diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4086 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 6.961584633853541,
+  "eval_steps": 500,
+  "global_step": 5799,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.012004801920768308,
+      "grad_norm": 0.02549879066646099,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.9862,
+      "step": 10
+    },
+    {
+      "epoch": 0.024009603841536616,
+      "grad_norm": 0.03309444710612297,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.9386,
+      "step": 20
+    },
+    {
+      "epoch": 0.03601440576230492,
+      "grad_norm": 0.02672477439045906,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.9425,
+      "step": 30
+    },
+    {
+      "epoch": 0.04801920768307323,
+      "grad_norm": 0.02686227485537529,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.9175,
+      "step": 40
+    },
+    {
+      "epoch": 0.060024009603841535,
+      "grad_norm": 0.025844795629382133,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.9963,
+      "step": 50
+    },
+    {
+      "epoch": 0.07202881152460984,
+      "grad_norm": 0.032747991383075714,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.9103,
+      "step": 60
+    },
+    {
+      "epoch": 0.08403361344537816,
+      "grad_norm": 0.03013288602232933,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.9381,
+      "step": 70
+    },
+    {
+      "epoch": 0.09603841536614646,
+      "grad_norm": 0.03312341496348381,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.9105,
+      "step": 80
+    },
+    {
+      "epoch": 0.10804321728691477,
+      "grad_norm": 0.03535911813378334,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.9662,
+      "step": 90
+    },
+    {
+      "epoch": 0.12004801920768307,
+      "grad_norm": 0.03163569048047066,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8886,
+      "step": 100
+    },
+    {
+      "epoch": 0.13205282112845138,
+      "grad_norm": 0.03126745671033859,
+      "learning_rate": 4.4e-06,
+      "loss": 0.9677,
+      "step": 110
+    },
+    {
+      "epoch": 0.14405762304921968,
+      "grad_norm": 0.03188377246260643,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.9096,
+      "step": 120
+    },
+    {
+      "epoch": 0.15606242496998798,
+      "grad_norm": 0.03766326606273651,
+      "learning_rate": 5.2e-06,
+      "loss": 0.9868,
+      "step": 130
+    },
+    {
+      "epoch": 0.16806722689075632,
+      "grad_norm": 0.03288175165653229,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.9614,
+      "step": 140
+    },
+    {
+      "epoch": 0.18007202881152462,
+      "grad_norm": 0.03540439158678055,
+      "learning_rate": 6e-06,
+      "loss": 0.9471,
+      "step": 150
+    },
+    {
+      "epoch": 0.19207683073229292,
+      "grad_norm": 0.0429881289601326,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.8901,
+      "step": 160
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 0.04951930418610573,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.9403,
+      "step": 170
+    },
+    {
+      "epoch": 0.21608643457382953,
+      "grad_norm": 0.049168169498443604,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.8943,
+      "step": 180
+    },
+    {
+      "epoch": 0.22809123649459784,
+      "grad_norm": 0.04803888127207756,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.8852,
+      "step": 190
+    },
+    {
+      "epoch": 0.24009603841536614,
+      "grad_norm": 0.04201197251677513,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.8882,
+      "step": 200
+    },
+    {
+      "epoch": 0.25210084033613445,
+      "grad_norm": 0.04449041560292244,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.8777,
+      "step": 210
+    },
+    {
+      "epoch": 0.26410564225690275,
+      "grad_norm": 0.036004405468702316,
+      "learning_rate": 8.8e-06,
+      "loss": 0.8496,
+      "step": 220
+    },
+    {
+      "epoch": 0.27611044417767105,
+      "grad_norm": 0.04503042995929718,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.8712,
+      "step": 230
+    },
+    {
+      "epoch": 0.28811524609843936,
+      "grad_norm": 0.04312640801072121,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.8459,
+      "step": 240
+    },
+    {
+      "epoch": 0.30012004801920766,
+      "grad_norm": 0.03977839648723602,
+      "learning_rate": 1e-05,
+      "loss": 0.8705,
+      "step": 250
+    },
+    {
+      "epoch": 0.31212484993997597,
+      "grad_norm": 0.039520714432001114,
+      "learning_rate": 9.999919867416569e-06,
+      "loss": 0.8751,
+      "step": 260
+    },
+    {
+      "epoch": 0.3241296518607443,
+      "grad_norm": 0.037476055324077606,
+      "learning_rate": 9.999679472234764e-06,
+      "loss": 0.8137,
+      "step": 270
+    },
+    {
+      "epoch": 0.33613445378151263,
+      "grad_norm": 0.0433887243270874,
+      "learning_rate": 9.999278822159985e-06,
+      "loss": 0.8147,
+      "step": 280
+    },
+    {
+      "epoch": 0.34813925570228094,
+      "grad_norm": 0.04283862188458443,
+      "learning_rate": 9.998717930034278e-06,
+      "loss": 0.8642,
+      "step": 290
+    },
+    {
+      "epoch": 0.36014405762304924,
+      "grad_norm": 0.03704555705189705,
+      "learning_rate": 9.997996813835938e-06,
+      "loss": 0.8481,
+      "step": 300
+    },
+    {
+      "epoch": 0.37214885954381755,
+      "grad_norm": 0.038123104721307755,
+      "learning_rate": 9.997115496678926e-06,
+      "loss": 0.8328,
+      "step": 310
+    },
+    {
+      "epoch": 0.38415366146458585,
+      "grad_norm": 0.03972741588950157,
+      "learning_rate": 9.996074006812133e-06,
+      "loss": 0.8846,
+      "step": 320
+    },
+    {
+      "epoch": 0.39615846338535415,
+      "grad_norm": 0.03823426738381386,
+      "learning_rate": 9.994872377618467e-06,
+      "loss": 0.8571,
+      "step": 330
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.03882453590631485,
+      "learning_rate": 9.993510647613787e-06,
+      "loss": 0.8664,
+      "step": 340
+    },
+    {
+      "epoch": 0.42016806722689076,
+      "grad_norm": 0.040137797594070435,
+      "learning_rate": 9.991988860445672e-06,
+      "loss": 0.8386,
+      "step": 350
+    },
+    {
+      "epoch": 0.43217286914765907,
+      "grad_norm": 0.0372915081679821,
+      "learning_rate": 9.990307064892015e-06,
+      "loss": 0.8799,
+      "step": 360
+    },
+    {
+      "epoch": 0.44417767106842737,
+      "grad_norm": 0.03684937208890915,
+      "learning_rate": 9.988465314859469e-06,
+      "loss": 0.8352,
+      "step": 370
+    },
+    {
+      "epoch": 0.4561824729891957,
+      "grad_norm": 0.03629535436630249,
+      "learning_rate": 9.986463669381705e-06,
+      "loss": 0.8084,
+      "step": 380
+    },
+    {
+      "epoch": 0.468187274909964,
+      "grad_norm": 0.034760646522045135,
+      "learning_rate": 9.984302192617534e-06,
+      "loss": 0.8363,
+      "step": 390
+    },
+    {
+      "epoch": 0.4801920768307323,
+      "grad_norm": 0.033850379288196564,
+      "learning_rate": 9.981980953848844e-06,
+      "loss": 0.8209,
+      "step": 400
+    },
+    {
+      "epoch": 0.4921968787515006,
+      "grad_norm": 0.031105060130357742,
+      "learning_rate": 9.979500027478377e-06,
+      "loss": 0.8589,
+      "step": 410
+    },
+    {
+      "epoch": 0.5042016806722689,
+      "grad_norm": 0.034456316381692886,
+      "learning_rate": 9.97685949302735e-06,
+      "loss": 0.8599,
+      "step": 420
+    },
+    {
+      "epoch": 0.5162064825930373,
+      "grad_norm": 0.03859219700098038,
+      "learning_rate": 9.974059435132901e-06,
+      "loss": 0.8862,
+      "step": 430
+    },
+    {
+      "epoch": 0.5282112845138055,
+      "grad_norm": 0.031010838225483894,
+      "learning_rate": 9.97109994354538e-06,
+      "loss": 0.8565,
+      "step": 440
+    },
+    {
+      "epoch": 0.5402160864345739,
+      "grad_norm": 0.03447699174284935,
+      "learning_rate": 9.967981113125466e-06,
+      "loss": 0.8691,
+      "step": 450
+    },
+    {
+      "epoch": 0.5522208883553421,
+      "grad_norm": 0.033097319304943085,
+      "learning_rate": 9.964703043841139e-06,
+      "loss": 0.8003,
+      "step": 460
+    },
+    {
+      "epoch": 0.5642256902761105,
+      "grad_norm": 0.03505665436387062,
+      "learning_rate": 9.961265840764463e-06,
+      "loss": 0.863,
+      "step": 470
+    },
+    {
+      "epoch": 0.5762304921968787,
+      "grad_norm": 0.04121314734220505,
+      "learning_rate": 9.957669614068222e-06,
+      "loss": 0.8208,
+      "step": 480
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 0.03139583393931389,
+      "learning_rate": 9.953914479022388e-06,
+      "loss": 0.8637,
+      "step": 490
+    },
+    {
+      "epoch": 0.6002400960384153,
+      "grad_norm": 0.03294196352362633,
+      "learning_rate": 9.950000555990434e-06,
+      "loss": 0.804,
+      "step": 500
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 0.03441837430000305,
+      "learning_rate": 9.945927970425462e-06,
+      "loss": 0.842,
+      "step": 510
+    },
+    {
+      "epoch": 0.6242496998799519,
+      "grad_norm": 0.03234728053212166,
+      "learning_rate": 9.941696852866195e-06,
+      "loss": 0.8194,
+      "step": 520
+    },
+    {
+      "epoch": 0.6362545018007203,
+      "grad_norm": 0.03413551673293114,
+      "learning_rate": 9.937307338932783e-06,
+      "loss": 0.7902,
+      "step": 530
+    },
+    {
+      "epoch": 0.6482593037214885,
+      "grad_norm": 0.03235715255141258,
+      "learning_rate": 9.932759569322467e-06,
+      "loss": 0.8163,
+      "step": 540
+    },
+    {
+      "epoch": 0.6602641056422569,
+      "grad_norm": 0.03173699602484703,
+      "learning_rate": 9.928053689805054e-06,
+      "loss": 0.8364,
+      "step": 550
+    },
+    {
+      "epoch": 0.6722689075630253,
+      "grad_norm": 0.035562604665756226,
+      "learning_rate": 9.923189851218259e-06,
+      "loss": 0.8336,
+      "step": 560
+    },
+    {
+      "epoch": 0.6842737094837935,
+      "grad_norm": 0.03246847167611122,
+      "learning_rate": 9.918168209462862e-06,
+      "loss": 0.8354,
+      "step": 570
+    },
+    {
+      "epoch": 0.6962785114045619,
+      "grad_norm": 0.03735557198524475,
+      "learning_rate": 9.912988925497716e-06,
+      "loss": 0.8705,
+      "step": 580
+    },
+    {
+      "epoch": 0.7082833133253301,
+      "grad_norm": 0.04147843271493912,
+      "learning_rate": 9.907652165334579e-06,
+      "loss": 0.8296,
+      "step": 590
+    },
+    {
+      "epoch": 0.7202881152460985,
+      "grad_norm": 0.03415181115269661,
+      "learning_rate": 9.902158100032804e-06,
+      "loss": 0.8191,
+      "step": 600
+    },
+    {
+      "epoch": 0.7322929171668667,
+      "grad_norm": 0.03535843268036842,
+      "learning_rate": 9.89650690569385e-06,
+      "loss": 0.8365,
+      "step": 610
+    },
+    {
+      "epoch": 0.7442977190876351,
+      "grad_norm": 0.036314673721790314,
+      "learning_rate": 9.890698763455637e-06,
+      "loss": 0.8575,
+      "step": 620
+    },
+    {
+      "epoch": 0.7563025210084033,
+      "grad_norm": 0.033384788781404495,
+      "learning_rate": 9.884733859486745e-06,
+      "loss": 0.8736,
+      "step": 630
+    },
+    {
+      "epoch": 0.7683073229291717,
+      "grad_norm": 0.036313604563474655,
+      "learning_rate": 9.878612384980435e-06,
+      "loss": 0.8104,
+      "step": 640
+    },
+    {
+      "epoch": 0.78031212484994,
+      "grad_norm": 0.03539718687534332,
+      "learning_rate": 9.872334536148538e-06,
+      "loss": 0.8535,
+      "step": 650
+    },
+    {
+      "epoch": 0.7923169267707083,
+      "grad_norm": 0.040795594453811646,
+      "learning_rate": 9.86590051421515e-06,
+      "loss": 0.859,
+      "step": 660
+    },
+    {
+      "epoch": 0.8043217286914766,
+      "grad_norm": 0.03459113836288452,
+      "learning_rate": 9.85931052541019e-06,
+      "loss": 0.8637,
+      "step": 670
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.03594481945037842,
+      "learning_rate": 9.852564780962793e-06,
+      "loss": 0.8462,
+      "step": 680
+    },
+    {
+      "epoch": 0.8283313325330132,
+      "grad_norm": 0.03311854600906372,
+      "learning_rate": 9.845663497094526e-06,
+      "loss": 0.8201,
+      "step": 690
+    },
+    {
+      "epoch": 0.8403361344537815,
+      "grad_norm": 0.04675761237740517,
+      "learning_rate": 9.838606895012474e-06,
+      "loss": 0.8724,
+      "step": 700
+    },
+    {
+      "epoch": 0.8523409363745498,
+      "grad_norm": 0.035599809139966965,
+      "learning_rate": 9.831395200902137e-06,
+      "loss": 0.7937,
+      "step": 710
+    },
+    {
+      "epoch": 0.8643457382953181,
+      "grad_norm": 0.036689773201942444,
+      "learning_rate": 9.82402864592019e-06,
+      "loss": 0.8179,
+      "step": 720
+    },
+    {
+      "epoch": 0.8763505402160864,
+      "grad_norm": 0.041255176067352295,
+      "learning_rate": 9.816507466187064e-06,
+      "loss": 0.8175,
+      "step": 730
+    },
+    {
+      "epoch": 0.8883553421368547,
+      "grad_norm": 0.03711531683802605,
+      "learning_rate": 9.808831902779382e-06,
+      "loss": 0.8199,
+      "step": 740
+    },
+    {
+      "epoch": 0.9003601440576231,
+      "grad_norm": 0.04047127068042755,
+      "learning_rate": 9.801002201722237e-06,
+      "loss": 0.8439,
+      "step": 750
+    },
+    {
+      "epoch": 0.9123649459783914,
+      "grad_norm": 0.03690866753458977,
+      "learning_rate": 9.793018613981298e-06,
+      "loss": 0.8206,
+      "step": 760
+    },
+    {
+      "epoch": 0.9243697478991597,
+      "grad_norm": 0.03672650083899498,
+      "learning_rate": 9.784881395454767e-06,
+      "loss": 0.8057,
+      "step": 770
+    },
+    {
+      "epoch": 0.936374549819928,
+      "grad_norm": 0.0367751307785511,
+      "learning_rate": 9.776590806965183e-06,
+      "loss": 0.8021,
+      "step": 780
+    },
+    {
+      "epoch": 0.9483793517406963,
+      "grad_norm": 0.03809288516640663,
+      "learning_rate": 9.768147114251056e-06,
+      "loss": 0.841,
+      "step": 790
+    },
+    {
+      "epoch": 0.9603841536614646,
+      "grad_norm": 0.038219958543777466,
+      "learning_rate": 9.75955058795835e-06,
+      "loss": 0.8226,
+      "step": 800
+    },
+    {
+      "epoch": 0.9723889555822329,
+      "grad_norm": 0.03921416401863098,
+      "learning_rate": 9.750801503631809e-06,
+      "loss": 0.8298,
+      "step": 810
+    },
+    {
+      "epoch": 0.9843937575030012,
+      "grad_norm": 0.03702569007873535,
+      "learning_rate": 9.741900141706122e-06,
+      "loss": 0.8547,
+      "step": 820
+    },
+    {
+      "epoch": 0.9963985594237695,
+      "grad_norm": 0.041355907917022705,
+      "learning_rate": 9.732846787496943e-06,
+      "loss": 0.8252,
+      "step": 830
+    },
+    {
+      "epoch": 1.0084033613445378,
+      "grad_norm": 0.039719488471746445,
+      "learning_rate": 9.723641731191737e-06,
+      "loss": 0.8116,
+      "step": 840
+    },
+    {
+      "epoch": 1.0204081632653061,
+      "grad_norm": 0.041531551629304886,
+      "learning_rate": 9.71428526784048e-06,
+      "loss": 0.8268,
+      "step": 850
+    },
+    {
+      "epoch": 1.0324129651860745,
+      "grad_norm": 0.03957758843898773,
+      "learning_rate": 9.704777697346204e-06,
+      "loss": 0.8147,
+      "step": 860
+    },
+    {
+      "epoch": 1.0444177671068426,
+      "grad_norm": 0.040645621716976166,
+      "learning_rate": 9.695119324455383e-06,
+      "loss": 0.8487,
+      "step": 870
+    },
+    {
+      "epoch": 1.056422569027611,
+      "grad_norm": 0.037387099117040634,
+      "learning_rate": 9.685310458748165e-06,
+      "loss": 0.8016,
+      "step": 880
+    },
+    {
+      "epoch": 1.0684273709483794,
+      "grad_norm": 0.0387730598449707,
+      "learning_rate": 9.675351414628452e-06,
+      "loss": 0.8389,
+      "step": 890
+    },
+    {
+      "epoch": 1.0804321728691477,
+      "grad_norm": 0.03654830902814865,
+      "learning_rate": 9.665242511313816e-06,
+      "loss": 0.8571,
+      "step": 900
+    },
+    {
+      "epoch": 1.092436974789916,
+      "grad_norm": 0.03688422963023186,
+      "learning_rate": 9.654984072825273e-06,
+      "loss": 0.8148,
+      "step": 910
+    },
+    {
+      "epoch": 1.1044417767106842,
+      "grad_norm": 0.04102282598614693,
+      "learning_rate": 9.644576427976893e-06,
+      "loss": 0.8252,
+      "step": 920
+    },
+    {
+      "epoch": 1.1164465786314526,
+      "grad_norm": 0.04032592847943306,
+      "learning_rate": 9.634019910365265e-06,
+      "loss": 0.8362,
+      "step": 930
+    },
+    {
+      "epoch": 1.128451380552221,
+      "grad_norm": 0.038995206356048584,
+      "learning_rate": 9.6233148583588e-06,
+      "loss": 0.8071,
+      "step": 940
+    },
+    {
+      "epoch": 1.140456182472989,
+      "grad_norm": 0.04222116991877556,
+      "learning_rate": 9.612461615086888e-06,
+      "loss": 0.8162,
+      "step": 950
+    },
+    {
+      "epoch": 1.1524609843937574,
+      "grad_norm": 0.03999168053269386,
+      "learning_rate": 9.601460528428894e-06,
+      "loss": 0.8124,
+      "step": 960
+    },
+    {
+      "epoch": 1.1644657863145258,
+      "grad_norm": 0.04248590022325516,
+      "learning_rate": 9.590311951003022e-06,
+      "loss": 0.809,
+      "step": 970
+    },
+    {
+      "epoch": 1.1764705882352942,
+      "grad_norm": 0.041043415665626526,
+      "learning_rate": 9.57901624015499e-06,
+      "loss": 0.7999,
+      "step": 980
+    },
+    {
+      "epoch": 1.1884753901560625,
+      "grad_norm": 0.037975288927555084,
+      "learning_rate": 9.5675737579466e-06,
+      "loss": 0.8839,
+      "step": 990
+    },
+    {
+      "epoch": 1.2004801920768307,
+      "grad_norm": 0.038373079150915146,
+      "learning_rate": 9.555984871144112e-06,
+      "loss": 0.8297,
+      "step": 1000
+    },
+    {
+      "epoch": 1.212484993997599,
+      "grad_norm": 0.03899591043591499,
+      "learning_rate": 9.544249951206504e-06,
+      "loss": 0.7856,
+      "step": 1010
+    },
+    {
+      "epoch": 1.2244897959183674,
+      "grad_norm": 0.0428166463971138,
+      "learning_rate": 9.532369374273555e-06,
+      "loss": 0.8018,
+      "step": 1020
+    },
+    {
+      "epoch": 1.2364945978391357,
+      "grad_norm": 0.04245438799262047,
+      "learning_rate": 9.520343521153796e-06,
+      "loss": 0.8404,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2484993997599039,
+      "grad_norm": 0.04125148430466652,
+      "learning_rate": 9.508172777312297e-06,
+      "loss": 0.8264,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2605042016806722,
+      "grad_norm": 0.04240383952856064,
+      "learning_rate": 9.495857532858314e-06,
+      "loss": 0.8442,
+      "step": 1050
+    },
+    {
+      "epoch": 1.2725090036014406,
+      "grad_norm": 0.04034110903739929,
+      "learning_rate": 9.483398182532792e-06,
+      "loss": 0.8459,
+      "step": 1060
+    },
+    {
+      "epoch": 1.284513805522209,
+      "grad_norm": 0.045645780861377716,
+      "learning_rate": 9.470795125695704e-06,
+      "loss": 0.8316,
+      "step": 1070
+    },
+    {
+      "epoch": 1.296518607442977,
+      "grad_norm": 0.0422181636095047,
+      "learning_rate": 9.458048766313246e-06,
+      "loss": 0.8144,
+      "step": 1080
+    },
+    {
+      "epoch": 1.3085234093637454,
+      "grad_norm": 0.04134783521294594,
+      "learning_rate": 9.445159512944906e-06,
+      "loss": 0.8307,
+      "step": 1090
+    },
+    {
+      "epoch": 1.3205282112845138,
+      "grad_norm": 0.04174504429101944,
+      "learning_rate": 9.43212777873035e-06,
+      "loss": 0.8179,
+      "step": 1100
+    },
+    {
+      "epoch": 1.3325330132052822,
+      "grad_norm": 0.0407400019466877,
+      "learning_rate": 9.418953981376188e-06,
+      "loss": 0.8141,
+      "step": 1110
+    },
+    {
+      "epoch": 1.3445378151260505,
+      "grad_norm": 0.04575135558843613,
+      "learning_rate": 9.405638543142589e-06,
+      "loss": 0.7888,
+      "step": 1120
+    },
+    {
+      "epoch": 1.3565426170468187,
+      "grad_norm": 0.042352344840765,
+      "learning_rate": 9.392181890829739e-06,
+      "loss": 0.8257,
+      "step": 1130
+    },
+    {
+      "epoch": 1.368547418967587,
+      "grad_norm": 0.046745285391807556,
+      "learning_rate": 9.378584455764161e-06,
+      "loss": 0.816,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3805522208883554,
+      "grad_norm": 0.04341897740960121,
+      "learning_rate": 9.364846673784897e-06,
+      "loss": 0.7988,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3925570228091235,
+      "grad_norm": 0.04738551750779152,
+      "learning_rate": 9.350968985229532e-06,
+      "loss": 0.8764,
+      "step": 1160
+    },
+    {
+      "epoch": 1.4045618247298919,
+      "grad_norm": 0.03800031542778015,
+      "learning_rate": 9.336951834920081e-06,
+      "loss": 0.8228,
+      "step": 1170
+    },
+    {
+      "epoch": 1.4165666266506602,
+      "grad_norm": 0.04292312636971474,
+      "learning_rate": 9.322795672148726e-06,
+      "loss": 0.7931,
+      "step": 1180
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.04134088754653931,
+      "learning_rate": 9.30850095066343e-06,
+      "loss": 0.8238,
+      "step": 1190
+    },
+    {
+      "epoch": 1.440576230492197,
+      "grad_norm": 0.04374649375677109,
+      "learning_rate": 9.294068128653373e-06,
+      "loss": 0.8494,
+      "step": 1200
+    },
+    {
+      "epoch": 1.452581032412965,
+      "grad_norm": 0.04475456848740578,
+      "learning_rate": 9.279497668734286e-06,
+      "loss": 0.8213,
+      "step": 1210
+    },
+    {
+      "epoch": 1.4645858343337335,
+      "grad_norm": 0.04506712034344673,
+      "learning_rate": 9.264790037933604e-06,
+      "loss": 0.8117,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4765906362545018,
+      "grad_norm": 0.042217791080474854,
+      "learning_rate": 9.249945707675506e-06,
+      "loss": 0.8387,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4885954381752702,
+      "grad_norm": 0.043338458985090256,
+      "learning_rate": 9.234965153765809e-06,
+      "loss": 0.8345,
+      "step": 1240
+    },
+    {
+      "epoch": 1.5006002400960385,
+      "grad_norm": 0.04206259921193123,
+      "learning_rate": 9.219848856376706e-06,
+      "loss": 0.8062,
+      "step": 1250
+    },
+    {
+      "epoch": 1.5126050420168067,
+      "grad_norm": 0.0476699024438858,
+      "learning_rate": 9.204597300031379e-06,
+      "loss": 0.8626,
+      "step": 1260
+    },
+    {
+      "epoch": 1.524609843937575,
+      "grad_norm": 0.04465340077877045,
+      "learning_rate": 9.189210973588476e-06,
+      "loss": 0.8379,
+      "step": 1270
+    },
+    {
+      "epoch": 1.5366146458583434,
+      "grad_norm": 0.04375738650560379,
+      "learning_rate": 9.173690370226432e-06,
+      "loss": 0.7904,
+      "step": 1280
+    },
+    {
+      "epoch": 1.5486194477791115,
+      "grad_norm": 0.04561934620141983,
+      "learning_rate": 9.158035987427662e-06,
+      "loss": 0.8152,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5606242496998801,
+      "grad_norm": 0.04491221904754639,
+      "learning_rate": 9.142248326962624e-06,
+      "loss": 0.7844,
+      "step": 1300
+    },
+    {
+      "epoch": 1.5726290516206483,
+      "grad_norm": 0.04250409081578255,
+      "learning_rate": 9.12632789487372e-06,
+      "loss": 0.8156,
+      "step": 1310
+    },
+    {
+      "epoch": 1.5846338535414166,
+      "grad_norm": 0.04542950913310051,
+      "learning_rate": 9.110275201459099e-06,
+      "loss": 0.8496,
+      "step": 1320
+    },
+    {
+      "epoch": 1.596638655462185,
+      "grad_norm": 0.04421227052807808,
+      "learning_rate": 9.09409076125627e-06,
+      "loss": 0.817,
+      "step": 1330
+    },
+    {
+      "epoch": 1.6086434573829531,
+      "grad_norm": 0.04881924018263817,
+      "learning_rate": 9.07777509302564e-06,
+      "loss": 0.8047,
+      "step": 1340
+    },
+    {
+      "epoch": 1.6206482593037215,
+      "grad_norm": 0.04575929790735245,
+      "learning_rate": 9.061328719733867e-06,
+      "loss": 0.818,
+      "step": 1350
+    },
+    {
+      "epoch": 1.6326530612244898,
+      "grad_norm": 0.047302693128585815,
+      "learning_rate": 9.044752168537104e-06,
+      "loss": 0.8255,
+      "step": 1360
+    },
+    {
+      "epoch": 1.644657863145258,
+      "grad_norm": 0.048905596137046814,
+      "learning_rate": 9.028045970764095e-06,
+      "loss": 0.8033,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6566626650660266,
+      "grad_norm": 0.04502958804368973,
+      "learning_rate": 9.01121066189916e-06,
+      "loss": 0.8292,
+      "step": 1380
+    },
+    {
+      "epoch": 1.6686674669867947,
+      "grad_norm": 0.04705483838915825,
+      "learning_rate": 8.994246781565012e-06,
+      "loss": 0.8343,
+      "step": 1390
+    },
+    {
+      "epoch": 1.680672268907563,
+      "grad_norm": 0.04515758529305458,
+      "learning_rate": 8.977154873505478e-06,
+      "loss": 0.8458,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6926770708283314,
+      "grad_norm": 0.05474863573908806,
+      "learning_rate": 8.959935485568054e-06,
+      "loss": 0.8044,
+      "step": 1410
+    },
+    {
+      "epoch": 1.7046818727490995,
+      "grad_norm": 0.044155530631542206,
+      "learning_rate": 8.942589169686356e-06,
+      "loss": 0.8499,
+      "step": 1420
+    },
+    {
+      "epoch": 1.716686674669868,
+      "grad_norm": 0.049613580107688904,
+      "learning_rate": 8.925116481862426e-06,
+      "loss": 0.8643,
+      "step": 1430
+    },
+    {
+      "epoch": 1.7286914765906363,
+      "grad_norm": 0.05003758892416954,
+      "learning_rate": 8.907517982148913e-06,
+      "loss": 0.8944,
+      "step": 1440
+    },
+    {
+      "epoch": 1.7406962785114044,
+      "grad_norm": 0.048006169497966766,
+      "learning_rate": 8.889794234631112e-06,
+      "loss": 0.8063,
+      "step": 1450
+    },
+    {
+      "epoch": 1.752701080432173,
+      "grad_norm": 0.04439476877450943,
+      "learning_rate": 8.871945807408895e-06,
+      "loss": 0.8679,
+      "step": 1460
+    },
+    {
+      "epoch": 1.7647058823529411,
+      "grad_norm": 0.04884785786271095,
+      "learning_rate": 8.853973272578495e-06,
+      "loss": 0.8433,
+      "step": 1470
+    },
+    {
+      "epoch": 1.7767106842737095,
+      "grad_norm": 0.050317153334617615,
+      "learning_rate": 8.835877206214172e-06,
+      "loss": 0.8394,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7887154861944778,
+      "grad_norm": 0.04775743559002876,
+      "learning_rate": 8.817658188349745e-06,
+      "loss": 0.8309,
+      "step": 1490
+    },
+    {
+      "epoch": 1.800720288115246,
+      "grad_norm": 0.04834270477294922,
+      "learning_rate": 8.79931680296e-06,
+      "loss": 0.8414,
+      "step": 1500
+    },
+    {
+      "epoch": 1.8127250900360146,
+      "grad_norm": 0.04563102871179581,
+      "learning_rate": 8.780853637941976e-06,
+      "loss": 0.8934,
+      "step": 1510
+    },
+    {
+      "epoch": 1.8247298919567827,
+      "grad_norm": 0.04749139025807381,
+      "learning_rate": 8.762269285096117e-06,
+      "loss": 0.8946,
+      "step": 1520
+    },
+    {
+      "epoch": 1.836734693877551,
+      "grad_norm": 0.04547303169965744,
+      "learning_rate": 8.743564340107305e-06,
+      "loss": 0.7862,
+      "step": 1530
+    },
+    {
+      "epoch": 1.8487394957983194,
+      "grad_norm": 0.043644532561302185,
+      "learning_rate": 8.724739402525765e-06,
+      "loss": 0.8096,
+      "step": 1540
+    },
+    {
+      "epoch": 1.8607442977190876,
+      "grad_norm": 0.050219349563121796,
+      "learning_rate": 8.705795075747854e-06,
+      "loss": 0.798,
+      "step": 1550
+    },
+    {
+      "epoch": 1.872749099639856,
+      "grad_norm": 0.04704947769641876,
+      "learning_rate": 8.686731966996707e-06,
+      "loss": 0.809,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8847539015606243,
+      "grad_norm": 0.05036606639623642,
+      "learning_rate": 8.667550687302784e-06,
+      "loss": 0.8156,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8967587034813924,
+      "grad_norm": 0.05243535712361336,
+      "learning_rate": 8.648251851484284e-06,
+      "loss": 0.772,
+      "step": 1580
+    },
+    {
+      "epoch": 1.908763505402161,
+      "grad_norm": 0.0510157085955143,
+      "learning_rate": 8.628836078127435e-06,
+      "loss": 0.836,
+      "step": 1590
+    },
+    {
+      "epoch": 1.9207683073229291,
+      "grad_norm": 0.04631471633911133,
+      "learning_rate": 8.609303989566673e-06,
+      "loss": 0.8281,
+      "step": 1600
+    },
+    {
+      "epoch": 1.9327731092436975,
+      "grad_norm": 0.04962284117937088,
+      "learning_rate": 8.589656211864678e-06,
+      "loss": 0.831,
+      "step": 1610
+    },
+    {
+      "epoch": 1.9447779111644659,
+      "grad_norm": 0.054218314588069916,
+      "learning_rate": 8.569893374792328e-06,
+      "loss": 0.7904,
+      "step": 1620
+    },
+    {
+      "epoch": 1.956782713085234,
+      "grad_norm": 0.05222290754318237,
+      "learning_rate": 8.5500161118085e-06,
+      "loss": 0.8141,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9687875150060024,
+      "grad_norm": 0.04537142068147659,
+      "learning_rate": 8.530025060039763e-06,
+      "loss": 0.8605,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9807923169267707,
+      "grad_norm": 0.05147700011730194,
+      "learning_rate": 8.509920860259973e-06,
+      "loss": 0.8549,
+      "step": 1650
+    },
+    {
+      "epoch": 1.9927971188475389,
+      "grad_norm": 0.052020113915205,
+      "learning_rate": 8.489704156869711e-06,
+      "loss": 0.8093,
+      "step": 1660
+    },
+    {
+      "epoch": 2.0048019207683074,
+      "grad_norm": 0.05193515121936798,
+      "learning_rate": 8.469375597875648e-06,
+      "loss": 0.8481,
+      "step": 1670
+    },
+    {
+      "epoch": 2.0168067226890756,
+      "grad_norm": 0.04633639007806778,
+      "learning_rate": 8.448935834869763e-06,
+      "loss": 0.8627,
+      "step": 1680
+    },
+    {
+      "epoch": 2.028811524609844,
+      "grad_norm": 0.05221736803650856,
+      "learning_rate": 8.428385523008463e-06,
+      "loss": 0.8339,
+      "step": 1690
+    },
+    {
+      "epoch": 2.0408163265306123,
+      "grad_norm": 0.04951602965593338,
+      "learning_rate": 8.40772532099158e-06,
+      "loss": 0.8483,
+      "step": 1700
+    },
+    {
+      "epoch": 2.0528211284513804,
+      "grad_norm": 0.053726278245449066,
+      "learning_rate": 8.386955891041255e-06,
+      "loss": 0.8191,
+      "step": 1710
+    },
+    {
+      "epoch": 2.064825930372149,
+      "grad_norm": 0.04682547226548195,
+      "learning_rate": 8.366077898880725e-06,
+      "loss": 0.8037,
+      "step": 1720
+    },
+    {
+      "epoch": 2.076830732292917,
+      "grad_norm": 0.04982489347457886,
+      "learning_rate": 8.345092013712966e-06,
+      "loss": 0.7827,
+      "step": 1730
+    },
+    {
+      "epoch": 2.0888355342136853,
+      "grad_norm": 0.05112074315547943,
+      "learning_rate": 8.323998908199256e-06,
+      "loss": 0.8163,
+      "step": 1740
+    },
+    {
+      "epoch": 2.100840336134454,
+      "grad_norm": 0.05606475844979286,
+      "learning_rate": 8.302799258437611e-06,
+      "loss": 0.8447,
+      "step": 1750
+    },
+    {
+      "epoch": 2.112845138055222,
+      "grad_norm": 0.05271725356578827,
+      "learning_rate": 8.281493743941112e-06,
+      "loss": 0.8125,
+      "step": 1760
+    },
+    {
+      "epoch": 2.1248499399759906,
+      "grad_norm": 0.05468921363353729,
+      "learning_rate": 8.260083047616126e-06,
+      "loss": 0.8326,
+      "step": 1770
+    },
+    {
+      "epoch": 2.1368547418967587,
+      "grad_norm": 0.05409565195441246,
+      "learning_rate": 8.238567855740413e-06,
+      "loss": 0.8082,
+      "step": 1780
+    },
+    {
+      "epoch": 2.148859543817527,
+      "grad_norm": 0.05312246084213257,
+      "learning_rate": 8.216948857941143e-06,
+      "loss": 0.8137,
+      "step": 1790
+    },
+    {
+      "epoch": 2.1608643457382954,
+      "grad_norm": 0.05429564416408539,
+      "learning_rate": 8.19522674717277e-06,
+      "loss": 0.8161,
+      "step": 1800
+    },
+    {
+      "epoch": 2.1728691476590636,
+      "grad_norm": 0.049361683428287506,
+      "learning_rate": 8.173402219694838e-06,
+      "loss": 0.8547,
+      "step": 1810
+    },
+    {
+      "epoch": 2.184873949579832,
+      "grad_norm": 0.050867076963186264,
+      "learning_rate": 8.151475975049653e-06,
+      "loss": 0.788,
+      "step": 1820
+    },
+    {
+      "epoch": 2.1968787515006003,
+      "grad_norm": 0.051545899361371994,
+      "learning_rate": 8.129448716039863e-06,
+      "loss": 0.8079,
+      "step": 1830
+    },
+    {
+      "epoch": 2.2088835534213684,
+      "grad_norm": 0.05476554483175278,
+      "learning_rate": 8.107321148705941e-06,
+      "loss": 0.8123,
+      "step": 1840
+    },
+    {
+      "epoch": 2.220888355342137,
+      "grad_norm": 0.05693934112787247,
+      "learning_rate": 8.085093982303539e-06,
+      "loss": 0.815,
+      "step": 1850
+    },
+    {
+      "epoch": 2.232893157262905,
+      "grad_norm": 0.057461485266685486,
+      "learning_rate": 8.062767929280764e-06,
+      "loss": 0.7947,
+      "step": 1860
+    },
+    {
+      "epoch": 2.2448979591836733,
+      "grad_norm": 0.057199206203222275,
+      "learning_rate": 8.040343705255338e-06,
+      "loss": 0.815,
+      "step": 1870
+    },
+    {
+      "epoch": 2.256902761104442,
+      "grad_norm": 0.05683717131614685,
+      "learning_rate": 8.017822028991666e-06,
+      "loss": 0.8979,
+      "step": 1880
+    },
+    {
+      "epoch": 2.26890756302521,
+      "grad_norm": 0.05384090170264244,
+      "learning_rate": 7.995203622377783e-06,
+      "loss": 0.8424,
+      "step": 1890
+    },
+    {
+      "epoch": 2.280912364945978,
+      "grad_norm": 0.05016244202852249,
+      "learning_rate": 7.972489210402232e-06,
+      "loss": 0.7551,
+      "step": 1900
+    },
+    {
+      "epoch": 2.2929171668667467,
+      "grad_norm": 0.05437694489955902,
+      "learning_rate": 7.949679521130823e-06,
+      "loss": 0.8446,
+      "step": 1910
+    },
+    {
+      "epoch": 2.304921968787515,
+      "grad_norm": 0.055492572486400604,
+      "learning_rate": 7.926775285683281e-06,
+      "loss": 0.8182,
+      "step": 1920
+    },
+    {
+      "epoch": 2.3169267707082835,
+      "grad_norm": 0.05502455681562424,
+      "learning_rate": 7.903777238209835e-06,
+      "loss": 0.8181,
+      "step": 1930
+    },
+    {
+      "epoch": 2.3289315726290516,
+      "grad_norm": 0.05110107734799385,
+      "learning_rate": 7.880686115867664e-06,
+      "loss": 0.8521,
+      "step": 1940
+    },
+    {
+      "epoch": 2.3409363745498197,
+      "grad_norm": 0.0539153516292572,
+      "learning_rate": 7.857502658797284e-06,
+      "loss": 0.8272,
+      "step": 1950
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 0.05461420118808746,
+      "learning_rate": 7.83422761009882e-06,
+      "loss": 0.8413,
+      "step": 1960
+    },
+    {
+      "epoch": 2.3649459783913565,
+      "grad_norm": 0.05569763854146004,
+      "learning_rate": 7.810861715808184e-06,
+      "loss": 0.8687,
+      "step": 1970
+    },
+    {
+      "epoch": 2.376950780312125,
+      "grad_norm": 0.055215220898389816,
+      "learning_rate": 7.787405724873162e-06,
+      "loss": 0.8196,
+      "step": 1980
+    },
+    {
+      "epoch": 2.388955582232893,
+      "grad_norm": 0.06094464287161827,
+      "learning_rate": 7.76386038912942e-06,
+      "loss": 0.7921,
+      "step": 1990
+    },
+    {
+      "epoch": 2.4009603841536613,
+      "grad_norm": 0.05856522545218468,
+      "learning_rate": 7.740226463276385e-06,
+      "loss": 0.8332,
+      "step": 2000
+    },
+    {
+      "epoch": 2.41296518607443,
+      "grad_norm": 0.05313005670905113,
+      "learning_rate": 7.716504704853076e-06,
+      "loss": 0.8076,
+      "step": 2010
+    },
+    {
+      "epoch": 2.424969987995198,
+      "grad_norm": 0.05347991734743118,
+      "learning_rate": 7.692695874213802e-06,
+      "loss": 0.8328,
+      "step": 2020
+    },
+    {
+      "epoch": 2.4369747899159666,
+      "grad_norm": 0.058754704892635345,
+      "learning_rate": 7.668800734503812e-06,
+      "loss": 0.8445,
+      "step": 2030
+    },
+    {
+      "epoch": 2.4489795918367347,
+      "grad_norm": 0.05653975158929825,
+      "learning_rate": 7.644820051634813e-06,
+      "loss": 0.8378,
+      "step": 2040
+    },
+    {
+      "epoch": 2.460984393757503,
+      "grad_norm": 0.05895698443055153,
+      "learning_rate": 7.620754594260433e-06,
+      "loss": 0.8214,
+      "step": 2050
+    },
+    {
+      "epoch": 2.4729891956782715,
+      "grad_norm": 0.056038543581962585,
+      "learning_rate": 7.596605133751583e-06,
+      "loss": 0.792,
+      "step": 2060
+    },
+    {
+      "epoch": 2.4849939975990396,
+      "grad_norm": 0.0582413524389267,
+      "learning_rate": 7.5723724441717225e-06,
+      "loss": 0.8292,
+      "step": 2070
+    },
+    {
+      "epoch": 2.4969987995198077,
+      "grad_norm": 0.05528351292014122,
+      "learning_rate": 7.548057302252063e-06,
+      "loss": 0.7871,
+      "step": 2080
+    },
+    {
+      "epoch": 2.5090036014405763,
+      "grad_norm": 0.05144486203789711,
+      "learning_rate": 7.523660487366658e-06,
+      "loss": 0.7738,
+      "step": 2090
+    },
+    {
+      "epoch": 2.5210084033613445,
+      "grad_norm": 0.05947820469737053,
+      "learning_rate": 7.4991827815074306e-06,
+      "loss": 0.7955,
+      "step": 2100
+    },
+    {
+      "epoch": 2.5330132052821126,
+      "grad_norm": 0.05376136675477028,
+      "learning_rate": 7.474624969259101e-06,
+      "loss": 0.8145,
+      "step": 2110
+    },
+    {
+      "epoch": 2.545018007202881,
+      "grad_norm": 0.05993662402033806,
+      "learning_rate": 7.449987837774049e-06,
+      "loss": 0.8357,
+      "step": 2120
+    },
+    {
+      "epoch": 2.5570228091236493,
+      "grad_norm": 0.05299674719572067,
+      "learning_rate": 7.425272176747068e-06,
+      "loss": 0.8185,
+      "step": 2130
+    },
+    {
+      "epoch": 2.569027611044418,
+      "grad_norm": 0.06037457659840584,
+      "learning_rate": 7.400478778390068e-06,
+      "loss": 0.8327,
+      "step": 2140
+    },
+    {
+      "epoch": 2.581032412965186,
+      "grad_norm": 0.054685529321432114,
+      "learning_rate": 7.375608437406674e-06,
+      "loss": 0.7932,
+      "step": 2150
+    },
+    {
+      "epoch": 2.593037214885954,
+      "grad_norm": 0.061256833374500275,
+      "learning_rate": 7.350661950966755e-06,
+      "loss": 0.8395,
+      "step": 2160
+    },
+    {
+      "epoch": 2.6050420168067228,
+      "grad_norm": 0.056471362709999084,
+      "learning_rate": 7.325640118680872e-06,
+      "loss": 0.7806,
+      "step": 2170
+    },
+    {
+      "epoch": 2.617046818727491,
+      "grad_norm": 0.0574592649936676,
+      "learning_rate": 7.300543742574654e-06,
+      "loss": 0.8243,
+      "step": 2180
+    },
+    {
+      "epoch": 2.6290516206482595,
+      "grad_norm": 0.05507262796163559,
+      "learning_rate": 7.275373627063079e-06,
+      "loss": 0.8201,
+      "step": 2190
+    },
+    {
+      "epoch": 2.6410564225690276,
+      "grad_norm": 0.05754077062010765,
+      "learning_rate": 7.2501305789247e-06,
+      "loss": 0.8003,
+      "step": 2200
+    },
+    {
+      "epoch": 2.6530612244897958,
+      "grad_norm": 0.055398665368556976,
+      "learning_rate": 7.224815407275782e-06,
+      "loss": 0.8291,
+      "step": 2210
+    },
+    {
+      "epoch": 2.6650660264105643,
+      "grad_norm": 0.057937685400247574,
+      "learning_rate": 7.199428923544366e-06,
+      "loss": 0.7827,
+      "step": 2220
+    },
+    {
+      "epoch": 2.6770708283313325,
+      "grad_norm": 0.05905699357390404,
+      "learning_rate": 7.173971941444262e-06,
+      "loss": 0.8482,
+      "step": 2230
+    },
+    {
+      "epoch": 2.689075630252101,
+      "grad_norm": 0.05891605094075203,
+      "learning_rate": 7.148445276948968e-06,
+      "loss": 0.848,
+      "step": 2240
+    },
+    {
+      "epoch": 2.701080432172869,
+      "grad_norm": 0.05597429722547531,
+      "learning_rate": 7.122849748265512e-06,
+      "loss": 0.8288,
+      "step": 2250
+    },
+    {
+      "epoch": 2.7130852340936373,
+      "grad_norm": 0.0589454285800457,
+      "learning_rate": 7.097186175808229e-06,
+      "loss": 0.8711,
+      "step": 2260
+    },
+    {
+      "epoch": 2.725090036014406,
+      "grad_norm": 0.056011561304330826,
+      "learning_rate": 7.071455382172465e-06,
+      "loss": 0.7916,
+      "step": 2270
+    },
+    {
+      "epoch": 2.737094837935174,
+      "grad_norm": 0.06104077771306038,
+      "learning_rate": 7.045658192108204e-06,
+      "loss": 0.8378,
+      "step": 2280
+    },
+    {
+      "epoch": 2.7490996398559426,
+      "grad_norm": 0.060269393026828766,
+      "learning_rate": 7.019795432493644e-06,
+      "loss": 0.8296,
+      "step": 2290
+    },
+    {
+      "epoch": 2.7611044417767108,
+      "grad_norm": 0.05378441885113716,
+      "learning_rate": 6.9938679323086785e-06,
+      "loss": 0.8312,
+      "step": 2300
+    },
+    {
+      "epoch": 2.773109243697479,
+      "grad_norm": 0.05314185470342636,
+      "learning_rate": 6.967876522608339e-06,
+      "loss": 0.8127,
+      "step": 2310
+    },
+    {
+      "epoch": 2.785114045618247,
+      "grad_norm": 0.056222010403871536,
+      "learning_rate": 6.941822036496146e-06,
+      "loss": 0.8065,
+      "step": 2320
+    },
+    {
+      "epoch": 2.7971188475390156,
+      "grad_norm": 0.05297733470797539,
+      "learning_rate": 6.915705309097413e-06,
+      "loss": 0.776,
+      "step": 2330
+    },
+    {
+      "epoch": 2.8091236494597838,
+      "grad_norm": 0.05919932574033737,
+      "learning_rate": 6.889527177532476e-06,
+      "loss": 0.8115,
+      "step": 2340
+    },
+    {
+      "epoch": 2.8211284513805523,
+      "grad_norm": 0.05913585424423218,
+      "learning_rate": 6.863288480889858e-06,
+      "loss": 0.8291,
+      "step": 2350
+    },
+    {
+      "epoch": 2.8331332533013205,
+      "grad_norm": 0.05690290406346321,
+      "learning_rate": 6.836990060199379e-06,
+      "loss": 0.7931,
+      "step": 2360
+    },
+    {
+      "epoch": 2.8451380552220886,
+      "grad_norm": 0.058122627437114716,
+      "learning_rate": 6.810632758405195e-06,
+      "loss": 0.8436,
+      "step": 2370
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.0562375970184803,
+      "learning_rate": 6.7842174203387804e-06,
+      "loss": 0.8147,
+      "step": 2380
+    },
+    {
+      "epoch": 2.8691476590636253,
+      "grad_norm": 0.06252553313970566,
+      "learning_rate": 6.757744892691847e-06,
+      "loss": 0.8408,
+      "step": 2390
+    },
+    {
+      "epoch": 2.881152460984394,
+      "grad_norm": 0.06220783293247223,
+      "learning_rate": 6.731216023989207e-06,
+      "loss": 0.8358,
+      "step": 2400
+    },
+    {
+      "epoch": 2.893157262905162,
+      "grad_norm": 0.05928228050470352,
+      "learning_rate": 6.704631664561574e-06,
+      "loss": 0.835,
+      "step": 2410
+    },
+    {
+      "epoch": 2.90516206482593,
+      "grad_norm": 0.06004854291677475,
+      "learning_rate": 6.67799266651831e-06,
+      "loss": 0.8417,
+      "step": 2420
+    },
+    {
+      "epoch": 2.917166866746699,
+      "grad_norm": 0.05948818475008011,
+      "learning_rate": 6.651299883720106e-06,
+      "loss": 0.8026,
+      "step": 2430
+    },
+    {
+      "epoch": 2.929171668667467,
+      "grad_norm": 0.05721951648592949,
+      "learning_rate": 6.62455417175162e-06,
+      "loss": 0.8143,
+      "step": 2440
+    },
+    {
+      "epoch": 2.9411764705882355,
+      "grad_norm": 0.060878921300172806,
+      "learning_rate": 6.597756387894051e-06,
+      "loss": 0.7804,
+      "step": 2450
+    },
+    {
+      "epoch": 2.9531812725090036,
+      "grad_norm": 0.06494616717100143,
+      "learning_rate": 6.57090739109766e-06,
+      "loss": 0.8528,
+      "step": 2460
+    },
+    {
+      "epoch": 2.965186074429772,
+      "grad_norm": 0.061719998717308044,
+      "learning_rate": 6.544008041954236e-06,
+      "loss": 0.8341,
+      "step": 2470
+    },
+    {
+      "epoch": 2.9771908763505404,
+      "grad_norm": 0.06074754521250725,
+      "learning_rate": 6.517059202669516e-06,
+      "loss": 0.8163,
+      "step": 2480
+    },
+    {
+      "epoch": 2.9891956782713085,
+      "grad_norm": 0.05759371444582939,
+      "learning_rate": 6.4900617370355435e-06,
+      "loss": 0.797,
+      "step": 2490
+    },
+    {
+      "epoch": 3.0012004801920766,
+      "grad_norm": 0.06447486579418182,
+      "learning_rate": 6.4630165104029875e-06,
+      "loss": 0.8478,
+      "step": 2500
+    },
+    {
+      "epoch": 3.013205282112845,
+      "grad_norm": 0.05972028896212578,
+      "learning_rate": 6.435924389653398e-06,
+      "loss": 0.7434,
+      "step": 2510
+    },
+    {
+      "epoch": 3.0252100840336134,
+      "grad_norm": 0.06719402968883514,
+      "learning_rate": 6.408786243171426e-06,
+      "loss": 0.8223,
+      "step": 2520
+    },
+    {
+      "epoch": 3.037214885954382,
+      "grad_norm": 0.06629668921232224,
+      "learning_rate": 6.3816029408169846e-06,
+      "loss": 0.8201,
+      "step": 2530
+    },
+    {
+      "epoch": 3.04921968787515,
+      "grad_norm": 0.06306376308202744,
+      "learning_rate": 6.354375353897376e-06,
+      "loss": 0.8387,
+      "step": 2540
+    },
+    {
+      "epoch": 3.061224489795918,
+      "grad_norm": 0.05803842842578888,
+      "learning_rate": 6.3271043551393485e-06,
+      "loss": 0.8517,
+      "step": 2550
+    },
+    {
+      "epoch": 3.073229291716687,
+      "grad_norm": 0.05941968038678169,
+      "learning_rate": 6.299790818661135e-06,
+      "loss": 0.8243,
+      "step": 2560
+    },
+    {
+      "epoch": 3.085234093637455,
+      "grad_norm": 0.06141958758234978,
+      "learning_rate": 6.272435619944436e-06,
+      "loss": 0.8409,
+      "step": 2570
+    },
+    {
+      "epoch": 3.097238895558223,
+      "grad_norm": 0.05685436725616455,
+      "learning_rate": 6.245039635806345e-06,
+      "loss": 0.8159,
+      "step": 2580
+    },
+    {
+      "epoch": 3.1092436974789917,
+      "grad_norm": 0.058166082948446274,
+      "learning_rate": 6.2176037443712575e-06,
+      "loss": 0.8486,
+      "step": 2590
+    },
+    {
+      "epoch": 3.12124849939976,
+      "grad_norm": 0.05954616516828537,
+      "learning_rate": 6.190128825042716e-06,
+      "loss": 0.8417,
+      "step": 2600
+    },
+    {
+      "epoch": 3.1332533013205284,
+      "grad_norm": 0.05970808118581772,
+      "learning_rate": 6.1626157584752285e-06,
+      "loss": 0.7822,
+      "step": 2610
+    },
+    {
+      "epoch": 3.1452581032412965,
+      "grad_norm": 0.06996554881334305,
+      "learning_rate": 6.135065426546034e-06,
+      "loss": 0.8074,
+      "step": 2620
+    },
+    {
+      "epoch": 3.1572629051620646,
+      "grad_norm": 0.0612507127225399,
+      "learning_rate": 6.107478712326843e-06,
+      "loss": 0.819,
+      "step": 2630
+    },
+    {
+      "epoch": 3.1692677070828332,
+      "grad_norm": 0.06266125291585922,
+      "learning_rate": 6.079856500055526e-06,
+      "loss": 0.8005,
+      "step": 2640
+    },
+    {
+      "epoch": 3.1812725090036014,
+      "grad_norm": 0.06757565587759018,
+      "learning_rate": 6.052199675107773e-06,
+      "loss": 0.8025,
+      "step": 2650
+    },
+    {
+      "epoch": 3.19327731092437,
+      "grad_norm": 0.060953736305236816,
+      "learning_rate": 6.024509123968722e-06,
+      "loss": 0.8195,
+      "step": 2660
+    },
+    {
+      "epoch": 3.205282112845138,
+      "grad_norm": 0.074392169713974,
+      "learning_rate": 5.996785734204528e-06,
+      "loss": 0.8411,
+      "step": 2670
+    },
+    {
+      "epoch": 3.2172869147659062,
+      "grad_norm": 0.06586959213018417,
+      "learning_rate": 5.969030394433928e-06,
+      "loss": 0.8246,
+      "step": 2680
+    },
+    {
+      "epoch": 3.229291716686675,
+      "grad_norm": 0.06062298268079758,
+      "learning_rate": 5.941243994299758e-06,
+      "loss": 0.8361,
+      "step": 2690
+    },
+    {
+      "epoch": 3.241296518607443,
+      "grad_norm": 0.06378082185983658,
+      "learning_rate": 5.913427424440428e-06,
+      "loss": 0.8121,
+      "step": 2700
+    },
+    {
+      "epoch": 3.2533013205282115,
+      "grad_norm": 0.061464712023735046,
+      "learning_rate": 5.885581576461376e-06,
+      "loss": 0.813,
+      "step": 2710
+    },
+    {
+      "epoch": 3.2653061224489797,
+      "grad_norm": 0.06229786574840546,
+      "learning_rate": 5.857707342906501e-06,
+      "loss": 0.7757,
+      "step": 2720
+    },
+    {
+      "epoch": 3.277310924369748,
+      "grad_norm": 0.0691327154636383,
+      "learning_rate": 5.82980561722954e-06,
+      "loss": 0.8063,
+      "step": 2730
+    },
+    {
+      "epoch": 3.2893157262905164,
+      "grad_norm": 0.06736263632774353,
+      "learning_rate": 5.8018772937654366e-06,
+      "loss": 0.8352,
+      "step": 2740
+    },
+    {
+      "epoch": 3.3013205282112845,
+      "grad_norm": 0.06093461439013481,
+      "learning_rate": 5.773923267701676e-06,
+      "loss": 0.8216,
+      "step": 2750
+    },
+    {
+      "epoch": 3.3133253301320527,
+      "grad_norm": 0.06377444416284561,
+      "learning_rate": 5.745944435049586e-06,
+      "loss": 0.8095,
+      "step": 2760
+    },
+    {
+      "epoch": 3.3253301320528212,
+      "grad_norm": 0.06320587545633316,
+      "learning_rate": 5.717941692615626e-06,
+      "loss": 0.828,
+      "step": 2770
+    },
+    {
+      "epoch": 3.3373349339735894,
+      "grad_norm": 0.06745033711194992,
+      "learning_rate": 5.689915937972634e-06,
+      "loss": 0.8277,
+      "step": 2780
+    },
+    {
+      "epoch": 3.3493397358943575,
+      "grad_norm": 0.06507708877325058,
+      "learning_rate": 5.661868069431056e-06,
+      "loss": 0.8177,
+      "step": 2790
+    },
+    {
+      "epoch": 3.361344537815126,
+      "grad_norm": 0.0653938576579094,
+      "learning_rate": 5.633798986010159e-06,
+      "loss": 0.8302,
+      "step": 2800
+    },
+    {
+      "epoch": 3.3733493397358942,
+      "grad_norm": 0.06798886507749557,
+      "learning_rate": 5.605709587409216e-06,
+      "loss": 0.7741,
+      "step": 2810
+    },
+    {
+      "epoch": 3.385354141656663,
+      "grad_norm": 0.06364308297634125,
+      "learning_rate": 5.577600773978651e-06,
+      "loss": 0.7791,
+      "step": 2820
+    },
+    {
+      "epoch": 3.397358943577431,
+      "grad_norm": 0.06712163984775543,
+      "learning_rate": 5.5494734466912005e-06,
+      "loss": 0.8316,
+      "step": 2830
+    },
+    {
+      "epoch": 3.409363745498199,
+      "grad_norm": 0.06373037397861481,
+      "learning_rate": 5.521328507113027e-06,
+      "loss": 0.7897,
+      "step": 2840
+    },
+    {
+      "epoch": 3.4213685474189677,
+      "grad_norm": 0.06734546273946762,
+      "learning_rate": 5.493166857374818e-06,
+      "loss": 0.8325,
+      "step": 2850
+    },
+    {
+      "epoch": 3.433373349339736,
+      "grad_norm": 0.0662597045302391,
+      "learning_rate": 5.464989400142868e-06,
+      "loss": 0.8304,
+      "step": 2860
+    },
+    {
+      "epoch": 3.4453781512605044,
+      "grad_norm": 0.06525112688541412,
+      "learning_rate": 5.436797038590159e-06,
+      "loss": 0.869,
+      "step": 2870
+    },
+    {
+      "epoch": 3.4573829531812725,
+      "grad_norm": 0.06612803786993027,
+      "learning_rate": 5.408590676367395e-06,
+      "loss": 0.8372,
+      "step": 2880
+    },
+    {
+      "epoch": 3.4693877551020407,
+      "grad_norm": 0.07390609383583069,
+      "learning_rate": 5.380371217574046e-06,
+      "loss": 0.8011,
+      "step": 2890
+    },
+    {
+      "epoch": 3.4813925570228093,
+      "grad_norm": 0.06323985010385513,
+      "learning_rate": 5.352139566729366e-06,
+      "loss": 0.7972,
+      "step": 2900
+    },
+    {
+      "epoch": 3.4933973589435774,
+      "grad_norm": 0.06164005026221275,
+      "learning_rate": 5.323896628743401e-06,
+      "loss": 0.8099,
+      "step": 2910
+    },
+    {
+      "epoch": 3.505402160864346,
+      "grad_norm": 0.06231848523020744,
+      "learning_rate": 5.295643308887986e-06,
+      "loss": 0.8172,
+      "step": 2920
+    },
+    {
+      "epoch": 3.517406962785114,
+      "grad_norm": 0.06743638217449188,
+      "learning_rate": 5.2673805127677255e-06,
+      "loss": 0.7886,
+      "step": 2930
+    },
+    {
+      "epoch": 3.5294117647058822,
+      "grad_norm": 0.0683428943157196,
+      "learning_rate": 5.239109146290965e-06,
+      "loss": 0.8155,
+      "step": 2940
+    },
+    {
+      "epoch": 3.5414165666266504,
+      "grad_norm": 0.06549340486526489,
+      "learning_rate": 5.210830115640758e-06,
+      "loss": 0.7952,
+      "step": 2950
+    },
+    {
+      "epoch": 3.553421368547419,
+      "grad_norm": 0.07066480070352554,
+      "learning_rate": 5.1825443272458194e-06,
+      "loss": 0.8514,
+      "step": 2960
+    },
+    {
+      "epoch": 3.5654261704681876,
+      "grad_norm": 0.0690867155790329,
+      "learning_rate": 5.154252687751469e-06,
+      "loss": 0.837,
+      "step": 2970
+    },
+    {
+      "epoch": 3.5774309723889557,
+      "grad_norm": 0.07441065460443497,
+      "learning_rate": 5.125956103990567e-06,
+      "loss": 0.7944,
+      "step": 2980
+    },
+    {
+      "epoch": 3.589435774309724,
+      "grad_norm": 0.06854850053787231,
+      "learning_rate": 5.0976554829544626e-06,
+      "loss": 0.8022,
+      "step": 2990
+    },
+    {
+      "epoch": 3.601440576230492,
+      "grad_norm": 0.06906180083751678,
+      "learning_rate": 5.069351731763903e-06,
+      "loss": 0.7914,
+      "step": 3000
+    },
+    {
+      "epoch": 3.6134453781512605,
+      "grad_norm": 0.06705962121486664,
+      "learning_rate": 5.041045757639972e-06,
+      "loss": 0.8043,
+      "step": 3010
+    },
+    {
+      "epoch": 3.6254501800720287,
+      "grad_norm": 0.07068238407373428,
+      "learning_rate": 5.0127384678750005e-06,
+      "loss": 0.8737,
+      "step": 3020
+    },
+    {
+      "epoch": 3.6374549819927973,
+      "grad_norm": 0.07324163615703583,
+      "learning_rate": 4.984430769803493e-06,
+      "loss": 0.8224,
+      "step": 3030
+    },
+    {
+      "epoch": 3.6494597839135654,
+      "grad_norm": 0.06777600198984146,
+      "learning_rate": 4.95612357077304e-06,
+      "loss": 0.8424,
+      "step": 3040
+    },
+    {
+      "epoch": 3.6614645858343335,
+      "grad_norm": 0.07339445501565933,
+      "learning_rate": 4.927817778115239e-06,
+      "loss": 0.8345,
+      "step": 3050
+    },
+    {
+      "epoch": 3.673469387755102,
+      "grad_norm": 0.07379196584224701,
+      "learning_rate": 4.899514299116603e-06,
+      "loss": 0.7919,
+      "step": 3060
+    },
+    {
+      "epoch": 3.6854741896758703,
+      "grad_norm": 0.06788371503353119,
+      "learning_rate": 4.871214040989492e-06,
+      "loss": 0.7914,
+      "step": 3070
+    },
+    {
+      "epoch": 3.697478991596639,
+      "grad_norm": 0.07233492285013199,
+      "learning_rate": 4.8429179108430246e-06,
+      "loss": 0.8388,
+      "step": 3080
+    },
+    {
+      "epoch": 3.709483793517407,
+      "grad_norm": 0.07236211746931076,
+      "learning_rate": 4.814626815654e-06,
+      "loss": 0.8619,
+      "step": 3090
+    },
+    {
+      "epoch": 3.721488595438175,
+      "grad_norm": 0.06984652578830719,
+      "learning_rate": 4.7863416622378426e-06,
+      "loss": 0.8038,
+      "step": 3100
+    },
+    {
+      "epoch": 3.7334933973589437,
+      "grad_norm": 0.06518954783678055,
+      "learning_rate": 4.7580633572195156e-06,
+      "loss": 0.8219,
+      "step": 3110
+    },
+    {
+      "epoch": 3.745498199279712,
+      "grad_norm": 0.06867209821939468,
+      "learning_rate": 4.729792807004472e-06,
+      "loss": 0.852,
+      "step": 3120
+    },
+    {
+      "epoch": 3.7575030012004804,
+      "grad_norm": 0.06609649956226349,
+      "learning_rate": 4.701530917749604e-06,
+      "loss": 0.803,
+      "step": 3130
+    },
+    {
+      "epoch": 3.7695078031212486,
+      "grad_norm": 0.06952491402626038,
+      "learning_rate": 4.673278595334191e-06,
+      "loss": 0.8405,
+      "step": 3140
+    },
+    {
+      "epoch": 3.7815126050420167,
+      "grad_norm": 0.0671219453215599,
+      "learning_rate": 4.645036745330863e-06,
+      "loss": 0.7897,
+      "step": 3150
+    },
+    {
+      "epoch": 3.7935174069627853,
+      "grad_norm": 0.06863446533679962,
+      "learning_rate": 4.616806272976584e-06,
+      "loss": 0.8096,
+      "step": 3160
+    },
+    {
+      "epoch": 3.8055222088835534,
+      "grad_norm": 0.07853756099939346,
+      "learning_rate": 4.588588083143625e-06,
+      "loss": 0.8143,
+      "step": 3170
+    },
+    {
+      "epoch": 3.817527010804322,
+      "grad_norm": 0.07341323047876358,
+      "learning_rate": 4.56038308031057e-06,
+      "loss": 0.8079,
+      "step": 3180
+    },
+    {
+      "epoch": 3.82953181272509,
+      "grad_norm": 0.06746179610490799,
+      "learning_rate": 4.5321921685333105e-06,
+      "loss": 0.8208,
+      "step": 3190
+    },
+    {
+      "epoch": 3.8415366146458583,
+      "grad_norm": 0.07139655202627182,
+      "learning_rate": 4.504016251416085e-06,
+      "loss": 0.8224,
+      "step": 3200
+    },
+    {
+      "epoch": 3.8535414165666264,
+      "grad_norm": 0.0728779211640358,
+      "learning_rate": 4.475856232082506e-06,
+      "loss": 0.8249,
+      "step": 3210
+    },
+    {
+      "epoch": 3.865546218487395,
+      "grad_norm": 0.06635065376758575,
+      "learning_rate": 4.447713013146612e-06,
+      "loss": 0.8219,
+      "step": 3220
+    },
+    {
+      "epoch": 3.877551020408163,
+      "grad_norm": 0.06454911828041077,
+      "learning_rate": 4.419587496683939e-06,
+      "loss": 0.8435,
+      "step": 3230
+    },
+    {
+      "epoch": 3.8895558223289317,
+      "grad_norm": 0.07131163030862808,
+      "learning_rate": 4.391480584202605e-06,
+      "loss": 0.8159,
+      "step": 3240
+    },
+    {
+      "epoch": 3.9015606242497,
+      "grad_norm": 0.07186614722013474,
+      "learning_rate": 4.363393176614411e-06,
+      "loss": 0.7746,
+      "step": 3250
+    },
+    {
+      "epoch": 3.913565426170468,
+      "grad_norm": 0.07490645349025726,
+      "learning_rate": 4.335326174205976e-06,
+      "loss": 0.8523,
+      "step": 3260
+    },
+    {
+      "epoch": 3.9255702280912366,
+      "grad_norm": 0.07318370789289474,
+      "learning_rate": 4.3072804766098605e-06,
+      "loss": 0.8141,
+      "step": 3270
+    },
+    {
+      "epoch": 3.9375750300120047,
+      "grad_norm": 0.07088599354028702,
+      "learning_rate": 4.279256982775744e-06,
+      "loss": 0.7967,
+      "step": 3280
+    },
+    {
+      "epoch": 3.9495798319327733,
+      "grad_norm": 0.07442790269851685,
+      "learning_rate": 4.251256590941615e-06,
+      "loss": 0.8148,
+      "step": 3290
+    },
+    {
+      "epoch": 3.9615846338535414,
+      "grad_norm": 0.07056485116481781,
+      "learning_rate": 4.223280198604962e-06,
+      "loss": 0.8109,
+      "step": 3300
+    },
+    {
+      "epoch": 3.9735894357743096,
+      "grad_norm": 0.07370737940073013,
+      "learning_rate": 4.195328702494024e-06,
+      "loss": 0.7949,
+      "step": 3310
+    },
+    {
+      "epoch": 3.985594237695078,
+      "grad_norm": 0.07084988802671432,
+      "learning_rate": 4.167402998539041e-06,
+      "loss": 0.8082,
+      "step": 3320
+    },
+    {
+      "epoch": 3.9975990396158463,
+      "grad_norm": 0.06937604397535324,
+      "learning_rate": 4.13950398184353e-06,
+      "loss": 0.8696,
+      "step": 3330
+    },
+    {
+      "epoch": 4.009603841536615,
+      "grad_norm": 0.07246571779251099,
+      "learning_rate": 4.111632546655608e-06,
+      "loss": 0.8017,
+      "step": 3340
+    },
+    {
+      "epoch": 4.021608643457383,
+      "grad_norm": 0.07561132311820984,
+      "learning_rate": 4.083789586339315e-06,
+      "loss": 0.801,
+      "step": 3350
+    },
+    {
+      "epoch": 4.033613445378151,
+      "grad_norm": 0.07453512400388718,
+      "learning_rate": 4.055975993345986e-06,
+      "loss": 0.8244,
+      "step": 3360
+    },
+    {
+      "epoch": 4.045618247298919,
+      "grad_norm": 0.0704115629196167,
+      "learning_rate": 4.028192659185649e-06,
+      "loss": 0.8294,
+      "step": 3370
+    },
+    {
+      "epoch": 4.057623049219688,
+      "grad_norm": 0.06777546554803848,
+      "learning_rate": 4.000440474398439e-06,
+      "loss": 0.8299,
+      "step": 3380
+    },
+    {
+      "epoch": 4.069627851140456,
+      "grad_norm": 0.08122120797634125,
+      "learning_rate": 3.972720328526059e-06,
+      "loss": 0.7945,
+      "step": 3390
+    },
+    {
+      "epoch": 4.081632653061225,
+      "grad_norm": 0.0788722038269043,
+      "learning_rate": 3.945033110083273e-06,
+      "loss": 0.8077,
+      "step": 3400
+    },
+    {
+      "epoch": 4.093637454981993,
+      "grad_norm": 0.0742715448141098,
+      "learning_rate": 3.917379706529416e-06,
+      "loss": 0.8234,
+      "step": 3410
+    },
+    {
+      "epoch": 4.105642256902761,
+      "grad_norm": 0.07386495918035507,
+      "learning_rate": 3.889761004239958e-06,
+      "loss": 0.8523,
+      "step": 3420
+    },
+    {
+      "epoch": 4.117647058823529,
+      "grad_norm": 0.07414324581623077,
+      "learning_rate": 3.862177888478082e-06,
+      "loss": 0.8185,
+      "step": 3430
+    },
+    {
+      "epoch": 4.129651860744298,
+      "grad_norm": 0.07434631884098053,
+      "learning_rate": 3.834631243366318e-06,
+      "loss": 0.8613,
+      "step": 3440
+    },
+    {
+      "epoch": 4.141656662665066,
+      "grad_norm": 0.07150658965110779,
+      "learning_rate": 3.8071219518582032e-06,
+      "loss": 0.7821,
+      "step": 3450
+    },
+    {
+      "epoch": 4.153661464585834,
+      "grad_norm": 0.06888636946678162,
+      "learning_rate": 3.7796508957099744e-06,
+      "loss": 0.7794,
+      "step": 3460
+    },
+    {
+      "epoch": 4.165666266506602,
+      "grad_norm": 0.07071466743946075,
+      "learning_rate": 3.752218955452311e-06,
+      "loss": 0.8367,
+      "step": 3470
+    },
+    {
+      "epoch": 4.177671068427371,
+      "grad_norm": 0.07348918169736862,
+      "learning_rate": 3.7248270103621128e-06,
+      "loss": 0.8085,
+      "step": 3480
+    },
+    {
+      "epoch": 4.18967587034814,
+      "grad_norm": 0.07582999765872955,
+      "learning_rate": 3.697475938434305e-06,
+      "loss": 0.837,
+      "step": 3490
+    },
+    {
+      "epoch": 4.201680672268908,
+      "grad_norm": 0.06943266838788986,
+      "learning_rate": 3.6701666163537125e-06,
+      "loss": 0.8058,
+      "step": 3500
+    },
+    {
+      "epoch": 4.213685474189676,
+      "grad_norm": 0.08077062666416168,
+      "learning_rate": 3.6428999194669478e-06,
+      "loss": 0.8466,
+      "step": 3510
+    },
+    {
+      "epoch": 4.225690276110444,
+      "grad_norm": 0.07306414097547531,
+      "learning_rate": 3.6156767217543507e-06,
+      "loss": 0.8098,
+      "step": 3520
+    },
+    {
+      "epoch": 4.237695078031212,
+      "grad_norm": 0.07767776399850845,
+      "learning_rate": 3.588497895801994e-06,
+      "loss": 0.8227,
+      "step": 3530
+    },
+    {
+      "epoch": 4.249699879951981,
+      "grad_norm": 0.06969287246465683,
+      "learning_rate": 3.5613643127736884e-06,
+      "loss": 0.8048,
+      "step": 3540
+    },
+    {
+      "epoch": 4.261704681872749,
+      "grad_norm": 0.08369370549917221,
+      "learning_rate": 3.534276842383074e-06,
+      "loss": 0.8361,
+      "step": 3550
+    },
+    {
+      "epoch": 4.2737094837935174,
+      "grad_norm": 0.07350631803274155,
+      "learning_rate": 3.5072363528657488e-06,
+      "loss": 0.8241,
+      "step": 3560
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.09979286789894104,
+      "learning_rate": 3.480243710951422e-06,
+      "loss": 0.8196,
+      "step": 3570
+    },
+    {
+      "epoch": 4.297719087635054,
+      "grad_norm": 0.07020698487758636,
+      "learning_rate": 3.4532997818361456e-06,
+      "loss": 0.8213,
+      "step": 3580
+    },
+    {
+      "epoch": 4.309723889555823,
+      "grad_norm": 0.0725114569067955,
+      "learning_rate": 3.426405429154581e-06,
+      "loss": 0.8232,
+      "step": 3590
+    },
+    {
+      "epoch": 4.321728691476591,
+      "grad_norm": 0.07283388823270798,
+      "learning_rate": 3.3995615149523102e-06,
+      "loss": 0.855,
+      "step": 3600
+    },
+    {
+      "epoch": 4.333733493397359,
+      "grad_norm": 0.0759081020951271,
+      "learning_rate": 3.372768899658212e-06,
+      "loss": 0.8116,
+      "step": 3610
+    },
+    {
+      "epoch": 4.345738295318127,
+      "grad_norm": 0.07164635509252548,
+      "learning_rate": 3.3460284420568802e-06,
+      "loss": 0.8147,
+      "step": 3620
+    },
+    {
+      "epoch": 4.357743097238895,
+      "grad_norm": 0.07087879627943039,
+      "learning_rate": 3.3193409992610897e-06,
+      "loss": 0.7556,
+      "step": 3630
+    },
+    {
+      "epoch": 4.369747899159664,
+      "grad_norm": 0.07007823884487152,
+      "learning_rate": 3.2927074266843394e-06,
+      "loss": 0.8307,
+      "step": 3640
+    },
+    {
+      "epoch": 4.3817527010804325,
+      "grad_norm": 0.07446113228797913,
+      "learning_rate": 3.26612857801342e-06,
+      "loss": 0.8155,
+      "step": 3650
+    },
+    {
+      "epoch": 4.393757503001201,
+      "grad_norm": 0.07006789743900299,
+      "learning_rate": 3.2396053051810515e-06,
+      "loss": 0.8118,
+      "step": 3660
+    },
+    {
+      "epoch": 4.405762304921969,
+      "grad_norm": 0.07662977278232574,
+      "learning_rate": 3.213138458338587e-06,
+      "loss": 0.8394,
+      "step": 3670
+    },
+    {
+      "epoch": 4.417767106842737,
+      "grad_norm": 0.08047199994325638,
+      "learning_rate": 3.186728885828748e-06,
+      "loss": 0.8321,
+      "step": 3680
+    },
+    {
+      "epoch": 4.429771908763505,
+      "grad_norm": 0.08414141833782196,
+      "learning_rate": 3.160377434158448e-06,
+      "loss": 0.7842,
+      "step": 3690
+    },
+    {
+      "epoch": 4.441776710684274,
+      "grad_norm": 0.06629359722137451,
+      "learning_rate": 3.134084947971644e-06,
+      "loss": 0.7735,
+      "step": 3700
+    },
+    {
+      "epoch": 4.453781512605042,
+      "grad_norm": 0.07507917284965515,
+      "learning_rate": 3.1078522700222713e-06,
+      "loss": 0.8031,
+      "step": 3710
+    },
+    {
+      "epoch": 4.46578631452581,
+      "grad_norm": 0.07528281211853027,
+      "learning_rate": 3.0816802411472362e-06,
+      "loss": 0.817,
+      "step": 3720
+    },
+    {
+      "epoch": 4.4777911164465785,
+      "grad_norm": 0.07056977599859238,
+      "learning_rate": 3.05556970023945e-06,
+      "loss": 0.8379,
+      "step": 3730
+    },
+    {
+      "epoch": 4.489795918367347,
+      "grad_norm": 0.07075227051973343,
+      "learning_rate": 3.029521484220953e-06,
+      "loss": 0.7759,
+      "step": 3740
+    },
+    {
+      "epoch": 4.501800720288116,
+      "grad_norm": 0.08120349049568176,
+      "learning_rate": 3.003536428016084e-06,
+      "loss": 0.812,
+      "step": 3750
+    },
+    {
+      "epoch": 4.513805522208884,
+      "grad_norm": 0.0773065984249115,
+      "learning_rate": 2.9776153645247135e-06,
+      "loss": 0.7805,
+      "step": 3760
+    },
+    {
+      "epoch": 4.525810324129652,
+      "grad_norm": 0.08427418768405914,
+      "learning_rate": 2.951759124595558e-06,
+      "loss": 0.8452,
+      "step": 3770
+    },
+    {
+      "epoch": 4.53781512605042,
+      "grad_norm": 0.07738559693098068,
+      "learning_rate": 2.9259685369995372e-06,
+      "loss": 0.8379,
+      "step": 3780
+    },
+    {
+      "epoch": 4.549819927971188,
+      "grad_norm": 0.07159377634525299,
+      "learning_rate": 2.900244428403214e-06,
+      "loss": 0.8248,
+      "step": 3790
+    },
+    {
+      "epoch": 4.561824729891956,
+      "grad_norm": 0.08146795630455017,
+      "learning_rate": 2.874587623342305e-06,
+      "loss": 0.7939,
+      "step": 3800
+    },
+    {
+      "epoch": 4.573829531812725,
+      "grad_norm": 0.06939620524644852,
+      "learning_rate": 2.8489989441952357e-06,
+      "loss": 0.8221,
+      "step": 3810
+    },
+    {
+      "epoch": 4.5858343337334935,
+      "grad_norm": 0.0722632110118866,
+      "learning_rate": 2.8234792111567915e-06,
+      "loss": 0.8351,
+      "step": 3820
+    },
+    {
+      "epoch": 4.597839135654262,
+      "grad_norm": 0.07618506252765656,
+      "learning_rate": 2.7980292422118282e-06,
+      "loss": 0.8298,
+      "step": 3830
+    },
+    {
+      "epoch": 4.60984393757503,
+      "grad_norm": 0.07517799735069275,
+      "learning_rate": 2.7726498531090507e-06,
+      "loss": 0.8025,
+      "step": 3840
+    },
+    {
+      "epoch": 4.621848739495798,
+      "grad_norm": 0.07529459148645401,
+      "learning_rate": 2.747341857334866e-06,
+      "loss": 0.8004,
+      "step": 3850
+    },
+    {
+      "epoch": 4.633853541416567,
+      "grad_norm": 0.07275144010782242,
+      "learning_rate": 2.7221060660873057e-06,
+      "loss": 0.8117,
+      "step": 3860
+    },
+    {
+      "epoch": 4.645858343337335,
+      "grad_norm": 0.0774557814002037,
+      "learning_rate": 2.6969432882500266e-06,
+      "loss": 0.846,
+      "step": 3870
+    },
+    {
+      "epoch": 4.657863145258103,
+      "grad_norm": 0.07570381462574005,
+      "learning_rate": 2.6718543303663912e-06,
+      "loss": 0.8258,
+      "step": 3880
+    },
+    {
+      "epoch": 4.669867947178871,
+      "grad_norm": 0.07305650413036346,
+      "learning_rate": 2.646839996613599e-06,
+      "loss": 0.8034,
+      "step": 3890
+    },
+    {
+      "epoch": 4.6818727490996395,
+      "grad_norm": 0.0736539289355278,
+      "learning_rate": 2.621901088776927e-06,
+      "loss": 0.7702,
+      "step": 3900
+    },
+    {
+      "epoch": 4.6938775510204085,
+      "grad_norm": 0.07207903265953064,
+      "learning_rate": 2.597038406224022e-06,
+      "loss": 0.7824,
+      "step": 3910
+    },
+    {
+      "epoch": 4.705882352941177,
+      "grad_norm": 0.07334660738706589,
+      "learning_rate": 2.5722527458792743e-06,
+      "loss": 0.8283,
+      "step": 3920
+    },
+    {
+      "epoch": 4.717887154861945,
+      "grad_norm": 0.06604504585266113,
+      "learning_rate": 2.5475449021982855e-06,
+      "loss": 0.8122,
+      "step": 3930
+    },
+    {
+      "epoch": 4.729891956782713,
+      "grad_norm": 0.07602079212665558,
+      "learning_rate": 2.5229156671423915e-06,
+      "loss": 0.7998,
+      "step": 3940
+    },
+    {
+      "epoch": 4.741896758703481,
+      "grad_norm": 0.08058540523052216,
+      "learning_rate": 2.498365830153285e-06,
+      "loss": 0.8088,
+      "step": 3950
+    },
+    {
+      "epoch": 4.75390156062425,
+      "grad_norm": 0.07190197706222534,
+      "learning_rate": 2.4738961781277097e-06,
+      "loss": 0.8067,
+      "step": 3960
+    },
+    {
+      "epoch": 4.765906362545018,
+      "grad_norm": 0.07564514130353928,
+      "learning_rate": 2.4495074953922425e-06,
+      "loss": 0.8065,
+      "step": 3970
+    },
+    {
+      "epoch": 4.777911164465786,
+      "grad_norm": 0.0754808560013771,
+      "learning_rate": 2.425200563678141e-06,
+      "loss": 0.8003,
+      "step": 3980
+    },
+    {
+      "epoch": 4.7899159663865545,
+      "grad_norm": 0.07958831638097763,
+      "learning_rate": 2.400976162096302e-06,
+      "loss": 0.8669,
+      "step": 3990
+    },
+    {
+      "epoch": 4.801920768307323,
+      "grad_norm": 0.07733064144849777,
+      "learning_rate": 2.376835067112276e-06,
+      "loss": 0.8245,
+      "step": 4000
+    },
+    {
+      "epoch": 4.813925570228092,
+      "grad_norm": 0.07412835210561752,
+      "learning_rate": 2.352778052521384e-06,
+      "loss": 0.8312,
+      "step": 4010
+    },
+    {
+      "epoch": 4.82593037214886,
+      "grad_norm": 0.07761096954345703,
+      "learning_rate": 2.3288058894239202e-06,
+      "loss": 0.8025,
+      "step": 4020
+    },
+    {
+      "epoch": 4.837935174069628,
+      "grad_norm": 0.06937912851572037,
+      "learning_rate": 2.3049193462004277e-06,
+      "loss": 0.8182,
+      "step": 4030
+    },
+    {
+      "epoch": 4.849939975990396,
+      "grad_norm": 0.07850659638643265,
+      "learning_rate": 2.2811191884870755e-06,
+      "loss": 0.8176,
+      "step": 4040
+    },
+    {
+      "epoch": 4.861944777911164,
+      "grad_norm": 0.0807727575302124,
+      "learning_rate": 2.2574061791511114e-06,
+      "loss": 0.8194,
+      "step": 4050
+    },
+    {
+      "epoch": 4.873949579831933,
+      "grad_norm": 0.07483908534049988,
+      "learning_rate": 2.2337810782664138e-06,
+      "loss": 0.7926,
+      "step": 4060
+    },
+    {
+      "epoch": 4.885954381752701,
+      "grad_norm": 0.08230816572904587,
+      "learning_rate": 2.2102446430891323e-06,
+      "loss": 0.758,
+      "step": 4070
+    },
+    {
+      "epoch": 4.8979591836734695,
+      "grad_norm": 0.0813629999756813,
+      "learning_rate": 2.186797628033406e-06,
+      "loss": 0.7916,
+      "step": 4080
+    },
+    {
+      "epoch": 4.909963985594238,
+      "grad_norm": 0.0788528099656105,
+      "learning_rate": 2.163440784647193e-06,
+      "loss": 0.8624,
+      "step": 4090
+    },
+    {
+      "epoch": 4.921968787515006,
+      "grad_norm": 0.07894974946975708,
+      "learning_rate": 2.140174861588175e-06,
+      "loss": 0.8324,
+      "step": 4100
+    },
+    {
+      "epoch": 4.933973589435775,
+      "grad_norm": 0.08242552727460861,
+      "learning_rate": 2.1170006045997577e-06,
+      "loss": 0.8335,
+      "step": 4110
+    },
+    {
+      "epoch": 4.945978391356543,
+      "grad_norm": 0.08206847310066223,
+      "learning_rate": 2.0939187564871768e-06,
+      "loss": 0.7888,
+      "step": 4120
+    },
+    {
+      "epoch": 4.957983193277311,
+      "grad_norm": 0.07950755953788757,
+      "learning_rate": 2.070930057093676e-06,
+      "loss": 0.815,
+      "step": 4130
+    },
+    {
+      "epoch": 4.969987995198079,
+      "grad_norm": 0.07717156410217285,
+      "learning_rate": 2.0480352432768063e-06,
+      "loss": 0.8451,
+      "step": 4140
+    },
+    {
+      "epoch": 4.981992797118847,
+      "grad_norm": 0.08424907177686691,
+      "learning_rate": 2.0252350488848015e-06,
+      "loss": 0.821,
+      "step": 4150
+    },
+    {
+      "epoch": 4.9939975990396155,
+      "grad_norm": 0.06911070644855499,
+      "learning_rate": 2.002530204733051e-06,
+      "loss": 0.8038,
+      "step": 4160
+    },
+    {
+      "epoch": 5.0060024009603845,
+      "grad_norm": 0.0768696665763855,
+      "learning_rate": 1.9799214385806804e-06,
+      "loss": 0.8111,
+      "step": 4170
+    },
+    {
+      "epoch": 5.018007202881153,
+      "grad_norm": 0.07615645974874496,
+      "learning_rate": 1.9574094751072287e-06,
+      "loss": 0.8087,
+      "step": 4180
+    },
+    {
+      "epoch": 5.030012004801921,
+      "grad_norm": 0.08393856883049011,
+      "learning_rate": 1.93499503588941e-06,
+      "loss": 0.8233,
+      "step": 4190
+    },
+    {
+      "epoch": 5.042016806722689,
+      "grad_norm": 0.07502923160791397,
+      "learning_rate": 1.9126788393779935e-06,
+      "loss": 0.836,
+      "step": 4200
+    },
+    {
+      "epoch": 5.054021608643457,
+      "grad_norm": 0.07897184789180756,
+      "learning_rate": 1.8904616008747728e-06,
+      "loss": 0.7748,
+      "step": 4210
+    },
+    {
+      "epoch": 5.066026410564226,
+      "grad_norm": 0.0756894201040268,
+      "learning_rate": 1.8683440325096314e-06,
+      "loss": 0.7697,
+      "step": 4220
+    },
+    {
+      "epoch": 5.078031212484994,
+      "grad_norm": 0.07758835703134537,
+      "learning_rate": 1.8463268432177301e-06,
+      "loss": 0.809,
+      "step": 4230
+    },
+    {
+      "epoch": 5.090036014405762,
+      "grad_norm": 0.08467192202806473,
+      "learning_rate": 1.8244107387167697e-06,
+      "loss": 0.81,
+      "step": 4240
+    },
+    {
+      "epoch": 5.1020408163265305,
+      "grad_norm": 0.07697489112615585,
+      "learning_rate": 1.8025964214843783e-06,
+      "loss": 0.836,
+      "step": 4250
+    },
+    {
+      "epoch": 5.114045618247299,
+      "grad_norm": 0.07720308005809784,
+      "learning_rate": 1.780884590735595e-06,
+      "loss": 0.8154,
+      "step": 4260
+    },
+    {
+      "epoch": 5.126050420168067,
+      "grad_norm": 0.0763421505689621,
+      "learning_rate": 1.7592759424004558e-06,
+      "loss": 0.8531,
+      "step": 4270
+    },
+    {
+      "epoch": 5.138055222088836,
+      "grad_norm": 0.08317775279283524,
+      "learning_rate": 1.7377711691016885e-06,
+      "loss": 0.7965,
+      "step": 4280
+    },
+    {
+      "epoch": 5.150060024009604,
+      "grad_norm": 0.08028178662061691,
+      "learning_rate": 1.7163709601325078e-06,
+      "loss": 0.7991,
+      "step": 4290
+    },
+    {
+      "epoch": 5.162064825930372,
+      "grad_norm": 0.07481025159358978,
+      "learning_rate": 1.695076001434524e-06,
+      "loss": 0.7828,
+      "step": 4300
+    },
+    {
+      "epoch": 5.17406962785114,
+      "grad_norm": 0.07857973873615265,
+      "learning_rate": 1.673886975575763e-06,
+      "loss": 0.8107,
+      "step": 4310
+    },
+    {
+      "epoch": 5.186074429771908,
+      "grad_norm": 0.08358076214790344,
+      "learning_rate": 1.6528045617287736e-06,
+      "loss": 0.8227,
+      "step": 4320
+    },
+    {
+      "epoch": 5.198079231692677,
+      "grad_norm": 0.08463145047426224,
+      "learning_rate": 1.6318294356488723e-06,
+      "loss": 0.8329,
+      "step": 4330
+    },
+    {
+      "epoch": 5.2100840336134455,
+      "grad_norm": 0.08062751591205597,
+      "learning_rate": 1.6109622696524775e-06,
+      "loss": 0.8005,
+      "step": 4340
+    },
+    {
+      "epoch": 5.222088835534214,
+      "grad_norm": 0.07316137105226517,
+      "learning_rate": 1.5902037325955533e-06,
+      "loss": 0.7609,
+      "step": 4350
+    },
+    {
+      "epoch": 5.234093637454982,
+      "grad_norm": 0.08057930320501328,
+      "learning_rate": 1.5695544898521841e-06,
+      "loss": 0.8192,
+      "step": 4360
+    },
+    {
+      "epoch": 5.24609843937575,
+      "grad_norm": 0.07895936816930771,
+      "learning_rate": 1.5490152032932338e-06,
+      "loss": 0.836,
+      "step": 4370
+    },
+    {
+      "epoch": 5.258103241296519,
+      "grad_norm": 0.0799291655421257,
+      "learning_rate": 1.5285865312651417e-06,
+      "loss": 0.8181,
+      "step": 4380
+    },
+    {
+      "epoch": 5.270108043217287,
+      "grad_norm": 0.07720459997653961,
+      "learning_rate": 1.508269128568815e-06,
+      "loss": 0.786,
+      "step": 4390
+    },
+    {
+      "epoch": 5.282112845138055,
+      "grad_norm": 0.08408314734697342,
+      "learning_rate": 1.4880636464386388e-06,
+      "loss": 0.7868,
+      "step": 4400
+    },
+    {
+      "epoch": 5.294117647058823,
+      "grad_norm": 0.08219724148511887,
+      "learning_rate": 1.4679707325216057e-06,
+      "loss": 0.8084,
+      "step": 4410
+    },
+    {
+      "epoch": 5.3061224489795915,
+      "grad_norm": 0.07679635286331177,
+      "learning_rate": 1.447991030856558e-06,
+      "loss": 0.829,
+      "step": 4420
+    },
+    {
+      "epoch": 5.3181272509003605,
+      "grad_norm": 0.07849011570215225,
+      "learning_rate": 1.4281251818535368e-06,
+      "loss": 0.7877,
+      "step": 4430
+    },
+    {
+      "epoch": 5.330132052821129,
+      "grad_norm": 0.08212018013000488,
+      "learning_rate": 1.4083738222732656e-06,
+      "loss": 0.8482,
+      "step": 4440
+    },
+    {
+      "epoch": 5.342136854741897,
+      "grad_norm": 0.07957368344068527,
+      "learning_rate": 1.388737585206733e-06,
+      "loss": 0.7826,
+      "step": 4450
+    },
+    {
+      "epoch": 5.354141656662665,
+      "grad_norm": 0.07521677762269974,
+      "learning_rate": 1.3692171000548987e-06,
+      "loss": 0.7929,
+      "step": 4460
+    },
+    {
+      "epoch": 5.366146458583433,
+      "grad_norm": 0.08291694521903992,
+      "learning_rate": 1.349812992508527e-06,
+      "loss": 0.7702,
+      "step": 4470
+    },
+    {
+      "epoch": 5.378151260504202,
+      "grad_norm": 0.08579208701848984,
+      "learning_rate": 1.3305258845281232e-06,
+      "loss": 0.7703,
+      "step": 4480
+    },
+    {
+      "epoch": 5.39015606242497,
+      "grad_norm": 0.07739798724651337,
+      "learning_rate": 1.311356394324002e-06,
+      "loss": 0.8072,
+      "step": 4490
+    },
+    {
+      "epoch": 5.402160864345738,
+      "grad_norm": 0.07492494583129883,
+      "learning_rate": 1.2923051363364736e-06,
+      "loss": 0.824,
+      "step": 4500
+    },
+    {
+      "epoch": 5.4141656662665065,
+      "grad_norm": 0.0814451351761818,
+      "learning_rate": 1.2733727212161478e-06,
+      "loss": 0.8135,
+      "step": 4510
+    },
+    {
+      "epoch": 5.426170468187275,
+      "grad_norm": 0.0821041613817215,
+      "learning_rate": 1.2545597558043559e-06,
+      "loss": 0.8094,
+      "step": 4520
+    },
+    {
+      "epoch": 5.438175270108043,
+      "grad_norm": 0.08446614444255829,
+      "learning_rate": 1.2358668431137082e-06,
+      "loss": 0.799,
+      "step": 4530
+    },
+    {
+      "epoch": 5.450180072028812,
+      "grad_norm": 0.08138373494148254,
+      "learning_rate": 1.2172945823087562e-06,
+      "loss": 0.8269,
+      "step": 4540
+    },
+    {
+      "epoch": 5.46218487394958,
+      "grad_norm": 0.08267953246831894,
+      "learning_rate": 1.1988435686867988e-06,
+      "loss": 0.7813,
+      "step": 4550
+    },
+    {
+      "epoch": 5.474189675870348,
+      "grad_norm": 0.08228688687086105,
+      "learning_rate": 1.1805143936587881e-06,
+      "loss": 0.7929,
+      "step": 4560
+    },
+    {
+      "epoch": 5.486194477791116,
+      "grad_norm": 0.07604286819696426,
+      "learning_rate": 1.1623076447303839e-06,
+      "loss": 0.7689,
+      "step": 4570
+    },
+    {
+      "epoch": 5.498199279711884,
+      "grad_norm": 0.08142969757318497,
+      "learning_rate": 1.1442239054831194e-06,
+      "loss": 0.8335,
+      "step": 4580
+    },
+    {
+      "epoch": 5.510204081632653,
+      "grad_norm": 0.07931126654148102,
+      "learning_rate": 1.1262637555556905e-06,
+      "loss": 0.8089,
+      "step": 4590
+    },
+    {
+      "epoch": 5.5222088835534215,
+      "grad_norm": 0.07624116539955139,
+      "learning_rate": 1.108427770625381e-06,
+      "loss": 0.8337,
+      "step": 4600
+    },
+    {
+      "epoch": 5.53421368547419,
+      "grad_norm": 0.07351070642471313,
+      "learning_rate": 1.0907165223896138e-06,
+      "loss": 0.8302,
+      "step": 4610
+    },
+    {
+      "epoch": 5.546218487394958,
+      "grad_norm": 0.08688168227672577,
+      "learning_rate": 1.0731305785476147e-06,
+      "loss": 0.8255,
+      "step": 4620
+    },
+    {
+      "epoch": 5.558223289315726,
+      "grad_norm": 0.08103056997060776,
+      "learning_rate": 1.0556705027822356e-06,
+      "loss": 0.7938,
+      "step": 4630
+    },
+    {
+      "epoch": 5.570228091236495,
+      "grad_norm": 0.08347798883914948,
+      "learning_rate": 1.0383368547418638e-06,
+      "loss": 0.8092,
+      "step": 4640
+    },
+    {
+      "epoch": 5.582232893157263,
+      "grad_norm": 0.07991039007902145,
+      "learning_rate": 1.0211301900224973e-06,
+      "loss": 0.8539,
+      "step": 4650
+    },
+    {
+      "epoch": 5.594237695078031,
+      "grad_norm": 0.07759847491979599,
+      "learning_rate": 1.0040510601499375e-06,
+      "loss": 0.8124,
+      "step": 4660
+    },
+    {
+      "epoch": 5.606242496998799,
+      "grad_norm": 0.08721420913934708,
+      "learning_rate": 9.871000125621017e-07,
+      "loss": 0.7767,
+      "step": 4670
+    },
+    {
+      "epoch": 5.6182472989195675,
+      "grad_norm": 0.08467090129852295,
+      "learning_rate": 9.702775905914847e-07,
+      "loss": 0.8203,
+      "step": 4680
+    },
+    {
+      "epoch": 5.630252100840336,
+      "grad_norm": 0.0824965164065361,
+      "learning_rate": 9.535843334477413e-07,
+      "loss": 0.8575,
+      "step": 4690
+    },
+    {
+      "epoch": 5.642256902761105,
+      "grad_norm": 0.07862916588783264,
+      "learning_rate": 9.370207762003963e-07,
+      "loss": 0.806,
+      "step": 4700
+    },
+    {
+      "epoch": 5.654261704681873,
+      "grad_norm": 0.07857292890548706,
+      "learning_rate": 9.205874497617063e-07,
+      "loss": 0.7803,
+      "step": 4710
+    },
+    {
+      "epoch": 5.666266506602641,
+      "grad_norm": 0.08207172900438309,
+      "learning_rate": 9.042848808696303e-07,
+      "loss": 0.815,
+      "step": 4720
+    },
+    {
+      "epoch": 5.678271308523409,
+      "grad_norm": 0.07979180663824081,
+      "learning_rate": 8.881135920709521e-07,
+      "loss": 0.8307,
+      "step": 4730
+    },
+    {
+      "epoch": 5.690276110444177,
+      "grad_norm": 0.07480289041996002,
+      "learning_rate": 8.720741017045325e-07,
+      "loss": 0.8415,
+      "step": 4740
+    },
+    {
+      "epoch": 5.702280912364946,
+      "grad_norm": 0.08345764875411987,
+      "learning_rate": 8.561669238846926e-07,
+      "loss": 0.8413,
+      "step": 4750
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.08144385367631912,
+      "learning_rate": 8.403925684847319e-07,
+      "loss": 0.8402,
+      "step": 4760
+    },
+    {
+      "epoch": 5.7262905162064826,
+      "grad_norm": 0.07664457708597183,
+      "learning_rate": 8.24751541120592e-07,
+      "loss": 0.8109,
+      "step": 4770
+    },
+    {
+      "epoch": 5.738295318127251,
+      "grad_norm": 0.07589486986398697,
+      "learning_rate": 8.092443431346447e-07,
+      "loss": 0.7965,
+      "step": 4780
+    },
+    {
+      "epoch": 5.750300120048019,
+      "grad_norm": 0.07511530071496964,
+      "learning_rate": 7.938714715796225e-07,
+      "loss": 0.8361,
+      "step": 4790
+    },
+    {
+      "epoch": 5.762304921968788,
+      "grad_norm": 0.07903044670820236,
+      "learning_rate": 7.786334192026929e-07,
+      "loss": 0.8345,
+      "step": 4800
+    },
+    {
+      "epoch": 5.774309723889556,
+      "grad_norm": 0.07598318159580231,
+      "learning_rate": 7.635306744296555e-07,
+      "loss": 0.7717,
+      "step": 4810
+    },
+    {
+      "epoch": 5.786314525810324,
+      "grad_norm": 0.08294142037630081,
+      "learning_rate": 7.485637213492952e-07,
+      "loss": 0.8084,
+      "step": 4820
+    },
+    {
+      "epoch": 5.798319327731092,
+      "grad_norm": 0.07683376967906952,
+      "learning_rate": 7.337330396978565e-07,
+      "loss": 0.8092,
+      "step": 4830
+    },
+    {
+      "epoch": 5.81032412965186,
+      "grad_norm": 0.08927746117115021,
+      "learning_rate": 7.190391048436729e-07,
+      "loss": 0.8369,
+      "step": 4840
+    },
+    {
+      "epoch": 5.822328931572629,
+      "grad_norm": 0.0817866250872612,
+      "learning_rate": 7.044823877719303e-07,
+      "loss": 0.8217,
+      "step": 4850
+    },
+    {
+      "epoch": 5.834333733493398,
+      "grad_norm": 0.08067484200000763,
+      "learning_rate": 6.900633550695651e-07,
+      "loss": 0.8313,
+      "step": 4860
+    },
+    {
+      "epoch": 5.846338535414166,
+      "grad_norm": 0.07704111933708191,
+      "learning_rate": 6.757824689103143e-07,
+      "loss": 0.8036,
+      "step": 4870
+    },
+    {
+      "epoch": 5.858343337334934,
+      "grad_norm": 0.07738319784402847,
+      "learning_rate": 6.616401870398997e-07,
+      "loss": 0.839,
+      "step": 4880
+    },
+    {
+      "epoch": 5.870348139255702,
+      "grad_norm": 0.07391506433486938,
+      "learning_rate": 6.476369627613522e-07,
+      "loss": 0.828,
+      "step": 4890
+    },
+    {
+      "epoch": 5.882352941176471,
+      "grad_norm": 0.08492206037044525,
+      "learning_rate": 6.337732449204886e-07,
+      "loss": 0.8255,
+      "step": 4900
+    },
+    {
+      "epoch": 5.894357743097239,
+      "grad_norm": 0.08384674787521362,
+      "learning_rate": 6.200494778915184e-07,
+      "loss": 0.7918,
+      "step": 4910
+    },
+    {
+      "epoch": 5.906362545018007,
+      "grad_norm": 0.09019272029399872,
+      "learning_rate": 6.064661015628043e-07,
+      "loss": 0.8298,
+      "step": 4920
+    },
+    {
+      "epoch": 5.918367346938775,
+      "grad_norm": 0.08015299588441849,
+      "learning_rate": 5.930235513227622e-07,
+      "loss": 0.854,
+      "step": 4930
+    },
+    {
+      "epoch": 5.930372148859544,
+      "grad_norm": 0.0807531401515007,
+      "learning_rate": 5.797222580459027e-07,
+      "loss": 0.8293,
+      "step": 4940
+    },
+    {
+      "epoch": 5.942376950780313,
+      "grad_norm": 0.08178652077913284,
+      "learning_rate": 5.665626480790221e-07,
+      "loss": 0.8492,
+      "step": 4950
+    },
+    {
+      "epoch": 5.954381752701081,
+      "grad_norm": 0.08085218816995621,
+      "learning_rate": 5.535451432275402e-07,
+      "loss": 0.82,
+      "step": 4960
+    },
+    {
+      "epoch": 5.966386554621849,
+      "grad_norm": 0.07937034964561462,
+      "learning_rate": 5.406701607419718e-07,
+      "loss": 0.8534,
+      "step": 4970
+    },
+    {
+      "epoch": 5.978391356542617,
+      "grad_norm": 0.07996084541082382,
+      "learning_rate": 5.27938113304562e-07,
+      "loss": 0.8082,
+      "step": 4980
+    },
+    {
+      "epoch": 5.990396158463385,
+      "grad_norm": 0.07725504040718079,
+      "learning_rate": 5.153494090160522e-07,
+      "loss": 0.8334,
+      "step": 4990
+    },
+    {
+      "epoch": 6.002400960384153,
+      "grad_norm": 0.08056992292404175,
+      "learning_rate": 5.029044513825992e-07,
+      "loss": 0.8501,
+      "step": 5000
+    },
+    {
+      "epoch": 6.014405762304922,
+      "grad_norm": 0.08236386626958847,
+      "learning_rate": 4.906036393028485e-07,
+      "loss": 0.8664,
+      "step": 5010
+    },
+    {
+      "epoch": 6.02641056422569,
+      "grad_norm": 0.08499747514724731,
+      "learning_rate": 4.784473670551376e-07,
+      "loss": 0.8105,
+      "step": 5020
+    },
+    {
+      "epoch": 6.038415366146459,
+      "grad_norm": 0.07700789719820023,
+      "learning_rate": 4.66436024284867e-07,
+      "loss": 0.7831,
+      "step": 5030
+    },
+    {
+      "epoch": 6.050420168067227,
+      "grad_norm": 0.07973281294107437,
+      "learning_rate": 4.54569995992008e-07,
+      "loss": 0.8325,
+      "step": 5040
+    },
+    {
+      "epoch": 6.062424969987995,
+      "grad_norm": 0.08407416939735413,
+      "learning_rate": 4.428496625187606e-07,
+      "loss": 0.8581,
+      "step": 5050
+    },
+    {
+      "epoch": 6.074429771908764,
+      "grad_norm": 0.08415237069129944,
+      "learning_rate": 4.312753995373664e-07,
+      "loss": 0.811,
+      "step": 5060
+    },
+    {
+      "epoch": 6.086434573829532,
+      "grad_norm": 0.08722592890262604,
+      "learning_rate": 4.198475780380623e-07,
+      "loss": 0.8302,
+      "step": 5070
+    },
+    {
+      "epoch": 6.0984393757503,
+      "grad_norm": 0.07541859149932861,
+      "learning_rate": 4.0856656431719043e-07,
+      "loss": 0.7965,
+      "step": 5080
+    },
+    {
+      "epoch": 6.110444177671068,
+      "grad_norm": 0.08787133544683456,
+      "learning_rate": 3.974327199654632e-07,
+      "loss": 0.8184,
+      "step": 5090
+    },
+    {
+      "epoch": 6.122448979591836,
+      "grad_norm": 0.08842716366052628,
+      "learning_rate": 3.8644640185636184e-07,
+      "loss": 0.8253,
+      "step": 5100
+    },
+    {
+      "epoch": 6.1344537815126055,
+      "grad_norm": 0.07971406728029251,
+      "learning_rate": 3.756079621347097e-07,
+      "loss": 0.8103,
+      "step": 5110
+    },
+    {
+      "epoch": 6.146458583433374,
+      "grad_norm": 0.079940065741539,
+      "learning_rate": 3.6491774820537616e-07,
+      "loss": 0.7491,
+      "step": 5120
+    },
+    {
+      "epoch": 6.158463385354142,
+      "grad_norm": 0.08296643197536469,
+      "learning_rate": 3.5437610272214484e-07,
+      "loss": 0.7997,
+      "step": 5130
+    },
+    {
+      "epoch": 6.17046818727491,
+      "grad_norm": 0.08030758798122406,
+      "learning_rate": 3.4398336357673146e-07,
+      "loss": 0.8038,
+      "step": 5140
+    },
+    {
+      "epoch": 6.182472989195678,
+      "grad_norm": 0.09067036956548691,
+      "learning_rate": 3.337398638879491e-07,
+      "loss": 0.8053,
+      "step": 5150
+    },
+    {
+      "epoch": 6.194477791116446,
+      "grad_norm": 0.08207783848047256,
+      "learning_rate": 3.2364593199103456e-07,
+      "loss": 0.7983,
+      "step": 5160
+    },
+    {
+      "epoch": 6.206482593037215,
+      "grad_norm": 0.0827014297246933,
+      "learning_rate": 3.1370189142712673e-07,
+      "loss": 0.812,
+      "step": 5170
+    },
+    {
+      "epoch": 6.218487394957983,
+      "grad_norm": 0.08526603877544403,
+      "learning_rate": 3.0390806093288707e-07,
+      "loss": 0.8212,
+      "step": 5180
+    },
+    {
+      "epoch": 6.230492196878751,
+      "grad_norm": 0.0812913328409195,
+      "learning_rate": 2.942647544302918e-07,
+      "loss": 0.7901,
+      "step": 5190
+    },
+    {
+      "epoch": 6.24249699879952,
+      "grad_norm": 0.07846683263778687,
+      "learning_rate": 2.8477228101656764e-07,
+      "loss": 0.7913,
+      "step": 5200
+    },
+    {
+      "epoch": 6.254501800720288,
+      "grad_norm": 0.08263833075761795,
+      "learning_rate": 2.7543094495427913e-07,
+      "loss": 0.8711,
+      "step": 5210
+    },
+    {
+      "epoch": 6.266506602641057,
+      "grad_norm": 0.07612549513578415,
+      "learning_rate": 2.6624104566158383e-07,
+      "loss": 0.8158,
+      "step": 5220
+    },
+    {
+      "epoch": 6.278511404561825,
+      "grad_norm": 0.07775219529867172,
+      "learning_rate": 2.57202877702632e-07,
+      "loss": 0.8169,
+      "step": 5230
+    },
+    {
+      "epoch": 6.290516206482593,
+      "grad_norm": 0.07751983404159546,
+      "learning_rate": 2.4831673077812034e-07,
+      "loss": 0.838,
+      "step": 5240
+    },
+    {
+      "epoch": 6.302521008403361,
+      "grad_norm": 0.0775042176246643,
+      "learning_rate": 2.395828897160152e-07,
+      "loss": 0.7884,
+      "step": 5250
+    },
+    {
+      "epoch": 6.314525810324129,
+      "grad_norm": 0.08516095578670502,
+      "learning_rate": 2.310016344624144e-07,
+      "loss": 0.7672,
+      "step": 5260
+    },
+    {
+      "epoch": 6.326530612244898,
+      "grad_norm": 0.07730519026517868,
+      "learning_rate": 2.225732400725783e-07,
+      "loss": 0.814,
+      "step": 5270
+    },
+    {
+      "epoch": 6.3385354141656665,
+      "grad_norm": 0.08212218433618546,
+      "learning_rate": 2.1429797670211416e-07,
+      "loss": 0.8149,
+      "step": 5280
+    },
+    {
+      "epoch": 6.350540216086435,
+      "grad_norm": 0.08002620935440063,
+      "learning_rate": 2.0617610959831613e-07,
+      "loss": 0.825,
+      "step": 5290
+    },
+    {
+      "epoch": 6.362545018007203,
+      "grad_norm": 0.08387167006731033,
+      "learning_rate": 1.982078990916597e-07,
+      "loss": 0.801,
+      "step": 5300
+    },
+    {
+      "epoch": 6.374549819927971,
+      "grad_norm": 0.07701138406991959,
+      "learning_rate": 1.9039360058746347e-07,
+      "loss": 0.844,
+      "step": 5310
+    },
+    {
+      "epoch": 6.38655462184874,
+      "grad_norm": 0.08538337796926498,
+      "learning_rate": 1.827334645576967e-07,
+      "loss": 0.8474,
+      "step": 5320
+    },
+    {
+      "epoch": 6.398559423769508,
+      "grad_norm": 0.08097740262746811,
+      "learning_rate": 1.7522773653295687e-07,
+      "loss": 0.7721,
+      "step": 5330
+    },
+    {
+      "epoch": 6.410564225690276,
+      "grad_norm": 0.08466079086065292,
+      "learning_rate": 1.6787665709459387e-07,
+      "loss": 0.8222,
+      "step": 5340
+    },
+    {
+      "epoch": 6.422569027611044,
+      "grad_norm": 0.07942117750644684,
+      "learning_rate": 1.6068046186700271e-07,
+      "loss": 0.8039,
+      "step": 5350
+    },
+    {
+      "epoch": 6.4345738295318124,
+      "grad_norm": 0.08644157648086548,
+      "learning_rate": 1.5363938151006897e-07,
+      "loss": 0.8215,
+      "step": 5360
+    },
+    {
+      "epoch": 6.4465786314525815,
+      "grad_norm": 0.07266967743635178,
+      "learning_rate": 1.4675364171177652e-07,
+      "loss": 0.801,
+      "step": 5370
+    },
+    {
+      "epoch": 6.45858343337335,
+      "grad_norm": 0.07520122081041336,
+      "learning_rate": 1.400234631809727e-07,
+      "loss": 0.8319,
+      "step": 5380
+    },
+    {
+      "epoch": 6.470588235294118,
+      "grad_norm": 0.08132954686880112,
+      "learning_rate": 1.3344906164029447e-07,
+      "loss": 0.8102,
+      "step": 5390
+    },
+    {
+      "epoch": 6.482593037214886,
+      "grad_norm": 0.07629051059484482,
+      "learning_rate": 1.2703064781925345e-07,
+      "loss": 0.8258,
+      "step": 5400
+    },
+    {
+      "epoch": 6.494597839135654,
+      "grad_norm": 0.08365026861429214,
+      "learning_rate": 1.2076842744748295e-07,
+      "loss": 0.8448,
+      "step": 5410
+    },
+    {
+      "epoch": 6.506602641056423,
+      "grad_norm": 0.07794215530157089,
+      "learning_rate": 1.146626012481411e-07,
+      "loss": 0.7922,
+      "step": 5420
+    },
+    {
+      "epoch": 6.518607442977191,
+      "grad_norm": 0.08589312434196472,
+      "learning_rate": 1.0871336493147866e-07,
+      "loss": 0.7865,
+      "step": 5430
+    },
+    {
+      "epoch": 6.530612244897959,
+      "grad_norm": 0.08183544129133224,
+      "learning_rate": 1.0292090918856634e-07,
+      "loss": 0.8062,
+      "step": 5440
+    },
+    {
+      "epoch": 6.5426170468187275,
+      "grad_norm": 0.08870046585798264,
+      "learning_rate": 9.728541968518135e-08,
+      "loss": 0.806,
+      "step": 5450
+    },
+    {
+      "epoch": 6.554621848739496,
+      "grad_norm": 0.08433903008699417,
+      "learning_rate": 9.180707705585545e-08,
+      "loss": 0.8488,
+      "step": 5460
+    },
+    {
+      "epoch": 6.566626650660264,
+      "grad_norm": 0.07865394651889801,
+      "learning_rate": 8.648605689808965e-08,
+      "loss": 0.8068,
+      "step": 5470
+    },
+    {
+      "epoch": 6.578631452581033,
+      "grad_norm": 0.08350817114114761,
+      "learning_rate": 8.132252976671972e-08,
+      "loss": 0.8063,
+      "step": 5480
+    },
+    {
+      "epoch": 6.590636254501801,
+      "grad_norm": 0.08394873887300491,
+      "learning_rate": 7.631666116845349e-08,
+      "loss": 0.815,
+      "step": 5490
+    },
+    {
+      "epoch": 6.602641056422569,
+      "grad_norm": 0.08824454993009567,
+      "learning_rate": 7.146861155656381e-08,
+      "loss": 0.8267,
+      "step": 5500
+    },
+    {
+      "epoch": 6.614645858343337,
+      "grad_norm": 0.08010157197713852,
+      "learning_rate": 6.677853632574616e-08,
+      "loss": 0.8069,
+      "step": 5510
+    },
+    {
+      "epoch": 6.626650660264105,
+      "grad_norm": 0.07938225567340851,
+      "learning_rate": 6.224658580713971e-08,
+      "loss": 0.773,
+      "step": 5520
+    },
+    {
+      "epoch": 6.6386554621848735,
+      "grad_norm": 0.0801059678196907,
+      "learning_rate": 5.787290526350464e-08,
+      "loss": 0.7957,
+      "step": 5530
+    },
+    {
+      "epoch": 6.6506602641056425,
+      "grad_norm": 0.07904977351427078,
+      "learning_rate": 5.365763488457021e-08,
+      "loss": 0.7944,
+      "step": 5540
+    },
+    {
+      "epoch": 6.662665066026411,
+      "grad_norm": 0.0811835378408432,
+      "learning_rate": 4.96009097825384e-08,
+      "loss": 0.8019,
+      "step": 5550
+    },
+    {
+      "epoch": 6.674669867947179,
+      "grad_norm": 0.07907839119434357,
+      "learning_rate": 4.5702859987753476e-08,
+      "loss": 0.802,
+      "step": 5560
+    },
+    {
+      "epoch": 6.686674669867947,
+      "grad_norm": 0.08255333453416824,
+      "learning_rate": 4.196361044453645e-08,
+      "loss": 0.7887,
+      "step": 5570
+    },
+    {
+      "epoch": 6.698679471788715,
+      "grad_norm": 0.07908835262060165,
+      "learning_rate": 3.838328100717714e-08,
+      "loss": 0.8445,
+      "step": 5580
+    },
+    {
+      "epoch": 6.710684273709484,
+      "grad_norm": 0.08409386873245239,
+      "learning_rate": 3.49619864360945e-08,
+      "loss": 0.8007,
+      "step": 5590
+    },
+    {
+      "epoch": 6.722689075630252,
+      "grad_norm": 0.0717078447341919,
+      "learning_rate": 3.169983639415841e-08,
+      "loss": 0.8396,
+      "step": 5600
+    },
+    {
+      "epoch": 6.73469387755102,
+      "grad_norm": 0.08110038191080093,
+      "learning_rate": 2.8596935443172545e-08,
+      "loss": 0.8113,
+      "step": 5610
+    },
+    {
+      "epoch": 6.7466986794717885,
+      "grad_norm": 0.07791571319103241,
+      "learning_rate": 2.5653383040524228e-08,
+      "loss": 0.8015,
+      "step": 5620
+    },
+    {
+      "epoch": 6.758703481392557,
+      "grad_norm": 0.0744730681180954,
+      "learning_rate": 2.2869273535997548e-08,
+      "loss": 0.8109,
+      "step": 5630
+    },
+    {
+      "epoch": 6.770708283313326,
+      "grad_norm": 0.08296327292919159,
+      "learning_rate": 2.0244696168746912e-08,
+      "loss": 0.8064,
+      "step": 5640
+    },
+    {
+      "epoch": 6.782713085234094,
+      "grad_norm": 0.07782948017120361,
+      "learning_rate": 1.7779735064438196e-08,
+      "loss": 0.8212,
+      "step": 5650
+    },
+    {
+      "epoch": 6.794717887154862,
+      "grad_norm": 0.08207013458013535,
+      "learning_rate": 1.5474469232552026e-08,
+      "loss": 0.8189,
+      "step": 5660
+    },
+    {
+      "epoch": 6.80672268907563,
+      "grad_norm": 0.08035318553447723,
+      "learning_rate": 1.332897256385135e-08,
+      "loss": 0.7571,
+      "step": 5670
+    },
+    {
+      "epoch": 6.818727490996398,
+      "grad_norm": 0.0768115296959877,
+      "learning_rate": 1.1343313828012237e-08,
+      "loss": 0.8449,
+      "step": 5680
+    },
+    {
+      "epoch": 6.830732292917167,
+      "grad_norm": 0.08141221851110458,
+      "learning_rate": 9.517556671420624e-09,
+      "loss": 0.8121,
+      "step": 5690
+    },
+    {
+      "epoch": 6.842737094837935,
+      "grad_norm": 0.08427503705024719,
+      "learning_rate": 7.851759615131182e-09,
+      "loss": 0.8023,
+      "step": 5700
+    },
+    {
+      "epoch": 6.8547418967587035,
+      "grad_norm": 0.07593482732772827,
+      "learning_rate": 6.345976052993252e-09,
+      "loss": 0.8071,
+      "step": 5710
+    },
+    {
+      "epoch": 6.866746698679472,
+      "grad_norm": 0.07990961521863937,
+      "learning_rate": 5.0002542499366645e-09,
+      "loss": 0.8203,
+      "step": 5720
+    },
+    {
+      "epoch": 6.87875150060024,
+      "grad_norm": 0.07622530311346054,
+      "learning_rate": 3.814637340427418e-09,
+      "loss": 0.8464,
+      "step": 5730
+    },
+    {
+      "epoch": 6.890756302521009,
+      "grad_norm": 0.08417756855487823,
+      "learning_rate": 2.7891633270848985e-09,
+      "loss": 0.8526,
+      "step": 5740
+    },
+    {
+      "epoch": 6.902761104441777,
+      "grad_norm": 0.08052320778369904,
+      "learning_rate": 1.9238650794611848e-09,
+      "loss": 0.8285,
+      "step": 5750
+    },
+    {
+      "epoch": 6.914765906362545,
+      "grad_norm": 0.08085649460554123,
+      "learning_rate": 1.2187703329891166e-09,
+      "loss": 0.8666,
+      "step": 5760
+    },
+    {
+      "epoch": 6.926770708283313,
+      "grad_norm": 0.08147014677524567,
+      "learning_rate": 6.739016880957794e-10,
+      "loss": 0.8447,
+      "step": 5770
+    },
+    {
+      "epoch": 6.938775510204081,
+      "grad_norm": 0.08438652008771896,
+      "learning_rate": 2.89276609472533e-10,
+      "loss": 0.8355,
+      "step": 5780
+    },
+    {
+      "epoch": 6.95078031212485,
+      "grad_norm": 0.08426124602556229,
+      "learning_rate": 6.490742552101026e-11,
+      "loss": 0.8033,
+      "step": 5790
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 5799,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 7,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.653539147717018e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}