
+      "loss": 0.0606,
+      "step": 21640
+    },
+    {
+      "epoch": 0.03937446434133023,
+      "grad_norm": 0.021362677216529846,
+      "learning_rate": 0.0002,
+      "loss": 0.0175,
+      "step": 21650
+    },
+    {
+      "epoch": 0.03939265116088743,
+      "grad_norm": 0.27973899245262146,
+      "learning_rate": 0.0002,
+      "loss": 0.1641,
+      "step": 21660
+    },
+    {
+      "epoch": 0.039410837980444625,
+      "grad_norm": 0.09090718626976013,
+      "learning_rate": 0.0002,
+      "loss": 0.073,
+      "step": 21670
+    },
+    {
+      "epoch": 0.03942902480000182,
+      "grad_norm": 0.13408254086971283,
+      "learning_rate": 0.0002,
+      "loss": 0.0769,
+      "step": 21680
+    },
+    {
+      "epoch": 0.03944721161955902,
+      "grad_norm": 0.2530055046081543,
+      "learning_rate": 0.0002,
+      "loss": 0.0729,
+      "step": 21690
+    },
+    {
+      "epoch": 0.039465398439116214,
+      "grad_norm": 0.027523871511220932,
+      "learning_rate": 0.0002,
+      "loss": 0.017,
+      "step": 21700
+    },
+    {
+      "epoch": 0.03948358525867341,
+      "grad_norm": 0.2520642578601837,
+      "learning_rate": 0.0002,
+      "loss": 0.1804,
+      "step": 21710
+    },
+    {
+      "epoch": 0.03950177207823061,
+      "grad_norm": 0.11017465591430664,
+      "learning_rate": 0.0002,
+      "loss": 0.0767,
+      "step": 21720
+    },
+    {
+      "epoch": 0.0395199588977878,
+      "grad_norm": 0.05129052326083183,
+      "learning_rate": 0.0002,
+      "loss": 0.0723,
+      "step": 21730
+    },
+    {
+      "epoch": 0.039538145717345,
+      "grad_norm": 0.1846659779548645,
+      "learning_rate": 0.0002,
+      "loss": 0.0619,
+      "step": 21740
+    },
+    {
+      "epoch": 0.039556332536902196,
+      "grad_norm": 0.014305013231933117,
+      "learning_rate": 0.0002,
+      "loss": 0.0171,
+      "step": 21750
+    },
+    {
+      "epoch": 0.03957451935645939,
+      "grad_norm": 0.21667814254760742,
+      "learning_rate": 0.0002,
+      "loss": 0.157,
+      "step": 21760
+    },
+    {
+      "epoch": 0.03959270617601659,
+      "grad_norm": 0.21456903219223022,
+      "learning_rate": 0.0002,
+      "loss": 0.0803,
+      "step": 21770
+    },
+    {
+      "epoch": 0.039610892995573785,
+      "grad_norm": 0.03621416166424751,
+      "learning_rate": 0.0002,
+      "loss": 0.0796,
+      "step": 21780
+    },
+    {
+      "epoch": 0.03962907981513098,
+      "grad_norm": 0.20819205045700073,
+      "learning_rate": 0.0002,
+      "loss": 0.0633,
+      "step": 21790
+    },
+    {
+      "epoch": 0.03964726663468818,
+      "grad_norm": 0.06860963255167007,
+      "learning_rate": 0.0002,
+      "loss": 0.0172,
+      "step": 21800
+    },
+    {
+      "epoch": 0.039665453454245374,
+      "grad_norm": 0.2568039894104004,
+      "learning_rate": 0.0002,
+      "loss": 0.134,
+      "step": 21810
+    },
+    {
+      "epoch": 0.03968364027380257,
+      "grad_norm": 0.08747372031211853,
+      "learning_rate": 0.0002,
+      "loss": 0.0753,
+      "step": 21820
+    },
+    {
+      "epoch": 0.039701827093359766,
+      "grad_norm": 0.13403570652008057,
+      "learning_rate": 0.0002,
+      "loss": 0.0807,
+      "step": 21830
+    },
+    {
+      "epoch": 0.03972001391291696,
+      "grad_norm": 0.20756667852401733,
+      "learning_rate": 0.0002,
+      "loss": 0.0625,
+      "step": 21840
+    },
+    {
+      "epoch": 0.03973820073247416,
+      "grad_norm": 0.03678170591592789,
+      "learning_rate": 0.0002,
+      "loss": 0.019,
+      "step": 21850
+    },
+    {
+      "epoch": 0.039756387552031355,
+      "grad_norm": 0.1847693920135498,
+      "learning_rate": 0.0002,
+      "loss": 0.1385,
+      "step": 21860
+    },
+    {
+      "epoch": 0.03977457437158855,
+      "grad_norm": 0.1627635508775711,
+      "learning_rate": 0.0002,
+      "loss": 0.0765,
+      "step": 21870
+    },
+    {
+      "epoch": 0.03979276119114575,
+      "grad_norm": 0.0535571426153183,
+      "learning_rate": 0.0002,
+      "loss": 0.0741,
+      "step": 21880
+    },
+    {
+      "epoch": 0.039810948010702944,
+      "grad_norm": 0.3128276765346527,
+      "learning_rate": 0.0002,
+      "loss": 0.0598,
+      "step": 21890
+    },
+    {
+      "epoch": 0.03982913483026014,
+      "grad_norm": 0.03369860351085663,
+      "learning_rate": 0.0002,
+      "loss": 0.0217,
+      "step": 21900
+    },
+    {
+      "epoch": 0.03984732164981734,
+      "grad_norm": 0.1962599903345108,
+      "learning_rate": 0.0002,
+      "loss": 0.1319,
+      "step": 21910
+    },
+    {
+      "epoch": 0.03986550846937453,
+      "grad_norm": 0.1397421509027481,
+      "learning_rate": 0.0002,
+      "loss": 0.068,
+      "step": 21920
+    },
+    {
+      "epoch": 0.03988369528893173,
+      "grad_norm": 0.10252605378627777,
+      "learning_rate": 0.0002,
+      "loss": 0.0736,
+      "step": 21930
+    },
+    {
+      "epoch": 0.039901882108488926,
+      "grad_norm": 0.22179432213306427,
+      "learning_rate": 0.0002,
+      "loss": 0.0625,
+      "step": 21940
+    },
+    {
+      "epoch": 0.03992006892804612,
+      "grad_norm": 0.06068069487810135,
+      "learning_rate": 0.0002,
+      "loss": 0.0242,
+      "step": 21950
+    },
+    {
+      "epoch": 0.03993825574760332,
+      "grad_norm": 0.20243950188159943,
+      "learning_rate": 0.0002,
+      "loss": 0.143,
+      "step": 21960
+    },
+    {
+      "epoch": 0.039956442567160515,
+      "grad_norm": 0.11786511540412903,
+      "learning_rate": 0.0002,
+      "loss": 0.0779,
+      "step": 21970
+    },
+    {
+      "epoch": 0.03997462938671771,
+      "grad_norm": 0.08299421519041061,
+      "learning_rate": 0.0002,
+      "loss": 0.0774,
+      "step": 21980
+    },
+    {
+      "epoch": 0.03999281620627491,
+      "grad_norm": 0.2844075858592987,
+      "learning_rate": 0.0002,
+      "loss": 0.0711,
+      "step": 21990
+    },
+    {
+      "epoch": 0.040011003025832104,
+      "grad_norm": 0.034433312714099884,
+      "learning_rate": 0.0002,
+      "loss": 0.0217,
+      "step": 22000
+    },
+    {
+      "epoch": 0.0400291898453893,
+      "grad_norm": 0.3878481388092041,
+      "learning_rate": 0.0002,
+      "loss": 0.1525,
+      "step": 22010
+    },
+    {
+      "epoch": 0.040047376664946496,
+      "grad_norm": 0.16157971322536469,
+      "learning_rate": 0.0002,
+      "loss": 0.0788,
+      "step": 22020
+    },
+    {
+      "epoch": 0.04006556348450369,
+      "grad_norm": 0.10347063094377518,
+      "learning_rate": 0.0002,
+      "loss": 0.0809,
+      "step": 22030
+    },
+    {
+      "epoch": 0.04008375030406089,
+      "grad_norm": 0.20982638001441956,
+      "learning_rate": 0.0002,
+      "loss": 0.0662,
+      "step": 22040
+    },
+    {
+      "epoch": 0.040101937123618085,
+      "grad_norm": 5.856126308441162,
+      "learning_rate": 0.0002,
+      "loss": 0.0578,
+      "step": 22050
+    },
+    {
+      "epoch": 0.04012012394317528,
+      "grad_norm": 0.21289357542991638,
+      "learning_rate": 0.0002,
+      "loss": 0.1257,
+      "step": 22060
+    },
+    {
+      "epoch": 0.04013831076273248,
+      "grad_norm": 0.040848907083272934,
+      "learning_rate": 0.0002,
+      "loss": 0.0783,
+      "step": 22070
+    },
+    {
+      "epoch": 0.040156497582289674,
+      "grad_norm": 0.056517478078603745,
+      "learning_rate": 0.0002,
+      "loss": 0.0693,
+      "step": 22080
+    },
+    {
+      "epoch": 0.04017468440184687,
+      "grad_norm": 0.274312287569046,
+      "learning_rate": 0.0002,
+      "loss": 0.0685,
+      "step": 22090
+    },
+    {
+      "epoch": 0.04019287122140407,
+      "grad_norm": 0.06353340297937393,
+      "learning_rate": 0.0002,
+      "loss": 0.0263,
+      "step": 22100
+    },
+    {
+      "epoch": 0.04021105804096126,
+      "grad_norm": 0.287201464176178,
+      "learning_rate": 0.0002,
+      "loss": 0.1425,
+      "step": 22110
+    },
+    {
+      "epoch": 0.04022924486051846,
+      "grad_norm": 0.0990116223692894,
+      "learning_rate": 0.0002,
+      "loss": 0.0732,
+      "step": 22120
+    },
+    {
+      "epoch": 0.040247431680075656,
+      "grad_norm": 0.03471527248620987,
+      "learning_rate": 0.0002,
+      "loss": 0.0806,
+      "step": 22130
+    },
+    {
+      "epoch": 0.04026561849963285,
+      "grad_norm": 0.16411902010440826,
+      "learning_rate": 0.0002,
+      "loss": 0.0646,
+      "step": 22140
+    },
+    {
+      "epoch": 0.04028380531919005,
+      "grad_norm": 0.032927367836236954,
+      "learning_rate": 0.0002,
+      "loss": 0.0225,
+      "step": 22150
+    },
+    {
+      "epoch": 0.040301992138747245,
+      "grad_norm": 0.31128716468811035,
+      "learning_rate": 0.0002,
+      "loss": 0.1227,
+      "step": 22160
+    },
+    {
+      "epoch": 0.04032017895830444,
+      "grad_norm": 0.14056596159934998,
+      "learning_rate": 0.0002,
+      "loss": 0.0866,
+      "step": 22170
+    },
+    {
+      "epoch": 0.04033836577786164,
+      "grad_norm": 0.10555677115917206,
+      "learning_rate": 0.0002,
+      "loss": 0.0785,
+      "step": 22180
+    },
+    {
+      "epoch": 0.040356552597418834,
+      "grad_norm": 0.25597816705703735,
+      "learning_rate": 0.0002,
+      "loss": 0.0667,
+      "step": 22190
+    },
+    {
+      "epoch": 0.04037473941697603,
+      "grad_norm": 0.04694845899939537,
+      "learning_rate": 0.0002,
+      "loss": 0.021,
+      "step": 22200
+    },
+    {
+      "epoch": 0.040392926236533226,
+      "grad_norm": 0.2536766529083252,
+      "learning_rate": 0.0002,
+      "loss": 0.1485,
+      "step": 22210
+    },
+    {
+      "epoch": 0.04041111305609042,
+      "grad_norm": 0.0536673367023468,
+      "learning_rate": 0.0002,
+      "loss": 0.0737,
+      "step": 22220
+    },
+    {
+      "epoch": 0.04042929987564762,
+      "grad_norm": 0.13121111690998077,
+      "learning_rate": 0.0002,
+      "loss": 0.0793,
+      "step": 22230
+    },
+    {
+      "epoch": 0.040447486695204815,
+      "grad_norm": 0.23850645124912262,
+      "learning_rate": 0.0002,
+      "loss": 0.0698,
+      "step": 22240
+    },
+    {
+      "epoch": 0.04046567351476201,
+      "grad_norm": 0.04178560525178909,
+      "learning_rate": 0.0002,
+      "loss": 0.0212,
+      "step": 22250
+    },
+    {
+      "epoch": 0.04048386033431921,
+      "grad_norm": 0.42834579944610596,
+      "learning_rate": 0.0002,
+      "loss": 0.1352,
+      "step": 22260
+    },
+    {
+      "epoch": 0.040502047153876404,
+      "grad_norm": 0.050178542733192444,
+      "learning_rate": 0.0002,
+      "loss": 0.0853,
+      "step": 22270
+    },
+    {
+      "epoch": 0.0405202339734336,
+      "grad_norm": 0.042758237570524216,
+      "learning_rate": 0.0002,
+      "loss": 0.0709,
+      "step": 22280
+    },
+    {
+      "epoch": 0.0405384207929908,
+      "grad_norm": 0.2604416012763977,
+      "learning_rate": 0.0002,
+      "loss": 0.0643,
+      "step": 22290
+    },
+    {
+      "epoch": 0.04055660761254799,
+      "grad_norm": 0.06166388466954231,
+      "learning_rate": 0.0002,
+      "loss": 0.0236,
+      "step": 22300
+    },
+    {
+      "epoch": 0.04057479443210519,
+      "grad_norm": 0.2337518334388733,
+      "learning_rate": 0.0002,
+      "loss": 0.132,
+      "step": 22310
+    },
+    {
+      "epoch": 0.040592981251662386,
+      "grad_norm": 0.15794694423675537,
+      "learning_rate": 0.0002,
+      "loss": 0.0739,
+      "step": 22320
+    },
+    {
+      "epoch": 0.04061116807121959,
+      "grad_norm": 0.12059915065765381,
+      "learning_rate": 0.0002,
+      "loss": 0.0743,
+      "step": 22330
+    },
+    {
+      "epoch": 0.040629354890776785,
+      "grad_norm": 0.25351977348327637,
+      "learning_rate": 0.0002,
+      "loss": 0.065,
+      "step": 22340
+    },
+    {
+      "epoch": 0.04064754171033398,
+      "grad_norm": 0.03265364468097687,
+      "learning_rate": 0.0002,
+      "loss": 0.02,
+      "step": 22350
+    },
+    {
+      "epoch": 0.04066572852989118,
+      "grad_norm": 0.22959749400615692,
+      "learning_rate": 0.0002,
+      "loss": 0.1278,
+      "step": 22360
+    },
+    {
+      "epoch": 0.040683915349448374,
+      "grad_norm": 0.11381889134645462,
+      "learning_rate": 0.0002,
+      "loss": 0.0823,
+      "step": 22370
+    },
+    {
+      "epoch": 0.04070210216900557,
+      "grad_norm": 0.03541165217757225,
+      "learning_rate": 0.0002,
+      "loss": 0.0809,
+      "step": 22380
+    },
+    {
+      "epoch": 0.04072028898856277,
+      "grad_norm": 0.20604047179222107,
+      "learning_rate": 0.0002,
+      "loss": 0.0693,
+      "step": 22390
+    },
+    {
+      "epoch": 0.04073847580811996,
+      "grad_norm": 0.051576532423496246,
+      "learning_rate": 0.0002,
+      "loss": 0.0213,
+      "step": 22400
+    },
+    {
+      "epoch": 0.04075666262767716,
+      "grad_norm": 0.208265483379364,
+      "learning_rate": 0.0002,
+      "loss": 0.1203,
+      "step": 22410
+    },
+    {
+      "epoch": 0.040774849447234356,
+      "grad_norm": 0.14376410841941833,
+      "learning_rate": 0.0002,
+      "loss": 0.0832,
+      "step": 22420
+    },
+    {
+      "epoch": 0.04079303626679155,
+      "grad_norm": 0.0634629875421524,
+      "learning_rate": 0.0002,
+      "loss": 0.0797,
+      "step": 22430
+    },
+    {
+      "epoch": 0.04081122308634875,
+      "grad_norm": 0.22782418131828308,
+      "learning_rate": 0.0002,
+      "loss": 0.0594,
+      "step": 22440
+    },
+    {
+      "epoch": 0.040829409905905945,
+      "grad_norm": 0.034153662621974945,
+      "learning_rate": 0.0002,
+      "loss": 0.0197,
+      "step": 22450
+    },
+    {
+      "epoch": 0.04084759672546314,
+      "grad_norm": 0.22994177043437958,
+      "learning_rate": 0.0002,
+      "loss": 0.1276,
+      "step": 22460
+    },
+    {
+      "epoch": 0.04086578354502034,
+      "grad_norm": 0.37397289276123047,
+      "learning_rate": 0.0002,
+      "loss": 0.0794,
+      "step": 22470
+    },
+    {
+      "epoch": 0.040883970364577533,
+      "grad_norm": 0.03585643321275711,
+      "learning_rate": 0.0002,
+      "loss": 0.0765,
+      "step": 22480
+    },
+    {
+      "epoch": 0.04090215718413473,
+      "grad_norm": 0.2266087681055069,
+      "learning_rate": 0.0002,
+      "loss": 0.0661,
+      "step": 22490
+    },
+    {
+      "epoch": 0.040920344003691926,
+      "grad_norm": 0.03867397829890251,
+      "learning_rate": 0.0002,
+      "loss": 0.0241,
+      "step": 22500
+    },
+    {
+      "epoch": 0.04093853082324912,
+      "grad_norm": 0.23483702540397644,
+      "learning_rate": 0.0002,
+      "loss": 0.1442,
+      "step": 22510
+    },
+    {
+      "epoch": 0.04095671764280632,
+      "grad_norm": 0.11447428911924362,
+      "learning_rate": 0.0002,
+      "loss": 0.0759,
+      "step": 22520
+    },
+    {
+      "epoch": 0.040974904462363515,
+      "grad_norm": 0.1060417965054512,
+      "learning_rate": 0.0002,
+      "loss": 0.0792,
+      "step": 22530
+    },
+    {
+      "epoch": 0.04099309128192071,
+      "grad_norm": 0.1915966123342514,
+      "learning_rate": 0.0002,
+      "loss": 0.0679,
+      "step": 22540
+    },
+    {
+      "epoch": 0.04101127810147791,
+      "grad_norm": 0.05328527092933655,
+      "learning_rate": 0.0002,
+      "loss": 0.0211,
+      "step": 22550
+    },
+    {
+      "epoch": 0.041029464921035104,
+      "grad_norm": 0.31612515449523926,
+      "learning_rate": 0.0002,
+      "loss": 0.1395,
+      "step": 22560
+    },
+    {
+      "epoch": 0.0410476517405923,
+      "grad_norm": 0.1860841065645218,
+      "learning_rate": 0.0002,
+      "loss": 0.0704,
+      "step": 22570
+    },
+    {
+      "epoch": 0.0410658385601495,
+      "grad_norm": 0.11183702945709229,
+      "learning_rate": 0.0002,
+      "loss": 0.073,
+      "step": 22580
+    },
+    {
+      "epoch": 0.04108402537970669,
+      "grad_norm": 0.2028307020664215,
+      "learning_rate": 0.0002,
+      "loss": 0.0592,
+      "step": 22590
+    },
+    {
+      "epoch": 0.04110221219926389,
+      "grad_norm": 0.032915905117988586,
+      "learning_rate": 0.0002,
+      "loss": 0.0211,
+      "step": 22600
+    },
+    {
+      "epoch": 0.041120399018821086,
+      "grad_norm": 0.2932131588459015,
+      "learning_rate": 0.0002,
+      "loss": 0.1542,
+      "step": 22610
+    },
+    {
+      "epoch": 0.04113858583837828,
+      "grad_norm": 0.08883325010538101,
+      "learning_rate": 0.0002,
+      "loss": 0.079,
+      "step": 22620
+    },
+    {
+      "epoch": 0.04115677265793548,
+      "grad_norm": 0.07874555885791779,
+      "learning_rate": 0.0002,
+      "loss": 0.0801,
+      "step": 22630
+    },
+    {
+      "epoch": 0.041174959477492674,
+      "grad_norm": 0.13785040378570557,
+      "learning_rate": 0.0002,
+      "loss": 0.0636,
+      "step": 22640
+    },
+    {
+      "epoch": 0.04119314629704987,
+      "grad_norm": 0.0321812778711319,
+      "learning_rate": 0.0002,
+      "loss": 0.0208,
+      "step": 22650
+    },
+    {
+      "epoch": 0.04121133311660707,
+      "grad_norm": 0.142785906791687,
+      "learning_rate": 0.0002,
+      "loss": 0.1292,
+      "step": 22660
+    },
+    {
+      "epoch": 0.04122951993616426,
+      "grad_norm": 0.15572668612003326,
+      "learning_rate": 0.0002,
+      "loss": 0.0774,
+      "step": 22670
+    },
+    {
+      "epoch": 0.04124770675572146,
+      "grad_norm": 0.033191781491041183,
+      "learning_rate": 0.0002,
+      "loss": 0.0805,
+      "step": 22680
+    },
+    {
+      "epoch": 0.041265893575278656,
+      "grad_norm": 0.23840776085853577,
+      "learning_rate": 0.0002,
+      "loss": 0.06,
+      "step": 22690
+    },
+    {
+      "epoch": 0.04128408039483585,
+      "grad_norm": 0.05943412706255913,
+      "learning_rate": 0.0002,
+      "loss": 0.0215,
+      "step": 22700
+    },
+    {
+      "epoch": 0.04130226721439305,
+      "grad_norm": 0.05142183229327202,
+      "learning_rate": 0.0002,
+      "loss": 0.1181,
+      "step": 22710
+    },
+    {
+      "epoch": 0.041320454033950245,
+      "grad_norm": 0.1583058387041092,
+      "learning_rate": 0.0002,
+      "loss": 0.0822,
+      "step": 22720
+    },
+    {
+      "epoch": 0.04133864085350744,
+      "grad_norm": 0.035809941589832306,
+      "learning_rate": 0.0002,
+      "loss": 0.0723,
+      "step": 22730
+    },
+    {
+      "epoch": 0.04135682767306464,
+      "grad_norm": 0.24066607654094696,
+      "learning_rate": 0.0002,
+      "loss": 0.0621,
+      "step": 22740
+    },
+    {
+      "epoch": 0.041375014492621834,
+      "grad_norm": 0.0327225998044014,
+      "learning_rate": 0.0002,
+      "loss": 0.0264,
+      "step": 22750
+    },
+    {
+      "epoch": 0.04139320131217903,
+      "grad_norm": 0.16599033772945404,
+      "learning_rate": 0.0002,
+      "loss": 0.1082,
+      "step": 22760
+    },
+    {
+      "epoch": 0.041411388131736226,
+      "grad_norm": 0.18834830820560455,
+      "learning_rate": 0.0002,
+      "loss": 0.0767,
+      "step": 22770
+    },
+    {
+      "epoch": 0.04142957495129342,
+      "grad_norm": 0.04162973538041115,
+      "learning_rate": 0.0002,
+      "loss": 0.0821,
+      "step": 22780
+    },
+    {
+      "epoch": 0.04144776177085062,
+      "grad_norm": 0.21065399050712585,
+      "learning_rate": 0.0002,
+      "loss": 0.0591,
+      "step": 22790
+    },
+    {
+      "epoch": 0.041465948590407815,
+      "grad_norm": 0.03744394704699516,
+      "learning_rate": 0.0002,
+      "loss": 0.0199,
+      "step": 22800
+    },
+    {
+      "epoch": 0.04148413540996501,
+      "grad_norm": 0.30440911650657654,
+      "learning_rate": 0.0002,
+      "loss": 0.1321,
+      "step": 22810
+    },
+    {
+      "epoch": 0.04150232222952221,
+      "grad_norm": 0.07215052098035812,
+      "learning_rate": 0.0002,
+      "loss": 0.077,
+      "step": 22820
+    },
+    {
+      "epoch": 0.041520509049079404,
+      "grad_norm": 0.0822744220495224,
+      "learning_rate": 0.0002,
+      "loss": 0.0695,
+      "step": 22830
+    },
+    {
+      "epoch": 0.0415386958686366,
+      "grad_norm": 0.20610104501247406,
+      "learning_rate": 0.0002,
+      "loss": 0.0668,
+      "step": 22840
+    },
+    {
+      "epoch": 0.0415568826881938,
+      "grad_norm": 0.05089128017425537,
+      "learning_rate": 0.0002,
+      "loss": 0.0275,
+      "step": 22850
+    },
+    {
+      "epoch": 0.04157506950775099,
+      "grad_norm": 0.23365797102451324,
+      "learning_rate": 0.0002,
+      "loss": 0.1308,
+      "step": 22860
+    },
+    {
+      "epoch": 0.04159325632730819,
+      "grad_norm": 0.03983612358570099,
+      "learning_rate": 0.0002,
+      "loss": 0.0738,
+      "step": 22870
+    },
+    {
+      "epoch": 0.041611443146865386,
+      "grad_norm": 0.12472117692232132,
+      "learning_rate": 0.0002,
+      "loss": 0.082,
+      "step": 22880
+    },
+    {
+      "epoch": 0.04162962996642258,
+      "grad_norm": 0.19599118828773499,
+      "learning_rate": 0.0002,
+      "loss": 0.0614,
+      "step": 22890
+    },
+    {
+      "epoch": 0.04164781678597978,
+      "grad_norm": 0.04077763110399246,
+      "learning_rate": 0.0002,
+      "loss": 0.0298,
+      "step": 22900
+    },
+    {
+      "epoch": 0.041666003605536975,
+      "grad_norm": 0.3027828633785248,
+      "learning_rate": 0.0002,
+      "loss": 0.1294,
+      "step": 22910
+    },
+    {
+      "epoch": 0.04168419042509417,
+      "grad_norm": 0.1551598757505417,
+      "learning_rate": 0.0002,
+      "loss": 0.0716,
+      "step": 22920
+    },
+    {
+      "epoch": 0.04170237724465137,
+      "grad_norm": 0.06512947380542755,
+      "learning_rate": 0.0002,
+      "loss": 0.071,
+      "step": 22930
+    },
+    {
+      "epoch": 0.041720564064208564,
+      "grad_norm": 0.2486017346382141,
+      "learning_rate": 0.0002,
+      "loss": 0.0726,
+      "step": 22940
+    },
+    {
+      "epoch": 0.04173875088376576,
+      "grad_norm": 0.0658118799328804,
+      "learning_rate": 0.0002,
+      "loss": 0.0211,
+      "step": 22950
+    },
+    {
+      "epoch": 0.041756937703322956,
+      "grad_norm": 0.18327641487121582,
+      "learning_rate": 0.0002,
+      "loss": 0.1307,
+      "step": 22960
+    },
+    {
+      "epoch": 0.04177512452288015,
+      "grad_norm": 0.06218123063445091,
+      "learning_rate": 0.0002,
+      "loss": 0.0839,
+      "step": 22970
+    },
+    {
+      "epoch": 0.04179331134243735,
+      "grad_norm": 0.07085203379392624,
+      "learning_rate": 0.0002,
+      "loss": 0.0786,
+      "step": 22980
+    },
+    {
+      "epoch": 0.041811498161994545,
+      "grad_norm": 0.19552426040172577,
+      "learning_rate": 0.0002,
+      "loss": 0.0646,
+      "step": 22990
+    },
+    {
+      "epoch": 0.04182968498155174,
+      "grad_norm": 0.06710335612297058,
+      "learning_rate": 0.0002,
+      "loss": 0.0252,
+      "step": 23000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 100000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 250,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0398767809662812e+19,
+  "train_batch_size": 24,
+  "trial_name": null,
+  "trial_params": null
+}