diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8267 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 11757,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0025516713447307987,
+      "grad_norm": 67.57911682128906,
+      "learning_rate": 5.102040816326531e-08,
+      "loss": 4.2312,
+      "step": 10
+    },
+    {
+      "epoch": 0.0051033426894615975,
+      "grad_norm": 66.49089813232422,
+      "learning_rate": 1.0204081632653062e-07,
+      "loss": 4.2064,
+      "step": 20
+    },
+    {
+      "epoch": 0.007655014034192396,
+      "grad_norm": 63.74226379394531,
+      "learning_rate": 1.5306122448979592e-07,
+      "loss": 4.1169,
+      "step": 30
+    },
+    {
+      "epoch": 0.010206685378923195,
+      "grad_norm": 67.60637664794922,
+      "learning_rate": 2.0408163265306124e-07,
+      "loss": 3.9093,
+      "step": 40
+    },
+    {
+      "epoch": 0.012758356723653993,
+      "grad_norm": 58.32143783569336,
+      "learning_rate": 2.551020408163265e-07,
+      "loss": 3.2288,
+      "step": 50
+    },
+    {
+      "epoch": 0.015310028068384792,
+      "grad_norm": 44.05619430541992,
+      "learning_rate": 3.0612244897959183e-07,
+      "loss": 2.2825,
+      "step": 60
+    },
+    {
+      "epoch": 0.017861699413115592,
+      "grad_norm": 28.302391052246094,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 1.3651,
+      "step": 70
+    },
+    {
+      "epoch": 0.02041337075784639,
+      "grad_norm": 23.326255798339844,
+      "learning_rate": 4.081632653061225e-07,
+      "loss": 0.8761,
+      "step": 80
+    },
+    {
+      "epoch": 0.022965042102577188,
+      "grad_norm": 25.92929458618164,
+      "learning_rate": 4.5918367346938775e-07,
+      "loss": 0.5584,
+      "step": 90
+    },
+    {
+      "epoch": 0.025516713447307986,
+      "grad_norm": 18.867393493652344,
+      "learning_rate": 5.10204081632653e-07,
+      "loss": 0.4095,
+      "step": 100
+    },
+    {
+      "epoch": 0.028068384792038787,
+      "grad_norm": 26.22024917602539,
+      "learning_rate": 5.612244897959184e-07,
+      "loss": 0.321,
+      "step": 110
+    },
+    {
+      "epoch": 0.030620056136769585,
+      "grad_norm": 17.609697341918945,
+      "learning_rate": 6.122448979591837e-07,
+      "loss": 0.2597,
+      "step": 120
+    },
+    {
+      "epoch": 0.03317172748150038,
+      "grad_norm": 13.896285057067871,
+      "learning_rate": 6.63265306122449e-07,
+      "loss": 0.2442,
+      "step": 130
+    },
+    {
+      "epoch": 0.035723398826231184,
+      "grad_norm": 17.147130966186523,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.2128,
+      "step": 140
+    },
+    {
+      "epoch": 0.03827507017096198,
+      "grad_norm": 16.05860137939453,
+      "learning_rate": 7.653061224489796e-07,
+      "loss": 0.2197,
+      "step": 150
+    },
+    {
+      "epoch": 0.04082674151569278,
+      "grad_norm": 11.800621032714844,
+      "learning_rate": 8.16326530612245e-07,
+      "loss": 0.1864,
+      "step": 160
+    },
+    {
+      "epoch": 0.04337841286042358,
+      "grad_norm": 15.409507751464844,
+      "learning_rate": 8.673469387755102e-07,
+      "loss": 0.183,
+      "step": 170
+    },
+    {
+      "epoch": 0.045930084205154376,
+      "grad_norm": 20.847375869750977,
+      "learning_rate": 9.183673469387755e-07,
+      "loss": 0.1753,
+      "step": 180
+    },
+    {
+      "epoch": 0.048481755549885173,
+      "grad_norm": 22.60802459716797,
+      "learning_rate": 9.69387755102041e-07,
+      "loss": 0.1571,
+      "step": 190
+    },
+    {
+      "epoch": 0.05103342689461597,
+      "grad_norm": 14.373199462890625,
+      "learning_rate": 1.020408163265306e-06,
+      "loss": 0.1538,
+      "step": 200
+    },
+    {
+      "epoch": 0.05358509823934677,
+      "grad_norm": 10.351411819458008,
+      "learning_rate": 1.0714285714285716e-06,
+      "loss": 0.1412,
+      "step": 210
+    },
+    {
+      "epoch": 0.056136769584077574,
+      "grad_norm": 18.28008270263672,
+      "learning_rate": 1.122448979591837e-06,
+      "loss": 0.1428,
+      "step": 220
+    },
+    {
+      "epoch": 0.05868844092880837,
+      "grad_norm": 10.837668418884277,
+      "learning_rate": 1.173469387755102e-06,
+      "loss": 0.1383,
+      "step": 230
+    },
+    {
+      "epoch": 0.06124011227353917,
+      "grad_norm": 14.021323204040527,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 0.1261,
+      "step": 240
+    },
+    {
+      "epoch": 0.06379178361826997,
+      "grad_norm": 20.093788146972656,
+      "learning_rate": 1.2755102040816327e-06,
+      "loss": 0.1283,
+      "step": 250
+    },
+    {
+      "epoch": 0.06634345496300076,
+      "grad_norm": 12.026954650878906,
+      "learning_rate": 1.326530612244898e-06,
+      "loss": 0.1357,
+      "step": 260
+    },
+    {
+      "epoch": 0.06889512630773156,
+      "grad_norm": 15.393950462341309,
+      "learning_rate": 1.3775510204081633e-06,
+      "loss": 0.1188,
+      "step": 270
+    },
+    {
+      "epoch": 0.07144679765246237,
+      "grad_norm": 9.086231231689453,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.1205,
+      "step": 280
+    },
+    {
+      "epoch": 0.07399846899719316,
+      "grad_norm": 14.759052276611328,
+      "learning_rate": 1.479591836734694e-06,
+      "loss": 0.1095,
+      "step": 290
+    },
+    {
+      "epoch": 0.07655014034192396,
+      "grad_norm": 16.71165657043457,
+      "learning_rate": 1.5306122448979593e-06,
+      "loss": 0.1091,
+      "step": 300
+    },
+    {
+      "epoch": 0.07910181168665475,
+      "grad_norm": 9.48653507232666,
+      "learning_rate": 1.5816326530612246e-06,
+      "loss": 0.112,
+      "step": 310
+    },
+    {
+      "epoch": 0.08165348303138556,
+      "grad_norm": 13.855453491210938,
+      "learning_rate": 1.63265306122449e-06,
+      "loss": 0.1131,
+      "step": 320
+    },
+    {
+      "epoch": 0.08420515437611635,
+      "grad_norm": 8.177651405334473,
+      "learning_rate": 1.683673469387755e-06,
+      "loss": 0.1034,
+      "step": 330
+    },
+    {
+      "epoch": 0.08675682572084716,
+      "grad_norm": 10.370412826538086,
+      "learning_rate": 1.7346938775510204e-06,
+      "loss": 0.1044,
+      "step": 340
+    },
+    {
+      "epoch": 0.08930849706557796,
+      "grad_norm": 8.438699722290039,
+      "learning_rate": 1.7857142857142857e-06,
+      "loss": 0.0935,
+      "step": 350
+    },
+    {
+      "epoch": 0.09186016841030875,
+      "grad_norm": 8.10465145111084,
+      "learning_rate": 1.836734693877551e-06,
+      "loss": 0.09,
+      "step": 360
+    },
+    {
+      "epoch": 0.09441183975503956,
+      "grad_norm": 8.28003215789795,
+      "learning_rate": 1.8877551020408165e-06,
+      "loss": 0.0926,
+      "step": 370
+    },
+    {
+      "epoch": 0.09696351109977035,
+      "grad_norm": 8.460153579711914,
+      "learning_rate": 1.938775510204082e-06,
+      "loss": 0.0999,
+      "step": 380
+    },
+    {
+      "epoch": 0.09951518244450115,
+      "grad_norm": 8.236766815185547,
+      "learning_rate": 1.989795918367347e-06,
+      "loss": 0.0859,
+      "step": 390
+    },
+    {
+      "epoch": 0.10206685378923194,
+      "grad_norm": 8.402081489562988,
+      "learning_rate": 2.040816326530612e-06,
+      "loss": 0.0878,
+      "step": 400
+    },
+    {
+      "epoch": 0.10461852513396275,
+      "grad_norm": 24.110572814941406,
+      "learning_rate": 2.0918367346938776e-06,
+      "loss": 0.0936,
+      "step": 410
+    },
+    {
+      "epoch": 0.10717019647869354,
+      "grad_norm": 7.079232215881348,
+      "learning_rate": 2.142857142857143e-06,
+      "loss": 0.0881,
+      "step": 420
+    },
+    {
+      "epoch": 0.10972186782342434,
+      "grad_norm": 9.083630561828613,
+      "learning_rate": 2.1938775510204083e-06,
+      "loss": 0.0874,
+      "step": 430
+    },
+    {
+      "epoch": 0.11227353916815515,
+      "grad_norm": 7.967429161071777,
+      "learning_rate": 2.244897959183674e-06,
+      "loss": 0.0932,
+      "step": 440
+    },
+    {
+      "epoch": 0.11482521051288594,
+      "grad_norm": 9.130484580993652,
+      "learning_rate": 2.295918367346939e-06,
+      "loss": 0.0862,
+      "step": 450
+    },
+    {
+      "epoch": 0.11737688185761674,
+      "grad_norm": 6.145804405212402,
+      "learning_rate": 2.346938775510204e-06,
+      "loss": 0.0906,
+      "step": 460
+    },
+    {
+      "epoch": 0.11992855320234753,
+      "grad_norm": 9.372196197509766,
+      "learning_rate": 2.3979591836734696e-06,
+      "loss": 0.0964,
+      "step": 470
+    },
+    {
+      "epoch": 0.12248022454707834,
+      "grad_norm": 11.447351455688477,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 0.0798,
+      "step": 480
+    },
+    {
+      "epoch": 0.12503189589180913,
+      "grad_norm": 8.37991714477539,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0919,
+      "step": 490
+    },
+    {
+      "epoch": 0.12758356723653994,
+      "grad_norm": 6.319784164428711,
+      "learning_rate": 2.5510204081632653e-06,
+      "loss": 0.0794,
+      "step": 500
+    },
+    {
+      "epoch": 0.13013523858127074,
+      "grad_norm": 5.678624629974365,
+      "learning_rate": 2.6020408163265304e-06,
+      "loss": 0.0728,
+      "step": 510
+    },
+    {
+      "epoch": 0.13268690992600152,
+      "grad_norm": 7.462113857269287,
+      "learning_rate": 2.653061224489796e-06,
+      "loss": 0.0754,
+      "step": 520
+    },
+    {
+      "epoch": 0.13523858127073232,
+      "grad_norm": 4.428163528442383,
+      "learning_rate": 2.7040816326530615e-06,
+      "loss": 0.0873,
+      "step": 530
+    },
+    {
+      "epoch": 0.13779025261546313,
+      "grad_norm": 6.491887092590332,
+      "learning_rate": 2.7551020408163266e-06,
+      "loss": 0.071,
+      "step": 540
+    },
+    {
+      "epoch": 0.14034192396019393,
+      "grad_norm": 5.226602554321289,
+      "learning_rate": 2.806122448979592e-06,
+      "loss": 0.0737,
+      "step": 550
+    },
+    {
+      "epoch": 0.14289359530492474,
+      "grad_norm": 5.6451191902160645,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.0802,
+      "step": 560
+    },
+    {
+      "epoch": 0.1454452666496555,
+      "grad_norm": 6.9333600997924805,
+      "learning_rate": 2.9081632653061224e-06,
+      "loss": 0.0773,
+      "step": 570
+    },
+    {
+      "epoch": 0.14799693799438632,
+      "grad_norm": 5.240703582763672,
+      "learning_rate": 2.959183673469388e-06,
+      "loss": 0.0748,
+      "step": 580
+    },
+    {
+      "epoch": 0.15054860933911712,
+      "grad_norm": 7.25963020324707,
+      "learning_rate": 3.010204081632653e-06,
+      "loss": 0.0755,
+      "step": 590
+    },
+    {
+      "epoch": 0.15310028068384793,
+      "grad_norm": 6.730240821838379,
+      "learning_rate": 3.0612244897959185e-06,
+      "loss": 0.0724,
+      "step": 600
+    },
+    {
+      "epoch": 0.15565195202857873,
+      "grad_norm": 5.880996227264404,
+      "learning_rate": 3.1122448979591837e-06,
+      "loss": 0.0774,
+      "step": 610
+    },
+    {
+      "epoch": 0.1582036233733095,
+      "grad_norm": 3.661736488342285,
+      "learning_rate": 3.163265306122449e-06,
+      "loss": 0.0739,
+      "step": 620
+    },
+    {
+      "epoch": 0.16075529471804031,
+      "grad_norm": 6.331421375274658,
+      "learning_rate": 3.2142857142857143e-06,
+      "loss": 0.0741,
+      "step": 630
+    },
+    {
+      "epoch": 0.16330696606277112,
+      "grad_norm": 3.5448763370513916,
+      "learning_rate": 3.26530612244898e-06,
+      "loss": 0.0733,
+      "step": 640
+    },
+    {
+      "epoch": 0.16585863740750192,
+      "grad_norm": 4.7668023109436035,
+      "learning_rate": 3.316326530612245e-06,
+      "loss": 0.0649,
+      "step": 650
+    },
+    {
+      "epoch": 0.1684103087522327,
+      "grad_norm": 4.21407413482666,
+      "learning_rate": 3.36734693877551e-06,
+      "loss": 0.071,
+      "step": 660
+    },
+    {
+      "epoch": 0.1709619800969635,
+      "grad_norm": 6.080636501312256,
+      "learning_rate": 3.4183673469387756e-06,
+      "loss": 0.0753,
+      "step": 670
+    },
+    {
+      "epoch": 0.1735136514416943,
+      "grad_norm": 3.409601926803589,
+      "learning_rate": 3.4693877551020407e-06,
+      "loss": 0.0755,
+      "step": 680
+    },
+    {
+      "epoch": 0.17606532278642512,
+      "grad_norm": 4.414758682250977,
+      "learning_rate": 3.5204081632653062e-06,
+      "loss": 0.069,
+      "step": 690
+    },
+    {
+      "epoch": 0.17861699413115592,
+      "grad_norm": 6.209197044372559,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 0.0862,
+      "step": 700
+    },
+    {
+      "epoch": 0.1811686654758867,
+      "grad_norm": 7.26833963394165,
+      "learning_rate": 3.6224489795918365e-06,
+      "loss": 0.0699,
+      "step": 710
+    },
+    {
+      "epoch": 0.1837203368206175,
+      "grad_norm": 4.285154342651367,
+      "learning_rate": 3.673469387755102e-06,
+      "loss": 0.0655,
+      "step": 720
+    },
+    {
+      "epoch": 0.1862720081653483,
+      "grad_norm": 3.7479612827301025,
+      "learning_rate": 3.724489795918367e-06,
+      "loss": 0.0605,
+      "step": 730
+    },
+    {
+      "epoch": 0.1888236795100791,
+      "grad_norm": 4.82859468460083,
+      "learning_rate": 3.775510204081633e-06,
+      "loss": 0.074,
+      "step": 740
+    },
+    {
+      "epoch": 0.1913753508548099,
+      "grad_norm": 4.887997150421143,
+      "learning_rate": 3.826530612244898e-06,
+      "loss": 0.0738,
+      "step": 750
+    },
+    {
+      "epoch": 0.1939270221995407,
+      "grad_norm": 6.674945831298828,
+      "learning_rate": 3.877551020408164e-06,
+      "loss": 0.0584,
+      "step": 760
+    },
+    {
+      "epoch": 0.1964786935442715,
+      "grad_norm": 4.140785217285156,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 0.0608,
+      "step": 770
+    },
+    {
+      "epoch": 0.1990303648890023,
+      "grad_norm": 5.594793319702148,
+      "learning_rate": 3.979591836734694e-06,
+      "loss": 0.066,
+      "step": 780
+    },
+    {
+      "epoch": 0.2015820362337331,
+      "grad_norm": 8.363929748535156,
+      "learning_rate": 4.030612244897959e-06,
+      "loss": 0.0752,
+      "step": 790
+    },
+    {
+      "epoch": 0.20413370757846389,
+      "grad_norm": 4.264303684234619,
+      "learning_rate": 4.081632653061224e-06,
+      "loss": 0.0662,
+      "step": 800
+    },
+    {
+      "epoch": 0.2066853789231947,
+      "grad_norm": 3.919865131378174,
+      "learning_rate": 4.13265306122449e-06,
+      "loss": 0.0707,
+      "step": 810
+    },
+    {
+      "epoch": 0.2092370502679255,
+      "grad_norm": 3.484290838241577,
+      "learning_rate": 4.183673469387755e-06,
+      "loss": 0.0671,
+      "step": 820
+    },
+    {
+      "epoch": 0.2117887216126563,
+      "grad_norm": 4.169510841369629,
+      "learning_rate": 4.23469387755102e-06,
+      "loss": 0.0642,
+      "step": 830
+    },
+    {
+      "epoch": 0.21434039295738708,
+      "grad_norm": 2.740579843521118,
+      "learning_rate": 4.285714285714286e-06,
+      "loss": 0.0588,
+      "step": 840
+    },
+    {
+      "epoch": 0.21689206430211788,
+      "grad_norm": 3.434375047683716,
+      "learning_rate": 4.3367346938775506e-06,
+      "loss": 0.0537,
+      "step": 850
+    },
+    {
+      "epoch": 0.2194437356468487,
+      "grad_norm": 4.923778533935547,
+      "learning_rate": 4.3877551020408165e-06,
+      "loss": 0.0702,
+      "step": 860
+    },
+    {
+      "epoch": 0.2219954069915795,
+      "grad_norm": 3.2125160694122314,
+      "learning_rate": 4.438775510204082e-06,
+      "loss": 0.0618,
+      "step": 870
+    },
+    {
+      "epoch": 0.2245470783363103,
+      "grad_norm": 3.649038076400757,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 0.0641,
+      "step": 880
+    },
+    {
+      "epoch": 0.22709874968104107,
+      "grad_norm": 2.1178414821624756,
+      "learning_rate": 4.540816326530613e-06,
+      "loss": 0.0544,
+      "step": 890
+    },
+    {
+      "epoch": 0.22965042102577188,
+      "grad_norm": 5.1373467445373535,
+      "learning_rate": 4.591836734693878e-06,
+      "loss": 0.0574,
+      "step": 900
+    },
+    {
+      "epoch": 0.23220209237050268,
+      "grad_norm": 3.6157491207122803,
+      "learning_rate": 4.642857142857143e-06,
+      "loss": 0.0617,
+      "step": 910
+    },
+    {
+      "epoch": 0.2347537637152335,
+      "grad_norm": 4.86325216293335,
+      "learning_rate": 4.693877551020408e-06,
+      "loss": 0.055,
+      "step": 920
+    },
+    {
+      "epoch": 0.23730543505996426,
+      "grad_norm": 2.5493531227111816,
+      "learning_rate": 4.744897959183674e-06,
+      "loss": 0.0628,
+      "step": 930
+    },
+    {
+      "epoch": 0.23985710640469507,
+      "grad_norm": 2.850733518600464,
+      "learning_rate": 4.795918367346939e-06,
+      "loss": 0.0621,
+      "step": 940
+    },
+    {
+      "epoch": 0.24240877774942587,
+      "grad_norm": 3.4511430263519287,
+      "learning_rate": 4.846938775510204e-06,
+      "loss": 0.0574,
+      "step": 950
+    },
+    {
+      "epoch": 0.24496044909415668,
+      "grad_norm": 3.961998224258423,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 0.0522,
+      "step": 960
+    },
+    {
+      "epoch": 0.24751212043888748,
+      "grad_norm": 3.7193942070007324,
+      "learning_rate": 4.9489795918367345e-06,
+      "loss": 0.059,
+      "step": 970
+    },
+    {
+      "epoch": 0.25006379178361826,
+      "grad_norm": 2.4739599227905273,
+      "learning_rate": 5e-06,
+      "loss": 0.062,
+      "step": 980
+    },
+    {
+      "epoch": 0.2526154631283491,
+      "grad_norm": 5.526200771331787,
+      "learning_rate": 5.0510204081632655e-06,
+      "loss": 0.0613,
+      "step": 990
+    },
+    {
+      "epoch": 0.25516713447307987,
+      "grad_norm": 2.617518186569214,
+      "learning_rate": 5.102040816326531e-06,
+      "loss": 0.0605,
+      "step": 1000
+    },
+    {
+      "epoch": 0.25771880581781065,
+      "grad_norm": 2.276796340942383,
+      "learning_rate": 5.153061224489796e-06,
+      "loss": 0.0509,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2602704771625415,
+      "grad_norm": 1.4227871894836426,
+      "learning_rate": 5.204081632653061e-06,
+      "loss": 0.0562,
+      "step": 1020
+    },
+    {
+      "epoch": 0.26282214850727226,
+      "grad_norm": 4.203145503997803,
+      "learning_rate": 5.255102040816327e-06,
+      "loss": 0.0588,
+      "step": 1030
+    },
+    {
+      "epoch": 0.26537381985200303,
+      "grad_norm": 2.96553111076355,
+      "learning_rate": 5.306122448979592e-06,
+      "loss": 0.0578,
+      "step": 1040
+    },
+    {
+      "epoch": 0.26792549119673387,
+      "grad_norm": 2.9146676063537598,
+      "learning_rate": 5.357142857142858e-06,
+      "loss": 0.0616,
+      "step": 1050
+    },
+    {
+      "epoch": 0.27047716254146464,
+      "grad_norm": 2.5181937217712402,
+      "learning_rate": 5.408163265306123e-06,
+      "loss": 0.057,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2730288338861955,
+      "grad_norm": 2.5414464473724365,
+      "learning_rate": 5.459183673469387e-06,
+      "loss": 0.0567,
+      "step": 1070
+    },
+    {
+      "epoch": 0.27558050523092625,
+      "grad_norm": 1.901694893836975,
+      "learning_rate": 5.510204081632653e-06,
+      "loss": 0.0515,
+      "step": 1080
+    },
+    {
+      "epoch": 0.27813217657565703,
+      "grad_norm": 3.047863245010376,
+      "learning_rate": 5.561224489795918e-06,
+      "loss": 0.052,
+      "step": 1090
+    },
+    {
+      "epoch": 0.28068384792038786,
+      "grad_norm": 3.3792757987976074,
+      "learning_rate": 5.612244897959184e-06,
+      "loss": 0.065,
+      "step": 1100
+    },
+    {
+      "epoch": 0.28323551926511864,
+      "grad_norm": 4.105442047119141,
+      "learning_rate": 5.663265306122449e-06,
+      "loss": 0.0537,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2857871906098495,
+      "grad_norm": 4.40492582321167,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.0529,
+      "step": 1120
+    },
+    {
+      "epoch": 0.28833886195458025,
+      "grad_norm": 3.5336432456970215,
+      "learning_rate": 5.76530612244898e-06,
+      "loss": 0.0497,
+      "step": 1130
+    },
+    {
+      "epoch": 0.290890533299311,
+      "grad_norm": 3.989068031311035,
+      "learning_rate": 5.816326530612245e-06,
+      "loss": 0.0575,
+      "step": 1140
+    },
+    {
+      "epoch": 0.29344220464404186,
+      "grad_norm": 1.8923063278198242,
+      "learning_rate": 5.867346938775511e-06,
+      "loss": 0.0495,
+      "step": 1150
+    },
+    {
+      "epoch": 0.29599387598877264,
+      "grad_norm": 3.7669458389282227,
+      "learning_rate": 5.918367346938776e-06,
+      "loss": 0.0489,
+      "step": 1160
+    },
+    {
+      "epoch": 0.29854554733350347,
+      "grad_norm": 2.3980700969696045,
+      "learning_rate": 5.969387755102041e-06,
+      "loss": 0.0523,
+      "step": 1170
+    },
+    {
+      "epoch": 0.30109721867823425,
+      "grad_norm": 1.5095144510269165,
+      "learning_rate": 5.999997884283333e-06,
+      "loss": 0.0514,
+      "step": 1180
+    },
+    {
+      "epoch": 0.303648890022965,
+      "grad_norm": 3.0456111431121826,
+      "learning_rate": 5.9999740825051035e-06,
+      "loss": 0.0577,
+      "step": 1190
+    },
+    {
+      "epoch": 0.30620056136769586,
+      "grad_norm": 2.5419769287109375,
+      "learning_rate": 5.999923834513332e-06,
+      "loss": 0.0508,
+      "step": 1200
+    },
+    {
+      "epoch": 0.30875223271242663,
+      "grad_norm": 4.042941093444824,
+      "learning_rate": 5.999847140750981e-06,
+      "loss": 0.0578,
+      "step": 1210
+    },
+    {
+      "epoch": 0.31130390405715747,
+      "grad_norm": 3.173715591430664,
+      "learning_rate": 5.9997440018941426e-06,
+      "loss": 0.0586,
+      "step": 1220
+    },
+    {
+      "epoch": 0.31385557540188824,
+      "grad_norm": 2.7003302574157715,
+      "learning_rate": 5.9996144188520335e-06,
+      "loss": 0.0571,
+      "step": 1230
+    },
+    {
+      "epoch": 0.316407246746619,
+      "grad_norm": 3.698715925216675,
+      "learning_rate": 5.999458392766994e-06,
+      "loss": 0.0481,
+      "step": 1240
+    },
+    {
+      "epoch": 0.31895891809134985,
+      "grad_norm": 2.736983060836792,
+      "learning_rate": 5.999275925014465e-06,
+      "loss": 0.0511,
+      "step": 1250
+    },
+    {
+      "epoch": 0.32151058943608063,
+      "grad_norm": 1.4954791069030762,
+      "learning_rate": 5.999067017202991e-06,
+      "loss": 0.0496,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3240622607808114,
+      "grad_norm": 2.0082359313964844,
+      "learning_rate": 5.998831671174193e-06,
+      "loss": 0.0486,
+      "step": 1270
+    },
+    {
+      "epoch": 0.32661393212554224,
+      "grad_norm": 7.479063510894775,
+      "learning_rate": 5.998569889002759e-06,
+      "loss": 0.0614,
+      "step": 1280
+    },
+    {
+      "epoch": 0.329165603470273,
+      "grad_norm": 2.5666940212249756,
+      "learning_rate": 5.998281672996425e-06,
+      "loss": 0.0474,
+      "step": 1290
+    },
+    {
+      "epoch": 0.33171727481500385,
+      "grad_norm": 3.8249595165252686,
+      "learning_rate": 5.997967025695955e-06,
+      "loss": 0.0461,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3342689461597346,
+      "grad_norm": 2.665925979614258,
+      "learning_rate": 5.997625949875115e-06,
+      "loss": 0.0486,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3368206175044654,
+      "grad_norm": 1.3909707069396973,
+      "learning_rate": 5.997258448540651e-06,
+      "loss": 0.0546,
+      "step": 1320
+    },
+    {
+      "epoch": 0.33937228884919624,
+      "grad_norm": 2.5048632621765137,
+      "learning_rate": 5.996864524932266e-06,
+      "loss": 0.0462,
+      "step": 1330
+    },
+    {
+      "epoch": 0.341923960193927,
+      "grad_norm": 2.33251690864563,
+      "learning_rate": 5.996444182522586e-06,
+      "loss": 0.0521,
+      "step": 1340
+    },
+    {
+      "epoch": 0.34447563153865784,
+      "grad_norm": 1.7945542335510254,
+      "learning_rate": 5.99599742501713e-06,
+      "loss": 0.0474,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3470273028833886,
+      "grad_norm": 2.4074184894561768,
+      "learning_rate": 5.995524256354281e-06,
+      "loss": 0.0483,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3495789742281194,
+      "grad_norm": 2.026684522628784,
+      "learning_rate": 5.995024680705248e-06,
+      "loss": 0.0492,
+      "step": 1370
+    },
+    {
+      "epoch": 0.35213064557285023,
+      "grad_norm": 2.6286892890930176,
+      "learning_rate": 5.994498702474031e-06,
+      "loss": 0.0468,
+      "step": 1380
+    },
+    {
+      "epoch": 0.354682316917581,
+      "grad_norm": 1.3679580688476562,
+      "learning_rate": 5.993946326297379e-06,
+      "loss": 0.0472,
+      "step": 1390
+    },
+    {
+      "epoch": 0.35723398826231184,
+      "grad_norm": 2.495452642440796,
+      "learning_rate": 5.9933675570447556e-06,
+      "loss": 0.0449,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3597856596070426,
+      "grad_norm": 2.3190741539001465,
+      "learning_rate": 5.99276239981829e-06,
+      "loss": 0.0481,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3623373309517734,
+      "grad_norm": 1.6283559799194336,
+      "learning_rate": 5.9921308599527325e-06,
+      "loss": 0.0546,
+      "step": 1420
+    },
+    {
+      "epoch": 0.36488900229650423,
+      "grad_norm": 1.187363624572754,
+      "learning_rate": 5.991472943015412e-06,
+      "loss": 0.0431,
+      "step": 1430
+    },
+    {
+      "epoch": 0.367440673641235,
+      "grad_norm": 1.6949318647384644,
+      "learning_rate": 5.990788654806182e-06,
+      "loss": 0.0477,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3699923449859658,
+      "grad_norm": 1.8516268730163574,
+      "learning_rate": 5.9900780013573714e-06,
+      "loss": 0.0414,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3725440163306966,
+      "grad_norm": 2.0654797554016113,
+      "learning_rate": 5.989340988933732e-06,
+      "loss": 0.0384,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3750956876754274,
+      "grad_norm": 1.335979700088501,
+      "learning_rate": 5.988577624032384e-06,
+      "loss": 0.0405,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3776473590201582,
+      "grad_norm": 3.470012664794922,
+      "learning_rate": 5.987787913382754e-06,
+      "loss": 0.0464,
+      "step": 1480
+    },
+    {
+      "epoch": 0.380199030364889,
+      "grad_norm": 1.7210215330123901,
+      "learning_rate": 5.9869718639465234e-06,
+      "loss": 0.051,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3827507017096198,
+      "grad_norm": 2.3212034702301025,
+      "learning_rate": 5.986129482917558e-06,
+      "loss": 0.0474,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3853023730543506,
+      "grad_norm": 3.0615479946136475,
+      "learning_rate": 5.985260777721852e-06,
+      "loss": 0.044,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3878540443990814,
+      "grad_norm": 1.6388686895370483,
+      "learning_rate": 5.9843657560174604e-06,
+      "loss": 0.046,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3904057157438122,
+      "grad_norm": 2.6303327083587646,
+      "learning_rate": 5.983444425694428e-06,
+      "loss": 0.0435,
+      "step": 1530
+    },
+    {
+      "epoch": 0.392957387088543,
+      "grad_norm": 1.6358684301376343,
+      "learning_rate": 5.982496794874726e-06,
+      "loss": 0.0382,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3955090584332738,
+      "grad_norm": 1.609343409538269,
+      "learning_rate": 5.981522871912175e-06,
+      "loss": 0.0411,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3980607297780046,
+      "grad_norm": 2.244525194168091,
+      "learning_rate": 5.980522665392373e-06,
+      "loss": 0.0473,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4006124011227354,
+      "grad_norm": 3.230689525604248,
+      "learning_rate": 5.979496184132622e-06,
+      "loss": 0.0452,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4031640724674662,
+      "grad_norm": 3.2919723987579346,
+      "learning_rate": 5.978443437181849e-06,
+      "loss": 0.0449,
+      "step": 1580
+    },
+    {
+      "epoch": 0.405715743812197,
+      "grad_norm": 1.8723994493484497,
+      "learning_rate": 5.977364433820522e-06,
+      "loss": 0.0408,
+      "step": 1590
+    },
+    {
+      "epoch": 0.40826741515692777,
+      "grad_norm": 3.2039082050323486,
+      "learning_rate": 5.976259183560576e-06,
+      "loss": 0.0465,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4108190865016586,
+      "grad_norm": 1.4034531116485596,
+      "learning_rate": 5.975127696145324e-06,
+      "loss": 0.0417,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4133707578463894,
+      "grad_norm": 2.3331546783447266,
+      "learning_rate": 5.97396998154937e-06,
+      "loss": 0.0418,
+      "step": 1620
+    },
+    {
+      "epoch": 0.41592242919112016,
+      "grad_norm": 1.5517454147338867,
+      "learning_rate": 5.972786049978524e-06,
+      "loss": 0.0428,
+      "step": 1630
+    },
+    {
+      "epoch": 0.418474100535851,
+      "grad_norm": 3.171250104904175,
+      "learning_rate": 5.971575911869713e-06,
+      "loss": 0.0408,
+      "step": 1640
+    },
+    {
+      "epoch": 0.42102577188058177,
+      "grad_norm": 2.458007574081421,
+      "learning_rate": 5.9703395778908865e-06,
+      "loss": 0.0472,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4235774432253126,
+      "grad_norm": 1.4991718530654907,
+      "learning_rate": 5.969077058940923e-06,
+      "loss": 0.0408,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4261291145700434,
+      "grad_norm": 2.1240978240966797,
+      "learning_rate": 5.967788366149531e-06,
+      "loss": 0.0414,
+      "step": 1670
+    },
+    {
+      "epoch": 0.42868078591477415,
+      "grad_norm": 3.0107381343841553,
+      "learning_rate": 5.96647351087716e-06,
+      "loss": 0.04,
+      "step": 1680
+    },
+    {
+      "epoch": 0.431232457259505,
+      "grad_norm": 2.087351083755493,
+      "learning_rate": 5.965132504714891e-06,
+      "loss": 0.0412,
+      "step": 1690
+    },
+    {
+      "epoch": 0.43378412860423576,
+      "grad_norm": 1.0149552822113037,
+      "learning_rate": 5.963765359484337e-06,
+      "loss": 0.0424,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4363357999489666,
+      "grad_norm": 2.0345070362091064,
+      "learning_rate": 5.962372087237539e-06,
+      "loss": 0.035,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4388874712936974,
+      "grad_norm": 2.1437127590179443,
+      "learning_rate": 5.9609527002568646e-06,
+      "loss": 0.0358,
+      "step": 1720
+    },
+    {
+      "epoch": 0.44143914263842815,
+      "grad_norm": 2.7861342430114746,
+      "learning_rate": 5.9595072110548906e-06,
+      "loss": 0.0403,
+      "step": 1730
+    },
+    {
+      "epoch": 0.443990813983159,
+      "grad_norm": 1.211660385131836,
+      "learning_rate": 5.958035632374299e-06,
+      "loss": 0.043,
+      "step": 1740
+    },
+    {
+      "epoch": 0.44654248532788976,
+      "grad_norm": 3.840953826904297,
+      "learning_rate": 5.956537977187764e-06,
+      "loss": 0.0393,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4490941566726206,
+      "grad_norm": 1.265952706336975,
+      "learning_rate": 5.955014258697835e-06,
+      "loss": 0.043,
+      "step": 1760
+    },
+    {
+      "epoch": 0.45164582801735137,
+      "grad_norm": 4.032707214355469,
+      "learning_rate": 5.953464490336823e-06,
+      "loss": 0.0323,
+      "step": 1770
+    },
+    {
+      "epoch": 0.45419749936208215,
+      "grad_norm": 1.410612940788269,
+      "learning_rate": 5.951888685766683e-06,
+      "loss": 0.0381,
+      "step": 1780
+    },
+    {
+      "epoch": 0.456749170706813,
+      "grad_norm": 1.7868704795837402,
+      "learning_rate": 5.950286858878886e-06,
+      "loss": 0.0417,
+      "step": 1790
+    },
+    {
+      "epoch": 0.45930084205154376,
+      "grad_norm": 3.2607359886169434,
+      "learning_rate": 5.948659023794309e-06,
+      "loss": 0.0388,
+      "step": 1800
+    },
+    {
+      "epoch": 0.46185251339627453,
+      "grad_norm": 1.940967321395874,
+      "learning_rate": 5.9470051948631e-06,
+      "loss": 0.0407,
+      "step": 1810
+    },
+    {
+      "epoch": 0.46440418474100537,
+      "grad_norm": 1.7820950746536255,
+      "learning_rate": 5.945325386664556e-06,
+      "loss": 0.0351,
+      "step": 1820
+    },
+    {
+      "epoch": 0.46695585608573614,
+      "grad_norm": 3.1679279804229736,
+      "learning_rate": 5.943619614006993e-06,
+      "loss": 0.0401,
+      "step": 1830
+    },
+    {
+      "epoch": 0.469507527430467,
+      "grad_norm": 0.912773609161377,
+      "learning_rate": 5.941887891927618e-06,
+      "loss": 0.0311,
+      "step": 1840
+    },
+    {
+      "epoch": 0.47205919877519775,
+      "grad_norm": 2.1303391456604004,
+      "learning_rate": 5.9401302356923925e-06,
+      "loss": 0.0398,
+      "step": 1850
+    },
+    {
+      "epoch": 0.47461087011992853,
+      "grad_norm": 1.0795629024505615,
+      "learning_rate": 5.938346660795902e-06,
+      "loss": 0.0384,
+      "step": 1860
+    },
+    {
+      "epoch": 0.47716254146465936,
+      "grad_norm": 1.5061334371566772,
+      "learning_rate": 5.936537182961215e-06,
+      "loss": 0.032,
+      "step": 1870
+    },
+    {
+      "epoch": 0.47971421280939014,
+      "grad_norm": 1.0948927402496338,
+      "learning_rate": 5.93470181813975e-06,
+      "loss": 0.0378,
+      "step": 1880
+    },
+    {
+      "epoch": 0.48226588415412097,
+      "grad_norm": 2.728391647338867,
+      "learning_rate": 5.9328405825111296e-06,
+      "loss": 0.0351,
+      "step": 1890
+    },
+    {
+      "epoch": 0.48481755549885175,
+      "grad_norm": 2.6559503078460693,
+      "learning_rate": 5.93095349248304e-06,
+      "loss": 0.0321,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4873692268435825,
+      "grad_norm": 0.8435110449790955,
+      "learning_rate": 5.929040564691087e-06,
+      "loss": 0.0371,
+      "step": 1910
+    },
+    {
+      "epoch": 0.48992089818831336,
+      "grad_norm": 1.813321828842163,
+      "learning_rate": 5.92710181599865e-06,
+      "loss": 0.0374,
+      "step": 1920
+    },
+    {
+      "epoch": 0.49247256953304414,
+      "grad_norm": 1.7775193452835083,
+      "learning_rate": 5.925137263496729e-06,
+      "loss": 0.0367,
+      "step": 1930
+    },
+    {
+      "epoch": 0.49502424087777497,
+      "grad_norm": 1.2854493856430054,
+      "learning_rate": 5.9231469245038e-06,
+      "loss": 0.0415,
+      "step": 1940
+    },
+    {
+      "epoch": 0.49757591222250575,
+      "grad_norm": 1.0430928468704224,
+      "learning_rate": 5.921130816565655e-06,
+      "loss": 0.0412,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5001275835672365,
+      "grad_norm": 1.0425087213516235,
+      "learning_rate": 5.919088957455258e-06,
+      "loss": 0.0394,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5026792549119673,
+      "grad_norm": 1.738433837890625,
+      "learning_rate": 5.917021365172577e-06,
+      "loss": 0.0314,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5052309262566982,
+      "grad_norm": 0.9304326176643372,
+      "learning_rate": 5.914928057944431e-06,
+      "loss": 0.0393,
+      "step": 1980
+    },
+    {
+      "epoch": 0.507782597601429,
+      "grad_norm": 2.4667181968688965,
+      "learning_rate": 5.912809054224332e-06,
+      "loss": 0.0375,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5103342689461597,
+      "grad_norm": 1.0782585144042969,
+      "learning_rate": 5.910664372692313e-06,
+      "loss": 0.0336,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5128859402908905,
+      "grad_norm": 1.5189335346221924,
+      "learning_rate": 5.9084940322547754e-06,
+      "loss": 0.0366,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5154376116356213,
+      "grad_norm": 2.4441661834716797,
+      "learning_rate": 5.9062980520443115e-06,
+      "loss": 0.0341,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5179892829803522,
+      "grad_norm": 1.2969788312911987,
+      "learning_rate": 5.904076451419543e-06,
+      "loss": 0.0309,
+      "step": 2030
+    },
+    {
+      "epoch": 0.520540954325083,
+      "grad_norm": 1.3563958406448364,
+      "learning_rate": 5.901829249964948e-06,
+      "loss": 0.0276,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5230926256698137,
+      "grad_norm": 2.4599099159240723,
+      "learning_rate": 5.899556467490687e-06,
+      "loss": 0.0328,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5256442970145445,
+      "grad_norm": 2.01973557472229,
+      "learning_rate": 5.897258124032429e-06,
+      "loss": 0.038,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5281959683592753,
+      "grad_norm": 1.541785717010498,
+      "learning_rate": 5.894934239851178e-06,
+      "loss": 0.0342,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5307476397040061,
+      "grad_norm": 1.1081935167312622,
+      "learning_rate": 5.892584835433089e-06,
+      "loss": 0.033,
+      "step": 2080
+    },
+    {
+      "epoch": 0.533299311048737,
+      "grad_norm": 1.496882677078247,
+      "learning_rate": 5.890209931489292e-06,
+      "loss": 0.029,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5358509823934677,
+      "grad_norm": 12.85782241821289,
+      "learning_rate": 5.887809548955706e-06,
+      "loss": 0.0404,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5384026537381985,
+      "grad_norm": 1.9677023887634277,
+      "learning_rate": 5.885383708992856e-06,
+      "loss": 0.0346,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5409543250829293,
+      "grad_norm": 1.9975990056991577,
+      "learning_rate": 5.88293243298569e-06,
+      "loss": 0.0364,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5435059964276601,
+      "grad_norm": 1.0146771669387817,
+      "learning_rate": 5.880455742543382e-06,
+      "loss": 0.0303,
+      "step": 2130
+    },
+    {
+      "epoch": 0.546057667772391,
+      "grad_norm": 2.213881731033325,
+      "learning_rate": 5.877953659499151e-06,
+      "loss": 0.0319,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5486093391171217,
+      "grad_norm": 1.3682197332382202,
+      "learning_rate": 5.875426205910057e-06,
+      "loss": 0.0343,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5511610104618525,
+      "grad_norm": 1.53465735912323,
+      "learning_rate": 5.872873404056825e-06,
+      "loss": 0.0316,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5537126818065833,
+      "grad_norm": 1.061879277229309,
+      "learning_rate": 5.870295276443624e-06,
+      "loss": 0.0317,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5562643531513141,
+      "grad_norm": 1.099205493927002,
+      "learning_rate": 5.867691845797893e-06,
+      "loss": 0.0312,
+      "step": 2180
+    },
+    {
+      "epoch": 0.558816024496045,
+      "grad_norm": 0.7490290999412537,
+      "learning_rate": 5.8650631350701205e-06,
+      "loss": 0.0269,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5613676958407757,
+      "grad_norm": 1.189103603363037,
+      "learning_rate": 5.862409167433658e-06,
+      "loss": 0.0312,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5639193671855065,
+      "grad_norm": 1.3587863445281982,
+      "learning_rate": 5.859729966284505e-06,
+      "loss": 0.0323,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5664710385302373,
+      "grad_norm": 0.6706628203392029,
+      "learning_rate": 5.857025555241108e-06,
+      "loss": 0.0296,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5690227098749681,
+      "grad_norm": 7.282918930053711,
+      "learning_rate": 5.85429595814415e-06,
+      "loss": 0.0297,
+      "step": 2230
+    },
+    {
+      "epoch": 0.571574381219699,
+      "grad_norm": 1.1752480268478394,
+      "learning_rate": 5.85154119905634e-06,
+      "loss": 0.0288,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5741260525644297,
+      "grad_norm": 1.0217607021331787,
+      "learning_rate": 5.848761302262205e-06,
+      "loss": 0.0309,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5766777239091605,
+      "grad_norm": 2.6321234703063965,
+      "learning_rate": 5.845956292267872e-06,
+      "loss": 0.035,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5792293952538913,
+      "grad_norm": 0.8494094610214233,
+      "learning_rate": 5.8431261938008515e-06,
+      "loss": 0.031,
+      "step": 2270
+    },
+    {
+      "epoch": 0.581781066598622,
+      "grad_norm": 1.3556350469589233,
+      "learning_rate": 5.840271031809821e-06,
+      "loss": 0.0305,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5843327379433529,
+      "grad_norm": 0.7412778735160828,
+      "learning_rate": 5.8373908314644075e-06,
+      "loss": 0.0262,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5868844092880837,
+      "grad_norm": 1.4868544340133667,
+      "learning_rate": 5.83448561815496e-06,
+      "loss": 0.0293,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5894360806328145,
+      "grad_norm": 0.9419906735420227,
+      "learning_rate": 5.831555417492332e-06,
+      "loss": 0.0335,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5919877519775453,
+      "grad_norm": 1.1363625526428223,
+      "learning_rate": 5.8286002553076485e-06,
+      "loss": 0.0336,
+      "step": 2320
+    },
+    {
+      "epoch": 0.594539423322276,
+      "grad_norm": 2.777545213699341,
+      "learning_rate": 5.8256201576520865e-06,
+      "loss": 0.0306,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5970910946670069,
+      "grad_norm": 1.014387845993042,
+      "learning_rate": 5.822615150796639e-06,
+      "loss": 0.0319,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5996427660117377,
+      "grad_norm": 1.114518404006958,
+      "learning_rate": 5.8195852612318845e-06,
+      "loss": 0.0329,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6021944373564685,
+      "grad_norm": 1.7633538246154785,
+      "learning_rate": 5.816530515667758e-06,
+      "loss": 0.0306,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6047461087011993,
+      "grad_norm": 1.5200434923171997,
+      "learning_rate": 5.813450941033309e-06,
+      "loss": 0.0296,
+      "step": 2370
+    },
+    {
+      "epoch": 0.60729778004593,
+      "grad_norm": 0.8720308542251587,
+      "learning_rate": 5.810346564476468e-06,
+      "loss": 0.0282,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6098494513906609,
+      "grad_norm": 1.8047078847885132,
+      "learning_rate": 5.807217413363805e-06,
+      "loss": 0.039,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6124011227353917,
+      "grad_norm": 1.6201497316360474,
+      "learning_rate": 5.804063515280296e-06,
+      "loss": 0.0307,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6149527940801225,
+      "grad_norm": 1.6760756969451904,
+      "learning_rate": 5.800884898029062e-06,
+      "loss": 0.0254,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6175044654248533,
+      "grad_norm": 3.973919153213501,
+      "learning_rate": 5.7976815896311476e-06,
+      "loss": 0.0303,
+      "step": 2420
+    },
+    {
+      "epoch": 0.620056136769584,
+      "grad_norm": 0.60660719871521,
+      "learning_rate": 5.794453618325253e-06,
+      "loss": 0.0292,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6226078081143149,
+      "grad_norm": 0.8211406469345093,
+      "learning_rate": 5.791201012567498e-06,
+      "loss": 0.0283,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6251594794590457,
+      "grad_norm": 0.6944069862365723,
+      "learning_rate": 5.787923801031164e-06,
+      "loss": 0.0336,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6277111508037765,
+      "grad_norm": 1.1745727062225342,
+      "learning_rate": 5.78462201260645e-06,
+      "loss": 0.0284,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6302628221485073,
+      "grad_norm": 1.3961211442947388,
+      "learning_rate": 5.781295676400203e-06,
+      "loss": 0.0327,
+      "step": 2470
+    },
+    {
+      "epoch": 0.632814493493238,
+      "grad_norm": 1.1352729797363281,
+      "learning_rate": 5.777944821735679e-06,
+      "loss": 0.0302,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6353661648379688,
+      "grad_norm": 1.200098991394043,
+      "learning_rate": 5.774569478152271e-06,
+      "loss": 0.0287,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6379178361826997,
+      "grad_norm": 1.50564706325531,
+      "learning_rate": 5.771169675405256e-06,
+      "loss": 0.0301,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6404695075274305,
+      "grad_norm": 1.6297107934951782,
+      "learning_rate": 5.767745443465529e-06,
+      "loss": 0.0313,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6430211788721613,
+      "grad_norm": 0.7969597578048706,
+      "learning_rate": 5.76429681251934e-06,
+      "loss": 0.0281,
+      "step": 2520
+    },
+    {
+      "epoch": 0.645572850216892,
+      "grad_norm": 0.9594812393188477,
+      "learning_rate": 5.7608238129680294e-06,
+      "loss": 0.0291,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6481245215616228,
+      "grad_norm": 1.628527283668518,
+      "learning_rate": 5.757326475427758e-06,
+      "loss": 0.0288,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6506761929063537,
+      "grad_norm": 1.2033168077468872,
+      "learning_rate": 5.753804830729234e-06,
+      "loss": 0.0271,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6532278642510845,
+      "grad_norm": 0.5983326435089111,
+      "learning_rate": 5.750258909917453e-06,
+      "loss": 0.0244,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6557795355958153,
+      "grad_norm": 0.9838243722915649,
+      "learning_rate": 5.746688744251408e-06,
+      "loss": 0.0251,
+      "step": 2570
+    },
+    {
+      "epoch": 0.658331206940546,
+      "grad_norm": 1.507983922958374,
+      "learning_rate": 5.743094365203827e-06,
+      "loss": 0.0258,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6608828782852768,
+      "grad_norm": 1.403834581375122,
+      "learning_rate": 5.739475804460888e-06,
+      "loss": 0.0254,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6634345496300077,
+      "grad_norm": 1.1948083639144897,
+      "learning_rate": 5.735833093921944e-06,
+      "loss": 0.0291,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6659862209747385,
+      "grad_norm": 3.1764698028564453,
+      "learning_rate": 5.732166265699239e-06,
+      "loss": 0.0333,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6685378923194693,
+      "grad_norm": 0.9992326498031616,
+      "learning_rate": 5.7284753521176264e-06,
+      "loss": 0.0286,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6710895636642,
+      "grad_norm": 0.8567758202552795,
+      "learning_rate": 5.724760385714284e-06,
+      "loss": 0.0266,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6736412350089308,
+      "grad_norm": 1.4158856868743896,
+      "learning_rate": 5.721021399238424e-06,
+      "loss": 0.0277,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6761929063536617,
+      "grad_norm": 0.7602766156196594,
+      "learning_rate": 5.717258425651013e-06,
+      "loss": 0.0316,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6787445776983925,
+      "grad_norm": 1.0612212419509888,
+      "learning_rate": 5.71347149812447e-06,
+      "loss": 0.0265,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6812962490431232,
+      "grad_norm": 2.0236854553222656,
+      "learning_rate": 5.709660650042382e-06,
+      "loss": 0.03,
+      "step": 2670
+    },
+    {
+      "epoch": 0.683847920387854,
+      "grad_norm": 2.617422342300415,
+      "learning_rate": 5.705825914999206e-06,
+      "loss": 0.0318,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6863995917325848,
+      "grad_norm": 0.8844764828681946,
+      "learning_rate": 5.701967326799974e-06,
+      "loss": 0.0244,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6889512630773157,
+      "grad_norm": 0.9346625804901123,
+      "learning_rate": 5.698084919459997e-06,
+      "loss": 0.0313,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6915029344220465,
+      "grad_norm": 1.3249640464782715,
+      "learning_rate": 5.694178727204562e-06,
+      "loss": 0.0247,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6940546057667772,
+      "grad_norm": 1.593597650527954,
+      "learning_rate": 5.69024878446863e-06,
+      "loss": 0.0294,
+      "step": 2720
+    },
+    {
+      "epoch": 0.696606277111508,
+      "grad_norm": 1.0783870220184326,
+      "learning_rate": 5.686295125896537e-06,
+      "loss": 0.0261,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6991579484562388,
+      "grad_norm": 1.3403818607330322,
+      "learning_rate": 5.682317786341682e-06,
+      "loss": 0.0259,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7017096198009697,
+      "grad_norm": 1.2575613260269165,
+      "learning_rate": 5.678316800866226e-06,
+      "loss": 0.0247,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7042612911457005,
+      "grad_norm": 0.5721168518066406,
+      "learning_rate": 5.674292204740779e-06,
+      "loss": 0.024,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7068129624904312,
+      "grad_norm": 0.49415379762649536,
+      "learning_rate": 5.67024403344409e-06,
+      "loss": 0.0265,
+      "step": 2770
+    },
+    {
+      "epoch": 0.709364633835162,
+      "grad_norm": 1.5040923357009888,
+      "learning_rate": 5.666172322662735e-06,
+      "loss": 0.0271,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7119163051798928,
+      "grad_norm": 1.4604111909866333,
+      "learning_rate": 5.662077108290802e-06,
+      "loss": 0.0252,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7144679765246237,
+      "grad_norm": 4.383986473083496,
+      "learning_rate": 5.657958426429573e-06,
+      "loss": 0.029,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7170196478693545,
+      "grad_norm": 1.1418787240982056,
+      "learning_rate": 5.65381631338721e-06,
+      "loss": 0.026,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7195713192140852,
+      "grad_norm": 1.178589105606079,
+      "learning_rate": 5.64965080567843e-06,
+      "loss": 0.0257,
+      "step": 2820
+    },
+    {
+      "epoch": 0.722122990558816,
+      "grad_norm": 2.3111040592193604,
+      "learning_rate": 5.6454619400241845e-06,
+      "loss": 0.0287,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7246746619035468,
+      "grad_norm": 0.6986176371574402,
+      "learning_rate": 5.641249753351339e-06,
+      "loss": 0.0267,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7272263332482776,
+      "grad_norm": 0.7001646757125854,
+      "learning_rate": 5.637014282792342e-06,
+      "loss": 0.0249,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7297780045930085,
+      "grad_norm": 1.2817333936691284,
+      "learning_rate": 5.632755565684905e-06,
+      "loss": 0.0251,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7323296759377392,
+      "grad_norm": 0.9041908383369446,
+      "learning_rate": 5.628473639571665e-06,
+      "loss": 0.0252,
+      "step": 2870
+    },
+    {
+      "epoch": 0.73488134728247,
+      "grad_norm": 2.7807722091674805,
+      "learning_rate": 5.624168542199858e-06,
+      "loss": 0.0255,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7374330186272008,
+      "grad_norm": 1.0662809610366821,
+      "learning_rate": 5.619840311520989e-06,
+      "loss": 0.0332,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7399846899719316,
+      "grad_norm": 1.3166412115097046,
+      "learning_rate": 5.6154889856904894e-06,
+      "loss": 0.0264,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7425363613166625,
+      "grad_norm": 1.0728857517242432,
+      "learning_rate": 5.611114603067388e-06,
+      "loss": 0.0223,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7450880326613932,
+      "grad_norm": 1.4507665634155273,
+      "learning_rate": 5.606717202213974e-06,
+      "loss": 0.0274,
+      "step": 2920
+    },
+    {
+      "epoch": 0.747639704006124,
+      "grad_norm": 0.8330174088478088,
+      "learning_rate": 5.602296821895444e-06,
+      "loss": 0.0256,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7501913753508548,
+      "grad_norm": 0.7136677503585815,
+      "learning_rate": 5.597853501079579e-06,
+      "loss": 0.0265,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7527430466955856,
+      "grad_norm": 1.0832639932632446,
+      "learning_rate": 5.593387278936388e-06,
+      "loss": 0.0266,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7552947180403164,
+      "grad_norm": 0.5012922883033752,
+      "learning_rate": 5.588898194837764e-06,
+      "loss": 0.0271,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7578463893850472,
+      "grad_norm": 1.0677125453948975,
+      "learning_rate": 5.584386288357142e-06,
+      "loss": 0.0282,
+      "step": 2970
+    },
+    {
+      "epoch": 0.760398060729778,
+      "grad_norm": 1.0686490535736084,
+      "learning_rate": 5.579851599269146e-06,
+      "loss": 0.0283,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7629497320745088,
+      "grad_norm": 0.9940263032913208,
+      "learning_rate": 5.575294167549241e-06,
+      "loss": 0.0281,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7655014034192396,
+      "grad_norm": 1.4291343688964844,
+      "learning_rate": 5.5707140333733775e-06,
+      "loss": 0.0267,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7680530747639704,
+      "grad_norm": 0.6890590786933899,
+      "learning_rate": 5.566111237117639e-06,
+      "loss": 0.0231,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7706047461087012,
+      "grad_norm": 0.8809148073196411,
+      "learning_rate": 5.5614858193578896e-06,
+      "loss": 0.0279,
+      "step": 3020
+    },
+    {
+      "epoch": 0.773156417453432,
+      "grad_norm": 1.3962548971176147,
+      "learning_rate": 5.556837820869408e-06,
+      "loss": 0.0266,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7757080887981628,
+      "grad_norm": 0.9583940505981445,
+      "learning_rate": 5.552167282626535e-06,
+      "loss": 0.026,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7782597601428936,
+      "grad_norm": 0.7618580460548401,
+      "learning_rate": 5.54747424580231e-06,
+      "loss": 0.0228,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7808114314876244,
+      "grad_norm": 1.2238036394119263,
+      "learning_rate": 5.5427587517681095e-06,
+      "loss": 0.0296,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7833631028323552,
+      "grad_norm": 0.5979515314102173,
+      "learning_rate": 5.53802084209328e-06,
+      "loss": 0.0238,
+      "step": 3070
+    },
+    {
+      "epoch": 0.785914774177086,
+      "grad_norm": 0.6824269890785217,
+      "learning_rate": 5.533260558544772e-06,
+      "loss": 0.0227,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7884664455218168,
+      "grad_norm": 1.1425477266311646,
+      "learning_rate": 5.528477943086773e-06,
+      "loss": 0.0242,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7910181168665475,
+      "grad_norm": 0.8494576215744019,
+      "learning_rate": 5.523673037880338e-06,
+      "loss": 0.0225,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7935697882112784,
+      "grad_norm": 1.4294337034225464,
+      "learning_rate": 5.518845885283015e-06,
+      "loss": 0.0227,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7961214595560092,
+      "grad_norm": 1.3093550205230713,
+      "learning_rate": 5.513996527848475e-06,
+      "loss": 0.0229,
+      "step": 3120
+    },
+    {
+      "epoch": 0.79867313090074,
+      "grad_norm": 1.951343059539795,
+      "learning_rate": 5.509125008326135e-06,
+      "loss": 0.0277,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8012248022454708,
+      "grad_norm": 1.1987403631210327,
+      "learning_rate": 5.50423136966078e-06,
+      "loss": 0.0244,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8037764735902015,
+      "grad_norm": 0.8594563007354736,
+      "learning_rate": 5.499315654992189e-06,
+      "loss": 0.0262,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8063281449349324,
+      "grad_norm": 1.0187805891036987,
+      "learning_rate": 5.494377907654748e-06,
+      "loss": 0.0265,
+      "step": 3160
+    },
+    {
+      "epoch": 0.8088798162796632,
+      "grad_norm": 0.7313008308410645,
+      "learning_rate": 5.489418171177073e-06,
+      "loss": 0.0225,
+      "step": 3170
+    },
+    {
+      "epoch": 0.811431487624394,
+      "grad_norm": 1.7417691946029663,
+      "learning_rate": 5.484436489281627e-06,
+      "loss": 0.0237,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8139831589691248,
+      "grad_norm": 1.963599443435669,
+      "learning_rate": 5.4794329058843285e-06,
+      "loss": 0.0246,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8165348303138555,
+      "grad_norm": 1.622215986251831,
+      "learning_rate": 5.474407465094171e-06,
+      "loss": 0.025,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8190865016585863,
+      "grad_norm": 2.1439669132232666,
+      "learning_rate": 5.469360211212832e-06,
+      "loss": 0.023,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8216381730033172,
+      "grad_norm": 0.7606381177902222,
+      "learning_rate": 5.464291188734278e-06,
+      "loss": 0.0251,
+      "step": 3220
+    },
+    {
+      "epoch": 0.824189844348048,
+      "grad_norm": 2.439307451248169,
+      "learning_rate": 5.45920044234438e-06,
+      "loss": 0.0238,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8267415156927788,
+      "grad_norm": 1.7854689359664917,
+      "learning_rate": 5.454088016920514e-06,
+      "loss": 0.0262,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8292931870375095,
+      "grad_norm": 1.1369054317474365,
+      "learning_rate": 5.448953957531168e-06,
+      "loss": 0.0241,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8318448583822403,
+      "grad_norm": 1.2106547355651855,
+      "learning_rate": 5.4437983094355435e-06,
+      "loss": 0.0253,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8343965297269712,
+      "grad_norm": 0.7193657755851746,
+      "learning_rate": 5.438621118083155e-06,
+      "loss": 0.022,
+      "step": 3270
+    },
+    {
+      "epoch": 0.836948201071702,
+      "grad_norm": 1.0049270391464233,
+      "learning_rate": 5.433422429113436e-06,
+      "loss": 0.0255,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8394998724164328,
+      "grad_norm": 1.0903527736663818,
+      "learning_rate": 5.428202288355326e-06,
+      "loss": 0.0226,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8420515437611635,
+      "grad_norm": 1.3176894187927246,
+      "learning_rate": 5.422960741826874e-06,
+      "loss": 0.0242,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8446032151058943,
+      "grad_norm": 0.8586781024932861,
+      "learning_rate": 5.4176978357348355e-06,
+      "loss": 0.0219,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8471548864506252,
+      "grad_norm": 0.8540439605712891,
+      "learning_rate": 5.412413616474256e-06,
+      "loss": 0.0236,
+      "step": 3320
+    },
+    {
+      "epoch": 0.849706557795356,
+      "grad_norm": 0.4369836449623108,
+      "learning_rate": 5.407108130628066e-06,
+      "loss": 0.0201,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8522582291400868,
+      "grad_norm": 1.570779800415039,
+      "learning_rate": 5.401781424966676e-06,
+      "loss": 0.0247,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8548099004848175,
+      "grad_norm": 2.3272910118103027,
+      "learning_rate": 5.396433546447557e-06,
+      "loss": 0.0221,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8573615718295483,
+      "grad_norm": 1.004865288734436,
+      "learning_rate": 5.391064542214828e-06,
+      "loss": 0.0222,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8599132431742792,
+      "grad_norm": 0.9049104452133179,
+      "learning_rate": 5.385674459598841e-06,
+      "loss": 0.0263,
+      "step": 3370
+    },
+    {
+      "epoch": 0.86246491451901,
+      "grad_norm": 1.0357730388641357,
+      "learning_rate": 5.380263346115766e-06,
+      "loss": 0.0224,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8650165858637408,
+      "grad_norm": 2.943132162094116,
+      "learning_rate": 5.374831249467171e-06,
+      "loss": 0.0213,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8675682572084715,
+      "grad_norm": 0.7794683575630188,
+      "learning_rate": 5.369378217539595e-06,
+      "loss": 0.0261,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8701199285532023,
+      "grad_norm": 1.315697431564331,
+      "learning_rate": 5.36390429840414e-06,
+      "loss": 0.0243,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8726715998979332,
+      "grad_norm": 0.48824071884155273,
+      "learning_rate": 5.358409540316032e-06,
+      "loss": 0.0201,
+      "step": 3420
+    },
+    {
+      "epoch": 0.875223271242664,
+      "grad_norm": 1.1559224128723145,
+      "learning_rate": 5.3528939917142046e-06,
+      "loss": 0.0203,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8777749425873947,
+      "grad_norm": 1.6702337265014648,
+      "learning_rate": 5.34735770122087e-06,
+      "loss": 0.0225,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8803266139321255,
+      "grad_norm": 0.8018677234649658,
+      "learning_rate": 5.34180071764109e-06,
+      "loss": 0.0243,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8828782852768563,
+      "grad_norm": 1.211368203163147,
+      "learning_rate": 5.336223089962347e-06,
+      "loss": 0.0238,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8854299566215872,
+      "grad_norm": 0.7445624470710754,
+      "learning_rate": 5.33062486735411e-06,
+      "loss": 0.0221,
+      "step": 3470
+    },
+    {
+      "epoch": 0.887981627966318,
+      "grad_norm": 0.8386270403862,
+      "learning_rate": 5.3250060991674e-06,
+      "loss": 0.0201,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8905332993110487,
+      "grad_norm": 0.840062141418457,
+      "learning_rate": 5.319366834934361e-06,
+      "loss": 0.0227,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8930849706557795,
+      "grad_norm": 0.7337429523468018,
+      "learning_rate": 5.313707124367819e-06,
+      "loss": 0.0223,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8956366420005103,
+      "grad_norm": 1.0676528215408325,
+      "learning_rate": 5.30802701736084e-06,
+      "loss": 0.0231,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8981883133452412,
+      "grad_norm": 0.512732982635498,
+      "learning_rate": 5.3023265639863e-06,
+      "loss": 0.0238,
+      "step": 3520
+    },
+    {
+      "epoch": 0.900739984689972,
+      "grad_norm": 1.079931378364563,
+      "learning_rate": 5.296605814496435e-06,
+      "loss": 0.0217,
+      "step": 3530
+    },
+    {
+      "epoch": 0.9032916560347027,
+      "grad_norm": 1.2752869129180908,
+      "learning_rate": 5.290864819322402e-06,
+      "loss": 0.0214,
+      "step": 3540
+    },
+    {
+      "epoch": 0.9058433273794335,
+      "grad_norm": 1.5410728454589844,
+      "learning_rate": 5.2851036290738316e-06,
+      "loss": 0.0294,
+      "step": 3550
+    },
+    {
+      "epoch": 0.9083949987241643,
+      "grad_norm": 0.7965479493141174,
+      "learning_rate": 5.279322294538386e-06,
+      "loss": 0.0227,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9109466700688952,
+      "grad_norm": 0.8509993553161621,
+      "learning_rate": 5.273520866681309e-06,
+      "loss": 0.0214,
+      "step": 3570
+    },
+    {
+      "epoch": 0.913498341413626,
+      "grad_norm": 1.2866240739822388,
+      "learning_rate": 5.2676993966449736e-06,
+      "loss": 0.024,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9160500127583567,
+      "grad_norm": 1.3114237785339355,
+      "learning_rate": 5.261857935748437e-06,
+      "loss": 0.018,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9186016841030875,
+      "grad_norm": 1.0281429290771484,
+      "learning_rate": 5.255996535486987e-06,
+      "loss": 0.0217,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9211533554478183,
+      "grad_norm": 1.3615024089813232,
+      "learning_rate": 5.250115247531681e-06,
+      "loss": 0.0247,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9237050267925491,
+      "grad_norm": 0.7838472127914429,
+      "learning_rate": 5.244214123728902e-06,
+      "loss": 0.0238,
+      "step": 3620
+    },
+    {
+      "epoch": 0.92625669813728,
+      "grad_norm": 0.8573798537254333,
+      "learning_rate": 5.238293216099891e-06,
+      "loss": 0.0222,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9288083694820107,
+      "grad_norm": 1.4891736507415771,
+      "learning_rate": 5.2323525768402965e-06,
+      "loss": 0.0236,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9313600408267415,
+      "grad_norm": 0.6912400722503662,
+      "learning_rate": 5.226392258319708e-06,
+      "loss": 0.0208,
+      "step": 3650
+    },
+    {
+      "epoch": 0.9339117121714723,
+      "grad_norm": 0.5795527696609497,
+      "learning_rate": 5.220412313081197e-06,
+      "loss": 0.0203,
+      "step": 3660
+    },
+    {
+      "epoch": 0.9364633835162031,
+      "grad_norm": 1.261696457862854,
+      "learning_rate": 5.214412793840857e-06,
+      "loss": 0.0214,
+      "step": 3670
+    },
+    {
+      "epoch": 0.939015054860934,
+      "grad_norm": 0.7527537941932678,
+      "learning_rate": 5.208393753487333e-06,
+      "loss": 0.0224,
+      "step": 3680
+    },
+    {
+      "epoch": 0.9415667262056647,
+      "grad_norm": 0.8399683237075806,
+      "learning_rate": 5.20235524508136e-06,
+      "loss": 0.0243,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9441183975503955,
+      "grad_norm": 0.7504599094390869,
+      "learning_rate": 5.196297321855291e-06,
+      "loss": 0.0203,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9466700688951263,
+      "grad_norm": 0.9682078957557678,
+      "learning_rate": 5.190220037212632e-06,
+      "loss": 0.023,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9492217402398571,
+      "grad_norm": 0.9620463252067566,
+      "learning_rate": 5.184123444727567e-06,
+      "loss": 0.0217,
+      "step": 3720
+    },
+    {
+      "epoch": 0.951773411584588,
+      "grad_norm": 1.114750623703003,
+      "learning_rate": 5.178007598144491e-06,
+      "loss": 0.0216,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9543250829293187,
+      "grad_norm": 0.9930930137634277,
+      "learning_rate": 5.17187255137753e-06,
+      "loss": 0.0179,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9568767542740495,
+      "grad_norm": 0.584056556224823,
+      "learning_rate": 5.165718358510071e-06,
+      "loss": 0.0208,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9594284256187803,
+      "grad_norm": 0.9347606897354126,
+      "learning_rate": 5.159545073794281e-06,
+      "loss": 0.0195,
+      "step": 3760
+    },
+    {
+      "epoch": 0.961980096963511,
+      "grad_norm": 0.8671495318412781,
+      "learning_rate": 5.153352751650634e-06,
+      "loss": 0.0225,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9645317683082419,
+      "grad_norm": 0.8564977049827576,
+      "learning_rate": 5.147141446667424e-06,
+      "loss": 0.0227,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9670834396529727,
+      "grad_norm": 0.44186270236968994,
+      "learning_rate": 5.1409112136002915e-06,
+      "loss": 0.0166,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9696351109977035,
+      "grad_norm": 0.7479332685470581,
+      "learning_rate": 5.134662107371736e-06,
+      "loss": 0.0225,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9721867823424343,
+      "grad_norm": 0.8453100919723511,
+      "learning_rate": 5.128394183070631e-06,
+      "loss": 0.0175,
+      "step": 3810
+    },
+    {
+      "epoch": 0.974738453687165,
+      "grad_norm": 0.7815666198730469,
+      "learning_rate": 5.122107495951743e-06,
+      "loss": 0.022,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9772901250318959,
+      "grad_norm": 1.070888876914978,
+      "learning_rate": 5.115802101435242e-06,
+      "loss": 0.0226,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9798417963766267,
+      "grad_norm": 0.7922118902206421,
+      "learning_rate": 5.109478055106209e-06,
+      "loss": 0.0215,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9823934677213575,
+      "grad_norm": 0.7540226578712463,
+      "learning_rate": 5.103135412714155e-06,
+      "loss": 0.0188,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9849451390660883,
+      "grad_norm": 0.8961939811706543,
+      "learning_rate": 5.0967742301725186e-06,
+      "loss": 0.0208,
+      "step": 3860
+    },
+    {
+      "epoch": 0.987496810410819,
+      "grad_norm": 0.642845094203949,
+      "learning_rate": 5.090394563558184e-06,
+      "loss": 0.0242,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9900484817555499,
+      "grad_norm": 0.5724070072174072,
+      "learning_rate": 5.083996469110977e-06,
+      "loss": 0.0182,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9926001531002807,
+      "grad_norm": 0.4112820327281952,
+      "learning_rate": 5.077580003233176e-06,
+      "loss": 0.018,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9951518244450115,
+      "grad_norm": 1.535103440284729,
+      "learning_rate": 5.071145222489013e-06,
+      "loss": 0.0219,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9977034957897423,
+      "grad_norm": 1.7755980491638184,
+      "learning_rate": 5.064692183604172e-06,
+      "loss": 0.0243,
+      "step": 3910
+    },
+    {
+      "epoch": 1.000255167134473,
+      "grad_norm": 0.8194239735603333,
+      "learning_rate": 5.058220943465293e-06,
+      "loss": 0.0218,
+      "step": 3920
+    },
+    {
+      "epoch": 1.002806838479204,
+      "grad_norm": 1.6030933856964111,
+      "learning_rate": 5.051731559119469e-06,
+      "loss": 0.0244,
+      "step": 3930
+    },
+    {
+      "epoch": 1.0053585098239346,
+      "grad_norm": 0.6478273272514343,
+      "learning_rate": 5.045224087773742e-06,
+      "loss": 0.0199,
+      "step": 3940
+    },
+    {
+      "epoch": 1.0079101811686655,
+      "grad_norm": 0.4480527341365814,
+      "learning_rate": 5.038698586794598e-06,
+      "loss": 0.0213,
+      "step": 3950
+    },
+    {
+      "epoch": 1.0104618525133964,
+      "grad_norm": 1.267435908317566,
+      "learning_rate": 5.032155113707468e-06,
+      "loss": 0.0196,
+      "step": 3960
+    },
+    {
+      "epoch": 1.013013523858127,
+      "grad_norm": 0.7024682760238647,
+      "learning_rate": 5.02559372619621e-06,
+      "loss": 0.0198,
+      "step": 3970
+    },
+    {
+      "epoch": 1.015565195202858,
+      "grad_norm": 0.4587201774120331,
+      "learning_rate": 5.019014482102611e-06,
+      "loss": 0.0171,
+      "step": 3980
+    },
+    {
+      "epoch": 1.0181168665475886,
+      "grad_norm": 0.8392826318740845,
+      "learning_rate": 5.012417439425867e-06,
+      "loss": 0.0185,
+      "step": 3990
+    },
+    {
+      "epoch": 1.0206685378923195,
+      "grad_norm": 0.8017153739929199,
+      "learning_rate": 5.0058026563220825e-06,
+      "loss": 0.0189,
+      "step": 4000
+    },
+    {
+      "epoch": 1.0232202092370504,
+      "grad_norm": 0.6147284507751465,
+      "learning_rate": 4.99917019110375e-06,
+      "loss": 0.0185,
+      "step": 4010
+    },
+    {
+      "epoch": 1.025771880581781,
+      "grad_norm": 0.6263673305511475,
+      "learning_rate": 4.992520102239238e-06,
+      "loss": 0.0177,
+      "step": 4020
+    },
+    {
+      "epoch": 1.028323551926512,
+      "grad_norm": 0.7895020842552185,
+      "learning_rate": 4.985852448352276e-06,
+      "loss": 0.0213,
+      "step": 4030
+    },
+    {
+      "epoch": 1.0308752232712426,
+      "grad_norm": 0.9371950030326843,
+      "learning_rate": 4.9791672882214394e-06,
+      "loss": 0.0217,
+      "step": 4040
+    },
+    {
+      "epoch": 1.0334268946159735,
+      "grad_norm": 2.927732467651367,
+      "learning_rate": 4.972464680779627e-06,
+      "loss": 0.0219,
+      "step": 4050
+    },
+    {
+      "epoch": 1.0359785659607044,
+      "grad_norm": 0.7745063304901123,
+      "learning_rate": 4.965744685113544e-06,
+      "loss": 0.0189,
+      "step": 4060
+    },
+    {
+      "epoch": 1.038530237305435,
+      "grad_norm": 0.6315968036651611,
+      "learning_rate": 4.959007360463185e-06,
+      "loss": 0.0187,
+      "step": 4070
+    },
+    {
+      "epoch": 1.041081908650166,
+      "grad_norm": 0.3792322874069214,
+      "learning_rate": 4.952252766221303e-06,
+      "loss": 0.0214,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0436335799948966,
+      "grad_norm": 0.6313232779502869,
+      "learning_rate": 4.9454809619328935e-06,
+      "loss": 0.019,
+      "step": 4090
+    },
+    {
+      "epoch": 1.0461852513396275,
+      "grad_norm": 0.9274890422821045,
+      "learning_rate": 4.938692007294667e-06,
+      "loss": 0.0195,
+      "step": 4100
+    },
+    {
+      "epoch": 1.0487369226843581,
+      "grad_norm": 0.6988605856895447,
+      "learning_rate": 4.9318859621545215e-06,
+      "loss": 0.0175,
+      "step": 4110
+    },
+    {
+      "epoch": 1.051288594029089,
+      "grad_norm": 0.5787187814712524,
+      "learning_rate": 4.925062886511017e-06,
+      "loss": 0.0213,
+      "step": 4120
+    },
+    {
+      "epoch": 1.05384026537382,
+      "grad_norm": 0.7282422184944153,
+      "learning_rate": 4.918222840512843e-06,
+      "loss": 0.0206,
+      "step": 4130
+    },
+    {
+      "epoch": 1.0563919367185506,
+      "grad_norm": 0.8163778781890869,
+      "learning_rate": 4.911365884458295e-06,
+      "loss": 0.0193,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0589436080632815,
+      "grad_norm": 0.7783817052841187,
+      "learning_rate": 4.904492078794736e-06,
+      "loss": 0.0172,
+      "step": 4150
+    },
+    {
+      "epoch": 1.0614952794080121,
+      "grad_norm": 2.5214037895202637,
+      "learning_rate": 4.8976014841180664e-06,
+      "loss": 0.0207,
+      "step": 4160
+    },
+    {
+      "epoch": 1.064046950752743,
+      "grad_norm": 1.9668519496917725,
+      "learning_rate": 4.890694161172191e-06,
+      "loss": 0.0179,
+      "step": 4170
+    },
+    {
+      "epoch": 1.066598622097474,
+      "grad_norm": 0.6166751980781555,
+      "learning_rate": 4.883770170848479e-06,
+      "loss": 0.0187,
+      "step": 4180
+    },
+    {
+      "epoch": 1.0691502934422046,
+      "grad_norm": 0.41448432207107544,
+      "learning_rate": 4.876829574185239e-06,
+      "loss": 0.0188,
+      "step": 4190
+    },
+    {
+      "epoch": 1.0717019647869355,
+      "grad_norm": 1.5775961875915527,
+      "learning_rate": 4.869872432367162e-06,
+      "loss": 0.0199,
+      "step": 4200
+    },
+    {
+      "epoch": 1.0742536361316661,
+      "grad_norm": 0.5199787616729736,
+      "learning_rate": 4.862898806724798e-06,
+      "loss": 0.0169,
+      "step": 4210
+    },
+    {
+      "epoch": 1.076805307476397,
+      "grad_norm": 0.7429872155189514,
+      "learning_rate": 4.85590875873401e-06,
+      "loss": 0.0194,
+      "step": 4220
+    },
+    {
+      "epoch": 1.079356978821128,
+      "grad_norm": 0.6427246928215027,
+      "learning_rate": 4.848902350015431e-06,
+      "loss": 0.0182,
+      "step": 4230
+    },
+    {
+      "epoch": 1.0819086501658586,
+      "grad_norm": 0.4715655446052551,
+      "learning_rate": 4.8418796423339205e-06,
+      "loss": 0.0171,
+      "step": 4240
+    },
+    {
+      "epoch": 1.0844603215105895,
+      "grad_norm": 1.0023016929626465,
+      "learning_rate": 4.834840697598024e-06,
+      "loss": 0.0184,
+      "step": 4250
+    },
+    {
+      "epoch": 1.0870119928553201,
+      "grad_norm": 0.6949712038040161,
+      "learning_rate": 4.827785577859422e-06,
+      "loss": 0.0178,
+      "step": 4260
+    },
+    {
+      "epoch": 1.089563664200051,
+      "grad_norm": 1.3930619955062866,
+      "learning_rate": 4.820714345312386e-06,
+      "loss": 0.0191,
+      "step": 4270
+    },
+    {
+      "epoch": 1.092115335544782,
+      "grad_norm": 0.597090482711792,
+      "learning_rate": 4.813627062293229e-06,
+      "loss": 0.0197,
+      "step": 4280
+    },
+    {
+      "epoch": 1.0946670068895126,
+      "grad_norm": 0.7542505264282227,
+      "learning_rate": 4.80652379127976e-06,
+      "loss": 0.0194,
+      "step": 4290
+    },
+    {
+      "epoch": 1.0972186782342435,
+      "grad_norm": 0.7412875890731812,
+      "learning_rate": 4.799404594890724e-06,
+      "loss": 0.0177,
+      "step": 4300
+    },
+    {
+      "epoch": 1.0997703495789741,
+      "grad_norm": 0.8048778772354126,
+      "learning_rate": 4.792269535885262e-06,
+      "loss": 0.0203,
+      "step": 4310
+    },
+    {
+      "epoch": 1.102322020923705,
+      "grad_norm": 0.758451521396637,
+      "learning_rate": 4.7851186771623474e-06,
+      "loss": 0.0198,
+      "step": 4320
+    },
+    {
+      "epoch": 1.104873692268436,
+      "grad_norm": 0.5495659708976746,
+      "learning_rate": 4.777952081760237e-06,
+      "loss": 0.0192,
+      "step": 4330
+    },
+    {
+      "epoch": 1.1074253636131666,
+      "grad_norm": 0.3236237168312073,
+      "learning_rate": 4.770769812855914e-06,
+      "loss": 0.0183,
+      "step": 4340
+    },
+    {
+      "epoch": 1.1099770349578975,
+      "grad_norm": 0.42418259382247925,
+      "learning_rate": 4.763571933764533e-06,
+      "loss": 0.0184,
+      "step": 4350
+    },
+    {
+      "epoch": 1.1125287063026281,
+      "grad_norm": 1.0733011960983276,
+      "learning_rate": 4.756358507938854e-06,
+      "loss": 0.0181,
+      "step": 4360
+    },
+    {
+      "epoch": 1.115080377647359,
+      "grad_norm": 3.533398151397705,
+      "learning_rate": 4.749129598968697e-06,
+      "loss": 0.0168,
+      "step": 4370
+    },
+    {
+      "epoch": 1.11763204899209,
+      "grad_norm": 1.1660666465759277,
+      "learning_rate": 4.741885270580367e-06,
+      "loss": 0.0173,
+      "step": 4380
+    },
+    {
+      "epoch": 1.1201837203368206,
+      "grad_norm": 1.2423667907714844,
+      "learning_rate": 4.734625586636103e-06,
+      "loss": 0.0147,
+      "step": 4390
+    },
+    {
+      "epoch": 1.1227353916815515,
+      "grad_norm": 1.0858911275863647,
+      "learning_rate": 4.727350611133506e-06,
+      "loss": 0.0182,
+      "step": 4400
+    },
+    {
+      "epoch": 1.1252870630262821,
+      "grad_norm": 0.932620644569397,
+      "learning_rate": 4.720060408204983e-06,
+      "loss": 0.0227,
+      "step": 4410
+    },
+    {
+      "epoch": 1.127838734371013,
+      "grad_norm": 1.4234601259231567,
+      "learning_rate": 4.712755042117177e-06,
+      "loss": 0.0195,
+      "step": 4420
+    },
+    {
+      "epoch": 1.130390405715744,
+      "grad_norm": 0.46929091215133667,
+      "learning_rate": 4.705434577270402e-06,
+      "loss": 0.0177,
+      "step": 4430
+    },
+    {
+      "epoch": 1.1329420770604746,
+      "grad_norm": 1.3635838031768799,
+      "learning_rate": 4.698099078198074e-06,
+      "loss": 0.0227,
+      "step": 4440
+    },
+    {
+      "epoch": 1.1354937484052054,
+      "grad_norm": 0.6974967122077942,
+      "learning_rate": 4.690748609566143e-06,
+      "loss": 0.0172,
+      "step": 4450
+    },
+    {
+      "epoch": 1.1380454197499361,
+      "grad_norm": 0.6230807900428772,
+      "learning_rate": 4.683383236172523e-06,
+      "loss": 0.018,
+      "step": 4460
+    },
+    {
+      "epoch": 1.140597091094667,
+      "grad_norm": 0.4532192647457123,
+      "learning_rate": 4.676003022946522e-06,
+      "loss": 0.0157,
+      "step": 4470
+    },
+    {
+      "epoch": 1.143148762439398,
+      "grad_norm": 3.6540305614471436,
+      "learning_rate": 4.668608034948266e-06,
+      "loss": 0.0176,
+      "step": 4480
+    },
+    {
+      "epoch": 1.1457004337841286,
+      "grad_norm": 0.5863398313522339,
+      "learning_rate": 4.661198337368129e-06,
+      "loss": 0.0179,
+      "step": 4490
+    },
+    {
+      "epoch": 1.1482521051288594,
+      "grad_norm": 1.5659875869750977,
+      "learning_rate": 4.6537739955261566e-06,
+      "loss": 0.0199,
+      "step": 4500
+    },
+    {
+      "epoch": 1.1508037764735901,
+      "grad_norm": 1.0172702074050903,
+      "learning_rate": 4.646335074871491e-06,
+      "loss": 0.0184,
+      "step": 4510
+    },
+    {
+      "epoch": 1.153355447818321,
+      "grad_norm": 0.26044878363609314,
+      "learning_rate": 4.638881640981793e-06,
+      "loss": 0.016,
+      "step": 4520
+    },
+    {
+      "epoch": 1.1559071191630519,
+      "grad_norm": 1.071213722229004,
+      "learning_rate": 4.631413759562666e-06,
+      "loss": 0.0177,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1584587905077826,
+      "grad_norm": 1.1978178024291992,
+      "learning_rate": 4.623931496447074e-06,
+      "loss": 0.0188,
+      "step": 4540
+    },
+    {
+      "epoch": 1.1610104618525134,
+      "grad_norm": 2.745671272277832,
+      "learning_rate": 4.616434917594762e-06,
+      "loss": 0.0233,
+      "step": 4550
+    },
+    {
+      "epoch": 1.163562133197244,
+      "grad_norm": 0.8523492217063904,
+      "learning_rate": 4.6089240890916766e-06,
+      "loss": 0.0187,
+      "step": 4560
+    },
+    {
+      "epoch": 1.166113804541975,
+      "grad_norm": 1.745078444480896,
+      "learning_rate": 4.6013990771493815e-06,
+      "loss": 0.0223,
+      "step": 4570
+    },
+    {
+      "epoch": 1.1686654758867059,
+      "grad_norm": 0.6648458242416382,
+      "learning_rate": 4.593859948104475e-06,
+      "loss": 0.0185,
+      "step": 4580
+    },
+    {
+      "epoch": 1.1712171472314366,
+      "grad_norm": 0.840826690196991,
+      "learning_rate": 4.586306768418001e-06,
+      "loss": 0.019,
+      "step": 4590
+    },
+    {
+      "epoch": 1.1737688185761674,
+      "grad_norm": 1.2461105585098267,
+      "learning_rate": 4.57873960467487e-06,
+      "loss": 0.0175,
+      "step": 4600
+    },
+    {
+      "epoch": 1.176320489920898,
+      "grad_norm": 0.7605074048042297,
+      "learning_rate": 4.571158523583269e-06,
+      "loss": 0.0153,
+      "step": 4610
+    },
+    {
+      "epoch": 1.178872161265629,
+      "grad_norm": 0.8275895714759827,
+      "learning_rate": 4.563563591974071e-06,
+      "loss": 0.0175,
+      "step": 4620
+    },
+    {
+      "epoch": 1.1814238326103599,
+      "grad_norm": 0.8980046510696411,
+      "learning_rate": 4.555954876800248e-06,
+      "loss": 0.02,
+      "step": 4630
+    },
+    {
+      "epoch": 1.1839755039550905,
+      "grad_norm": 0.9885419011116028,
+      "learning_rate": 4.548332445136283e-06,
+      "loss": 0.017,
+      "step": 4640
+    },
+    {
+      "epoch": 1.1865271752998214,
+      "grad_norm": 0.5855565667152405,
+      "learning_rate": 4.540696364177575e-06,
+      "loss": 0.0174,
+      "step": 4650
+    },
+    {
+      "epoch": 1.189078846644552,
+      "grad_norm": 0.6320938467979431,
+      "learning_rate": 4.533046701239845e-06,
+      "loss": 0.0179,
+      "step": 4660
+    },
+    {
+      "epoch": 1.191630517989283,
+      "grad_norm": 0.8237224817276001,
+      "learning_rate": 4.525383523758553e-06,
+      "loss": 0.0188,
+      "step": 4670
+    },
+    {
+      "epoch": 1.1941821893340139,
+      "grad_norm": 0.49340859055519104,
+      "learning_rate": 4.51770689928829e-06,
+      "loss": 0.0159,
+      "step": 4680
+    },
+    {
+      "epoch": 1.1967338606787445,
+      "grad_norm": 0.900672972202301,
+      "learning_rate": 4.510016895502191e-06,
+      "loss": 0.0167,
+      "step": 4690
+    },
+    {
+      "epoch": 1.1992855320234754,
+      "grad_norm": 1.2280319929122925,
+      "learning_rate": 4.502313580191334e-06,
+      "loss": 0.0172,
+      "step": 4700
+    },
+    {
+      "epoch": 1.201837203368206,
+      "grad_norm": 1.2630693912506104,
+      "learning_rate": 4.494597021264149e-06,
+      "loss": 0.0183,
+      "step": 4710
+    },
+    {
+      "epoch": 1.204388874712937,
+      "grad_norm": 0.5751070380210876,
+      "learning_rate": 4.48686728674581e-06,
+      "loss": 0.0172,
+      "step": 4720
+    },
+    {
+      "epoch": 1.2069405460576679,
+      "grad_norm": 0.8484153151512146,
+      "learning_rate": 4.479124444777644e-06,
+      "loss": 0.0183,
+      "step": 4730
+    },
+    {
+      "epoch": 1.2094922174023985,
+      "grad_norm": 0.9844270944595337,
+      "learning_rate": 4.4713685636165244e-06,
+      "loss": 0.0203,
+      "step": 4740
+    },
+    {
+      "epoch": 1.2120438887471294,
+      "grad_norm": 0.38548630475997925,
+      "learning_rate": 4.463599711634269e-06,
+      "loss": 0.019,
+      "step": 4750
+    },
+    {
+      "epoch": 1.21459556009186,
+      "grad_norm": 0.5281028747558594,
+      "learning_rate": 4.455817957317046e-06,
+      "loss": 0.0176,
+      "step": 4760
+    },
+    {
+      "epoch": 1.217147231436591,
+      "grad_norm": 0.24428775906562805,
+      "learning_rate": 4.448023369264758e-06,
+      "loss": 0.0187,
+      "step": 4770
+    },
+    {
+      "epoch": 1.2196989027813219,
+      "grad_norm": 0.9260575771331787,
+      "learning_rate": 4.440216016190445e-06,
+      "loss": 0.0184,
+      "step": 4780
+    },
+    {
+      "epoch": 1.2222505741260525,
+      "grad_norm": 0.6549846529960632,
+      "learning_rate": 4.432395966919679e-06,
+      "loss": 0.0152,
+      "step": 4790
+    },
+    {
+      "epoch": 1.2248022454707834,
+      "grad_norm": 1.136384129524231,
+      "learning_rate": 4.424563290389952e-06,
+      "loss": 0.015,
+      "step": 4800
+    },
+    {
+      "epoch": 1.227353916815514,
+      "grad_norm": 1.2056633234024048,
+      "learning_rate": 4.416718055650073e-06,
+      "loss": 0.0161,
+      "step": 4810
+    },
+    {
+      "epoch": 1.229905588160245,
+      "grad_norm": 0.66669762134552,
+      "learning_rate": 4.408860331859556e-06,
+      "loss": 0.0202,
+      "step": 4820
+    },
+    {
+      "epoch": 1.2324572595049759,
+      "grad_norm": 0.6749626398086548,
+      "learning_rate": 4.400990188288015e-06,
+      "loss": 0.0207,
+      "step": 4830
+    },
+    {
+      "epoch": 1.2350089308497065,
+      "grad_norm": 2.977775812149048,
+      "learning_rate": 4.393107694314547e-06,
+      "loss": 0.0158,
+      "step": 4840
+    },
+    {
+      "epoch": 1.2375606021944374,
+      "grad_norm": 0.4771825671195984,
+      "learning_rate": 4.3852129194271264e-06,
+      "loss": 0.0181,
+      "step": 4850
+    },
+    {
+      "epoch": 1.240112273539168,
+      "grad_norm": 0.4319264888763428,
+      "learning_rate": 4.377305933221988e-06,
+      "loss": 0.0166,
+      "step": 4860
+    },
+    {
+      "epoch": 1.242663944883899,
+      "grad_norm": 0.7092644572257996,
+      "learning_rate": 4.369386805403016e-06,
+      "loss": 0.0162,
+      "step": 4870
+    },
+    {
+      "epoch": 1.2452156162286299,
+      "grad_norm": 1.278566598892212,
+      "learning_rate": 4.361455605781128e-06,
+      "loss": 0.0169,
+      "step": 4880
+    },
+    {
+      "epoch": 1.2477672875733605,
+      "grad_norm": 0.9039450287818909,
+      "learning_rate": 4.353512404273663e-06,
+      "loss": 0.0193,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2503189589180914,
+      "grad_norm": 0.6819228529930115,
+      "learning_rate": 4.3455572709037585e-06,
+      "loss": 0.0171,
+      "step": 4900
+    },
+    {
+      "epoch": 1.252870630262822,
+      "grad_norm": 0.428912878036499,
+      "learning_rate": 4.33759027579974e-06,
+      "loss": 0.0174,
+      "step": 4910
+    },
+    {
+      "epoch": 1.255422301607553,
+      "grad_norm": 1.0547304153442383,
+      "learning_rate": 4.329611489194497e-06,
+      "loss": 0.0142,
+      "step": 4920
+    },
+    {
+      "epoch": 1.2579739729522839,
+      "grad_norm": 0.7328991293907166,
+      "learning_rate": 4.321620981424871e-06,
+      "loss": 0.0165,
+      "step": 4930
+    },
+    {
+      "epoch": 1.2605256442970145,
+      "grad_norm": 1.1251776218414307,
+      "learning_rate": 4.313618822931027e-06,
+      "loss": 0.0155,
+      "step": 4940
+    },
+    {
+      "epoch": 1.2630773156417454,
+      "grad_norm": 0.44651368260383606,
+      "learning_rate": 4.30560508425584e-06,
+      "loss": 0.0161,
+      "step": 4950
+    },
+    {
+      "epoch": 1.265628986986476,
+      "grad_norm": 0.5329874753952026,
+      "learning_rate": 4.297579836044266e-06,
+      "loss": 0.0174,
+      "step": 4960
+    },
+    {
+      "epoch": 1.268180658331207,
+      "grad_norm": 0.8915277123451233,
+      "learning_rate": 4.289543149042729e-06,
+      "loss": 0.0179,
+      "step": 4970
+    },
+    {
+      "epoch": 1.2707323296759379,
+      "grad_norm": 1.3271677494049072,
+      "learning_rate": 4.281495094098486e-06,
+      "loss": 0.0178,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2732840010206685,
+      "grad_norm": 0.7584212422370911,
+      "learning_rate": 4.27343574215901e-06,
+      "loss": 0.0185,
+      "step": 4990
+    },
+    {
+      "epoch": 1.2758356723653994,
+      "grad_norm": 1.139729380607605,
+      "learning_rate": 4.265365164271361e-06,
+      "loss": 0.0187,
+      "step": 5000
+    },
+    {
+      "epoch": 1.27838734371013,
+      "grad_norm": 0.5593457221984863,
+      "learning_rate": 4.257283431581564e-06,
+      "loss": 0.0175,
+      "step": 5010
+    },
+    {
+      "epoch": 1.280939015054861,
+      "grad_norm": 0.9672866463661194,
+      "learning_rate": 4.249190615333977e-06,
+      "loss": 0.0173,
+      "step": 5020
+    },
+    {
+      "epoch": 1.2834906863995919,
+      "grad_norm": 0.8238097429275513,
+      "learning_rate": 4.241086786870664e-06,
+      "loss": 0.0183,
+      "step": 5030
+    },
+    {
+      "epoch": 1.2860423577443225,
+      "grad_norm": 0.30412089824676514,
+      "learning_rate": 4.232972017630771e-06,
+      "loss": 0.0164,
+      "step": 5040
+    },
+    {
+      "epoch": 1.2885940290890534,
+      "grad_norm": 1.4792495965957642,
+      "learning_rate": 4.224846379149886e-06,
+      "loss": 0.0163,
+      "step": 5050
+    },
+    {
+      "epoch": 1.291145700433784,
+      "grad_norm": 0.5357837677001953,
+      "learning_rate": 4.216709943059421e-06,
+      "loss": 0.0145,
+      "step": 5060
+    },
+    {
+      "epoch": 1.293697371778515,
+      "grad_norm": 0.684822142124176,
+      "learning_rate": 4.208562781085969e-06,
+      "loss": 0.0191,
+      "step": 5070
+    },
+    {
+      "epoch": 1.2962490431232458,
+      "grad_norm": 0.3603987693786621,
+      "learning_rate": 4.200404965050679e-06,
+      "loss": 0.0176,
+      "step": 5080
+    },
+    {
+      "epoch": 1.2988007144679765,
+      "grad_norm": 0.723609447479248,
+      "learning_rate": 4.192236566868622e-06,
+      "loss": 0.0198,
+      "step": 5090
+    },
+    {
+      "epoch": 1.3013523858127074,
+      "grad_norm": 0.43119895458221436,
+      "learning_rate": 4.184057658548154e-06,
+      "loss": 0.0155,
+      "step": 5100
+    },
+    {
+      "epoch": 1.303904057157438,
+      "grad_norm": 0.3890657126903534,
+      "learning_rate": 4.175868312190281e-06,
+      "loss": 0.014,
+      "step": 5110
+    },
+    {
+      "epoch": 1.306455728502169,
+      "grad_norm": 4.158993721008301,
+      "learning_rate": 4.167668599988031e-06,
+      "loss": 0.0182,
+      "step": 5120
+    },
+    {
+      "epoch": 1.3090073998468998,
+      "grad_norm": 0.766471266746521,
+      "learning_rate": 4.159458594225806e-06,
+      "loss": 0.0157,
+      "step": 5130
+    },
+    {
+      "epoch": 1.3115590711916305,
+      "grad_norm": 0.8652639389038086,
+      "learning_rate": 4.1512383672787536e-06,
+      "loss": 0.0131,
+      "step": 5140
+    },
+    {
+      "epoch": 1.3141107425363614,
+      "grad_norm": 0.40679681301116943,
+      "learning_rate": 4.143007991612126e-06,
+      "loss": 0.0182,
+      "step": 5150
+    },
+    {
+      "epoch": 1.316662413881092,
+      "grad_norm": 0.6223965883255005,
+      "learning_rate": 4.134767539780641e-06,
+      "loss": 0.0155,
+      "step": 5160
+    },
+    {
+      "epoch": 1.319214085225823,
+      "grad_norm": 0.5270498394966125,
+      "learning_rate": 4.126517084427841e-06,
+      "loss": 0.0149,
+      "step": 5170
+    },
+    {
+      "epoch": 1.3217657565705538,
+      "grad_norm": 0.554425060749054,
+      "learning_rate": 4.118256698285457e-06,
+      "loss": 0.0155,
+      "step": 5180
+    },
+    {
+      "epoch": 1.3243174279152845,
+      "grad_norm": 0.7247564196586609,
+      "learning_rate": 4.109986454172762e-06,
+      "loss": 0.0162,
+      "step": 5190
+    },
+    {
+      "epoch": 1.3268690992600152,
+      "grad_norm": 0.6803126335144043,
+      "learning_rate": 4.101706424995937e-06,
+      "loss": 0.0166,
+      "step": 5200
+    },
+    {
+      "epoch": 1.329420770604746,
+      "grad_norm": 1.2059119939804077,
+      "learning_rate": 4.093416683747414e-06,
+      "loss": 0.0171,
+      "step": 5210
+    },
+    {
+      "epoch": 1.331972441949477,
+      "grad_norm": 0.3244284689426422,
+      "learning_rate": 4.08511730350525e-06,
+      "loss": 0.0188,
+      "step": 5220
+    },
+    {
+      "epoch": 1.3345241132942078,
+      "grad_norm": 0.6303139925003052,
+      "learning_rate": 4.07680835743247e-06,
+      "loss": 0.0154,
+      "step": 5230
+    },
+    {
+      "epoch": 1.3370757846389385,
+      "grad_norm": 0.5484089851379395,
+      "learning_rate": 4.068489918776429e-06,
+      "loss": 0.0158,
+      "step": 5240
+    },
+    {
+      "epoch": 1.3396274559836692,
+      "grad_norm": 0.4632198214530945,
+      "learning_rate": 4.060162060868159e-06,
+      "loss": 0.0155,
+      "step": 5250
+    },
+    {
+      "epoch": 1.3421791273284,
+      "grad_norm": 0.6339452266693115,
+      "learning_rate": 4.051824857121733e-06,
+      "loss": 0.0178,
+      "step": 5260
+    },
+    {
+      "epoch": 1.344730798673131,
+      "grad_norm": 0.6535425782203674,
+      "learning_rate": 4.04347838103361e-06,
+      "loss": 0.0167,
+      "step": 5270
+    },
+    {
+      "epoch": 1.3472824700178618,
+      "grad_norm": 0.40795567631721497,
+      "learning_rate": 4.0351227061819856e-06,
+      "loss": 0.0159,
+      "step": 5280
+    },
+    {
+      "epoch": 1.3498341413625925,
+      "grad_norm": 0.9828789830207825,
+      "learning_rate": 4.026757906226153e-06,
+      "loss": 0.0148,
+      "step": 5290
+    },
+    {
+      "epoch": 1.3523858127073232,
+      "grad_norm": 0.3722200393676758,
+      "learning_rate": 4.0183840549058445e-06,
+      "loss": 0.0153,
+      "step": 5300
+    },
+    {
+      "epoch": 1.354937484052054,
+      "grad_norm": 0.7231210470199585,
+      "learning_rate": 4.010001226040581e-06,
+      "loss": 0.0183,
+      "step": 5310
+    },
+    {
+      "epoch": 1.357489155396785,
+      "grad_norm": 0.30369916558265686,
+      "learning_rate": 4.001609493529033e-06,
+      "loss": 0.0149,
+      "step": 5320
+    },
+    {
+      "epoch": 1.3600408267415158,
+      "grad_norm": 0.5192328691482544,
+      "learning_rate": 3.993208931348353e-06,
+      "loss": 0.0152,
+      "step": 5330
+    },
+    {
+      "epoch": 1.3625924980862465,
+      "grad_norm": 1.303830862045288,
+      "learning_rate": 3.984799613553535e-06,
+      "loss": 0.015,
+      "step": 5340
+    },
+    {
+      "epoch": 1.3651441694309772,
+      "grad_norm": 0.4460931122303009,
+      "learning_rate": 3.976381614276759e-06,
+      "loss": 0.015,
+      "step": 5350
+    },
+    {
+      "epoch": 1.367695840775708,
+      "grad_norm": 0.687340497970581,
+      "learning_rate": 3.967955007726734e-06,
+      "loss": 0.0184,
+      "step": 5360
+    },
+    {
+      "epoch": 1.370247512120439,
+      "grad_norm": 0.5577728748321533,
+      "learning_rate": 3.959519868188046e-06,
+      "loss": 0.0154,
+      "step": 5370
+    },
+    {
+      "epoch": 1.3727991834651696,
+      "grad_norm": 0.2629478871822357,
+      "learning_rate": 3.9510762700205065e-06,
+      "loss": 0.0143,
+      "step": 5380
+    },
+    {
+      "epoch": 1.3753508548099005,
+      "grad_norm": 0.48370009660720825,
+      "learning_rate": 3.942624287658492e-06,
+      "loss": 0.0161,
+      "step": 5390
+    },
+    {
+      "epoch": 1.3779025261546312,
+      "grad_norm": 0.7703590393066406,
+      "learning_rate": 3.934163995610287e-06,
+      "loss": 0.0135,
+      "step": 5400
+    },
+    {
+      "epoch": 1.380454197499362,
+      "grad_norm": 0.46242403984069824,
+      "learning_rate": 3.9256954684574355e-06,
+      "loss": 0.0175,
+      "step": 5410
+    },
+    {
+      "epoch": 1.383005868844093,
+      "grad_norm": 0.5483065843582153,
+      "learning_rate": 3.917218780854075e-06,
+      "loss": 0.015,
+      "step": 5420
+    },
+    {
+      "epoch": 1.3855575401888236,
+      "grad_norm": 0.3543405830860138,
+      "learning_rate": 3.9087340075262766e-06,
+      "loss": 0.0153,
+      "step": 5430
+    },
+    {
+      "epoch": 1.3881092115335545,
+      "grad_norm": 0.2908279001712799,
+      "learning_rate": 3.900241223271402e-06,
+      "loss": 0.0121,
+      "step": 5440
+    },
+    {
+      "epoch": 1.3906608828782852,
+      "grad_norm": 0.5572682619094849,
+      "learning_rate": 3.891740502957423e-06,
+      "loss": 0.0151,
+      "step": 5450
+    },
+    {
+      "epoch": 1.393212554223016,
+      "grad_norm": 0.6333874464035034,
+      "learning_rate": 3.883231921522275e-06,
+      "loss": 0.0168,
+      "step": 5460
+    },
+    {
+      "epoch": 1.395764225567747,
+      "grad_norm": 2.3678176403045654,
+      "learning_rate": 3.874715553973195e-06,
+      "loss": 0.0146,
+      "step": 5470
+    },
+    {
+      "epoch": 1.3983158969124776,
+      "grad_norm": 0.8847137093544006,
+      "learning_rate": 3.866191475386055e-06,
+      "loss": 0.0157,
+      "step": 5480
+    },
+    {
+      "epoch": 1.4008675682572085,
+      "grad_norm": 0.9330421686172485,
+      "learning_rate": 3.857659760904705e-06,
+      "loss": 0.0178,
+      "step": 5490
+    },
+    {
+      "epoch": 1.4034192396019392,
+      "grad_norm": 0.6955951452255249,
+      "learning_rate": 3.849120485740311e-06,
+      "loss": 0.0185,
+      "step": 5500
+    },
+    {
+      "epoch": 1.40597091094667,
+      "grad_norm": 0.7250450849533081,
+      "learning_rate": 3.840573725170686e-06,
+      "loss": 0.0132,
+      "step": 5510
+    },
+    {
+      "epoch": 1.408522582291401,
+      "grad_norm": 0.7257997989654541,
+      "learning_rate": 3.832019554539634e-06,
+      "loss": 0.0168,
+      "step": 5520
+    },
+    {
+      "epoch": 1.4110742536361316,
+      "grad_norm": 1.2495430707931519,
+      "learning_rate": 3.82345804925628e-06,
+      "loss": 0.0166,
+      "step": 5530
+    },
+    {
+      "epoch": 1.4136259249808625,
+      "grad_norm": 1.9423291683197021,
+      "learning_rate": 3.814889284794409e-06,
+      "loss": 0.0139,
+      "step": 5540
+    },
+    {
+      "epoch": 1.4161775963255931,
+      "grad_norm": 0.7038158774375916,
+      "learning_rate": 3.8063133366917976e-06,
+      "loss": 0.0178,
+      "step": 5550
+    },
+    {
+      "epoch": 1.418729267670324,
+      "grad_norm": 0.39670151472091675,
+      "learning_rate": 3.797730280549552e-06,
+      "loss": 0.0164,
+      "step": 5560
+    },
+    {
+      "epoch": 1.421280939015055,
+      "grad_norm": 1.0969537496566772,
+      "learning_rate": 3.7891401920314363e-06,
+      "loss": 0.0166,
+      "step": 5570
+    },
+    {
+      "epoch": 1.4238326103597856,
+      "grad_norm": 1.4217796325683594,
+      "learning_rate": 3.7805431468632105e-06,
+      "loss": 0.0176,
+      "step": 5580
+    },
+    {
+      "epoch": 1.4263842817045165,
+      "grad_norm": 0.8468742370605469,
+      "learning_rate": 3.77193922083196e-06,
+      "loss": 0.0147,
+      "step": 5590
+    },
+    {
+      "epoch": 1.4289359530492471,
+      "grad_norm": 0.4837898910045624,
+      "learning_rate": 3.763328489785428e-06,
+      "loss": 0.0167,
+      "step": 5600
+    },
+    {
+      "epoch": 1.431487624393978,
+      "grad_norm": 0.7742527723312378,
+      "learning_rate": 3.7547110296313486e-06,
+      "loss": 0.0151,
+      "step": 5610
+    },
+    {
+      "epoch": 1.434039295738709,
+      "grad_norm": 0.7425050735473633,
+      "learning_rate": 3.7460869163367744e-06,
+      "loss": 0.0137,
+      "step": 5620
+    },
+    {
+      "epoch": 1.4365909670834396,
+      "grad_norm": 0.8111595511436462,
+      "learning_rate": 3.7374562259274093e-06,
+      "loss": 0.0153,
+      "step": 5630
+    },
+    {
+      "epoch": 1.4391426384281705,
+      "grad_norm": 2.8571298122406006,
+      "learning_rate": 3.728819034486939e-06,
+      "loss": 0.0172,
+      "step": 5640
+    },
+    {
+      "epoch": 1.4416943097729011,
+      "grad_norm": 0.4270651042461395,
+      "learning_rate": 3.720175418156356e-06,
+      "loss": 0.0156,
+      "step": 5650
+    },
+    {
+      "epoch": 1.444245981117632,
+      "grad_norm": 0.715120792388916,
+      "learning_rate": 3.711525453133295e-06,
+      "loss": 0.015,
+      "step": 5660
+    },
+    {
+      "epoch": 1.446797652462363,
+      "grad_norm": 0.7231752872467041,
+      "learning_rate": 3.7028692156713536e-06,
+      "loss": 0.0133,
+      "step": 5670
+    },
+    {
+      "epoch": 1.4493493238070936,
+      "grad_norm": 2.8344600200653076,
+      "learning_rate": 3.694206782079427e-06,
+      "loss": 0.0147,
+      "step": 5680
+    },
+    {
+      "epoch": 1.4519009951518245,
+      "grad_norm": 0.8471620678901672,
+      "learning_rate": 3.68553822872103e-06,
+      "loss": 0.015,
+      "step": 5690
+    },
+    {
+      "epoch": 1.4544526664965551,
+      "grad_norm": 1.0441292524337769,
+      "learning_rate": 3.6768636320136284e-06,
+      "loss": 0.0147,
+      "step": 5700
+    },
+    {
+      "epoch": 1.457004337841286,
+      "grad_norm": 0.5474273562431335,
+      "learning_rate": 3.6681830684279616e-06,
+      "loss": 0.0144,
+      "step": 5710
+    },
+    {
+      "epoch": 1.459556009186017,
+      "grad_norm": 0.8801978826522827,
+      "learning_rate": 3.6594966144873693e-06,
+      "loss": 0.0145,
+      "step": 5720
+    },
+    {
+      "epoch": 1.4621076805307476,
+      "grad_norm": 0.8165146708488464,
+      "learning_rate": 3.6508043467671185e-06,
+      "loss": 0.0178,
+      "step": 5730
+    },
+    {
+      "epoch": 1.4646593518754785,
+      "grad_norm": 1.9424936771392822,
+      "learning_rate": 3.642106341893728e-06,
+      "loss": 0.0127,
+      "step": 5740
+    },
+    {
+      "epoch": 1.4672110232202091,
+      "grad_norm": 0.49900662899017334,
+      "learning_rate": 3.6334026765442917e-06,
+      "loss": 0.0134,
+      "step": 5750
+    },
+    {
+      "epoch": 1.46976269456494,
+      "grad_norm": 0.8361905217170715,
+      "learning_rate": 3.6246934274458025e-06,
+      "loss": 0.0148,
+      "step": 5760
+    },
+    {
+      "epoch": 1.472314365909671,
+      "grad_norm": 2.2165048122406006,
+      "learning_rate": 3.615978671374478e-06,
+      "loss": 0.0158,
+      "step": 5770
+    },
+    {
+      "epoch": 1.4748660372544016,
+      "grad_norm": 0.5519954562187195,
+      "learning_rate": 3.6072584851550814e-06,
+      "loss": 0.0163,
+      "step": 5780
+    },
+    {
+      "epoch": 1.4774177085991325,
+      "grad_norm": 0.5470237135887146,
+      "learning_rate": 3.598532945660246e-06,
+      "loss": 0.0162,
+      "step": 5790
+    },
+    {
+      "epoch": 1.4799693799438631,
+      "grad_norm": 0.6911956071853638,
+      "learning_rate": 3.589802129809797e-06,
+      "loss": 0.0156,
+      "step": 5800
+    },
+    {
+      "epoch": 1.482521051288594,
+      "grad_norm": 0.6574975848197937,
+      "learning_rate": 3.5810661145700722e-06,
+      "loss": 0.0163,
+      "step": 5810
+    },
+    {
+      "epoch": 1.485072722633325,
+      "grad_norm": 1.735755205154419,
+      "learning_rate": 3.5723249769532454e-06,
+      "loss": 0.0172,
+      "step": 5820
+    },
+    {
+      "epoch": 1.4876243939780556,
+      "grad_norm": 0.5131850838661194,
+      "learning_rate": 3.563578794016646e-06,
+      "loss": 0.0155,
+      "step": 5830
+    },
+    {
+      "epoch": 1.4901760653227865,
+      "grad_norm": 2.086827516555786,
+      "learning_rate": 3.554827642862081e-06,
+      "loss": 0.0157,
+      "step": 5840
+    },
+    {
+      "epoch": 1.4927277366675171,
+      "grad_norm": 1.4139020442962646,
+      "learning_rate": 3.546071600635154e-06,
+      "loss": 0.014,
+      "step": 5850
+    },
+    {
+      "epoch": 1.495279408012248,
+      "grad_norm": 0.5393047332763672,
+      "learning_rate": 3.537310744524587e-06,
+      "loss": 0.0177,
+      "step": 5860
+    },
+    {
+      "epoch": 1.497831079356979,
+      "grad_norm": 0.30612704157829285,
+      "learning_rate": 3.528545151761536e-06,
+      "loss": 0.0121,
+      "step": 5870
+    },
+    {
+      "epoch": 1.5003827507017096,
+      "grad_norm": 0.40969327092170715,
+      "learning_rate": 3.5197748996189154e-06,
+      "loss": 0.0139,
+      "step": 5880
+    },
+    {
+      "epoch": 1.5029344220464405,
+      "grad_norm": 0.7444477677345276,
+      "learning_rate": 3.5110000654107145e-06,
+      "loss": 0.0163,
+      "step": 5890
+    },
+    {
+      "epoch": 1.5054860933911711,
+      "grad_norm": 1.087607502937317,
+      "learning_rate": 3.5022207264913127e-06,
+      "loss": 0.0144,
+      "step": 5900
+    },
+    {
+      "epoch": 1.508037764735902,
+      "grad_norm": 0.7293539047241211,
+      "learning_rate": 3.493436960254803e-06,
+      "loss": 0.0152,
+      "step": 5910
+    },
+    {
+      "epoch": 1.510589436080633,
+      "grad_norm": 1.0831620693206787,
+      "learning_rate": 3.484648844134309e-06,
+      "loss": 0.0165,
+      "step": 5920
+    },
+    {
+      "epoch": 1.5131411074253636,
+      "grad_norm": 0.7094727754592896,
+      "learning_rate": 3.475856455601296e-06,
+      "loss": 0.0139,
+      "step": 5930
+    },
+    {
+      "epoch": 1.5156927787700945,
+      "grad_norm": 0.6312902569770813,
+      "learning_rate": 3.4670598721648956e-06,
+      "loss": 0.0154,
+      "step": 5940
+    },
+    {
+      "epoch": 1.5182444501148251,
+      "grad_norm": 0.8989167213439941,
+      "learning_rate": 3.4582591713712205e-06,
+      "loss": 0.0161,
+      "step": 5950
+    },
+    {
+      "epoch": 1.520796121459556,
+      "grad_norm": 0.4361861050128937,
+      "learning_rate": 3.449454430802677e-06,
+      "loss": 0.014,
+      "step": 5960
+    },
+    {
+      "epoch": 1.523347792804287,
+      "grad_norm": 0.8975860476493835,
+      "learning_rate": 3.4406457280772854e-06,
+      "loss": 0.0121,
+      "step": 5970
+    },
+    {
+      "epoch": 1.5258994641490176,
+      "grad_norm": 0.9588630795478821,
+      "learning_rate": 3.431833140847996e-06,
+      "loss": 0.0141,
+      "step": 5980
+    },
+    {
+      "epoch": 1.5284511354937484,
+      "grad_norm": 0.75777667760849,
+      "learning_rate": 3.4230167468019992e-06,
+      "loss": 0.0184,
+      "step": 5990
+    },
+    {
+      "epoch": 1.5310028068384791,
+      "grad_norm": 0.8160403966903687,
+      "learning_rate": 3.4141966236600466e-06,
+      "loss": 0.0152,
+      "step": 6000
+    },
+    {
+      "epoch": 1.53355447818321,
+      "grad_norm": 0.3753710091114044,
+      "learning_rate": 3.4053728491757654e-06,
+      "loss": 0.0138,
+      "step": 6010
+    },
+    {
+      "epoch": 1.536106149527941,
+      "grad_norm": 0.3643024265766144,
+      "learning_rate": 3.3965455011349656e-06,
+      "loss": 0.014,
+      "step": 6020
+    },
+    {
+      "epoch": 1.5386578208726716,
+      "grad_norm": 0.9779283404350281,
+      "learning_rate": 3.387714657354965e-06,
+      "loss": 0.0169,
+      "step": 6030
+    },
+    {
+      "epoch": 1.5412094922174024,
+      "grad_norm": 0.593742847442627,
+      "learning_rate": 3.378880395683898e-06,
+      "loss": 0.0158,
+      "step": 6040
+    },
+    {
+      "epoch": 1.543761163562133,
+      "grad_norm": 0.9815357327461243,
+      "learning_rate": 3.370042794000025e-06,
+      "loss": 0.0166,
+      "step": 6050
+    },
+    {
+      "epoch": 1.546312834906864,
+      "grad_norm": 0.4922243356704712,
+      "learning_rate": 3.3612019302110528e-06,
+      "loss": 0.0149,
+      "step": 6060
+    },
+    {
+      "epoch": 1.5488645062515949,
+      "grad_norm": 0.8108282685279846,
+      "learning_rate": 3.352357882253447e-06,
+      "loss": 0.0171,
+      "step": 6070
+    },
+    {
+      "epoch": 1.5514161775963256,
+      "grad_norm": 0.6546838283538818,
+      "learning_rate": 3.343510728091742e-06,
+      "loss": 0.0158,
+      "step": 6080
+    },
+    {
+      "epoch": 1.5539678489410564,
+      "grad_norm": 0.35426297783851624,
+      "learning_rate": 3.3346605457178506e-06,
+      "loss": 0.0128,
+      "step": 6090
+    },
+    {
+      "epoch": 1.556519520285787,
+      "grad_norm": 0.5806670188903809,
+      "learning_rate": 3.325807413150388e-06,
+      "loss": 0.0137,
+      "step": 6100
+    },
+    {
+      "epoch": 1.559071191630518,
+      "grad_norm": 0.43077072501182556,
+      "learning_rate": 3.3169514084339723e-06,
+      "loss": 0.0143,
+      "step": 6110
+    },
+    {
+      "epoch": 1.5616228629752489,
+      "grad_norm": 0.6155886650085449,
+      "learning_rate": 3.30809260963854e-06,
+      "loss": 0.0134,
+      "step": 6120
+    },
+    {
+      "epoch": 1.5641745343199795,
+      "grad_norm": 0.400890588760376,
+      "learning_rate": 3.2992310948586636e-06,
+      "loss": 0.0148,
+      "step": 6130
+    },
+    {
+      "epoch": 1.5667262056647104,
+      "grad_norm": 0.9763600826263428,
+      "learning_rate": 3.290366942212855e-06,
+      "loss": 0.012,
+      "step": 6140
+    },
+    {
+      "epoch": 1.569277877009441,
+      "grad_norm": 0.8423333764076233,
+      "learning_rate": 3.281500229842878e-06,
+      "loss": 0.0145,
+      "step": 6150
+    },
+    {
+      "epoch": 1.571829548354172,
+      "grad_norm": 0.7116320133209229,
+      "learning_rate": 3.2726310359130654e-06,
+      "loss": 0.0149,
+      "step": 6160
+    },
+    {
+      "epoch": 1.5743812196989029,
+      "grad_norm": 0.6901907920837402,
+      "learning_rate": 3.2637594386096266e-06,
+      "loss": 0.0138,
+      "step": 6170
+    },
+    {
+      "epoch": 1.5769328910436335,
+      "grad_norm": 0.2646922767162323,
+      "learning_rate": 3.2548855161399535e-06,
+      "loss": 0.0176,
+      "step": 6180
+    },
+    {
+      "epoch": 1.5794845623883644,
+      "grad_norm": 1.581405758857727,
+      "learning_rate": 3.24600934673194e-06,
+      "loss": 0.0148,
+      "step": 6190
+    },
+    {
+      "epoch": 1.582036233733095,
+      "grad_norm": 0.8435882925987244,
+      "learning_rate": 3.2371310086332856e-06,
+      "loss": 0.0131,
+      "step": 6200
+    },
+    {
+      "epoch": 1.584587905077826,
+      "grad_norm": 0.5191860198974609,
+      "learning_rate": 3.2282505801108067e-06,
+      "loss": 0.0142,
+      "step": 6210
+    },
+    {
+      "epoch": 1.5871395764225569,
+      "grad_norm": 0.8106646537780762,
+      "learning_rate": 3.219368139449752e-06,
+      "loss": 0.015,
+      "step": 6220
+    },
+    {
+      "epoch": 1.5896912477672875,
+      "grad_norm": 0.6545732617378235,
+      "learning_rate": 3.2104837649531035e-06,
+      "loss": 0.0139,
+      "step": 6230
+    },
+    {
+      "epoch": 1.5922429191120182,
+      "grad_norm": 0.6558523774147034,
+      "learning_rate": 3.2015975349408924e-06,
+      "loss": 0.0123,
+      "step": 6240
+    },
+    {
+      "epoch": 1.594794590456749,
+      "grad_norm": 0.5409281253814697,
+      "learning_rate": 3.192709527749509e-06,
+      "loss": 0.0135,
+      "step": 6250
+    },
+    {
+      "epoch": 1.59734626180148,
+      "grad_norm": 1.8452955484390259,
+      "learning_rate": 3.183819821731009e-06,
+      "loss": 0.0145,
+      "step": 6260
+    },
+    {
+      "epoch": 1.5998979331462109,
+      "grad_norm": 1.8067026138305664,
+      "learning_rate": 3.1749284952524226e-06,
+      "loss": 0.016,
+      "step": 6270
+    },
+    {
+      "epoch": 1.6024496044909415,
+      "grad_norm": 0.5791333913803101,
+      "learning_rate": 3.1660356266950677e-06,
+      "loss": 0.0152,
+      "step": 6280
+    },
+    {
+      "epoch": 1.6050012758356722,
+      "grad_norm": 0.9250419735908508,
+      "learning_rate": 3.157141294453855e-06,
+      "loss": 0.0156,
+      "step": 6290
+    },
+    {
+      "epoch": 1.607552947180403,
+      "grad_norm": 0.522924542427063,
+      "learning_rate": 3.1482455769365977e-06,
+      "loss": 0.0125,
+      "step": 6300
+    },
+    {
+      "epoch": 1.610104618525134,
+      "grad_norm": 0.4562835991382599,
+      "learning_rate": 3.139348552563323e-06,
+      "loss": 0.0143,
+      "step": 6310
+    },
+    {
+      "epoch": 1.6126562898698649,
+      "grad_norm": 0.5292758941650391,
+      "learning_rate": 3.1304502997655765e-06,
+      "loss": 0.0126,
+      "step": 6320
+    },
+    {
+      "epoch": 1.6152079612145955,
+      "grad_norm": 0.39967814087867737,
+      "learning_rate": 3.1215508969857334e-06,
+      "loss": 0.0123,
+      "step": 6330
+    },
+    {
+      "epoch": 1.6177596325593262,
+      "grad_norm": 0.47947466373443604,
+      "learning_rate": 3.1126504226763073e-06,
+      "loss": 0.014,
+      "step": 6340
+    },
+    {
+      "epoch": 1.620311303904057,
+      "grad_norm": 0.7605049014091492,
+      "learning_rate": 3.103748955299258e-06,
+      "loss": 0.0162,
+      "step": 6350
+    },
+    {
+      "epoch": 1.622862975248788,
+      "grad_norm": 0.7074354290962219,
+      "learning_rate": 3.0948465733252974e-06,
+      "loss": 0.0143,
+      "step": 6360
+    },
+    {
+      "epoch": 1.6254146465935189,
+      "grad_norm": 1.39603853225708,
+      "learning_rate": 3.0859433552332036e-06,
+      "loss": 0.0154,
+      "step": 6370
+    },
+    {
+      "epoch": 1.6279663179382495,
+      "grad_norm": 0.4325566291809082,
+      "learning_rate": 3.0770393795091224e-06,
+      "loss": 0.0166,
+      "step": 6380
+    },
+    {
+      "epoch": 1.6305179892829802,
+      "grad_norm": 0.5471740365028381,
+      "learning_rate": 3.0681347246458805e-06,
+      "loss": 0.012,
+      "step": 6390
+    },
+    {
+      "epoch": 1.633069660627711,
+      "grad_norm": 1.0500181913375854,
+      "learning_rate": 3.0592294691422905e-06,
+      "loss": 0.0163,
+      "step": 6400
+    },
+    {
+      "epoch": 1.635621331972442,
+      "grad_norm": 0.5767816305160522,
+      "learning_rate": 3.0503236915024603e-06,
+      "loss": 0.0126,
+      "step": 6410
+    },
+    {
+      "epoch": 1.6381730033171729,
+      "grad_norm": 1.0612276792526245,
+      "learning_rate": 3.0414174702351e-06,
+      "loss": 0.0137,
+      "step": 6420
+    },
+    {
+      "epoch": 1.6407246746619035,
+      "grad_norm": 0.5275915265083313,
+      "learning_rate": 3.0325108838528313e-06,
+      "loss": 0.0126,
+      "step": 6430
+    },
+    {
+      "epoch": 1.6432763460066342,
+      "grad_norm": 0.3712652027606964,
+      "learning_rate": 3.0236040108714945e-06,
+      "loss": 0.0134,
+      "step": 6440
+    },
+    {
+      "epoch": 1.645828017351365,
+      "grad_norm": 0.516525387763977,
+      "learning_rate": 3.014696929809456e-06,
+      "loss": 0.0134,
+      "step": 6450
+    },
+    {
+      "epoch": 1.648379688696096,
+      "grad_norm": 0.6508063077926636,
+      "learning_rate": 3.005789719186917e-06,
+      "loss": 0.016,
+      "step": 6460
+    },
+    {
+      "epoch": 1.6509313600408269,
+      "grad_norm": 0.25411614775657654,
+      "learning_rate": 2.9968824575252196e-06,
+      "loss": 0.0143,
+      "step": 6470
+    },
+    {
+      "epoch": 1.6534830313855575,
+      "grad_norm": 0.4073875844478607,
+      "learning_rate": 2.9879752233461587e-06,
+      "loss": 0.0124,
+      "step": 6480
+    },
+    {
+      "epoch": 1.6560347027302882,
+      "grad_norm": 0.8705873489379883,
+      "learning_rate": 2.9790680951712836e-06,
+      "loss": 0.0113,
+      "step": 6490
+    },
+    {
+      "epoch": 1.658586374075019,
+      "grad_norm": 0.9108631610870361,
+      "learning_rate": 2.9701611515212113e-06,
+      "loss": 0.0148,
+      "step": 6500
+    },
+    {
+      "epoch": 1.66113804541975,
+      "grad_norm": 0.5919849872589111,
+      "learning_rate": 2.961254470914931e-06,
+      "loss": 0.0131,
+      "step": 6510
+    },
+    {
+      "epoch": 1.6636897167644809,
+      "grad_norm": 0.5540559887886047,
+      "learning_rate": 2.952348131869112e-06,
+      "loss": 0.0127,
+      "step": 6520
+    },
+    {
+      "epoch": 1.6662413881092115,
+      "grad_norm": 0.23948337137699127,
+      "learning_rate": 2.9434422128974165e-06,
+      "loss": 0.0134,
+      "step": 6530
+    },
+    {
+      "epoch": 1.6687930594539422,
+      "grad_norm": 0.6661424040794373,
+      "learning_rate": 2.934536792509801e-06,
+      "loss": 0.0118,
+      "step": 6540
+    },
+    {
+      "epoch": 1.671344730798673,
+      "grad_norm": 0.2371857613325119,
+      "learning_rate": 2.925631949211825e-06,
+      "loss": 0.011,
+      "step": 6550
+    },
+    {
+      "epoch": 1.673896402143404,
+      "grad_norm": 0.9995185732841492,
+      "learning_rate": 2.916727761503963e-06,
+      "loss": 0.0133,
+      "step": 6560
+    },
+    {
+      "epoch": 1.6764480734881348,
+      "grad_norm": 0.30217620730400085,
+      "learning_rate": 2.9078243078809106e-06,
+      "loss": 0.0133,
+      "step": 6570
+    },
+    {
+      "epoch": 1.6789997448328655,
+      "grad_norm": 0.6044448018074036,
+      "learning_rate": 2.8989216668308897e-06,
+      "loss": 0.0132,
+      "step": 6580
+    },
+    {
+      "epoch": 1.6815514161775962,
+      "grad_norm": 0.6776731610298157,
+      "learning_rate": 2.8900199168349616e-06,
+      "loss": 0.0149,
+      "step": 6590
+    },
+    {
+      "epoch": 1.684103087522327,
+      "grad_norm": 0.5117316246032715,
+      "learning_rate": 2.881119136366332e-06,
+      "loss": 0.0146,
+      "step": 6600
+    },
+    {
+      "epoch": 1.686654758867058,
+      "grad_norm": 0.6031244993209839,
+      "learning_rate": 2.8722194038896565e-06,
+      "loss": 0.0122,
+      "step": 6610
+    },
+    {
+      "epoch": 1.6892064302117888,
+      "grad_norm": 0.5227133631706238,
+      "learning_rate": 2.863320797860357e-06,
+      "loss": 0.0164,
+      "step": 6620
+    },
+    {
+      "epoch": 1.6917581015565195,
+      "grad_norm": 0.5128726959228516,
+      "learning_rate": 2.854423396723922e-06,
+      "loss": 0.0129,
+      "step": 6630
+    },
+    {
+      "epoch": 1.6943097729012502,
+      "grad_norm": 0.677301824092865,
+      "learning_rate": 2.845527278915219e-06,
+      "loss": 0.0111,
+      "step": 6640
+    },
+    {
+      "epoch": 1.696861444245981,
+      "grad_norm": 1.8520700931549072,
+      "learning_rate": 2.836632522857803e-06,
+      "loss": 0.0128,
+      "step": 6650
+    },
+    {
+      "epoch": 1.699413115590712,
+      "grad_norm": 0.43956881761550903,
+      "learning_rate": 2.827739206963225e-06,
+      "loss": 0.0142,
+      "step": 6660
+    },
+    {
+      "epoch": 1.7019647869354428,
+      "grad_norm": 1.2026915550231934,
+      "learning_rate": 2.818847409630339e-06,
+      "loss": 0.0122,
+      "step": 6670
+    },
+    {
+      "epoch": 1.7045164582801735,
+      "grad_norm": 1.0047426223754883,
+      "learning_rate": 2.809957209244611e-06,
+      "loss": 0.0132,
+      "step": 6680
+    },
+    {
+      "epoch": 1.7070681296249042,
+      "grad_norm": 0.4368385672569275,
+      "learning_rate": 2.801068684177432e-06,
+      "loss": 0.0155,
+      "step": 6690
+    },
+    {
+      "epoch": 1.709619800969635,
+      "grad_norm": 2.1144449710845947,
+      "learning_rate": 2.792181912785422e-06,
+      "loss": 0.0133,
+      "step": 6700
+    },
+    {
+      "epoch": 1.712171472314366,
+      "grad_norm": 0.46182388067245483,
+      "learning_rate": 2.7832969734097437e-06,
+      "loss": 0.0132,
+      "step": 6710
+    },
+    {
+      "epoch": 1.7147231436590968,
+      "grad_norm": 1.1694209575653076,
+      "learning_rate": 2.774413944375408e-06,
+      "loss": 0.0126,
+      "step": 6720
+    },
+    {
+      "epoch": 1.7172748150038275,
+      "grad_norm": 0.6728443503379822,
+      "learning_rate": 2.765532903990586e-06,
+      "loss": 0.0125,
+      "step": 6730
+    },
+    {
+      "epoch": 1.7198264863485582,
+      "grad_norm": 0.3873930871486664,
+      "learning_rate": 2.7566539305459167e-06,
+      "loss": 0.0151,
+      "step": 6740
+    },
+    {
+      "epoch": 1.722378157693289,
+      "grad_norm": 0.5007622838020325,
+      "learning_rate": 2.747777102313819e-06,
+      "loss": 0.0114,
+      "step": 6750
+    },
+    {
+      "epoch": 1.72492982903802,
+      "grad_norm": 1.0474838018417358,
+      "learning_rate": 2.7389024975478e-06,
+      "loss": 0.0164,
+      "step": 6760
+    },
+    {
+      "epoch": 1.7274815003827508,
+      "grad_norm": 0.9246062636375427,
+      "learning_rate": 2.730030194481768e-06,
+      "loss": 0.0138,
+      "step": 6770
+    },
+    {
+      "epoch": 1.7300331717274815,
+      "grad_norm": 0.7463748455047607,
+      "learning_rate": 2.721160271329339e-06,
+      "loss": 0.0129,
+      "step": 6780
+    },
+    {
+      "epoch": 1.7325848430722122,
+      "grad_norm": 0.39589956402778625,
+      "learning_rate": 2.712292806283149e-06,
+      "loss": 0.012,
+      "step": 6790
+    },
+    {
+      "epoch": 1.735136514416943,
+      "grad_norm": 0.7912735342979431,
+      "learning_rate": 2.703427877514164e-06,
+      "loss": 0.0131,
+      "step": 6800
+    },
+    {
+      "epoch": 1.737688185761674,
+      "grad_norm": 0.9773954749107361,
+      "learning_rate": 2.694565563170992e-06,
+      "loss": 0.0119,
+      "step": 6810
+    },
+    {
+      "epoch": 1.7402398571064048,
+      "grad_norm": 0.45853710174560547,
+      "learning_rate": 2.685705941379194e-06,
+      "loss": 0.0135,
+      "step": 6820
+    },
+    {
+      "epoch": 1.7427915284511355,
+      "grad_norm": 0.4193810522556305,
+      "learning_rate": 2.676849090240596e-06,
+      "loss": 0.0117,
+      "step": 6830
+    },
+    {
+      "epoch": 1.7453431997958662,
+      "grad_norm": 0.5059391856193542,
+      "learning_rate": 2.6679950878325967e-06,
+      "loss": 0.0139,
+      "step": 6840
+    },
+    {
+      "epoch": 1.747894871140597,
+      "grad_norm": 1.798842191696167,
+      "learning_rate": 2.6591440122074846e-06,
+      "loss": 0.013,
+      "step": 6850
+    },
+    {
+      "epoch": 1.750446542485328,
+      "grad_norm": 0.6951261162757874,
+      "learning_rate": 2.6502959413917437e-06,
+      "loss": 0.0109,
+      "step": 6860
+    },
+    {
+      "epoch": 1.7529982138300588,
+      "grad_norm": 1.0176141262054443,
+      "learning_rate": 2.6414509533853728e-06,
+      "loss": 0.0121,
+      "step": 6870
+    },
+    {
+      "epoch": 1.7555498851747895,
+      "grad_norm": 1.5015063285827637,
+      "learning_rate": 2.632609126161191e-06,
+      "loss": 0.0099,
+      "step": 6880
+    },
+    {
+      "epoch": 1.7581015565195202,
+      "grad_norm": 0.5768877863883972,
+      "learning_rate": 2.6237705376641576e-06,
+      "loss": 0.0135,
+      "step": 6890
+    },
+    {
+      "epoch": 1.760653227864251,
+      "grad_norm": 2.051805257797241,
+      "learning_rate": 2.6149352658106775e-06,
+      "loss": 0.0136,
+      "step": 6900
+    },
+    {
+      "epoch": 1.763204899208982,
+      "grad_norm": 0.3947019577026367,
+      "learning_rate": 2.6061033884879193e-06,
+      "loss": 0.011,
+      "step": 6910
+    },
+    {
+      "epoch": 1.7657565705537128,
+      "grad_norm": 0.5471501350402832,
+      "learning_rate": 2.5972749835531266e-06,
+      "loss": 0.013,
+      "step": 6920
+    },
+    {
+      "epoch": 1.7683082418984435,
+      "grad_norm": 0.5119963884353638,
+      "learning_rate": 2.5884501288329326e-06,
+      "loss": 0.0112,
+      "step": 6930
+    },
+    {
+      "epoch": 1.7708599132431742,
+      "grad_norm": 1.8591612577438354,
+      "learning_rate": 2.5796289021226722e-06,
+      "loss": 0.0133,
+      "step": 6940
+    },
+    {
+      "epoch": 1.773411584587905,
+      "grad_norm": 0.6567859053611755,
+      "learning_rate": 2.570811381185701e-06,
+      "loss": 0.0134,
+      "step": 6950
+    },
+    {
+      "epoch": 1.775963255932636,
+      "grad_norm": 0.4754619002342224,
+      "learning_rate": 2.5619976437527034e-06,
+      "loss": 0.0135,
+      "step": 6960
+    },
+    {
+      "epoch": 1.7785149272773668,
+      "grad_norm": 0.66524338722229,
+      "learning_rate": 2.5531877675210114e-06,
+      "loss": 0.0155,
+      "step": 6970
+    },
+    {
+      "epoch": 1.7810665986220975,
+      "grad_norm": 0.5068429708480835,
+      "learning_rate": 2.54438183015392e-06,
+      "loss": 0.0124,
+      "step": 6980
+    },
+    {
+      "epoch": 1.7836182699668282,
+      "grad_norm": 0.700384795665741,
+      "learning_rate": 2.5355799092799975e-06,
+      "loss": 0.0116,
+      "step": 6990
+    },
+    {
+      "epoch": 1.786169941311559,
+      "grad_norm": 0.24284306168556213,
+      "learning_rate": 2.526782082492408e-06,
+      "loss": 0.0105,
+      "step": 7000
+    },
+    {
+      "epoch": 1.78872161265629,
+      "grad_norm": 0.7343897223472595,
+      "learning_rate": 2.5179884273482245e-06,
+      "loss": 0.0135,
+      "step": 7010
+    },
+    {
+      "epoch": 1.7912732840010208,
+      "grad_norm": 0.4073348641395569,
+      "learning_rate": 2.509199021367743e-06,
+      "loss": 0.011,
+      "step": 7020
+    },
+    {
+      "epoch": 1.7938249553457515,
+      "grad_norm": 0.6841785907745361,
+      "learning_rate": 2.5004139420338027e-06,
+      "loss": 0.0116,
+      "step": 7030
+    },
+    {
+      "epoch": 1.7963766266904821,
+      "grad_norm": 1.4202295541763306,
+      "learning_rate": 2.491633266791101e-06,
+      "loss": 0.0126,
+      "step": 7040
+    },
+    {
+      "epoch": 1.798928298035213,
+      "grad_norm": 0.4704054296016693,
+      "learning_rate": 2.482857073045508e-06,
+      "loss": 0.013,
+      "step": 7050
+    },
+    {
+      "epoch": 1.801479969379944,
+      "grad_norm": 0.650240957736969,
+      "learning_rate": 2.4740854381633936e-06,
+      "loss": 0.0113,
+      "step": 7060
+    },
+    {
+      "epoch": 1.8040316407246748,
+      "grad_norm": 0.936140239238739,
+      "learning_rate": 2.4653184394709347e-06,
+      "loss": 0.0129,
+      "step": 7070
+    },
+    {
+      "epoch": 1.8065833120694055,
+      "grad_norm": 0.6480603218078613,
+      "learning_rate": 2.4565561542534372e-06,
+      "loss": 0.0132,
+      "step": 7080
+    },
+    {
+      "epoch": 1.8091349834141361,
+      "grad_norm": 0.7110511064529419,
+      "learning_rate": 2.4477986597546585e-06,
+      "loss": 0.0124,
+      "step": 7090
+    },
+    {
+      "epoch": 1.811686654758867,
+      "grad_norm": 0.6329154372215271,
+      "learning_rate": 2.4390460331761213e-06,
+      "loss": 0.0114,
+      "step": 7100
+    },
+    {
+      "epoch": 1.814238326103598,
+      "grad_norm": 0.3106624484062195,
+      "learning_rate": 2.4302983516764366e-06,
+      "loss": 0.0104,
+      "step": 7110
+    },
+    {
+      "epoch": 1.8167899974483288,
+      "grad_norm": 0.49600428342819214,
+      "learning_rate": 2.421555692370622e-06,
+      "loss": 0.0139,
+      "step": 7120
+    },
+    {
+      "epoch": 1.8193416687930595,
+      "grad_norm": 0.7423809766769409,
+      "learning_rate": 2.4128181323294206e-06,
+      "loss": 0.0114,
+      "step": 7130
+    },
+    {
+      "epoch": 1.8218933401377901,
+      "grad_norm": 0.3624695837497711,
+      "learning_rate": 2.404085748578625e-06,
+      "loss": 0.0149,
+      "step": 7140
+    },
+    {
+      "epoch": 1.824445011482521,
+      "grad_norm": 0.2720579206943512,
+      "learning_rate": 2.3953586180983947e-06,
+      "loss": 0.01,
+      "step": 7150
+    },
+    {
+      "epoch": 1.826996682827252,
+      "grad_norm": 0.2844093441963196,
+      "learning_rate": 2.3866368178225797e-06,
+      "loss": 0.0119,
+      "step": 7160
+    },
+    {
+      "epoch": 1.8295483541719828,
+      "grad_norm": 1.0419397354125977,
+      "learning_rate": 2.3779204246380412e-06,
+      "loss": 0.0115,
+      "step": 7170
+    },
+    {
+      "epoch": 1.8321000255167135,
+      "grad_norm": 0.7186925411224365,
+      "learning_rate": 2.3692095153839766e-06,
+      "loss": 0.0136,
+      "step": 7180
+    },
+    {
+      "epoch": 1.8346516968614441,
+      "grad_norm": 1.3992877006530762,
+      "learning_rate": 2.360504166851238e-06,
+      "loss": 0.0123,
+      "step": 7190
+    },
+    {
+      "epoch": 1.837203368206175,
+      "grad_norm": 0.70779949426651,
+      "learning_rate": 2.351804455781655e-06,
+      "loss": 0.0119,
+      "step": 7200
+    },
+    {
+      "epoch": 1.839755039550906,
+      "grad_norm": 0.6416905522346497,
+      "learning_rate": 2.343110458867364e-06,
+      "loss": 0.012,
+      "step": 7210
+    },
+    {
+      "epoch": 1.8423067108956368,
+      "grad_norm": 0.5262870192527771,
+      "learning_rate": 2.3344222527501267e-06,
+      "loss": 0.0144,
+      "step": 7220
+    },
+    {
+      "epoch": 1.8448583822403675,
+      "grad_norm": 0.3885171711444855,
+      "learning_rate": 2.3257399140206547e-06,
+      "loss": 0.0142,
+      "step": 7230
+    },
+    {
+      "epoch": 1.8474100535850981,
+      "grad_norm": 0.6644590497016907,
+      "learning_rate": 2.3170635192179396e-06,
+      "loss": 0.0096,
+      "step": 7240
+    },
+    {
+      "epoch": 1.849961724929829,
+      "grad_norm": 0.5656073689460754,
+      "learning_rate": 2.308393144828571e-06,
+      "loss": 0.0111,
+      "step": 7250
+    },
+    {
+      "epoch": 1.85251339627456,
+      "grad_norm": 0.26087692379951477,
+      "learning_rate": 2.2997288672860663e-06,
+      "loss": 0.0101,
+      "step": 7260
+    },
+    {
+      "epoch": 1.8550650676192908,
+      "grad_norm": 0.6924827098846436,
+      "learning_rate": 2.2910707629701975e-06,
+      "loss": 0.013,
+      "step": 7270
+    },
+    {
+      "epoch": 1.8576167389640215,
+      "grad_norm": 0.7471089363098145,
+      "learning_rate": 2.282418908206315e-06,
+      "loss": 0.0098,
+      "step": 7280
+    },
+    {
+      "epoch": 1.8601684103087521,
+      "grad_norm": 1.0280992984771729,
+      "learning_rate": 2.2737733792646773e-06,
+      "loss": 0.014,
+      "step": 7290
+    },
+    {
+      "epoch": 1.862720081653483,
+      "grad_norm": 0.46094390749931335,
+      "learning_rate": 2.2651342523597784e-06,
+      "loss": 0.012,
+      "step": 7300
+    },
+    {
+      "epoch": 1.865271752998214,
+      "grad_norm": 0.7856783866882324,
+      "learning_rate": 2.2565016036496746e-06,
+      "loss": 0.0103,
+      "step": 7310
+    },
+    {
+      "epoch": 1.8678234243429448,
+      "grad_norm": 0.6690205931663513,
+      "learning_rate": 2.247875509235314e-06,
+      "loss": 0.0101,
+      "step": 7320
+    },
+    {
+      "epoch": 1.8703750956876755,
+      "grad_norm": 0.8430372476577759,
+      "learning_rate": 2.2392560451598635e-06,
+      "loss": 0.0143,
+      "step": 7330
+    },
+    {
+      "epoch": 1.8729267670324061,
+      "grad_norm": 0.931174635887146,
+      "learning_rate": 2.2306432874080436e-06,
+      "loss": 0.0133,
+      "step": 7340
+    },
+    {
+      "epoch": 1.875478438377137,
+      "grad_norm": 0.2438133805990219,
+      "learning_rate": 2.2220373119054524e-06,
+      "loss": 0.0125,
+      "step": 7350
+    },
+    {
+      "epoch": 1.878030109721868,
+      "grad_norm": 0.7844971418380737,
+      "learning_rate": 2.2134381945179023e-06,
+      "loss": 0.0119,
+      "step": 7360
+    },
+    {
+      "epoch": 1.8805817810665986,
+      "grad_norm": 0.3950898349285126,
+      "learning_rate": 2.204846011050746e-06,
+      "loss": 0.0136,
+      "step": 7370
+    },
+    {
+      "epoch": 1.8831334524113295,
+      "grad_norm": 0.7956571578979492,
+      "learning_rate": 2.1962608372482105e-06,
+      "loss": 0.0107,
+      "step": 7380
+    },
+    {
+      "epoch": 1.8856851237560601,
+      "grad_norm": 0.3765330910682678,
+      "learning_rate": 2.187682748792729e-06,
+      "loss": 0.0094,
+      "step": 7390
+    },
+    {
+      "epoch": 1.888236795100791,
+      "grad_norm": 0.5307440161705017,
+      "learning_rate": 2.1791118213042754e-06,
+      "loss": 0.0103,
+      "step": 7400
+    },
+    {
+      "epoch": 1.890788466445522,
+      "grad_norm": 0.5232400298118591,
+      "learning_rate": 2.1705481303396934e-06,
+      "loss": 0.0122,
+      "step": 7410
+    },
+    {
+      "epoch": 1.8933401377902526,
+      "grad_norm": 0.9073548913002014,
+      "learning_rate": 2.161991751392038e-06,
+      "loss": 0.0108,
+      "step": 7420
+    },
+    {
+      "epoch": 1.8958918091349835,
+      "grad_norm": 0.33009836077690125,
+      "learning_rate": 2.1534427598899006e-06,
+      "loss": 0.0112,
+      "step": 7430
+    },
+    {
+      "epoch": 1.8984434804797141,
+      "grad_norm": 0.528856635093689,
+      "learning_rate": 2.1449012311967517e-06,
+      "loss": 0.0103,
+      "step": 7440
+    },
+    {
+      "epoch": 1.900995151824445,
+      "grad_norm": 0.3639775514602661,
+      "learning_rate": 2.136367240610271e-06,
+      "loss": 0.0105,
+      "step": 7450
+    },
+    {
+      "epoch": 1.903546823169176,
+      "grad_norm": 0.4201386570930481,
+      "learning_rate": 2.1278408633616884e-06,
+      "loss": 0.0123,
+      "step": 7460
+    },
+    {
+      "epoch": 1.9060984945139066,
+      "grad_norm": 0.47134822607040405,
+      "learning_rate": 2.1193221746151174e-06,
+      "loss": 0.0109,
+      "step": 7470
+    },
+    {
+      "epoch": 1.9086501658586374,
+      "grad_norm": 1.3191474676132202,
+      "learning_rate": 2.1108112494668945e-06,
+      "loss": 0.0104,
+      "step": 7480
+    },
+    {
+      "epoch": 1.9112018372033681,
+      "grad_norm": 0.27142035961151123,
+      "learning_rate": 2.1023081629449156e-06,
+      "loss": 0.0097,
+      "step": 7490
+    },
+    {
+      "epoch": 1.913753508548099,
+      "grad_norm": 0.6145235300064087,
+      "learning_rate": 2.0938129900079753e-06,
+      "loss": 0.0104,
+      "step": 7500
+    },
+    {
+      "epoch": 1.91630517989283,
+      "grad_norm": 1.2480336427688599,
+      "learning_rate": 2.0853258055451077e-06,
+      "loss": 0.0126,
+      "step": 7510
+    },
+    {
+      "epoch": 1.9188568512375606,
+      "grad_norm": 1.0720759630203247,
+      "learning_rate": 2.0768466843749215e-06,
+      "loss": 0.0137,
+      "step": 7520
+    },
+    {
+      "epoch": 1.9214085225822914,
+      "grad_norm": 0.550995409488678,
+      "learning_rate": 2.068375701244946e-06,
+      "loss": 0.011,
+      "step": 7530
+    },
+    {
+      "epoch": 1.923960193927022,
+      "grad_norm": 0.5243539810180664,
+      "learning_rate": 2.0599129308309695e-06,
+      "loss": 0.0113,
+      "step": 7540
+    },
+    {
+      "epoch": 1.926511865271753,
+      "grad_norm": 0.43778249621391296,
+      "learning_rate": 2.051458447736382e-06,
+      "loss": 0.0149,
+      "step": 7550
+    },
+    {
+      "epoch": 1.9290635366164839,
+      "grad_norm": 0.5081802010536194,
+      "learning_rate": 2.0430123264915144e-06,
+      "loss": 0.011,
+      "step": 7560
+    },
+    {
+      "epoch": 1.9316152079612146,
+      "grad_norm": 0.6080982089042664,
+      "learning_rate": 2.034574641552987e-06,
+      "loss": 0.0133,
+      "step": 7570
+    },
+    {
+      "epoch": 1.9341668793059454,
+      "grad_norm": 0.5425201058387756,
+      "learning_rate": 2.0261454673030458e-06,
+      "loss": 0.013,
+      "step": 7580
+    },
+    {
+      "epoch": 1.936718550650676,
+      "grad_norm": 0.2850346863269806,
+      "learning_rate": 2.0177248780489145e-06,
+      "loss": 0.0116,
+      "step": 7590
+    },
+    {
+      "epoch": 1.939270221995407,
+      "grad_norm": 0.9347357749938965,
+      "learning_rate": 2.009312948022136e-06,
+      "loss": 0.0105,
+      "step": 7600
+    },
+    {
+      "epoch": 1.9418218933401379,
+      "grad_norm": 0.6447896957397461,
+      "learning_rate": 2.000909751377916e-06,
+      "loss": 0.0111,
+      "step": 7610
+    },
+    {
+      "epoch": 1.9443735646848685,
+      "grad_norm": 0.8438960313796997,
+      "learning_rate": 1.9925153621944724e-06,
+      "loss": 0.012,
+      "step": 7620
+    },
+    {
+      "epoch": 1.9469252360295994,
+      "grad_norm": 0.48641327023506165,
+      "learning_rate": 1.9841298544723804e-06,
+      "loss": 0.0112,
+      "step": 7630
+    },
+    {
+      "epoch": 1.94947690737433,
+      "grad_norm": 0.8980324268341064,
+      "learning_rate": 1.9757533021339214e-06,
+      "loss": 0.01,
+      "step": 7640
+    },
+    {
+      "epoch": 1.952028578719061,
+      "grad_norm": 0.6476429104804993,
+      "learning_rate": 1.96738577902243e-06,
+      "loss": 0.0125,
+      "step": 7650
+    },
+    {
+      "epoch": 1.9545802500637919,
+      "grad_norm": 0.34102582931518555,
+      "learning_rate": 1.9590273589016455e-06,
+      "loss": 0.0111,
+      "step": 7660
+    },
+    {
+      "epoch": 1.9571319214085225,
+      "grad_norm": 0.42341795563697815,
+      "learning_rate": 1.950678115455059e-06,
+      "loss": 0.01,
+      "step": 7670
+    },
+    {
+      "epoch": 1.9596835927532534,
+      "grad_norm": 0.6199698448181152,
+      "learning_rate": 1.9423381222852638e-06,
+      "loss": 0.0089,
+      "step": 7680
+    },
+    {
+      "epoch": 1.962235264097984,
+      "grad_norm": 1.1667146682739258,
+      "learning_rate": 1.93400745291331e-06,
+      "loss": 0.0144,
+      "step": 7690
+    },
+    {
+      "epoch": 1.964786935442715,
+      "grad_norm": 0.36817818880081177,
+      "learning_rate": 1.9256861807780505e-06,
+      "loss": 0.0111,
+      "step": 7700
+    },
+    {
+      "epoch": 1.9673386067874459,
+      "grad_norm": 0.34932783246040344,
+      "learning_rate": 1.9173743792354984e-06,
+      "loss": 0.0128,
+      "step": 7710
+    },
+    {
+      "epoch": 1.9698902781321765,
+      "grad_norm": 1.2853580713272095,
+      "learning_rate": 1.9090721215581833e-06,
+      "loss": 0.0123,
+      "step": 7720
+    },
+    {
+      "epoch": 1.9724419494769072,
+      "grad_norm": 0.9818376302719116,
+      "learning_rate": 1.9007794809344937e-06,
+      "loss": 0.0107,
+      "step": 7730
+    },
+    {
+      "epoch": 1.974993620821638,
+      "grad_norm": 0.5679308176040649,
+      "learning_rate": 1.892496530468045e-06,
+      "loss": 0.0111,
+      "step": 7740
+    },
+    {
+      "epoch": 1.977545292166369,
+      "grad_norm": 0.4230453073978424,
+      "learning_rate": 1.8842233431770254e-06,
+      "loss": 0.0113,
+      "step": 7750
+    },
+    {
+      "epoch": 1.9800969635110999,
+      "grad_norm": 0.7522900700569153,
+      "learning_rate": 1.8759599919935588e-06,
+      "loss": 0.0129,
+      "step": 7760
+    },
+    {
+      "epoch": 1.9826486348558305,
+      "grad_norm": 0.9982630610466003,
+      "learning_rate": 1.8677065497630554e-06,
+      "loss": 0.0101,
+      "step": 7770
+    },
+    {
+      "epoch": 1.9852003062005612,
+      "grad_norm": 0.9997731447219849,
+      "learning_rate": 1.8594630892435793e-06,
+      "loss": 0.0098,
+      "step": 7780
+    },
+    {
+      "epoch": 1.987751977545292,
+      "grad_norm": 0.8870327472686768,
+      "learning_rate": 1.851229683105195e-06,
+      "loss": 0.0107,
+      "step": 7790
+    },
+    {
+      "epoch": 1.990303648890023,
+      "grad_norm": 0.8995794057846069,
+      "learning_rate": 1.8430064039293358e-06,
+      "loss": 0.0109,
+      "step": 7800
+    },
+    {
+      "epoch": 1.9928553202347539,
+      "grad_norm": 0.4320782721042633,
+      "learning_rate": 1.8347933242081597e-06,
+      "loss": 0.0124,
+      "step": 7810
+    },
+    {
+      "epoch": 1.9954069915794845,
+      "grad_norm": 0.7484515309333801,
+      "learning_rate": 1.8265905163439137e-06,
+      "loss": 0.0104,
+      "step": 7820
+    },
+    {
+      "epoch": 1.9979586629242152,
+      "grad_norm": 0.7955275177955627,
+      "learning_rate": 1.8183980526482893e-06,
+      "loss": 0.0126,
+      "step": 7830
+    },
+    {
+      "epoch": 2.000510334268946,
+      "grad_norm": 0.43744155764579773,
+      "learning_rate": 1.8102160053417954e-06,
+      "loss": 0.0105,
+      "step": 7840
+    },
+    {
+      "epoch": 2.003062005613677,
+      "grad_norm": 0.35822105407714844,
+      "learning_rate": 1.8020444465531095e-06,
+      "loss": 0.0091,
+      "step": 7850
+    },
+    {
+      "epoch": 2.005613676958408,
+      "grad_norm": 0.23200540244579315,
+      "learning_rate": 1.793883448318451e-06,
+      "loss": 0.0105,
+      "step": 7860
+    },
+    {
+      "epoch": 2.0081653483031388,
+      "grad_norm": 0.8042770028114319,
+      "learning_rate": 1.7857330825809412e-06,
+      "loss": 0.0094,
+      "step": 7870
+    },
+    {
+      "epoch": 2.010717019647869,
+      "grad_norm": 0.8214767575263977,
+      "learning_rate": 1.7775934211899733e-06,
+      "loss": 0.0101,
+      "step": 7880
+    },
+    {
+      "epoch": 2.0132686909926,
+      "grad_norm": 0.41219642758369446,
+      "learning_rate": 1.7694645359005726e-06,
+      "loss": 0.01,
+      "step": 7890
+    },
+    {
+      "epoch": 2.015820362337331,
+      "grad_norm": 0.846447229385376,
+      "learning_rate": 1.7613464983727727e-06,
+      "loss": 0.0105,
+      "step": 7900
+    },
+    {
+      "epoch": 2.018372033682062,
+      "grad_norm": 0.4125836193561554,
+      "learning_rate": 1.753239380170976e-06,
+      "loss": 0.0098,
+      "step": 7910
+    },
+    {
+      "epoch": 2.0209237050267927,
+      "grad_norm": 0.33059266209602356,
+      "learning_rate": 1.7451432527633247e-06,
+      "loss": 0.01,
+      "step": 7920
+    },
+    {
+      "epoch": 2.023475376371523,
+      "grad_norm": 0.26018354296684265,
+      "learning_rate": 1.737058187521076e-06,
+      "loss": 0.0102,
+      "step": 7930
+    },
+    {
+      "epoch": 2.026027047716254,
+      "grad_norm": 0.9150725603103638,
+      "learning_rate": 1.7289842557179647e-06,
+      "loss": 0.0085,
+      "step": 7940
+    },
+    {
+      "epoch": 2.028578719060985,
+      "grad_norm": 0.6766268610954285,
+      "learning_rate": 1.7209215285295795e-06,
+      "loss": 0.0076,
+      "step": 7950
+    },
+    {
+      "epoch": 2.031130390405716,
+      "grad_norm": 0.7343044877052307,
+      "learning_rate": 1.7128700770327373e-06,
+      "loss": 0.0088,
+      "step": 7960
+    },
+    {
+      "epoch": 2.0336820617504467,
+      "grad_norm": 0.3323941230773926,
+      "learning_rate": 1.7048299722048546e-06,
+      "loss": 0.0087,
+      "step": 7970
+    },
+    {
+      "epoch": 2.036233733095177,
+      "grad_norm": 0.16270656883716583,
+      "learning_rate": 1.6968012849233174e-06,
+      "loss": 0.0068,
+      "step": 7980
+    },
+    {
+      "epoch": 2.038785404439908,
+      "grad_norm": 0.46309563517570496,
+      "learning_rate": 1.6887840859648656e-06,
+      "loss": 0.0093,
+      "step": 7990
+    },
+    {
+      "epoch": 2.041337075784639,
+      "grad_norm": 0.5606943368911743,
+      "learning_rate": 1.680778446004961e-06,
+      "loss": 0.0071,
+      "step": 8000
+    },
+    {
+      "epoch": 2.04388874712937,
+      "grad_norm": 0.6630533337593079,
+      "learning_rate": 1.672784435617166e-06,
+      "loss": 0.009,
+      "step": 8010
+    },
+    {
+      "epoch": 2.0464404184741007,
+      "grad_norm": 0.2969760596752167,
+      "learning_rate": 1.6648021252725266e-06,
+      "loss": 0.0073,
+      "step": 8020
+    },
+    {
+      "epoch": 2.048992089818831,
+      "grad_norm": 0.30385419726371765,
+      "learning_rate": 1.6568315853389458e-06,
+      "loss": 0.0115,
+      "step": 8030
+    },
+    {
+      "epoch": 2.051543761163562,
+      "grad_norm": 1.2547024488449097,
+      "learning_rate": 1.6488728860805627e-06,
+      "loss": 0.0089,
+      "step": 8040
+    },
+    {
+      "epoch": 2.054095432508293,
+      "grad_norm": 1.2566521167755127,
+      "learning_rate": 1.6409260976571383e-06,
+      "loss": 0.0099,
+      "step": 8050
+    },
+    {
+      "epoch": 2.056647103853024,
+      "grad_norm": 0.28552380204200745,
+      "learning_rate": 1.6329912901234318e-06,
+      "loss": 0.0088,
+      "step": 8060
+    },
+    {
+      "epoch": 2.0591987751977547,
+      "grad_norm": 0.6307502388954163,
+      "learning_rate": 1.6250685334285845e-06,
+      "loss": 0.0111,
+      "step": 8070
+    },
+    {
+      "epoch": 2.061750446542485,
+      "grad_norm": 0.7623242139816284,
+      "learning_rate": 1.6171578974155057e-06,
+      "loss": 0.0074,
+      "step": 8080
+    },
+    {
+      "epoch": 2.064302117887216,
+      "grad_norm": 0.3438488841056824,
+      "learning_rate": 1.6092594518202554e-06,
+      "loss": 0.0094,
+      "step": 8090
+    },
+    {
+      "epoch": 2.066853789231947,
+      "grad_norm": 0.6200628280639648,
+      "learning_rate": 1.6013732662714267e-06,
+      "loss": 0.0088,
+      "step": 8100
+    },
+    {
+      "epoch": 2.069405460576678,
+      "grad_norm": 0.407311350107193,
+      "learning_rate": 1.5934994102895367e-06,
+      "loss": 0.0077,
+      "step": 8110
+    },
+    {
+      "epoch": 2.0719571319214087,
+      "grad_norm": 0.7597397565841675,
+      "learning_rate": 1.5856379532864109e-06,
+      "loss": 0.0096,
+      "step": 8120
+    },
+    {
+      "epoch": 2.074508803266139,
+      "grad_norm": 0.30682581663131714,
+      "learning_rate": 1.577788964564568e-06,
+      "loss": 0.0072,
+      "step": 8130
+    },
+    {
+      "epoch": 2.07706047461087,
+      "grad_norm": 0.4774474501609802,
+      "learning_rate": 1.5699525133166222e-06,
+      "loss": 0.0075,
+      "step": 8140
+    },
+    {
+      "epoch": 2.079612145955601,
+      "grad_norm": 0.3659365773200989,
+      "learning_rate": 1.5621286686246547e-06,
+      "loss": 0.0108,
+      "step": 8150
+    },
+    {
+      "epoch": 2.082163817300332,
+      "grad_norm": 0.8173565864562988,
+      "learning_rate": 1.5543174994596167e-06,
+      "loss": 0.0096,
+      "step": 8160
+    },
+    {
+      "epoch": 2.0847154886450627,
+      "grad_norm": 0.5068744421005249,
+      "learning_rate": 1.5465190746807209e-06,
+      "loss": 0.0101,
+      "step": 8170
+    },
+    {
+      "epoch": 2.087267159989793,
+      "grad_norm": 1.1201165914535522,
+      "learning_rate": 1.5387334630348289e-06,
+      "loss": 0.0098,
+      "step": 8180
+    },
+    {
+      "epoch": 2.089818831334524,
+      "grad_norm": 0.5641067028045654,
+      "learning_rate": 1.5309607331558481e-06,
+      "loss": 0.0092,
+      "step": 8190
+    },
+    {
+      "epoch": 2.092370502679255,
+      "grad_norm": 0.6286213397979736,
+      "learning_rate": 1.5232009535641325e-06,
+      "loss": 0.0111,
+      "step": 8200
+    },
+    {
+      "epoch": 2.094922174023986,
+      "grad_norm": 0.3738521337509155,
+      "learning_rate": 1.5154541926658668e-06,
+      "loss": 0.008,
+      "step": 8210
+    },
+    {
+      "epoch": 2.0974738453687163,
+      "grad_norm": 0.9565776586532593,
+      "learning_rate": 1.507720518752471e-06,
+      "loss": 0.0084,
+      "step": 8220
+    },
+    {
+      "epoch": 2.100025516713447,
+      "grad_norm": 0.4144324064254761,
+      "learning_rate": 1.5000000000000007e-06,
+      "loss": 0.008,
+      "step": 8230
+    },
+    {
+      "epoch": 2.102577188058178,
+      "grad_norm": 0.4653526842594147,
+      "learning_rate": 1.4922927044685366e-06,
+      "loss": 0.0106,
+      "step": 8240
+    },
+    {
+      "epoch": 2.105128859402909,
+      "grad_norm": 0.2316306233406067,
+      "learning_rate": 1.4845987001015956e-06,
+      "loss": 0.0075,
+      "step": 8250
+    },
+    {
+      "epoch": 2.10768053074764,
+      "grad_norm": 0.2574253976345062,
+      "learning_rate": 1.4769180547255246e-06,
+      "loss": 0.01,
+      "step": 8260
+    },
+    {
+      "epoch": 2.1102322020923703,
+      "grad_norm": 0.5861005187034607,
+      "learning_rate": 1.469250836048904e-06,
+      "loss": 0.0087,
+      "step": 8270
+    },
+    {
+      "epoch": 2.112783873437101,
+      "grad_norm": 0.5676212310791016,
+      "learning_rate": 1.4615971116619504e-06,
+      "loss": 0.0097,
+      "step": 8280
+    },
+    {
+      "epoch": 2.115335544781832,
+      "grad_norm": 0.7331516146659851,
+      "learning_rate": 1.453956949035926e-06,
+      "loss": 0.0073,
+      "step": 8290
+    },
+    {
+      "epoch": 2.117887216126563,
+      "grad_norm": 0.6896494626998901,
+      "learning_rate": 1.446330415522534e-06,
+      "loss": 0.0083,
+      "step": 8300
+    },
+    {
+      "epoch": 2.120438887471294,
+      "grad_norm": 0.4225230813026428,
+      "learning_rate": 1.438717578353334e-06,
+      "loss": 0.0081,
+      "step": 8310
+    },
+    {
+      "epoch": 2.1229905588160243,
+      "grad_norm": 0.5864181518554688,
+      "learning_rate": 1.4311185046391469e-06,
+      "loss": 0.0083,
+      "step": 8320
+    },
+    {
+      "epoch": 2.125542230160755,
+      "grad_norm": 0.4726142883300781,
+      "learning_rate": 1.4235332613694585e-06,
+      "loss": 0.0091,
+      "step": 8330
+    },
+    {
+      "epoch": 2.128093901505486,
+      "grad_norm": 0.4282633364200592,
+      "learning_rate": 1.4159619154118332e-06,
+      "loss": 0.0101,
+      "step": 8340
+    },
+    {
+      "epoch": 2.130645572850217,
+      "grad_norm": 0.44410398602485657,
+      "learning_rate": 1.408404533511327e-06,
+      "loss": 0.0102,
+      "step": 8350
+    },
+    {
+      "epoch": 2.133197244194948,
+      "grad_norm": 0.5199018716812134,
+      "learning_rate": 1.4008611822898924e-06,
+      "loss": 0.0119,
+      "step": 8360
+    },
+    {
+      "epoch": 2.1357489155396783,
+      "grad_norm": 0.37344205379486084,
+      "learning_rate": 1.3933319282457967e-06,
+      "loss": 0.0104,
+      "step": 8370
+    },
+    {
+      "epoch": 2.138300586884409,
+      "grad_norm": 0.36821675300598145,
+      "learning_rate": 1.3858168377530349e-06,
+      "loss": 0.0072,
+      "step": 8380
+    },
+    {
+      "epoch": 2.14085225822914,
+      "grad_norm": 0.8769086003303528,
+      "learning_rate": 1.3783159770607406e-06,
+      "loss": 0.0083,
+      "step": 8390
+    },
+    {
+      "epoch": 2.143403929573871,
+      "grad_norm": 0.5919402837753296,
+      "learning_rate": 1.370829412292605e-06,
+      "loss": 0.0103,
+      "step": 8400
+    },
+    {
+      "epoch": 2.145955600918602,
+      "grad_norm": 1.004891037940979,
+      "learning_rate": 1.363357209446298e-06,
+      "loss": 0.0107,
+      "step": 8410
+    },
+    {
+      "epoch": 2.1485072722633323,
+      "grad_norm": 0.25619760155677795,
+      "learning_rate": 1.3558994343928764e-06,
+      "loss": 0.009,
+      "step": 8420
+    },
+    {
+      "epoch": 2.151058943608063,
+      "grad_norm": 0.2878786623477936,
+      "learning_rate": 1.348456152876214e-06,
+      "loss": 0.008,
+      "step": 8430
+    },
+    {
+      "epoch": 2.153610614952794,
+      "grad_norm": 0.6479395627975464,
+      "learning_rate": 1.3410274305124155e-06,
+      "loss": 0.0097,
+      "step": 8440
+    },
+    {
+      "epoch": 2.156162286297525,
+      "grad_norm": 0.42222821712493896,
+      "learning_rate": 1.3336133327892383e-06,
+      "loss": 0.009,
+      "step": 8450
+    },
+    {
+      "epoch": 2.158713957642256,
+      "grad_norm": 0.5351365804672241,
+      "learning_rate": 1.3262139250655168e-06,
+      "loss": 0.0114,
+      "step": 8460
+    },
+    {
+      "epoch": 2.1612656289869863,
+      "grad_norm": 0.4197307825088501,
+      "learning_rate": 1.3188292725705888e-06,
+      "loss": 0.0077,
+      "step": 8470
+    },
+    {
+      "epoch": 2.163817300331717,
+      "grad_norm": 0.7340565919876099,
+      "learning_rate": 1.3114594404037136e-06,
+      "loss": 0.0066,
+      "step": 8480
+    },
+    {
+      "epoch": 2.166368971676448,
+      "grad_norm": 0.9122597575187683,
+      "learning_rate": 1.3041044935335052e-06,
+      "loss": 0.0107,
+      "step": 8490
+    },
+    {
+      "epoch": 2.168920643021179,
+      "grad_norm": 0.8086641430854797,
+      "learning_rate": 1.2967644967973563e-06,
+      "loss": 0.0104,
+      "step": 8500
+    },
+    {
+      "epoch": 2.17147231436591,
+      "grad_norm": 0.7990751266479492,
+      "learning_rate": 1.2894395149008647e-06,
+      "loss": 0.0089,
+      "step": 8510
+    },
+    {
+      "epoch": 2.1740239857106403,
+      "grad_norm": 0.4715738892555237,
+      "learning_rate": 1.2821296124172686e-06,
+      "loss": 0.0079,
+      "step": 8520
+    },
+    {
+      "epoch": 2.176575657055371,
+      "grad_norm": 0.8491727709770203,
+      "learning_rate": 1.2748348537868711e-06,
+      "loss": 0.0106,
+      "step": 8530
+    },
+    {
+      "epoch": 2.179127328400102,
+      "grad_norm": 0.2744673788547516,
+      "learning_rate": 1.2675553033164747e-06,
+      "loss": 0.009,
+      "step": 8540
+    },
+    {
+      "epoch": 2.181678999744833,
+      "grad_norm": 0.8372253775596619,
+      "learning_rate": 1.2602910251788165e-06,
+      "loss": 0.0091,
+      "step": 8550
+    },
+    {
+      "epoch": 2.184230671089564,
+      "grad_norm": 0.6588044762611389,
+      "learning_rate": 1.2530420834120021e-06,
+      "loss": 0.0081,
+      "step": 8560
+    },
+    {
+      "epoch": 2.1867823424342943,
+      "grad_norm": 0.642224133014679,
+      "learning_rate": 1.245808541918935e-06,
+      "loss": 0.008,
+      "step": 8570
+    },
+    {
+      "epoch": 2.189334013779025,
+      "grad_norm": 0.6353108286857605,
+      "learning_rate": 1.2385904644667618e-06,
+      "loss": 0.0076,
+      "step": 8580
+    },
+    {
+      "epoch": 2.191885685123756,
+      "grad_norm": 0.8236711025238037,
+      "learning_rate": 1.2313879146863048e-06,
+      "loss": 0.01,
+      "step": 8590
+    },
+    {
+      "epoch": 2.194437356468487,
+      "grad_norm": 1.0461316108703613,
+      "learning_rate": 1.2242009560715007e-06,
+      "loss": 0.0106,
+      "step": 8600
+    },
+    {
+      "epoch": 2.196989027813218,
+      "grad_norm": 0.7238239049911499,
+      "learning_rate": 1.2170296519788455e-06,
+      "loss": 0.0082,
+      "step": 8610
+    },
+    {
+      "epoch": 2.1995406991579483,
+      "grad_norm": 0.45300090312957764,
+      "learning_rate": 1.209874065626833e-06,
+      "loss": 0.0079,
+      "step": 8620
+    },
+    {
+      "epoch": 2.202092370502679,
+      "grad_norm": 0.46156367659568787,
+      "learning_rate": 1.2027342600953938e-06,
+      "loss": 0.009,
+      "step": 8630
+    },
+    {
+      "epoch": 2.20464404184741,
+      "grad_norm": 0.5707991719245911,
+      "learning_rate": 1.195610298325348e-06,
+      "loss": 0.0068,
+      "step": 8640
+    },
+    {
+      "epoch": 2.207195713192141,
+      "grad_norm": 0.6443004608154297,
+      "learning_rate": 1.188502243117842e-06,
+      "loss": 0.0079,
+      "step": 8650
+    },
+    {
+      "epoch": 2.209747384536872,
+      "grad_norm": 0.5062205791473389,
+      "learning_rate": 1.1814101571337979e-06,
+      "loss": 0.0081,
+      "step": 8660
+    },
+    {
+      "epoch": 2.2122990558816022,
+      "grad_norm": 0.4682200849056244,
+      "learning_rate": 1.1743341028933635e-06,
+      "loss": 0.0099,
+      "step": 8670
+    },
+    {
+      "epoch": 2.214850727226333,
+      "grad_norm": 0.9355303049087524,
+      "learning_rate": 1.1672741427753596e-06,
+      "loss": 0.0103,
+      "step": 8680
+    },
+    {
+      "epoch": 2.217402398571064,
+      "grad_norm": 0.9519639611244202,
+      "learning_rate": 1.1602303390167257e-06,
+      "loss": 0.0091,
+      "step": 8690
+    },
+    {
+      "epoch": 2.219954069915795,
+      "grad_norm": 0.9006462693214417,
+      "learning_rate": 1.1532027537119796e-06,
+      "loss": 0.0104,
+      "step": 8700
+    },
+    {
+      "epoch": 2.222505741260526,
+      "grad_norm": 0.8201411366462708,
+      "learning_rate": 1.1461914488126636e-06,
+      "loss": 0.0101,
+      "step": 8710
+    },
+    {
+      "epoch": 2.2250574126052562,
+      "grad_norm": 0.4761882722377777,
+      "learning_rate": 1.1391964861267974e-06,
+      "loss": 0.0092,
+      "step": 8720
+    },
+    {
+      "epoch": 2.227609083949987,
+      "grad_norm": 1.0870357751846313,
+      "learning_rate": 1.132217927318344e-06,
+      "loss": 0.0079,
+      "step": 8730
+    },
+    {
+      "epoch": 2.230160755294718,
+      "grad_norm": 0.6677237153053284,
+      "learning_rate": 1.1252558339066518e-06,
+      "loss": 0.0086,
+      "step": 8740
+    },
+    {
+      "epoch": 2.232712426639449,
+      "grad_norm": 0.578660786151886,
+      "learning_rate": 1.1183102672659192e-06,
+      "loss": 0.0094,
+      "step": 8750
+    },
+    {
+      "epoch": 2.23526409798418,
+      "grad_norm": 0.4101839065551758,
+      "learning_rate": 1.111381288624657e-06,
+      "loss": 0.0085,
+      "step": 8760
+    },
+    {
+      "epoch": 2.2378157693289102,
+      "grad_norm": 0.3004243075847626,
+      "learning_rate": 1.1044689590651418e-06,
+      "loss": 0.0069,
+      "step": 8770
+    },
+    {
+      "epoch": 2.240367440673641,
+      "grad_norm": 1.1306524276733398,
+      "learning_rate": 1.0975733395228784e-06,
+      "loss": 0.0094,
+      "step": 8780
+    },
+    {
+      "epoch": 2.242919112018372,
+      "grad_norm": 0.6862215399742126,
+      "learning_rate": 1.0906944907860723e-06,
+      "loss": 0.0079,
+      "step": 8790
+    },
+    {
+      "epoch": 2.245470783363103,
+      "grad_norm": 0.5847616791725159,
+      "learning_rate": 1.0838324734950796e-06,
+      "loss": 0.0069,
+      "step": 8800
+    },
+    {
+      "epoch": 2.248022454707834,
+      "grad_norm": 0.3303333818912506,
+      "learning_rate": 1.0769873481418796e-06,
+      "loss": 0.0077,
+      "step": 8810
+    },
+    {
+      "epoch": 2.2505741260525642,
+      "grad_norm": 0.25424349308013916,
+      "learning_rate": 1.0701591750695445e-06,
+      "loss": 0.008,
+      "step": 8820
+    },
+    {
+      "epoch": 2.253125797397295,
+      "grad_norm": 0.4860968589782715,
+      "learning_rate": 1.0633480144717018e-06,
+      "loss": 0.0083,
+      "step": 8830
+    },
+    {
+      "epoch": 2.255677468742026,
+      "grad_norm": 0.9478049874305725,
+      "learning_rate": 1.0565539263920034e-06,
+      "loss": 0.0086,
+      "step": 8840
+    },
+    {
+      "epoch": 2.258229140086757,
+      "grad_norm": 0.6831640005111694,
+      "learning_rate": 1.0497769707236058e-06,
+      "loss": 0.0067,
+      "step": 8850
+    },
+    {
+      "epoch": 2.260780811431488,
+      "grad_norm": 0.41167324781417847,
+      "learning_rate": 1.0430172072086287e-06,
+      "loss": 0.0087,
+      "step": 8860
+    },
+    {
+      "epoch": 2.2633324827762182,
+      "grad_norm": 0.43332257866859436,
+      "learning_rate": 1.0362746954376356e-06,
+      "loss": 0.0083,
+      "step": 8870
+    },
+    {
+      "epoch": 2.265884154120949,
+      "grad_norm": 0.5903699994087219,
+      "learning_rate": 1.0295494948491105e-06,
+      "loss": 0.0084,
+      "step": 8880
+    },
+    {
+      "epoch": 2.26843582546568,
+      "grad_norm": 0.8941869735717773,
+      "learning_rate": 1.0228416647289285e-06,
+      "loss": 0.0091,
+      "step": 8890
+    },
+    {
+      "epoch": 2.270987496810411,
+      "grad_norm": 0.24812312424182892,
+      "learning_rate": 1.0161512642098346e-06,
+      "loss": 0.0072,
+      "step": 8900
+    },
+    {
+      "epoch": 2.273539168155142,
+      "grad_norm": 1.322016954421997,
+      "learning_rate": 1.0094783522709284e-06,
+      "loss": 0.01,
+      "step": 8910
+    },
+    {
+      "epoch": 2.2760908394998722,
+      "grad_norm": 0.4932226538658142,
+      "learning_rate": 1.0028229877371353e-06,
+      "loss": 0.008,
+      "step": 8920
+    },
+    {
+      "epoch": 2.278642510844603,
+      "grad_norm": 1.0810902118682861,
+      "learning_rate": 9.961852292786908e-07,
+      "loss": 0.0103,
+      "step": 8930
+    },
+    {
+      "epoch": 2.281194182189334,
+      "grad_norm": 0.3352721929550171,
+      "learning_rate": 9.895651354106285e-07,
+      "loss": 0.0078,
+      "step": 8940
+    },
+    {
+      "epoch": 2.283745853534065,
+      "grad_norm": 0.4848918914794922,
+      "learning_rate": 9.829627644922561e-07,
+      "loss": 0.0081,
+      "step": 8950
+    },
+    {
+      "epoch": 2.286297524878796,
+      "grad_norm": 0.585834801197052,
+      "learning_rate": 9.763781747266479e-07,
+      "loss": 0.0092,
+      "step": 8960
+    },
+    {
+      "epoch": 2.2888491962235262,
+      "grad_norm": 0.7620885372161865,
+      "learning_rate": 9.698114241601284e-07,
+      "loss": 0.0084,
+      "step": 8970
+    },
+    {
+      "epoch": 2.291400867568257,
+      "grad_norm": 0.7390540242195129,
+      "learning_rate": 9.632625706817594e-07,
+      "loss": 0.0098,
+      "step": 8980
+    },
+    {
+      "epoch": 2.293952538912988,
+      "grad_norm": 0.43986302614212036,
+      "learning_rate": 9.56731672022831e-07,
+      "loss": 0.0075,
+      "step": 8990
+    },
+    {
+      "epoch": 2.296504210257719,
+      "grad_norm": 0.8151856064796448,
+      "learning_rate": 9.502187857563568e-07,
+      "loss": 0.0077,
+      "step": 9000
+    },
+    {
+      "epoch": 2.29905588160245,
+      "grad_norm": 0.7944412231445312,
+      "learning_rate": 9.437239692965569e-07,
+      "loss": 0.0071,
+      "step": 9010
+    },
+    {
+      "epoch": 2.3016075529471802,
+      "grad_norm": 0.6182528734207153,
+      "learning_rate": 9.372472798983622e-07,
+      "loss": 0.0069,
+      "step": 9020
+    },
+    {
+      "epoch": 2.304159224291911,
+      "grad_norm": 0.47329604625701904,
+      "learning_rate": 9.307887746569027e-07,
+      "loss": 0.0115,
+      "step": 9030
+    },
+    {
+      "epoch": 2.306710895636642,
+      "grad_norm": 0.6696287393569946,
+      "learning_rate": 9.243485105070069e-07,
+      "loss": 0.0079,
+      "step": 9040
+    },
+    {
+      "epoch": 2.309262566981373,
+      "grad_norm": 0.7252456545829773,
+      "learning_rate": 9.179265442226967e-07,
+      "loss": 0.0085,
+      "step": 9050
+    },
+    {
+      "epoch": 2.3118142383261038,
+      "grad_norm": 0.405748188495636,
+      "learning_rate": 9.115229324166945e-07,
+      "loss": 0.0086,
+      "step": 9060
+    },
+    {
+      "epoch": 2.314365909670834,
+      "grad_norm": 0.2414780706167221,
+      "learning_rate": 9.051377315399145e-07,
+      "loss": 0.0071,
+      "step": 9070
+    },
+    {
+      "epoch": 2.316917581015565,
+      "grad_norm": 0.9178037643432617,
+      "learning_rate": 8.98770997880972e-07,
+      "loss": 0.008,
+      "step": 9080
+    },
+    {
+      "epoch": 2.319469252360296,
+      "grad_norm": 0.2835039496421814,
+      "learning_rate": 8.924227875656858e-07,
+      "loss": 0.0073,
+      "step": 9090
+    },
+    {
+      "epoch": 2.322020923705027,
+      "grad_norm": 1.6318212747573853,
+      "learning_rate": 8.860931565565804e-07,
+      "loss": 0.0078,
+      "step": 9100
+    },
+    {
+      "epoch": 2.3245725950497578,
+      "grad_norm": 0.8543625473976135,
+      "learning_rate": 8.797821606523942e-07,
+      "loss": 0.007,
+      "step": 9110
+    },
+    {
+      "epoch": 2.327124266394488,
+      "grad_norm": 0.4498523771762848,
+      "learning_rate": 8.734898554875923e-07,
+      "loss": 0.0083,
+      "step": 9120
+    },
+    {
+      "epoch": 2.329675937739219,
+      "grad_norm": 0.36660337448120117,
+      "learning_rate": 8.672162965318668e-07,
+      "loss": 0.008,
+      "step": 9130
+    },
+    {
+      "epoch": 2.33222760908395,
+      "grad_norm": 0.6560646891593933,
+      "learning_rate": 8.609615390896566e-07,
+      "loss": 0.0096,
+      "step": 9140
+    },
+    {
+      "epoch": 2.334779280428681,
+      "grad_norm": 0.532968282699585,
+      "learning_rate": 8.547256382996558e-07,
+      "loss": 0.0095,
+      "step": 9150
+    },
+    {
+      "epoch": 2.3373309517734118,
+      "grad_norm": 1.1277568340301514,
+      "learning_rate": 8.485086491343253e-07,
+      "loss": 0.0082,
+      "step": 9160
+    },
+    {
+      "epoch": 2.339882623118142,
+      "grad_norm": 0.6013373732566833,
+      "learning_rate": 8.42310626399415e-07,
+      "loss": 0.0096,
+      "step": 9170
+    },
+    {
+      "epoch": 2.342434294462873,
+      "grad_norm": 0.4544149339199066,
+      "learning_rate": 8.361316247334734e-07,
+      "loss": 0.0078,
+      "step": 9180
+    },
+    {
+      "epoch": 2.344985965807604,
+      "grad_norm": 0.3007356822490692,
+      "learning_rate": 8.299716986073698e-07,
+      "loss": 0.0071,
+      "step": 9190
+    },
+    {
+      "epoch": 2.347537637152335,
+      "grad_norm": 0.275519996881485,
+      "learning_rate": 8.238309023238144e-07,
+      "loss": 0.0107,
+      "step": 9200
+    },
+    {
+      "epoch": 2.3500893084970658,
+      "grad_norm": 0.516834020614624,
+      "learning_rate": 8.177092900168793e-07,
+      "loss": 0.0098,
+      "step": 9210
+    },
+    {
+      "epoch": 2.352640979841796,
+      "grad_norm": 0.6302925944328308,
+      "learning_rate": 8.116069156515177e-07,
+      "loss": 0.008,
+      "step": 9220
+    },
+    {
+      "epoch": 2.355192651186527,
+      "grad_norm": 0.2730187475681305,
+      "learning_rate": 8.055238330230943e-07,
+      "loss": 0.0087,
+      "step": 9230
+    },
+    {
+      "epoch": 2.357744322531258,
+      "grad_norm": 0.948367714881897,
+      "learning_rate": 7.994600957569057e-07,
+      "loss": 0.0084,
+      "step": 9240
+    },
+    {
+      "epoch": 2.360295993875989,
+      "grad_norm": 0.4010712504386902,
+      "learning_rate": 7.934157573077096e-07,
+      "loss": 0.0083,
+      "step": 9250
+    },
+    {
+      "epoch": 2.3628476652207198,
+      "grad_norm": 0.5648771524429321,
+      "learning_rate": 7.873908709592559e-07,
+      "loss": 0.0071,
+      "step": 9260
+    },
+    {
+      "epoch": 2.36539933656545,
+      "grad_norm": 1.0915343761444092,
+      "learning_rate": 7.813854898238146e-07,
+      "loss": 0.008,
+      "step": 9270
+    },
+    {
+      "epoch": 2.367951007910181,
+      "grad_norm": 0.3447641134262085,
+      "learning_rate": 7.753996668417053e-07,
+      "loss": 0.0057,
+      "step": 9280
+    },
+    {
+      "epoch": 2.370502679254912,
+      "grad_norm": 0.5241166949272156,
+      "learning_rate": 7.694334547808362e-07,
+      "loss": 0.0072,
+      "step": 9290
+    },
+    {
+      "epoch": 2.373054350599643,
+      "grad_norm": 0.875315248966217,
+      "learning_rate": 7.63486906236234e-07,
+      "loss": 0.0072,
+      "step": 9300
+    },
+    {
+      "epoch": 2.3756060219443738,
+      "grad_norm": 0.6171951293945312,
+      "learning_rate": 7.57560073629581e-07,
+      "loss": 0.0085,
+      "step": 9310
+    },
+    {
+      "epoch": 2.378157693289104,
+      "grad_norm": 0.4204012155532837,
+      "learning_rate": 7.51653009208757e-07,
+      "loss": 0.0057,
+      "step": 9320
+    },
+    {
+      "epoch": 2.380709364633835,
+      "grad_norm": 0.4826784133911133,
+      "learning_rate": 7.45765765047375e-07,
+      "loss": 0.0075,
+      "step": 9330
+    },
+    {
+      "epoch": 2.383261035978566,
+      "grad_norm": 0.6717602610588074,
+      "learning_rate": 7.398983930443199e-07,
+      "loss": 0.0076,
+      "step": 9340
+    },
+    {
+      "epoch": 2.385812707323297,
+      "grad_norm": 0.3673340082168579,
+      "learning_rate": 7.340509449232983e-07,
+      "loss": 0.0057,
+      "step": 9350
+    },
+    {
+      "epoch": 2.3883643786680278,
+      "grad_norm": 0.5914478898048401,
+      "learning_rate": 7.282234722323755e-07,
+      "loss": 0.0086,
+      "step": 9360
+    },
+    {
+      "epoch": 2.390916050012758,
+      "grad_norm": 0.41728970408439636,
+      "learning_rate": 7.224160263435226e-07,
+      "loss": 0.0092,
+      "step": 9370
+    },
+    {
+      "epoch": 2.393467721357489,
+      "grad_norm": 0.5362657308578491,
+      "learning_rate": 7.166286584521685e-07,
+      "loss": 0.0076,
+      "step": 9380
+    },
+    {
+      "epoch": 2.39601939270222,
+      "grad_norm": 0.5463047027587891,
+      "learning_rate": 7.108614195767432e-07,
+      "loss": 0.0094,
+      "step": 9390
+    },
+    {
+      "epoch": 2.398571064046951,
+      "grad_norm": 0.33496329188346863,
+      "learning_rate": 7.051143605582279e-07,
+      "loss": 0.0087,
+      "step": 9400
+    },
+    {
+      "epoch": 2.4011227353916818,
+      "grad_norm": 1.052742838859558,
+      "learning_rate": 6.99387532059712e-07,
+      "loss": 0.0087,
+      "step": 9410
+    },
+    {
+      "epoch": 2.403674406736412,
+      "grad_norm": 1.4576879739761353,
+      "learning_rate": 6.936809845659409e-07,
+      "loss": 0.0088,
+      "step": 9420
+    },
+    {
+      "epoch": 2.406226078081143,
+      "grad_norm": 0.5559976696968079,
+      "learning_rate": 6.879947683828726e-07,
+      "loss": 0.008,
+      "step": 9430
+    },
+    {
+      "epoch": 2.408777749425874,
+      "grad_norm": 0.45263102650642395,
+      "learning_rate": 6.823289336372369e-07,
+      "loss": 0.0083,
+      "step": 9440
+    },
+    {
+      "epoch": 2.411329420770605,
+      "grad_norm": 0.7428113222122192,
+      "learning_rate": 6.766835302760911e-07,
+      "loss": 0.0077,
+      "step": 9450
+    },
+    {
+      "epoch": 2.4138810921153357,
+      "grad_norm": 0.30585309863090515,
+      "learning_rate": 6.710586080663781e-07,
+      "loss": 0.0073,
+      "step": 9460
+    },
+    {
+      "epoch": 2.416432763460066,
+      "grad_norm": 0.5510261654853821,
+      "learning_rate": 6.654542165944923e-07,
+      "loss": 0.0076,
+      "step": 9470
+    },
+    {
+      "epoch": 2.418984434804797,
+      "grad_norm": 1.1228365898132324,
+      "learning_rate": 6.598704052658379e-07,
+      "loss": 0.0076,
+      "step": 9480
+    },
+    {
+      "epoch": 2.421536106149528,
+      "grad_norm": 0.3657705783843994,
+      "learning_rate": 6.543072233043939e-07,
+      "loss": 0.0097,
+      "step": 9490
+    },
+    {
+      "epoch": 2.424087777494259,
+      "grad_norm": 0.2254849523305893,
+      "learning_rate": 6.487647197522862e-07,
+      "loss": 0.0079,
+      "step": 9500
+    },
+    {
+      "epoch": 2.4266394488389897,
+      "grad_norm": 0.4132424294948578,
+      "learning_rate": 6.43242943469347e-07,
+      "loss": 0.0086,
+      "step": 9510
+    },
+    {
+      "epoch": 2.42919112018372,
+      "grad_norm": 0.5428745150566101,
+      "learning_rate": 6.377419431326867e-07,
+      "loss": 0.0077,
+      "step": 9520
+    },
+    {
+      "epoch": 2.431742791528451,
+      "grad_norm": 0.7623319625854492,
+      "learning_rate": 6.322617672362698e-07,
+      "loss": 0.0066,
+      "step": 9530
+    },
+    {
+      "epoch": 2.434294462873182,
+      "grad_norm": 0.4517093300819397,
+      "learning_rate": 6.2680246409048e-07,
+      "loss": 0.0059,
+      "step": 9540
+    },
+    {
+      "epoch": 2.436846134217913,
+      "grad_norm": 0.46577808260917664,
+      "learning_rate": 6.213640818216965e-07,
+      "loss": 0.0104,
+      "step": 9550
+    },
+    {
+      "epoch": 2.4393978055626437,
+      "grad_norm": 0.43454670906066895,
+      "learning_rate": 6.159466683718768e-07,
+      "loss": 0.0072,
+      "step": 9560
+    },
+    {
+      "epoch": 2.441949476907374,
+      "grad_norm": 0.5216269493103027,
+      "learning_rate": 6.105502714981227e-07,
+      "loss": 0.0068,
+      "step": 9570
+    },
+    {
+      "epoch": 2.444501148252105,
+      "grad_norm": 0.7452653050422668,
+      "learning_rate": 6.05174938772266e-07,
+      "loss": 0.0062,
+      "step": 9580
+    },
+    {
+      "epoch": 2.447052819596836,
+      "grad_norm": 0.6804510354995728,
+      "learning_rate": 5.998207175804498e-07,
+      "loss": 0.0065,
+      "step": 9590
+    },
+    {
+      "epoch": 2.449604490941567,
+      "grad_norm": 0.6842300891876221,
+      "learning_rate": 5.94487655122706e-07,
+      "loss": 0.0079,
+      "step": 9600
+    },
+    {
+      "epoch": 2.4521561622862977,
+      "grad_norm": 1.2596189975738525,
+      "learning_rate": 5.891757984125443e-07,
+      "loss": 0.008,
+      "step": 9610
+    },
+    {
+      "epoch": 2.454707833631028,
+      "grad_norm": 1.110859751701355,
+      "learning_rate": 5.838851942765357e-07,
+      "loss": 0.0077,
+      "step": 9620
+    },
+    {
+      "epoch": 2.457259504975759,
+      "grad_norm": 0.23649120330810547,
+      "learning_rate": 5.786158893538982e-07,
+      "loss": 0.008,
+      "step": 9630
+    },
+    {
+      "epoch": 2.45981117632049,
+      "grad_norm": 0.49982666969299316,
+      "learning_rate": 5.733679300960864e-07,
+      "loss": 0.0079,
+      "step": 9640
+    },
+    {
+      "epoch": 2.462362847665221,
+      "grad_norm": 1.2177993059158325,
+      "learning_rate": 5.681413627663851e-07,
+      "loss": 0.0078,
+      "step": 9650
+    },
+    {
+      "epoch": 2.4649145190099517,
+      "grad_norm": 0.22220726311206818,
+      "learning_rate": 5.629362334394964e-07,
+      "loss": 0.0069,
+      "step": 9660
+    },
+    {
+      "epoch": 2.467466190354682,
+      "grad_norm": 0.49332597851753235,
+      "learning_rate": 5.577525880011379e-07,
+      "loss": 0.0077,
+      "step": 9670
+    },
+    {
+      "epoch": 2.470017861699413,
+      "grad_norm": 1.0104281902313232,
+      "learning_rate": 5.525904721476366e-07,
+      "loss": 0.0086,
+      "step": 9680
+    },
+    {
+      "epoch": 2.472569533044144,
+      "grad_norm": 0.4739457964897156,
+      "learning_rate": 5.474499313855242e-07,
+      "loss": 0.0066,
+      "step": 9690
+    },
+    {
+      "epoch": 2.475121204388875,
+      "grad_norm": 0.38137850165367126,
+      "learning_rate": 5.423310110311376e-07,
+      "loss": 0.0065,
+      "step": 9700
+    },
+    {
+      "epoch": 2.4776728757336057,
+      "grad_norm": 0.4624011814594269,
+      "learning_rate": 5.372337562102222e-07,
+      "loss": 0.009,
+      "step": 9710
+    },
+    {
+      "epoch": 2.480224547078336,
+      "grad_norm": 0.46692025661468506,
+      "learning_rate": 5.32158211857527e-07,
+      "loss": 0.006,
+      "step": 9720
+    },
+    {
+      "epoch": 2.482776218423067,
+      "grad_norm": 1.0589154958724976,
+      "learning_rate": 5.271044227164167e-07,
+      "loss": 0.0084,
+      "step": 9730
+    },
+    {
+      "epoch": 2.485327889767798,
+      "grad_norm": 0.606287956237793,
+      "learning_rate": 5.22072433338472e-07,
+      "loss": 0.0072,
+      "step": 9740
+    },
+    {
+      "epoch": 2.487879561112529,
+      "grad_norm": 0.5379163026809692,
+      "learning_rate": 5.170622880830971e-07,
+      "loss": 0.0075,
+      "step": 9750
+    },
+    {
+      "epoch": 2.4904312324572597,
+      "grad_norm": 0.6357580423355103,
+      "learning_rate": 5.120740311171315e-07,
+      "loss": 0.0065,
+      "step": 9760
+    },
+    {
+      "epoch": 2.49298290380199,
+      "grad_norm": 1.1960883140563965,
+      "learning_rate": 5.071077064144577e-07,
+      "loss": 0.0078,
+      "step": 9770
+    },
+    {
+      "epoch": 2.495534575146721,
+      "grad_norm": 0.9033713340759277,
+      "learning_rate": 5.021633577556143e-07,
+      "loss": 0.007,
+      "step": 9780
+    },
+    {
+      "epoch": 2.498086246491452,
+      "grad_norm": 0.6243811845779419,
+      "learning_rate": 4.97241028727412e-07,
+      "loss": 0.0066,
+      "step": 9790
+    },
+    {
+      "epoch": 2.500637917836183,
+      "grad_norm": 0.4687057435512543,
+      "learning_rate": 4.923407627225478e-07,
+      "loss": 0.0076,
+      "step": 9800
+    },
+    {
+      "epoch": 2.5031895891809137,
+      "grad_norm": 0.3284855782985687,
+      "learning_rate": 4.874626029392212e-07,
+      "loss": 0.0063,
+      "step": 9810
+    },
+    {
+      "epoch": 2.505741260525644,
+      "grad_norm": 1.1885757446289062,
+      "learning_rate": 4.826065923807561e-07,
+      "loss": 0.0087,
+      "step": 9820
+    },
+    {
+      "epoch": 2.508292931870375,
+      "grad_norm": 1.5649694204330444,
+      "learning_rate": 4.777727738552193e-07,
+      "loss": 0.0097,
+      "step": 9830
+    },
+    {
+      "epoch": 2.510844603215106,
+      "grad_norm": 0.2750914692878723,
+      "learning_rate": 4.7296118997504413e-07,
+      "loss": 0.0075,
+      "step": 9840
+    },
+    {
+      "epoch": 2.513396274559837,
+      "grad_norm": 0.6689438819885254,
+      "learning_rate": 4.681718831566558e-07,
+      "loss": 0.0093,
+      "step": 9850
+    },
+    {
+      "epoch": 2.5159479459045677,
+      "grad_norm": 0.7268848419189453,
+      "learning_rate": 4.6340489562009657e-07,
+      "loss": 0.009,
+      "step": 9860
+    },
+    {
+      "epoch": 2.518499617249298,
+      "grad_norm": 0.2928844690322876,
+      "learning_rate": 4.5866026938865157e-07,
+      "loss": 0.0074,
+      "step": 9870
+    },
+    {
+      "epoch": 2.521051288594029,
+      "grad_norm": 0.3780820369720459,
+      "learning_rate": 4.5393804628848333e-07,
+      "loss": 0.0081,
+      "step": 9880
+    },
+    {
+      "epoch": 2.52360295993876,
+      "grad_norm": 0.5721049308776855,
+      "learning_rate": 4.492382679482574e-07,
+      "loss": 0.0073,
+      "step": 9890
+    },
+    {
+      "epoch": 2.526154631283491,
+      "grad_norm": 0.4491700530052185,
+      "learning_rate": 4.445609757987782e-07,
+      "loss": 0.0068,
+      "step": 9900
+    },
+    {
+      "epoch": 2.5287063026282217,
+      "grad_norm": 0.21525029838085175,
+      "learning_rate": 4.3990621107262476e-07,
+      "loss": 0.0074,
+      "step": 9910
+    },
+    {
+      "epoch": 2.531257973972952,
+      "grad_norm": 1.5350968837738037,
+      "learning_rate": 4.352740148037866e-07,
+      "loss": 0.0093,
+      "step": 9920
+    },
+    {
+      "epoch": 2.533809645317683,
+      "grad_norm": 0.6345850825309753,
+      "learning_rate": 4.3066442782729876e-07,
+      "loss": 0.0076,
+      "step": 9930
+    },
+    {
+      "epoch": 2.536361316662414,
+      "grad_norm": 0.23979954421520233,
+      "learning_rate": 4.260774907788877e-07,
+      "loss": 0.0074,
+      "step": 9940
+    },
+    {
+      "epoch": 2.538912988007145,
+      "grad_norm": 0.5163211226463318,
+      "learning_rate": 4.215132440946076e-07,
+      "loss": 0.0066,
+      "step": 9950
+    },
+    {
+      "epoch": 2.5414646593518757,
+      "grad_norm": 0.9633673429489136,
+      "learning_rate": 4.1697172801048587e-07,
+      "loss": 0.0072,
+      "step": 9960
+    },
+    {
+      "epoch": 2.544016330696606,
+      "grad_norm": 0.5062496066093445,
+      "learning_rate": 4.1245298256217114e-07,
+      "loss": 0.0071,
+      "step": 9970
+    },
+    {
+      "epoch": 2.546568002041337,
+      "grad_norm": 1.0265597105026245,
+      "learning_rate": 4.0795704758457666e-07,
+      "loss": 0.0084,
+      "step": 9980
+    },
+    {
+      "epoch": 2.549119673386068,
+      "grad_norm": 0.6191508173942566,
+      "learning_rate": 4.03483962711529e-07,
+      "loss": 0.0065,
+      "step": 9990
+    },
+    {
+      "epoch": 2.551671344730799,
+      "grad_norm": 0.36056914925575256,
+      "learning_rate": 3.9903376737542316e-07,
+      "loss": 0.0077,
+      "step": 10000
+    },
+    {
+      "epoch": 2.5542230160755297,
+      "grad_norm": 0.43993762135505676,
+      "learning_rate": 3.94606500806869e-07,
+      "loss": 0.0076,
+      "step": 10010
+    },
+    {
+      "epoch": 2.55677468742026,
+      "grad_norm": 0.5819681286811829,
+      "learning_rate": 3.902022020343492e-07,
+      "loss": 0.0081,
+      "step": 10020
+    },
+    {
+      "epoch": 2.559326358764991,
+      "grad_norm": 0.878422737121582,
+      "learning_rate": 3.8582090988387455e-07,
+      "loss": 0.0078,
+      "step": 10030
+    },
+    {
+      "epoch": 2.561878030109722,
+      "grad_norm": 0.5557768940925598,
+      "learning_rate": 3.814626629786423e-07,
+      "loss": 0.0064,
+      "step": 10040
+    },
+    {
+      "epoch": 2.564429701454453,
+      "grad_norm": 0.6650263667106628,
+      "learning_rate": 3.7712749973869264e-07,
+      "loss": 0.0087,
+      "step": 10050
+    },
+    {
+      "epoch": 2.5669813727991837,
+      "grad_norm": 1.7049756050109863,
+      "learning_rate": 3.728154583805745e-07,
+      "loss": 0.0091,
+      "step": 10060
+    },
+    {
+      "epoch": 2.569533044143914,
+      "grad_norm": 0.9691872596740723,
+      "learning_rate": 3.685265769170042e-07,
+      "loss": 0.0071,
+      "step": 10070
+    },
+    {
+      "epoch": 2.572084715488645,
+      "grad_norm": 0.47858548164367676,
+      "learning_rate": 3.64260893156533e-07,
+      "loss": 0.008,
+      "step": 10080
+    },
+    {
+      "epoch": 2.574636386833376,
+      "grad_norm": 0.4813213050365448,
+      "learning_rate": 3.600184447032141e-07,
+      "loss": 0.0066,
+      "step": 10090
+    },
+    {
+      "epoch": 2.577188058178107,
+      "grad_norm": 0.6237726807594299,
+      "learning_rate": 3.557992689562709e-07,
+      "loss": 0.0066,
+      "step": 10100
+    },
+    {
+      "epoch": 2.5797397295228377,
+      "grad_norm": 0.46429261565208435,
+      "learning_rate": 3.516034031097639e-07,
+      "loss": 0.0068,
+      "step": 10110
+    },
+    {
+      "epoch": 2.582291400867568,
+      "grad_norm": 0.524320125579834,
+      "learning_rate": 3.4743088415226895e-07,
+      "loss": 0.007,
+      "step": 10120
+    },
+    {
+      "epoch": 2.584843072212299,
+      "grad_norm": 0.6312581300735474,
+      "learning_rate": 3.432817488665456e-07,
+      "loss": 0.0076,
+      "step": 10130
+    },
+    {
+      "epoch": 2.58739474355703,
+      "grad_norm": 0.5773522257804871,
+      "learning_rate": 3.391560338292152e-07,
+      "loss": 0.0071,
+      "step": 10140
+    },
+    {
+      "epoch": 2.589946414901761,
+      "grad_norm": 0.2685536742210388,
+      "learning_rate": 3.3505377541043903e-07,
+      "loss": 0.0075,
+      "step": 10150
+    },
+    {
+      "epoch": 2.5924980862464917,
+      "grad_norm": 0.7439098358154297,
+      "learning_rate": 3.309750097735983e-07,
+      "loss": 0.0074,
+      "step": 10160
+    },
+    {
+      "epoch": 2.595049757591222,
+      "grad_norm": 0.8754897713661194,
+      "learning_rate": 3.2691977287497067e-07,
+      "loss": 0.006,
+      "step": 10170
+    },
+    {
+      "epoch": 2.597601428935953,
+      "grad_norm": 0.4401561915874481,
+      "learning_rate": 3.2288810046342035e-07,
+      "loss": 0.007,
+      "step": 10180
+    },
+    {
+      "epoch": 2.600153100280684,
+      "grad_norm": 0.4255291521549225,
+      "learning_rate": 3.1888002808007633e-07,
+      "loss": 0.0061,
+      "step": 10190
+    },
+    {
+      "epoch": 2.602704771625415,
+      "grad_norm": 0.1766541600227356,
+      "learning_rate": 3.1489559105802413e-07,
+      "loss": 0.0062,
+      "step": 10200
+    },
+    {
+      "epoch": 2.6052564429701457,
+      "grad_norm": 0.962647020816803,
+      "learning_rate": 3.1093482452199053e-07,
+      "loss": 0.0075,
+      "step": 10210
+    },
+    {
+      "epoch": 2.607808114314876,
+      "grad_norm": 0.5156567692756653,
+      "learning_rate": 3.069977633880369e-07,
+      "loss": 0.0065,
+      "step": 10220
+    },
+    {
+      "epoch": 2.610359785659607,
+      "grad_norm": 0.6798850297927856,
+      "learning_rate": 3.030844423632491e-07,
+      "loss": 0.007,
+      "step": 10230
+    },
+    {
+      "epoch": 2.612911457004338,
+      "grad_norm": 1.1052918434143066,
+      "learning_rate": 2.99194895945434e-07,
+      "loss": 0.0075,
+      "step": 10240
+    },
+    {
+      "epoch": 2.615463128349069,
+      "grad_norm": 0.40075254440307617,
+      "learning_rate": 2.9532915842281137e-07,
+      "loss": 0.0077,
+      "step": 10250
+    },
+    {
+      "epoch": 2.6180147996937997,
+      "grad_norm": 0.48784929513931274,
+      "learning_rate": 2.914872638737169e-07,
+      "loss": 0.0088,
+      "step": 10260
+    },
+    {
+      "epoch": 2.62056647103853,
+      "grad_norm": 1.6315809488296509,
+      "learning_rate": 2.8766924616629697e-07,
+      "loss": 0.0062,
+      "step": 10270
+    },
+    {
+      "epoch": 2.623118142383261,
+      "grad_norm": 0.5205185413360596,
+      "learning_rate": 2.838751389582135e-07,
+      "loss": 0.0054,
+      "step": 10280
+    },
+    {
+      "epoch": 2.625669813727992,
+      "grad_norm": 0.6846718192100525,
+      "learning_rate": 2.8010497569634475e-07,
+      "loss": 0.0064,
+      "step": 10290
+    },
+    {
+      "epoch": 2.628221485072723,
+      "grad_norm": 0.5500282645225525,
+      "learning_rate": 2.763587896164922e-07,
+      "loss": 0.0063,
+      "step": 10300
+    },
+    {
+      "epoch": 2.6307731564174537,
+      "grad_norm": 11.970605850219727,
+      "learning_rate": 2.726366137430867e-07,
+      "loss": 0.0072,
+      "step": 10310
+    },
+    {
+      "epoch": 2.633324827762184,
+      "grad_norm": 0.7035308480262756,
+      "learning_rate": 2.6893848088889836e-07,
+      "loss": 0.0063,
+      "step": 10320
+    },
+    {
+      "epoch": 2.635876499106915,
+      "grad_norm": 1.363316535949707,
+      "learning_rate": 2.65264423654745e-07,
+      "loss": 0.0078,
+      "step": 10330
+    },
+    {
+      "epoch": 2.638428170451646,
+      "grad_norm": 0.6791175007820129,
+      "learning_rate": 2.6161447442920737e-07,
+      "loss": 0.0066,
+      "step": 10340
+    },
+    {
+      "epoch": 2.640979841796377,
+      "grad_norm": 1.1736422777175903,
+      "learning_rate": 2.57988665388343e-07,
+      "loss": 0.0078,
+      "step": 10350
+    },
+    {
+      "epoch": 2.6435315131411077,
+      "grad_norm": 0.6160316467285156,
+      "learning_rate": 2.543870284954012e-07,
+      "loss": 0.0067,
+      "step": 10360
+    },
+    {
+      "epoch": 2.646083184485838,
+      "grad_norm": 1.1056572198867798,
+      "learning_rate": 2.508095955005415e-07,
+      "loss": 0.0075,
+      "step": 10370
+    },
+    {
+      "epoch": 2.648634855830569,
+      "grad_norm": 0.4946836233139038,
+      "learning_rate": 2.472563979405561e-07,
+      "loss": 0.0056,
+      "step": 10380
+    },
+    {
+      "epoch": 2.6511865271753,
+      "grad_norm": 0.4888780117034912,
+      "learning_rate": 2.437274671385891e-07,
+      "loss": 0.0073,
+      "step": 10390
+    },
+    {
+      "epoch": 2.6537381985200303,
+      "grad_norm": 0.7349345684051514,
+      "learning_rate": 2.402228342038619e-07,
+      "loss": 0.0067,
+      "step": 10400
+    },
+    {
+      "epoch": 2.6562898698647617,
+      "grad_norm": 1.420432448387146,
+      "learning_rate": 2.3674253003139888e-07,
+      "loss": 0.0097,
+      "step": 10410
+    },
+    {
+      "epoch": 2.658841541209492,
+      "grad_norm": 0.36828410625457764,
+      "learning_rate": 2.3328658530175373e-07,
+      "loss": 0.0052,
+      "step": 10420
+    },
+    {
+      "epoch": 2.661393212554223,
+      "grad_norm": 1.2144927978515625,
+      "learning_rate": 2.2985503048073997e-07,
+      "loss": 0.0071,
+      "step": 10430
+    },
+    {
+      "epoch": 2.663944883898954,
+      "grad_norm": 0.6407505869865417,
+      "learning_rate": 2.264478958191638e-07,
+      "loss": 0.0076,
+      "step": 10440
+    },
+    {
+      "epoch": 2.6664965552436843,
+      "grad_norm": 0.5099070072174072,
+      "learning_rate": 2.2306521135255398e-07,
+      "loss": 0.0066,
+      "step": 10450
+    },
+    {
+      "epoch": 2.6690482265884157,
+      "grad_norm": 0.44241103529930115,
+      "learning_rate": 2.197070069009003e-07,
+      "loss": 0.0078,
+      "step": 10460
+    },
+    {
+      "epoch": 2.671599897933146,
+      "grad_norm": 0.49996262788772583,
+      "learning_rate": 2.1637331206838983e-07,
+      "loss": 0.0081,
+      "step": 10470
+    },
+    {
+      "epoch": 2.674151569277877,
+      "grad_norm": 0.3228076696395874,
+      "learning_rate": 2.130641562431449e-07,
+      "loss": 0.006,
+      "step": 10480
+    },
+    {
+      "epoch": 2.676703240622608,
+      "grad_norm": 0.9245404601097107,
+      "learning_rate": 2.0977956859696345e-07,
+      "loss": 0.0086,
+      "step": 10490
+    },
+    {
+      "epoch": 2.6792549119673383,
+      "grad_norm": 0.51729816198349,
+      "learning_rate": 2.0651957808506604e-07,
+      "loss": 0.007,
+      "step": 10500
+    },
+    {
+      "epoch": 2.6818065833120697,
+      "grad_norm": 0.7117488384246826,
+      "learning_rate": 2.032842134458347e-07,
+      "loss": 0.0063,
+      "step": 10510
+    },
+    {
+      "epoch": 2.6843582546568,
+      "grad_norm": 0.7258452773094177,
+      "learning_rate": 2.000735032005644e-07,
+      "loss": 0.0083,
+      "step": 10520
+    },
+    {
+      "epoch": 2.686909926001531,
+      "grad_norm": 0.3727307617664337,
+      "learning_rate": 1.9688747565320964e-07,
+      "loss": 0.0068,
+      "step": 10530
+    },
+    {
+      "epoch": 2.689461597346262,
+      "grad_norm": 0.7492243647575378,
+      "learning_rate": 1.9372615889013434e-07,
+      "loss": 0.0059,
+      "step": 10540
+    },
+    {
+      "epoch": 2.6920132686909923,
+      "grad_norm": 0.6845988035202026,
+      "learning_rate": 1.9058958077986498e-07,
+      "loss": 0.0065,
+      "step": 10550
+    },
+    {
+      "epoch": 2.6945649400357237,
+      "grad_norm": 0.4264490604400635,
+      "learning_rate": 1.8747776897284586e-07,
+      "loss": 0.0074,
+      "step": 10560
+    },
+    {
+      "epoch": 2.697116611380454,
+      "grad_norm": 1.5105531215667725,
+      "learning_rate": 1.8439075090119318e-07,
+      "loss": 0.0064,
+      "step": 10570
+    },
+    {
+      "epoch": 2.699668282725185,
+      "grad_norm": 0.4452486038208008,
+      "learning_rate": 1.8132855377845537e-07,
+      "loss": 0.006,
+      "step": 10580
+    },
+    {
+      "epoch": 2.702219954069916,
+      "grad_norm": 0.40780559182167053,
+      "learning_rate": 1.7829120459937255e-07,
+      "loss": 0.0068,
+      "step": 10590
+    },
+    {
+      "epoch": 2.7047716254146463,
+      "grad_norm": 0.5933582782745361,
+      "learning_rate": 1.7527873013963734e-07,
+      "loss": 0.0061,
+      "step": 10600
+    },
+    {
+      "epoch": 2.7073232967593777,
+      "grad_norm": 0.36882108449935913,
+      "learning_rate": 1.7229115695565955e-07,
+      "loss": 0.0054,
+      "step": 10610
+    },
+    {
+      "epoch": 2.709874968104108,
+      "grad_norm": 0.6598870754241943,
+      "learning_rate": 1.6932851138433346e-07,
+      "loss": 0.0068,
+      "step": 10620
+    },
+    {
+      "epoch": 2.712426639448839,
+      "grad_norm": 0.6586048007011414,
+      "learning_rate": 1.6639081954280354e-07,
+      "loss": 0.0067,
+      "step": 10630
+    },
+    {
+      "epoch": 2.71497831079357,
+      "grad_norm": 1.0061739683151245,
+      "learning_rate": 1.6347810732823521e-07,
+      "loss": 0.0078,
+      "step": 10640
+    },
+    {
+      "epoch": 2.7175299821383003,
+      "grad_norm": 0.22482961416244507,
+      "learning_rate": 1.6059040041758765e-07,
+      "loss": 0.0058,
+      "step": 10650
+    },
+    {
+      "epoch": 2.7200816534830317,
+      "grad_norm": 0.737498939037323,
+      "learning_rate": 1.577277242673848e-07,
+      "loss": 0.007,
+      "step": 10660
+    },
+    {
+      "epoch": 2.722633324827762,
+      "grad_norm": 0.8647279739379883,
+      "learning_rate": 1.5489010411349237e-07,
+      "loss": 0.0068,
+      "step": 10670
+    },
+    {
+      "epoch": 2.725184996172493,
+      "grad_norm": 0.8353618383407593,
+      "learning_rate": 1.5207756497089643e-07,
+      "loss": 0.0073,
+      "step": 10680
+    },
+    {
+      "epoch": 2.727736667517224,
+      "grad_norm": 0.6162010431289673,
+      "learning_rate": 1.492901316334816e-07,
+      "loss": 0.0063,
+      "step": 10690
+    },
+    {
+      "epoch": 2.7302883388619543,
+      "grad_norm": 0.38633957505226135,
+      "learning_rate": 1.465278286738122e-07,
+      "loss": 0.0061,
+      "step": 10700
+    },
+    {
+      "epoch": 2.7328400102066857,
+      "grad_norm": 0.9637838006019592,
+      "learning_rate": 1.4379068044291744e-07,
+      "loss": 0.0077,
+      "step": 10710
+    },
+    {
+      "epoch": 2.735391681551416,
+      "grad_norm": 0.8239974975585938,
+      "learning_rate": 1.4107871107007387e-07,
+      "loss": 0.0069,
+      "step": 10720
+    },
+    {
+      "epoch": 2.737943352896147,
+      "grad_norm": 0.8153387308120728,
+      "learning_rate": 1.3839194446259494e-07,
+      "loss": 0.0077,
+      "step": 10730
+    },
+    {
+      "epoch": 2.740495024240878,
+      "grad_norm": 0.6699053645133972,
+      "learning_rate": 1.3573040430561988e-07,
+      "loss": 0.0058,
+      "step": 10740
+    },
+    {
+      "epoch": 2.7430466955856083,
+      "grad_norm": 0.5199264287948608,
+      "learning_rate": 1.330941140619051e-07,
+      "loss": 0.0074,
+      "step": 10750
+    },
+    {
+      "epoch": 2.745598366930339,
+      "grad_norm": 0.9827825427055359,
+      "learning_rate": 1.3048309697161563e-07,
+      "loss": 0.0048,
+      "step": 10760
+    },
+    {
+      "epoch": 2.74815003827507,
+      "grad_norm": 0.7316129207611084,
+      "learning_rate": 1.278973760521226e-07,
+      "loss": 0.007,
+      "step": 10770
+    },
+    {
+      "epoch": 2.750701709619801,
+      "grad_norm": 0.5252894759178162,
+      "learning_rate": 1.253369740977983e-07,
+      "loss": 0.0054,
+      "step": 10780
+    },
+    {
+      "epoch": 2.753253380964532,
+      "grad_norm": 0.8061007261276245,
+      "learning_rate": 1.228019136798174e-07,
+      "loss": 0.0056,
+      "step": 10790
+    },
+    {
+      "epoch": 2.7558050523092623,
+      "grad_norm": 0.6680382490158081,
+      "learning_rate": 1.2029221714595518e-07,
+      "loss": 0.0079,
+      "step": 10800
+    },
+    {
+      "epoch": 2.758356723653993,
+      "grad_norm": 0.6302357316017151,
+      "learning_rate": 1.1780790662039342e-07,
+      "loss": 0.0066,
+      "step": 10810
+    },
+    {
+      "epoch": 2.760908394998724,
+      "grad_norm": 0.35548314452171326,
+      "learning_rate": 1.1534900400352332e-07,
+      "loss": 0.0051,
+      "step": 10820
+    },
+    {
+      "epoch": 2.763460066343455,
+      "grad_norm": 0.7780895233154297,
+      "learning_rate": 1.1291553097175388e-07,
+      "loss": 0.0059,
+      "step": 10830
+    },
+    {
+      "epoch": 2.766011737688186,
+      "grad_norm": 0.5007578134536743,
+      "learning_rate": 1.1050750897731877e-07,
+      "loss": 0.0058,
+      "step": 10840
+    },
+    {
+      "epoch": 2.7685634090329163,
+      "grad_norm": 0.6232514381408691,
+      "learning_rate": 1.0812495924809007e-07,
+      "loss": 0.0066,
+      "step": 10850
+    },
+    {
+      "epoch": 2.771115080377647,
+      "grad_norm": 0.672356367111206,
+      "learning_rate": 1.0576790278738824e-07,
+      "loss": 0.0051,
+      "step": 10860
+    },
+    {
+      "epoch": 2.773666751722378,
+      "grad_norm": 0.34914809465408325,
+      "learning_rate": 1.0343636037379978e-07,
+      "loss": 0.0063,
+      "step": 10870
+    },
+    {
+      "epoch": 2.776218423067109,
+      "grad_norm": 1.2736538648605347,
+      "learning_rate": 1.011303525609908e-07,
+      "loss": 0.0059,
+      "step": 10880
+    },
+    {
+      "epoch": 2.77877009441184,
+      "grad_norm": 1.2168415784835815,
+      "learning_rate": 9.884989967752944e-08,
+      "loss": 0.0062,
+      "step": 10890
+    },
+    {
+      "epoch": 2.7813217657565703,
+      "grad_norm": 0.1484823226928711,
+      "learning_rate": 9.659502182670311e-08,
+      "loss": 0.0045,
+      "step": 10900
+    },
+    {
+      "epoch": 2.783873437101301,
+      "grad_norm": 0.7366049289703369,
+      "learning_rate": 9.436573888634481e-08,
+      "loss": 0.0069,
+      "step": 10910
+    },
+    {
+      "epoch": 2.786425108446032,
+      "grad_norm": 0.9276716113090515,
+      "learning_rate": 9.216207050865478e-08,
+      "loss": 0.0058,
+      "step": 10920
+    },
+    {
+      "epoch": 2.788976779790763,
+      "grad_norm": 1.392797589302063,
+      "learning_rate": 8.998403612002948e-08,
+      "loss": 0.0063,
+      "step": 10930
+    },
+    {
+      "epoch": 2.791528451135494,
+      "grad_norm": 0.40872761607170105,
+      "learning_rate": 8.783165492088785e-08,
+      "loss": 0.0064,
+      "step": 10940
+    },
+    {
+      "epoch": 2.7940801224802243,
+      "grad_norm": 0.910552442073822,
+      "learning_rate": 8.570494588550604e-08,
+      "loss": 0.007,
+      "step": 10950
+    },
+    {
+      "epoch": 2.796631793824955,
+      "grad_norm": 0.8210703730583191,
+      "learning_rate": 8.360392776184556e-08,
+      "loss": 0.0069,
+      "step": 10960
+    },
+    {
+      "epoch": 2.799183465169686,
+      "grad_norm": 0.6619951128959656,
+      "learning_rate": 8.152861907139176e-08,
+      "loss": 0.0068,
+      "step": 10970
+    },
+    {
+      "epoch": 2.801735136514417,
+      "grad_norm": 0.41350725293159485,
+      "learning_rate": 7.94790381089876e-08,
+      "loss": 0.0058,
+      "step": 10980
+    },
+    {
+      "epoch": 2.804286807859148,
+      "grad_norm": 0.4920821785926819,
+      "learning_rate": 7.74552029426745e-08,
+      "loss": 0.0057,
+      "step": 10990
+    },
+    {
+      "epoch": 2.8068384792038783,
+      "grad_norm": 0.4195524752140045,
+      "learning_rate": 7.545713141353305e-08,
+      "loss": 0.0048,
+      "step": 11000
+    },
+    {
+      "epoch": 2.809390150548609,
+      "grad_norm": 0.3614073395729065,
+      "learning_rate": 7.348484113552356e-08,
+      "loss": 0.0068,
+      "step": 11010
+    },
+    {
+      "epoch": 2.81194182189334,
+      "grad_norm": 0.5706660747528076,
+      "learning_rate": 7.153834949533278e-08,
+      "loss": 0.0062,
+      "step": 11020
+    },
+    {
+      "epoch": 2.814493493238071,
+      "grad_norm": 1.5906177759170532,
+      "learning_rate": 6.961767365222072e-08,
+      "loss": 0.0065,
+      "step": 11030
+    },
+    {
+      "epoch": 2.817045164582802,
+      "grad_norm": 0.7011880278587341,
+      "learning_rate": 6.77228305378671e-08,
+      "loss": 0.0065,
+      "step": 11040
+    },
+    {
+      "epoch": 2.8195968359275323,
+      "grad_norm": 0.9448215961456299,
+      "learning_rate": 6.585383685622548e-08,
+      "loss": 0.0072,
+      "step": 11050
+    },
+    {
+      "epoch": 2.822148507272263,
+      "grad_norm": 0.4587925374507904,
+      "learning_rate": 6.401070908337403e-08,
+      "loss": 0.0064,
+      "step": 11060
+    },
+    {
+      "epoch": 2.824700178616994,
+      "grad_norm": 0.1932133436203003,
+      "learning_rate": 6.219346346736932e-08,
+      "loss": 0.0054,
+      "step": 11070
+    },
+    {
+      "epoch": 2.827251849961725,
+      "grad_norm": 0.614255428314209,
+      "learning_rate": 6.040211602810475e-08,
+      "loss": 0.0055,
+      "step": 11080
+    },
+    {
+      "epoch": 2.829803521306456,
+      "grad_norm": 1.1453063488006592,
+      "learning_rate": 5.86366825571697e-08,
+      "loss": 0.0071,
+      "step": 11090
+    },
+    {
+      "epoch": 2.8323551926511863,
+      "grad_norm": 0.9590531587600708,
+      "learning_rate": 5.6897178617708267e-08,
+      "loss": 0.0066,
+      "step": 11100
+    },
+    {
+      "epoch": 2.834906863995917,
+      "grad_norm": 0.6798762083053589,
+      "learning_rate": 5.518361954428408e-08,
+      "loss": 0.0065,
+      "step": 11110
+    },
+    {
+      "epoch": 2.837458535340648,
+      "grad_norm": 0.6563116312026978,
+      "learning_rate": 5.349602044274404e-08,
+      "loss": 0.0084,
+      "step": 11120
+    },
+    {
+      "epoch": 2.840010206685379,
+      "grad_norm": 0.479773610830307,
+      "learning_rate": 5.1834396190085456e-08,
+      "loss": 0.0066,
+      "step": 11130
+    },
+    {
+      "epoch": 2.84256187803011,
+      "grad_norm": 0.5942010283470154,
+      "learning_rate": 5.019876143432445e-08,
+      "loss": 0.007,
+      "step": 11140
+    },
+    {
+      "epoch": 2.8451135493748403,
+      "grad_norm": 1.0438846349716187,
+      "learning_rate": 4.858913059436809e-08,
+      "loss": 0.0077,
+      "step": 11150
+    },
+    {
+      "epoch": 2.847665220719571,
+      "grad_norm": 0.9194878935813904,
+      "learning_rate": 4.7005517859885493e-08,
+      "loss": 0.0059,
+      "step": 11160
+    },
+    {
+      "epoch": 2.850216892064302,
+      "grad_norm": 1.6355482339859009,
+      "learning_rate": 4.544793719118423e-08,
+      "loss": 0.007,
+      "step": 11170
+    },
+    {
+      "epoch": 2.852768563409033,
+      "grad_norm": 0.43507319688796997,
+      "learning_rate": 4.3916402319087115e-08,
+      "loss": 0.0071,
+      "step": 11180
+    },
+    {
+      "epoch": 2.855320234753764,
+      "grad_norm": 0.5874977707862854,
+      "learning_rate": 4.241092674480962e-08,
+      "loss": 0.0066,
+      "step": 11190
+    },
+    {
+      "epoch": 2.8578719060984943,
+      "grad_norm": 0.5329188704490662,
+      "learning_rate": 4.0931523739842326e-08,
+      "loss": 0.0054,
+      "step": 11200
+    },
+    {
+      "epoch": 2.860423577443225,
+      "grad_norm": 0.5945627093315125,
+      "learning_rate": 3.9478206345834656e-08,
+      "loss": 0.0072,
+      "step": 11210
+    },
+    {
+      "epoch": 2.862975248787956,
+      "grad_norm": 0.42057672142982483,
+      "learning_rate": 3.805098737447732e-08,
+      "loss": 0.0051,
+      "step": 11220
+    },
+    {
+      "epoch": 2.865526920132687,
+      "grad_norm": 0.19661393761634827,
+      "learning_rate": 3.664987940739206e-08,
+      "loss": 0.006,
+      "step": 11230
+    },
+    {
+      "epoch": 2.868078591477418,
+      "grad_norm": 0.8557282090187073,
+      "learning_rate": 3.5274894796018756e-08,
+      "loss": 0.0056,
+      "step": 11240
+    },
+    {
+      "epoch": 2.8706302628221483,
+      "grad_norm": 0.543337881565094,
+      "learning_rate": 3.392604566150781e-08,
+      "loss": 0.0047,
+      "step": 11250
+    },
+    {
+      "epoch": 2.873181934166879,
+      "grad_norm": 0.9113593101501465,
+      "learning_rate": 3.260334389461261e-08,
+      "loss": 0.0066,
+      "step": 11260
+    },
+    {
+      "epoch": 2.87573360551161,
+      "grad_norm": 0.442043274641037,
+      "learning_rate": 3.130680115558493e-08,
+      "loss": 0.0058,
+      "step": 11270
+    },
+    {
+      "epoch": 2.878285276856341,
+      "grad_norm": 0.7696757316589355,
+      "learning_rate": 3.003642887407232e-08,
+      "loss": 0.0065,
+      "step": 11280
+    },
+    {
+      "epoch": 2.880836948201072,
+      "grad_norm": 0.9663532376289368,
+      "learning_rate": 2.879223824901722e-08,
+      "loss": 0.0066,
+      "step": 11290
+    },
+    {
+      "epoch": 2.8833886195458023,
+      "grad_norm": 0.40540120005607605,
+      "learning_rate": 2.7574240248558036e-08,
+      "loss": 0.0076,
+      "step": 11300
+    },
+    {
+      "epoch": 2.885940290890533,
+      "grad_norm": 0.1316661834716797,
+      "learning_rate": 2.6382445609932527e-08,
+      "loss": 0.0042,
+      "step": 11310
+    },
+    {
+      "epoch": 2.888491962235264,
+      "grad_norm": 0.5307180285453796,
+      "learning_rate": 2.5216864839382903e-08,
+      "loss": 0.005,
+      "step": 11320
+    },
+    {
+      "epoch": 2.891043633579995,
+      "grad_norm": 0.6010832190513611,
+      "learning_rate": 2.407750821206456e-08,
+      "loss": 0.0071,
+      "step": 11330
+    },
+    {
+      "epoch": 2.893595304924726,
+      "grad_norm": 0.35675570368766785,
+      "learning_rate": 2.296438577195348e-08,
+      "loss": 0.0057,
+      "step": 11340
+    },
+    {
+      "epoch": 2.8961469762694563,
+      "grad_norm": 0.5174592733383179,
+      "learning_rate": 2.1877507331758973e-08,
+      "loss": 0.0075,
+      "step": 11350
+    },
+    {
+      "epoch": 2.898698647614187,
+      "grad_norm": 2.2227537631988525,
+      "learning_rate": 2.0816882472837084e-08,
+      "loss": 0.0055,
+      "step": 11360
+    },
+    {
+      "epoch": 2.901250318958918,
+      "grad_norm": 0.5491740107536316,
+      "learning_rate": 1.9782520545105987e-08,
+      "loss": 0.0058,
+      "step": 11370
+    },
+    {
+      "epoch": 2.903801990303649,
+      "grad_norm": 0.6968376040458679,
+      "learning_rate": 1.8774430666962383e-08,
+      "loss": 0.0043,
+      "step": 11380
+    },
+    {
+      "epoch": 2.90635366164838,
+      "grad_norm": 0.35690832138061523,
+      "learning_rate": 1.779262172520424e-08,
+      "loss": 0.0066,
+      "step": 11390
+    },
+    {
+      "epoch": 2.9089053329931103,
+      "grad_norm": 0.6748278141021729,
+      "learning_rate": 1.683710237494851e-08,
+      "loss": 0.007,
+      "step": 11400
+    },
+    {
+      "epoch": 2.911457004337841,
+      "grad_norm": 0.7347134947776794,
+      "learning_rate": 1.590788103955787e-08,
+      "loss": 0.0078,
+      "step": 11410
+    },
+    {
+      "epoch": 2.914008675682572,
+      "grad_norm": 0.6468586325645447,
+      "learning_rate": 1.5004965910564772e-08,
+      "loss": 0.0068,
+      "step": 11420
+    },
+    {
+      "epoch": 2.916560347027303,
+      "grad_norm": 0.6974029541015625,
+      "learning_rate": 1.4128364947600503e-08,
+      "loss": 0.0069,
+      "step": 11430
+    },
+    {
+      "epoch": 2.919112018372034,
+      "grad_norm": 0.43353763222694397,
+      "learning_rate": 1.3278085878323576e-08,
+      "loss": 0.0043,
+      "step": 11440
+    },
+    {
+      "epoch": 2.9216636897167643,
+      "grad_norm": 0.5207705497741699,
+      "learning_rate": 1.2454136198353117e-08,
+      "loss": 0.0058,
+      "step": 11450
+    },
+    {
+      "epoch": 2.924215361061495,
+      "grad_norm": 0.6456037759780884,
+      "learning_rate": 1.1656523171200916e-08,
+      "loss": 0.0063,
+      "step": 11460
+    },
+    {
+      "epoch": 2.926767032406226,
+      "grad_norm": 0.5657256841659546,
+      "learning_rate": 1.0885253828209818e-08,
+      "loss": 0.0053,
+      "step": 11470
+    },
+    {
+      "epoch": 2.929318703750957,
+      "grad_norm": 0.7830215692520142,
+      "learning_rate": 1.0140334968489762e-08,
+      "loss": 0.0053,
+      "step": 11480
+    },
+    {
+      "epoch": 2.931870375095688,
+      "grad_norm": 0.30181020498275757,
+      "learning_rate": 9.421773158857838e-09,
+      "loss": 0.008,
+      "step": 11490
+    },
+    {
+      "epoch": 2.9344220464404183,
+      "grad_norm": 1.1031672954559326,
+      "learning_rate": 8.729574733782663e-09,
+      "loss": 0.0062,
+      "step": 11500
+    },
+    {
+      "epoch": 2.936973717785149,
+      "grad_norm": 0.9583958983421326,
+      "learning_rate": 8.063745795325761e-09,
+      "loss": 0.0062,
+      "step": 11510
+    },
+    {
+      "epoch": 2.93952538912988,
+      "grad_norm": 0.7751434445381165,
+      "learning_rate": 7.424292213089268e-09,
+      "loss": 0.0073,
+      "step": 11520
+    },
+    {
+      "epoch": 2.942077060474611,
+      "grad_norm": 0.9476333260536194,
+      "learning_rate": 6.811219624163645e-09,
+      "loss": 0.0046,
+      "step": 11530
+    },
+    {
+      "epoch": 2.944628731819342,
+      "grad_norm": 0.5230672359466553,
+      "learning_rate": 6.224533433079049e-09,
+      "loss": 0.0068,
+      "step": 11540
+    },
+    {
+      "epoch": 2.9471804031640723,
+      "grad_norm": 1.2709357738494873,
+      "learning_rate": 5.664238811755373e-09,
+      "loss": 0.0067,
+      "step": 11550
+    },
+    {
+      "epoch": 2.949732074508803,
+      "grad_norm": 0.8602284789085388,
+      "learning_rate": 5.130340699459946e-09,
+      "loss": 0.0074,
+      "step": 11560
+    },
+    {
+      "epoch": 2.952283745853534,
+      "grad_norm": 0.2690794765949249,
+      "learning_rate": 4.622843802760568e-09,
+      "loss": 0.0041,
+      "step": 11570
+    },
+    {
+      "epoch": 2.954835417198265,
+      "grad_norm": 0.5806980133056641,
+      "learning_rate": 4.141752595486547e-09,
+      "loss": 0.0059,
+      "step": 11580
+    },
+    {
+      "epoch": 2.957387088542996,
+      "grad_norm": 0.5464335680007935,
+      "learning_rate": 3.6870713186883954e-09,
+      "loss": 0.0073,
+      "step": 11590
+    },
+    {
+      "epoch": 2.9599387598877263,
+      "grad_norm": 0.3500746190547943,
+      "learning_rate": 3.258803980599856e-09,
+      "loss": 0.0058,
+      "step": 11600
+    },
+    {
+      "epoch": 2.962490431232457,
+      "grad_norm": 1.2113854885101318,
+      "learning_rate": 2.85695435660327e-09,
+      "loss": 0.0074,
+      "step": 11610
+    },
+    {
+      "epoch": 2.965042102577188,
+      "grad_norm": 0.7115603685379028,
+      "learning_rate": 2.481525989196598e-09,
+      "loss": 0.0069,
+      "step": 11620
+    },
+    {
+      "epoch": 2.967593773921919,
+      "grad_norm": 0.3272109031677246,
+      "learning_rate": 2.1325221879607835e-09,
+      "loss": 0.0084,
+      "step": 11630
+    },
+    {
+      "epoch": 2.97014544526665,
+      "grad_norm": 0.6895642280578613,
+      "learning_rate": 1.8099460295324389e-09,
+      "loss": 0.0052,
+      "step": 11640
+    },
+    {
+      "epoch": 2.9726971166113803,
+      "grad_norm": 0.5834820866584778,
+      "learning_rate": 1.5138003575755343e-09,
+      "loss": 0.0072,
+      "step": 11650
+    },
+    {
+      "epoch": 2.975248787956111,
+      "grad_norm": 1.0529255867004395,
+      "learning_rate": 1.2440877827567531e-09,
+      "loss": 0.006,
+      "step": 11660
+    },
+    {
+      "epoch": 2.977800459300842,
+      "grad_norm": 0.41195112466812134,
+      "learning_rate": 1.000810682721176e-09,
+      "loss": 0.0082,
+      "step": 11670
+    },
+    {
+      "epoch": 2.980352130645573,
+      "grad_norm": 0.7077150940895081,
+      "learning_rate": 7.839712020742962e-10,
+      "loss": 0.0054,
+      "step": 11680
+    },
+    {
+      "epoch": 2.982903801990304,
+      "grad_norm": 0.3624398112297058,
+      "learning_rate": 5.935712523600367e-10,
+      "loss": 0.005,
+      "step": 11690
+    },
+    {
+      "epoch": 2.9854554733350342,
+      "grad_norm": 0.6512740254402161,
+      "learning_rate": 4.296125120454297e-10,
+      "loss": 0.0066,
+      "step": 11700
+    },
+    {
+      "epoch": 2.988007144679765,
+      "grad_norm": 0.681259036064148,
+      "learning_rate": 2.9209642650562805e-10,
+      "loss": 0.0066,
+      "step": 11710
+    },
+    {
+      "epoch": 2.990558816024496,
+      "grad_norm": 0.49927040934562683,
+      "learning_rate": 1.810242080109159e-10,
+      "loss": 0.0078,
+      "step": 11720
+    },
+    {
+      "epoch": 2.993110487369227,
+      "grad_norm": 0.5806398391723633,
+      "learning_rate": 9.639683571638358e-11,
+      "loss": 0.0061,
+      "step": 11730
+    },
+    {
+      "epoch": 2.995662158713958,
+      "grad_norm": 0.5190763473510742,
+      "learning_rate": 3.8215055653267886e-11,
+      "loss": 0.0069,
+      "step": 11740
+    },
+    {
+      "epoch": 2.9982138300586882,
+      "grad_norm": 0.4141053855419159,
+      "learning_rate": 6.4793807212915056e-12,
+      "loss": 0.0058,
+      "step": 11750
+    },
+    {
+      "epoch": 3.0,
+      "step": 11757,
+      "total_flos": 6.099787643229307e+18,
+      "train_loss": 0.04448508482245479,
+      "train_runtime": 11326.9906,
+      "train_samples_per_second": 132.859,
+      "train_steps_per_second": 1.038
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 11757,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 6.099787643229307e+18,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}