diff --git "a/checkpoint-7500/trainer_state.json" "b/checkpoint-7500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-7500/trainer_state.json"
@@ -0,0 +1,5284 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.16464293064416546,
+  "eval_steps": 500,
+  "global_step": 7500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00021952390752555395,
+      "grad_norm": 129536.0,
+      "learning_rate": 1.99775e-05,
+      "loss": 11.4186,
+      "step": 10
+    },
+    {
+      "epoch": 0.0004390478150511079,
+      "grad_norm": 154.0,
+      "learning_rate": 1.9952500000000003e-05,
+      "loss": 6.1068,
+      "step": 20
+    },
+    {
+      "epoch": 0.0006585717225766619,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9927500000000002e-05,
+      "loss": 4.8048,
+      "step": 30
+    },
+    {
+      "epoch": 0.0008780956301022158,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.99025e-05,
+      "loss": 0.2014,
+      "step": 40
+    },
+    {
+      "epoch": 0.0010976195376277698,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.98775e-05,
+      "loss": 0.2177,
+      "step": 50
+    },
+    {
+      "epoch": 0.0013171434451533237,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.98525e-05,
+      "loss": 0.2286,
+      "step": 60
+    },
+    {
+      "epoch": 0.0015366673526788777,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9827500000000003e-05,
+      "loss": 0.1803,
+      "step": 70
+    },
+    {
+      "epoch": 0.0017561912602044316,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.9802500000000002e-05,
+      "loss": 0.1814,
+      "step": 80
+    },
+    {
+      "epoch": 0.0019757151677299856,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.97775e-05,
+      "loss": 0.1881,
+      "step": 90
+    },
+    {
+      "epoch": 0.0021952390752555395,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.97525e-05,
+      "loss": 0.1877,
+      "step": 100
+    },
+    {
+      "epoch": 0.0024147629827810935,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.97275e-05,
+      "loss": 0.1892,
+      "step": 110
+    },
+    {
+      "epoch": 0.0026342868903066474,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9702500000000003e-05,
+      "loss": 0.1921,
+      "step": 120
+    },
+    {
+      "epoch": 0.0028538107978322014,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9677500000000003e-05,
+      "loss": 0.2261,
+      "step": 130
+    },
+    {
+      "epoch": 0.0030733347053577553,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9652500000000002e-05,
+      "loss": 0.2294,
+      "step": 140
+    },
+    {
+      "epoch": 0.0032928586128833093,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.96275e-05,
+      "loss": 0.1871,
+      "step": 150
+    },
+    {
+      "epoch": 0.0035123825204088632,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.96025e-05,
+      "loss": 0.1888,
+      "step": 160
+    },
+    {
+      "epoch": 0.003731906427934417,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9577500000000004e-05,
+      "loss": 0.1925,
+      "step": 170
+    },
+    {
+      "epoch": 0.003951430335459971,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.95525e-05,
+      "loss": 0.1974,
+      "step": 180
+    },
+    {
+      "epoch": 0.004170954242985525,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.9527500000000002e-05,
+      "loss": 0.2109,
+      "step": 190
+    },
+    {
+      "epoch": 0.004390478150511079,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9502500000000002e-05,
+      "loss": 0.2015,
+      "step": 200
+    },
+    {
+      "epoch": 0.004610002058036633,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.94775e-05,
+      "loss": 0.1631,
+      "step": 210
+    },
+    {
+      "epoch": 0.004829525965562187,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.94525e-05,
+      "loss": 0.195,
+      "step": 220
+    },
+    {
+      "epoch": 0.005049049873087741,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.94275e-05,
+      "loss": 0.1882,
+      "step": 230
+    },
+    {
+      "epoch": 0.005268573780613295,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9402500000000003e-05,
+      "loss": 0.1956,
+      "step": 240
+    },
+    {
+      "epoch": 0.005488097688138849,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9377500000000002e-05,
+      "loss": 0.2236,
+      "step": 250
+    },
+    {
+      "epoch": 0.005707621595664403,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.93525e-05,
+      "loss": 0.1906,
+      "step": 260
+    },
+    {
+      "epoch": 0.005927145503189957,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.93275e-05,
+      "loss": 0.2174,
+      "step": 270
+    },
+    {
+      "epoch": 0.006146669410715511,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.93025e-05,
+      "loss": 0.1898,
+      "step": 280
+    },
+    {
+      "epoch": 0.006366193318241065,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9277500000000003e-05,
+      "loss": 0.2021,
+      "step": 290
+    },
+    {
+      "epoch": 0.0065857172257666186,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.92525e-05,
+      "loss": 0.192,
+      "step": 300
+    },
+    {
+      "epoch": 0.0068052411332921725,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9227500000000002e-05,
+      "loss": 0.2299,
+      "step": 310
+    },
+    {
+      "epoch": 0.0070247650408177265,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.92025e-05,
+      "loss": 0.2137,
+      "step": 320
+    },
+    {
+      "epoch": 0.00724428894834328,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.91775e-05,
+      "loss": 0.1918,
+      "step": 330
+    },
+    {
+      "epoch": 0.007463812855868834,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.91525e-05,
+      "loss": 0.2303,
+      "step": 340
+    },
+    {
+      "epoch": 0.007683336763394388,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.91275e-05,
+      "loss": 0.1957,
+      "step": 350
+    },
+    {
+      "epoch": 0.007902860670919942,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9102500000000002e-05,
+      "loss": 0.2029,
+      "step": 360
+    },
+    {
+      "epoch": 0.008122384578445496,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.90775e-05,
+      "loss": 0.219,
+      "step": 370
+    },
+    {
+      "epoch": 0.00834190848597105,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.90525e-05,
+      "loss": 0.2189,
+      "step": 380
+    },
+    {
+      "epoch": 0.008561432393496604,
+      "grad_norm": 1.125,
+      "learning_rate": 1.90275e-05,
+      "loss": 0.2014,
+      "step": 390
+    },
+    {
+      "epoch": 0.008780956301022158,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9002500000000003e-05,
+      "loss": 0.1757,
+      "step": 400
+    },
+    {
+      "epoch": 0.009000480208547712,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8977500000000003e-05,
+      "loss": 0.2146,
+      "step": 410
+    },
+    {
+      "epoch": 0.009220004116073266,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8952500000000002e-05,
+      "loss": 0.1767,
+      "step": 420
+    },
+    {
+      "epoch": 0.00943952802359882,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.89275e-05,
+      "loss": 0.1913,
+      "step": 430
+    },
+    {
+      "epoch": 0.009659051931124374,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.89025e-05,
+      "loss": 0.1938,
+      "step": 440
+    },
+    {
+      "epoch": 0.009878575838649928,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.8877500000000003e-05,
+      "loss": 0.1898,
+      "step": 450
+    },
+    {
+      "epoch": 0.010098099746175482,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.8852500000000003e-05,
+      "loss": 0.214,
+      "step": 460
+    },
+    {
+      "epoch": 0.010317623653701036,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8827500000000002e-05,
+      "loss": 0.191,
+      "step": 470
+    },
+    {
+      "epoch": 0.01053714756122659,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.88025e-05,
+      "loss": 0.207,
+      "step": 480
+    },
+    {
+      "epoch": 0.010756671468752144,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.87775e-05,
+      "loss": 0.2191,
+      "step": 490
+    },
+    {
+      "epoch": 0.010976195376277698,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.8752500000000004e-05,
+      "loss": 0.2271,
+      "step": 500
+    },
+    {
+      "epoch": 0.011195719283803252,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.87275e-05,
+      "loss": 0.1823,
+      "step": 510
+    },
+    {
+      "epoch": 0.011415243191328805,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.8702500000000003e-05,
+      "loss": 0.2054,
+      "step": 520
+    },
+    {
+      "epoch": 0.01163476709885436,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8677500000000002e-05,
+      "loss": 0.2131,
+      "step": 530
+    },
+    {
+      "epoch": 0.011854291006379913,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.86525e-05,
+      "loss": 0.1569,
+      "step": 540
+    },
+    {
+      "epoch": 0.012073814913905467,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.86275e-05,
+      "loss": 0.1877,
+      "step": 550
+    },
+    {
+      "epoch": 0.012293338821431021,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.86025e-05,
+      "loss": 0.2228,
+      "step": 560
+    },
+    {
+      "epoch": 0.012512862728956575,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.8577500000000003e-05,
+      "loss": 0.2096,
+      "step": 570
+    },
+    {
+      "epoch": 0.01273238663648213,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8552500000000002e-05,
+      "loss": 0.1861,
+      "step": 580
+    },
+    {
+      "epoch": 0.012951910544007683,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8527500000000002e-05,
+      "loss": 0.1832,
+      "step": 590
+    },
+    {
+      "epoch": 0.013171434451533237,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.85025e-05,
+      "loss": 0.2182,
+      "step": 600
+    },
+    {
+      "epoch": 0.013390958359058791,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.84775e-05,
+      "loss": 0.2285,
+      "step": 610
+    },
+    {
+      "epoch": 0.013610482266584345,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8452500000000003e-05,
+      "loss": 0.2024,
+      "step": 620
+    },
+    {
+      "epoch": 0.013830006174109899,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8427500000000003e-05,
+      "loss": 0.2115,
+      "step": 630
+    },
+    {
+      "epoch": 0.014049530081635453,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8402500000000002e-05,
+      "loss": 0.227,
+      "step": 640
+    },
+    {
+      "epoch": 0.014269053989161007,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.83775e-05,
+      "loss": 0.185,
+      "step": 650
+    },
+    {
+      "epoch": 0.01448857789668656,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.83525e-05,
+      "loss": 0.1754,
+      "step": 660
+    },
+    {
+      "epoch": 0.014708101804212115,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8327500000000004e-05,
+      "loss": 0.2198,
+      "step": 670
+    },
+    {
+      "epoch": 0.014927625711737669,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.83025e-05,
+      "loss": 0.1966,
+      "step": 680
+    },
+    {
+      "epoch": 0.015147149619263223,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.8277500000000002e-05,
+      "loss": 0.187,
+      "step": 690
+    },
+    {
+      "epoch": 0.015366673526788777,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.8252500000000002e-05,
+      "loss": 0.2088,
+      "step": 700
+    },
+    {
+      "epoch": 0.01558619743431433,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.82275e-05,
+      "loss": 0.2132,
+      "step": 710
+    },
+    {
+      "epoch": 0.015805721341839885,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.82025e-05,
+      "loss": 0.2149,
+      "step": 720
+    },
+    {
+      "epoch": 0.01602524524936544,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.81775e-05,
+      "loss": 0.1931,
+      "step": 730
+    },
+    {
+      "epoch": 0.016244769156890992,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8152500000000003e-05,
+      "loss": 0.1721,
+      "step": 740
+    },
+    {
+      "epoch": 0.016464293064416546,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8127500000000002e-05,
+      "loss": 0.1911,
+      "step": 750
+    },
+    {
+      "epoch": 0.0166838169719421,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.81025e-05,
+      "loss": 0.2294,
+      "step": 760
+    },
+    {
+      "epoch": 0.016903340879467654,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.80775e-05,
+      "loss": 0.1867,
+      "step": 770
+    },
+    {
+      "epoch": 0.017122864786993208,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.80525e-05,
+      "loss": 0.1997,
+      "step": 780
+    },
+    {
+      "epoch": 0.017342388694518762,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8027500000000003e-05,
+      "loss": 0.2095,
+      "step": 790
+    },
+    {
+      "epoch": 0.017561912602044316,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.80025e-05,
+      "loss": 0.1862,
+      "step": 800
+    },
+    {
+      "epoch": 0.01778143650956987,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7977500000000002e-05,
+      "loss": 0.1971,
+      "step": 810
+    },
+    {
+      "epoch": 0.018000960417095424,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.79525e-05,
+      "loss": 0.193,
+      "step": 820
+    },
+    {
+      "epoch": 0.018220484324620978,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.79275e-05,
+      "loss": 0.2195,
+      "step": 830
+    },
+    {
+      "epoch": 0.018440008232146532,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.79025e-05,
+      "loss": 0.1954,
+      "step": 840
+    },
+    {
+      "epoch": 0.018659532139672086,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.78775e-05,
+      "loss": 0.1546,
+      "step": 850
+    },
+    {
+      "epoch": 0.01887905604719764,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7852500000000002e-05,
+      "loss": 0.188,
+      "step": 860
+    },
+    {
+      "epoch": 0.019098579954723194,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.78275e-05,
+      "loss": 0.2128,
+      "step": 870
+    },
+    {
+      "epoch": 0.019318103862248748,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.78025e-05,
+      "loss": 0.1846,
+      "step": 880
+    },
+    {
+      "epoch": 0.0195376277697743,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.77775e-05,
+      "loss": 0.1938,
+      "step": 890
+    },
+    {
+      "epoch": 0.019757151677299856,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.77525e-05,
+      "loss": 0.1781,
+      "step": 900
+    },
+    {
+      "epoch": 0.01997667558482541,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.7727500000000003e-05,
+      "loss": 0.1848,
+      "step": 910
+    },
+    {
+      "epoch": 0.020196199492350964,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.7702500000000002e-05,
+      "loss": 0.1739,
+      "step": 920
+    },
+    {
+      "epoch": 0.020415723399876518,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.76775e-05,
+      "loss": 0.1878,
+      "step": 930
+    },
+    {
+      "epoch": 0.02063524730740207,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.76525e-05,
+      "loss": 0.2028,
+      "step": 940
+    },
+    {
+      "epoch": 0.020854771214927625,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.76275e-05,
+      "loss": 0.1897,
+      "step": 950
+    },
+    {
+      "epoch": 0.02107429512245318,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.7602500000000003e-05,
+      "loss": 0.1748,
+      "step": 960
+    },
+    {
+      "epoch": 0.021293819029978733,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.75775e-05,
+      "loss": 0.1997,
+      "step": 970
+    },
+    {
+      "epoch": 0.021513342937504287,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7552500000000002e-05,
+      "loss": 0.1989,
+      "step": 980
+    },
+    {
+      "epoch": 0.02173286684502984,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.75275e-05,
+      "loss": 0.1836,
+      "step": 990
+    },
+    {
+      "epoch": 0.021952390752555395,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.7502500000000004e-05,
+      "loss": 0.2198,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02217191466008095,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.74775e-05,
+      "loss": 0.1847,
+      "step": 1010
+    },
+    {
+      "epoch": 0.022391438567606503,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.7452500000000003e-05,
+      "loss": 0.1954,
+      "step": 1020
+    },
+    {
+      "epoch": 0.022610962475132057,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7427500000000002e-05,
+      "loss": 0.2007,
+      "step": 1030
+    },
+    {
+      "epoch": 0.02283048638265761,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.74025e-05,
+      "loss": 0.2068,
+      "step": 1040
+    },
+    {
+      "epoch": 0.023050010290183165,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.73775e-05,
+      "loss": 0.1898,
+      "step": 1050
+    },
+    {
+      "epoch": 0.02326953419770872,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.73525e-05,
+      "loss": 0.1661,
+      "step": 1060
+    },
+    {
+      "epoch": 0.023489058105234273,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7327500000000003e-05,
+      "loss": 0.1745,
+      "step": 1070
+    },
+    {
+      "epoch": 0.023708582012759827,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7302500000000002e-05,
+      "loss": 0.1691,
+      "step": 1080
+    },
+    {
+      "epoch": 0.02392810592028538,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7277500000000002e-05,
+      "loss": 0.204,
+      "step": 1090
+    },
+    {
+      "epoch": 0.024147629827810935,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.72525e-05,
+      "loss": 0.1775,
+      "step": 1100
+    },
+    {
+      "epoch": 0.02436715373533649,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.72275e-05,
+      "loss": 0.1953,
+      "step": 1110
+    },
+    {
+      "epoch": 0.024586677642862043,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7202500000000003e-05,
+      "loss": 0.1916,
+      "step": 1120
+    },
+    {
+      "epoch": 0.024806201550387597,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7177500000000003e-05,
+      "loss": 0.1762,
+      "step": 1130
+    },
+    {
+      "epoch": 0.02502572545791315,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7152500000000002e-05,
+      "loss": 0.2267,
+      "step": 1140
+    },
+    {
+      "epoch": 0.025245249365438704,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.71275e-05,
+      "loss": 0.1924,
+      "step": 1150
+    },
+    {
+      "epoch": 0.02546477327296426,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.71025e-05,
+      "loss": 0.2081,
+      "step": 1160
+    },
+    {
+      "epoch": 0.025684297180489812,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7077500000000004e-05,
+      "loss": 0.1693,
+      "step": 1170
+    },
+    {
+      "epoch": 0.025903821088015366,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.70525e-05,
+      "loss": 0.2278,
+      "step": 1180
+    },
+    {
+      "epoch": 0.02612334499554092,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7027500000000003e-05,
+      "loss": 0.1819,
+      "step": 1190
+    },
+    {
+      "epoch": 0.026342868903066474,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.7002500000000002e-05,
+      "loss": 0.147,
+      "step": 1200
+    },
+    {
+      "epoch": 0.026562392810592028,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.69775e-05,
+      "loss": 0.1857,
+      "step": 1210
+    },
+    {
+      "epoch": 0.026781916718117582,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.69525e-05,
+      "loss": 0.1665,
+      "step": 1220
+    },
+    {
+      "epoch": 0.027001440625643136,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.69275e-05,
+      "loss": 0.2032,
+      "step": 1230
+    },
+    {
+      "epoch": 0.02722096453316869,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6902500000000003e-05,
+      "loss": 0.182,
+      "step": 1240
+    },
+    {
+      "epoch": 0.027440488440694244,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6877500000000002e-05,
+      "loss": 0.1936,
+      "step": 1250
+    },
+    {
+      "epoch": 0.027660012348219798,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6852500000000002e-05,
+      "loss": 0.1971,
+      "step": 1260
+    },
+    {
+      "epoch": 0.027879536255745352,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.68275e-05,
+      "loss": 0.1824,
+      "step": 1270
+    },
+    {
+      "epoch": 0.028099060163270906,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.68025e-05,
+      "loss": 0.1731,
+      "step": 1280
+    },
+    {
+      "epoch": 0.02831858407079646,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6777500000000003e-05,
+      "loss": 0.1818,
+      "step": 1290
+    },
+    {
+      "epoch": 0.028538107978322014,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.67525e-05,
+      "loss": 0.1959,
+      "step": 1300
+    },
+    {
+      "epoch": 0.028757631885847568,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.6727500000000002e-05,
+      "loss": 0.1777,
+      "step": 1310
+    },
+    {
+      "epoch": 0.02897715579337312,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.67025e-05,
+      "loss": 0.1719,
+      "step": 1320
+    },
+    {
+      "epoch": 0.029196679700898676,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.66775e-05,
+      "loss": 0.1985,
+      "step": 1330
+    },
+    {
+      "epoch": 0.02941620360842423,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.66525e-05,
+      "loss": 0.1841,
+      "step": 1340
+    },
+    {
+      "epoch": 0.029635727515949783,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.66275e-05,
+      "loss": 0.2174,
+      "step": 1350
+    },
+    {
+      "epoch": 0.029855251423475337,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.6602500000000002e-05,
+      "loss": 0.1925,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03007477533100089,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6577500000000002e-05,
+      "loss": 0.1686,
+      "step": 1370
+    },
+    {
+      "epoch": 0.030294299238526445,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.65525e-05,
+      "loss": 0.1877,
+      "step": 1380
+    },
+    {
+      "epoch": 0.030513823146052,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.65275e-05,
+      "loss": 0.2048,
+      "step": 1390
+    },
+    {
+      "epoch": 0.030733347053577553,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.65025e-05,
+      "loss": 0.2031,
+      "step": 1400
+    },
+    {
+      "epoch": 0.030952870961103107,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.6477500000000003e-05,
+      "loss": 0.217,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03117239486862866,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6452500000000002e-05,
+      "loss": 0.1783,
+      "step": 1420
+    },
+    {
+      "epoch": 0.031391918776154215,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.64275e-05,
+      "loss": 0.1761,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03161144268367977,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.64025e-05,
+      "loss": 0.182,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03183096659120532,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.63775e-05,
+      "loss": 0.1732,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03205049049873088,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.6352500000000003e-05,
+      "loss": 0.1951,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03227001440625643,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.63275e-05,
+      "loss": 0.2427,
+      "step": 1470
+    },
+    {
+      "epoch": 0.032489538313781985,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6302500000000002e-05,
+      "loss": 0.1904,
+      "step": 1480
+    },
+    {
+      "epoch": 0.03270906222130754,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.62775e-05,
+      "loss": 0.1938,
+      "step": 1490
+    },
+    {
+      "epoch": 0.03292858612883309,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.62525e-05,
+      "loss": 0.1673,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03314811003635865,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.62275e-05,
+      "loss": 0.2362,
+      "step": 1510
+    },
+    {
+      "epoch": 0.0333676339438842,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.62025e-05,
+      "loss": 0.1716,
+      "step": 1520
+    },
+    {
+      "epoch": 0.033587157851409755,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.6177500000000002e-05,
+      "loss": 0.1872,
+      "step": 1530
+    },
+    {
+      "epoch": 0.03380668175893531,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.61525e-05,
+      "loss": 0.2068,
+      "step": 1540
+    },
+    {
+      "epoch": 0.03402620566646086,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.61275e-05,
+      "loss": 0.1902,
+      "step": 1550
+    },
+    {
+      "epoch": 0.034245729573986416,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.61025e-05,
+      "loss": 0.1866,
+      "step": 1560
+    },
+    {
+      "epoch": 0.03446525348151197,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.60775e-05,
+      "loss": 0.1694,
+      "step": 1570
+    },
+    {
+      "epoch": 0.034684777389037524,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6052500000000003e-05,
+      "loss": 0.1915,
+      "step": 1580
+    },
+    {
+      "epoch": 0.03490430129656308,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.60275e-05,
+      "loss": 0.2041,
+      "step": 1590
+    },
+    {
+      "epoch": 0.03512382520408863,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.60025e-05,
+      "loss": 0.2597,
+      "step": 1600
+    },
+    {
+      "epoch": 0.035343349111614186,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.59775e-05,
+      "loss": 0.2025,
+      "step": 1610
+    },
+    {
+      "epoch": 0.03556287301913974,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5952500000000004e-05,
+      "loss": 0.178,
+      "step": 1620
+    },
+    {
+      "epoch": 0.035782396926665294,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.5927500000000003e-05,
+      "loss": 0.2145,
+      "step": 1630
+    },
+    {
+      "epoch": 0.03600192083419085,
+      "grad_norm": 0.625,
+      "learning_rate": 1.5902500000000002e-05,
+      "loss": 0.2196,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0362214447417164,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5877500000000002e-05,
+      "loss": 0.236,
+      "step": 1650
+    },
+    {
+      "epoch": 0.036440968649241956,
+      "grad_norm": 1.0,
+      "learning_rate": 1.58525e-05,
+      "loss": 0.1927,
+      "step": 1660
+    },
+    {
+      "epoch": 0.03666049255676751,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5827500000000004e-05,
+      "loss": 0.1991,
+      "step": 1670
+    },
+    {
+      "epoch": 0.036880016464293064,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.58025e-05,
+      "loss": 0.2282,
+      "step": 1680
+    },
+    {
+      "epoch": 0.03709954037181862,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5777500000000003e-05,
+      "loss": 0.1799,
+      "step": 1690
+    },
+    {
+      "epoch": 0.03731906427934417,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5752500000000002e-05,
+      "loss": 0.1936,
+      "step": 1700
+    },
+    {
+      "epoch": 0.037538588186869726,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.57275e-05,
+      "loss": 0.1886,
+      "step": 1710
+    },
+    {
+      "epoch": 0.03775811209439528,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.57025e-05,
+      "loss": 0.1771,
+      "step": 1720
+    },
+    {
+      "epoch": 0.037977636001920834,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.56775e-05,
+      "loss": 0.1853,
+      "step": 1730
+    },
+    {
+      "epoch": 0.03819715990944639,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5652500000000003e-05,
+      "loss": 0.1987,
+      "step": 1740
+    },
+    {
+      "epoch": 0.03841668381697194,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5627500000000002e-05,
+      "loss": 0.2049,
+      "step": 1750
+    },
+    {
+      "epoch": 0.038636207724497496,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.5602500000000002e-05,
+      "loss": 0.1691,
+      "step": 1760
+    },
+    {
+      "epoch": 0.03885573163202305,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.55775e-05,
+      "loss": 0.1981,
+      "step": 1770
+    },
+    {
+      "epoch": 0.0390752555395486,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.55525e-05,
+      "loss": 0.1842,
+      "step": 1780
+    },
+    {
+      "epoch": 0.03929477944707416,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.5527500000000003e-05,
+      "loss": 0.1696,
+      "step": 1790
+    },
+    {
+      "epoch": 0.03951430335459971,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.55025e-05,
+      "loss": 0.2018,
+      "step": 1800
+    },
+    {
+      "epoch": 0.039733827262125265,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5477500000000002e-05,
+      "loss": 0.201,
+      "step": 1810
+    },
+    {
+      "epoch": 0.03995335116965082,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.54525e-05,
+      "loss": 0.1855,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04017287507717637,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.54275e-05,
+      "loss": 0.1679,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04039239898470193,
+      "grad_norm": 0.875,
+      "learning_rate": 1.54025e-05,
+      "loss": 0.1916,
+      "step": 1840
+    },
+    {
+      "epoch": 0.04061192289222748,
+      "grad_norm": 0.75,
+      "learning_rate": 1.53775e-05,
+      "loss": 0.1715,
+      "step": 1850
+    },
+    {
+      "epoch": 0.040831446799753035,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5352500000000003e-05,
+      "loss": 0.1899,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04105097070727859,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5327500000000002e-05,
+      "loss": 0.193,
+      "step": 1870
+    },
+    {
+      "epoch": 0.04127049461480414,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.53025e-05,
+      "loss": 0.1822,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0414900185223297,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.52775e-05,
+      "loss": 0.1502,
+      "step": 1890
+    },
+    {
+      "epoch": 0.04170954242985525,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.5252500000000002e-05,
+      "loss": 0.1909,
+      "step": 1900
+    },
+    {
+      "epoch": 0.041929066337380805,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5227500000000001e-05,
+      "loss": 0.1853,
+      "step": 1910
+    },
+    {
+      "epoch": 0.04214859024490636,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5202500000000002e-05,
+      "loss": 0.1905,
+      "step": 1920
+    },
+    {
+      "epoch": 0.04236811415243191,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.51775e-05,
+      "loss": 0.1877,
+      "step": 1930
+    },
+    {
+      "epoch": 0.04258763805995747,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5152500000000001e-05,
+      "loss": 0.1947,
+      "step": 1940
+    },
+    {
+      "epoch": 0.04280716196748302,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5127500000000002e-05,
+      "loss": 0.1985,
+      "step": 1950
+    },
+    {
+      "epoch": 0.043026685875008575,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5102500000000002e-05,
+      "loss": 0.1846,
+      "step": 1960
+    },
+    {
+      "epoch": 0.04324620978253413,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5077500000000001e-05,
+      "loss": 0.2175,
+      "step": 1970
+    },
+    {
+      "epoch": 0.04346573369005968,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.50525e-05,
+      "loss": 0.203,
+      "step": 1980
+    },
+    {
+      "epoch": 0.043685257597585236,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5027500000000001e-05,
+      "loss": 0.1688,
+      "step": 1990
+    },
+    {
+      "epoch": 0.04390478150511079,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.5002500000000002e-05,
+      "loss": 0.1951,
+      "step": 2000
+    },
+    {
+      "epoch": 0.044124305412636344,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.49775e-05,
+      "loss": 0.1931,
+      "step": 2010
+    },
+    {
+      "epoch": 0.0443438293201619,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4952500000000001e-05,
+      "loss": 0.2017,
+      "step": 2020
+    },
+    {
+      "epoch": 0.04456335322768745,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.49275e-05,
+      "loss": 0.2095,
+      "step": 2030
+    },
+    {
+      "epoch": 0.044782877135213006,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4902500000000002e-05,
+      "loss": 0.1814,
+      "step": 2040
+    },
+    {
+      "epoch": 0.04500240104273856,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.48775e-05,
+      "loss": 0.1912,
+      "step": 2050
+    },
+    {
+      "epoch": 0.045221924950264114,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.48525e-05,
+      "loss": 0.1799,
+      "step": 2060
+    },
+    {
+      "epoch": 0.04544144885778967,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4827500000000002e-05,
+      "loss": 0.1754,
+      "step": 2070
+    },
+    {
+      "epoch": 0.04566097276531522,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.4802500000000003e-05,
+      "loss": 0.2019,
+      "step": 2080
+    },
+    {
+      "epoch": 0.045880496672840776,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.47775e-05,
+      "loss": 0.1767,
+      "step": 2090
+    },
+    {
+      "epoch": 0.04610002058036633,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4752500000000001e-05,
+      "loss": 0.1762,
+      "step": 2100
+    },
+    {
+      "epoch": 0.046319544487891884,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4727500000000001e-05,
+      "loss": 0.2086,
+      "step": 2110
+    },
+    {
+      "epoch": 0.04653906839541744,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.4702500000000002e-05,
+      "loss": 0.2083,
+      "step": 2120
+    },
+    {
+      "epoch": 0.04675859230294299,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4677500000000003e-05,
+      "loss": 0.1774,
+      "step": 2130
+    },
+    {
+      "epoch": 0.046978116210468546,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.46525e-05,
+      "loss": 0.1908,
+      "step": 2140
+    },
+    {
+      "epoch": 0.0471976401179941,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4627500000000002e-05,
+      "loss": 0.2223,
+      "step": 2150
+    },
+    {
+      "epoch": 0.047417164025519654,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.4602500000000001e-05,
+      "loss": 0.2084,
+      "step": 2160
+    },
+    {
+      "epoch": 0.04763668793304521,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.4577500000000002e-05,
+      "loss": 0.2029,
+      "step": 2170
+    },
+    {
+      "epoch": 0.04785621184057076,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.45525e-05,
+      "loss": 0.1641,
+      "step": 2180
+    },
+    {
+      "epoch": 0.048075735748096315,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4527500000000001e-05,
+      "loss": 0.2053,
+      "step": 2190
+    },
+    {
+      "epoch": 0.04829525965562187,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4502500000000002e-05,
+      "loss": 0.1783,
+      "step": 2200
+    },
+    {
+      "epoch": 0.04851478356314742,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.4477500000000002e-05,
+      "loss": 0.1733,
+      "step": 2210
+    },
+    {
+      "epoch": 0.04873430747067298,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.4452500000000001e-05,
+      "loss": 0.1852,
+      "step": 2220
+    },
+    {
+      "epoch": 0.04895383137819853,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.44275e-05,
+      "loss": 0.1626,
+      "step": 2230
+    },
+    {
+      "epoch": 0.049173355285724085,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4402500000000001e-05,
+      "loss": 0.1804,
+      "step": 2240
+    },
+    {
+      "epoch": 0.04939287919324964,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4377500000000003e-05,
+      "loss": 0.1895,
+      "step": 2250
+    },
+    {
+      "epoch": 0.04961240310077519,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.43525e-05,
+      "loss": 0.1911,
+      "step": 2260
+    },
+    {
+      "epoch": 0.04983192700830075,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.4327500000000001e-05,
+      "loss": 0.1903,
+      "step": 2270
+    },
+    {
+      "epoch": 0.0500514509158263,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.43025e-05,
+      "loss": 0.2024,
+      "step": 2280
+    },
+    {
+      "epoch": 0.050270974823351855,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4277500000000002e-05,
+      "loss": 0.1669,
+      "step": 2290
+    },
+    {
+      "epoch": 0.05049049873087741,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.42525e-05,
+      "loss": 0.1819,
+      "step": 2300
+    },
+    {
+      "epoch": 0.05071002263840296,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.42275e-05,
+      "loss": 0.1808,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05092954654592852,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4202500000000002e-05,
+      "loss": 0.1809,
+      "step": 2320
+    },
+    {
+      "epoch": 0.05114907045345407,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.4177500000000001e-05,
+      "loss": 0.177,
+      "step": 2330
+    },
+    {
+      "epoch": 0.051368594360979625,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.41525e-05,
+      "loss": 0.1882,
+      "step": 2340
+    },
+    {
+      "epoch": 0.05158811826850518,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.41275e-05,
+      "loss": 0.1705,
+      "step": 2350
+    },
+    {
+      "epoch": 0.05180764217603073,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4102500000000001e-05,
+      "loss": 0.1671,
+      "step": 2360
+    },
+    {
+      "epoch": 0.05202716608355629,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4077500000000002e-05,
+      "loss": 0.1622,
+      "step": 2370
+    },
+    {
+      "epoch": 0.05224668999108184,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4052500000000001e-05,
+      "loss": 0.1847,
+      "step": 2380
+    },
+    {
+      "epoch": 0.052466213898607394,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.40275e-05,
+      "loss": 0.1789,
+      "step": 2390
+    },
+    {
+      "epoch": 0.05268573780613295,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4002500000000002e-05,
+      "loss": 0.1815,
+      "step": 2400
+    },
+    {
+      "epoch": 0.0529052617136585,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.3977500000000001e-05,
+      "loss": 0.2352,
+      "step": 2410
+    },
+    {
+      "epoch": 0.053124785621184056,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3952500000000002e-05,
+      "loss": 0.1681,
+      "step": 2420
+    },
+    {
+      "epoch": 0.05334430952870961,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.39275e-05,
+      "loss": 0.2157,
+      "step": 2430
+    },
+    {
+      "epoch": 0.053563833436235164,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3902500000000001e-05,
+      "loss": 0.1829,
+      "step": 2440
+    },
+    {
+      "epoch": 0.05378335734376072,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.3877500000000002e-05,
+      "loss": 0.1778,
+      "step": 2450
+    },
+    {
+      "epoch": 0.05400288125128627,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3852500000000002e-05,
+      "loss": 0.2,
+      "step": 2460
+    },
+    {
+      "epoch": 0.054222405158811826,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3827500000000001e-05,
+      "loss": 0.1975,
+      "step": 2470
+    },
+    {
+      "epoch": 0.05444192906633738,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.38025e-05,
+      "loss": 0.1732,
+      "step": 2480
+    },
+    {
+      "epoch": 0.054661452973862934,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3777500000000001e-05,
+      "loss": 0.2128,
+      "step": 2490
+    },
+    {
+      "epoch": 0.05488097688138849,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3752500000000003e-05,
+      "loss": 0.2146,
+      "step": 2500
+    },
+    {
+      "epoch": 0.05510050078891404,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.37275e-05,
+      "loss": 0.1893,
+      "step": 2510
+    },
+    {
+      "epoch": 0.055320024696439596,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3702500000000001e-05,
+      "loss": 0.2053,
+      "step": 2520
+    },
+    {
+      "epoch": 0.05553954860396515,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.36775e-05,
+      "loss": 0.1867,
+      "step": 2530
+    },
+    {
+      "epoch": 0.055759072511490704,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3652500000000002e-05,
+      "loss": 0.2008,
+      "step": 2540
+    },
+    {
+      "epoch": 0.05597859641901626,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.36275e-05,
+      "loss": 0.1968,
+      "step": 2550
+    },
+    {
+      "epoch": 0.05619812032654181,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.36025e-05,
+      "loss": 0.199,
+      "step": 2560
+    },
+    {
+      "epoch": 0.056417644234067366,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3577500000000002e-05,
+      "loss": 0.1708,
+      "step": 2570
+    },
+    {
+      "epoch": 0.05663716814159292,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3552500000000001e-05,
+      "loss": 0.1923,
+      "step": 2580
+    },
+    {
+      "epoch": 0.056856692049118474,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.35275e-05,
+      "loss": 0.2035,
+      "step": 2590
+    },
+    {
+      "epoch": 0.05707621595664403,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.35025e-05,
+      "loss": 0.2349,
+      "step": 2600
+    },
+    {
+      "epoch": 0.05729573986416958,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3477500000000001e-05,
+      "loss": 0.1965,
+      "step": 2610
+    },
+    {
+      "epoch": 0.057515263771695135,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3452500000000002e-05,
+      "loss": 0.1683,
+      "step": 2620
+    },
+    {
+      "epoch": 0.05773478767922069,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.3427500000000001e-05,
+      "loss": 0.2099,
+      "step": 2630
+    },
+    {
+      "epoch": 0.05795431158674624,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3402500000000001e-05,
+      "loss": 0.1802,
+      "step": 2640
+    },
+    {
+      "epoch": 0.0581738354942718,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.33775e-05,
+      "loss": 0.1875,
+      "step": 2650
+    },
+    {
+      "epoch": 0.05839335940179735,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3352500000000001e-05,
+      "loss": 0.1849,
+      "step": 2660
+    },
+    {
+      "epoch": 0.058612883309322905,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3327500000000002e-05,
+      "loss": 0.1549,
+      "step": 2670
+    },
+    {
+      "epoch": 0.05883240721684846,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.33025e-05,
+      "loss": 0.1936,
+      "step": 2680
+    },
+    {
+      "epoch": 0.05905193112437401,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3277500000000001e-05,
+      "loss": 0.1891,
+      "step": 2690
+    },
+    {
+      "epoch": 0.05927145503189957,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3252500000000002e-05,
+      "loss": 0.1879,
+      "step": 2700
+    },
+    {
+      "epoch": 0.05949097893942512,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3227500000000002e-05,
+      "loss": 0.1838,
+      "step": 2710
+    },
+    {
+      "epoch": 0.059710502846950675,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3202500000000001e-05,
+      "loss": 0.1807,
+      "step": 2720
+    },
+    {
+      "epoch": 0.05993002675447623,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.31775e-05,
+      "loss": 0.1519,
+      "step": 2730
+    },
+    {
+      "epoch": 0.06014955066200178,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.3152500000000002e-05,
+      "loss": 0.1843,
+      "step": 2740
+    },
+    {
+      "epoch": 0.06036907456952734,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3127500000000003e-05,
+      "loss": 0.2063,
+      "step": 2750
+    },
+    {
+      "epoch": 0.06058859847705289,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.31025e-05,
+      "loss": 0.1785,
+      "step": 2760
+    },
+    {
+      "epoch": 0.060808122384578445,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3077500000000001e-05,
+      "loss": 0.1707,
+      "step": 2770
+    },
+    {
+      "epoch": 0.061027646292104,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.30525e-05,
+      "loss": 0.1761,
+      "step": 2780
+    },
+    {
+      "epoch": 0.06124717019962955,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3027500000000002e-05,
+      "loss": 0.1782,
+      "step": 2790
+    },
+    {
+      "epoch": 0.061466694107155107,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.30025e-05,
+      "loss": 0.1988,
+      "step": 2800
+    },
+    {
+      "epoch": 0.06168621801468066,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.29775e-05,
+      "loss": 0.1701,
+      "step": 2810
+    },
+    {
+      "epoch": 0.061905741922206214,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2952500000000002e-05,
+      "loss": 0.2079,
+      "step": 2820
+    },
+    {
+      "epoch": 0.06212526582973177,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.2927500000000001e-05,
+      "loss": 0.1982,
+      "step": 2830
+    },
+    {
+      "epoch": 0.06234478973725732,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.29025e-05,
+      "loss": 0.1798,
+      "step": 2840
+    },
+    {
+      "epoch": 0.06256431364478288,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.28775e-05,
+      "loss": 0.1582,
+      "step": 2850
+    },
+    {
+      "epoch": 0.06278383755230843,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2852500000000001e-05,
+      "loss": 0.2233,
+      "step": 2860
+    },
+    {
+      "epoch": 0.06300336145983398,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2827500000000002e-05,
+      "loss": 0.1952,
+      "step": 2870
+    },
+    {
+      "epoch": 0.06322288536735954,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2802500000000002e-05,
+      "loss": 0.1854,
+      "step": 2880
+    },
+    {
+      "epoch": 0.06344240927488509,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2777500000000001e-05,
+      "loss": 0.205,
+      "step": 2890
+    },
+    {
+      "epoch": 0.06366193318241065,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.27525e-05,
+      "loss": 0.1899,
+      "step": 2900
+    },
+    {
+      "epoch": 0.0638814570899362,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2727500000000001e-05,
+      "loss": 0.1874,
+      "step": 2910
+    },
+    {
+      "epoch": 0.06410098099746175,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2702500000000002e-05,
+      "loss": 0.186,
+      "step": 2920
+    },
+    {
+      "epoch": 0.06432050490498731,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.26775e-05,
+      "loss": 0.1619,
+      "step": 2930
+    },
+    {
+      "epoch": 0.06454002881251286,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2652500000000001e-05,
+      "loss": 0.1959,
+      "step": 2940
+    },
+    {
+      "epoch": 0.06475955272003842,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.26275e-05,
+      "loss": 0.1919,
+      "step": 2950
+    },
+    {
+      "epoch": 0.06497907662756397,
+      "grad_norm": 1.25,
+      "learning_rate": 1.2602500000000002e-05,
+      "loss": 0.1998,
+      "step": 2960
+    },
+    {
+      "epoch": 0.06519860053508952,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.25775e-05,
+      "loss": 0.2058,
+      "step": 2970
+    },
+    {
+      "epoch": 0.06541812444261508,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.25525e-05,
+      "loss": 0.1861,
+      "step": 2980
+    },
+    {
+      "epoch": 0.06563764835014063,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2527500000000002e-05,
+      "loss": 0.2199,
+      "step": 2990
+    },
+    {
+      "epoch": 0.06585717225766619,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2502500000000003e-05,
+      "loss": 0.1804,
+      "step": 3000
+    },
+    {
+      "epoch": 0.06607669616519174,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.24775e-05,
+      "loss": 0.1742,
+      "step": 3010
+    },
+    {
+      "epoch": 0.0662962200727173,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2452500000000001e-05,
+      "loss": 0.1782,
+      "step": 3020
+    },
+    {
+      "epoch": 0.06651574398024285,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.2427500000000001e-05,
+      "loss": 0.1822,
+      "step": 3030
+    },
+    {
+      "epoch": 0.0667352678877684,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.2402500000000002e-05,
+      "loss": 0.1902,
+      "step": 3040
+    },
+    {
+      "epoch": 0.06695479179529396,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.23775e-05,
+      "loss": 0.1937,
+      "step": 3050
+    },
+    {
+      "epoch": 0.06717431570281951,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.23525e-05,
+      "loss": 0.1919,
+      "step": 3060
+    },
+    {
+      "epoch": 0.06739383961034506,
+      "grad_norm": 5.625,
+      "learning_rate": 1.2327500000000002e-05,
+      "loss": 0.1763,
+      "step": 3070
+    },
+    {
+      "epoch": 0.06761336351787062,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2302500000000001e-05,
+      "loss": 0.1956,
+      "step": 3080
+    },
+    {
+      "epoch": 0.06783288742539617,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.22775e-05,
+      "loss": 0.2159,
+      "step": 3090
+    },
+    {
+      "epoch": 0.06805241133292173,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.22525e-05,
+      "loss": 0.1775,
+      "step": 3100
+    },
+    {
+      "epoch": 0.06827193524044728,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2227500000000001e-05,
+      "loss": 0.1617,
+      "step": 3110
+    },
+    {
+      "epoch": 0.06849145914797283,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2202500000000002e-05,
+      "loss": 0.1725,
+      "step": 3120
+    },
+    {
+      "epoch": 0.06871098305549839,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2177500000000002e-05,
+      "loss": 0.2037,
+      "step": 3130
+    },
+    {
+      "epoch": 0.06893050696302394,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.2152500000000001e-05,
+      "loss": 0.21,
+      "step": 3140
+    },
+    {
+      "epoch": 0.0691500308705495,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.21275e-05,
+      "loss": 0.1816,
+      "step": 3150
+    },
+    {
+      "epoch": 0.06936955477807505,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2102500000000001e-05,
+      "loss": 0.1811,
+      "step": 3160
+    },
+    {
+      "epoch": 0.0695890786856006,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.2077500000000003e-05,
+      "loss": 0.1659,
+      "step": 3170
+    },
+    {
+      "epoch": 0.06980860259312616,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.20525e-05,
+      "loss": 0.1924,
+      "step": 3180
+    },
+    {
+      "epoch": 0.07002812650065171,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2027500000000001e-05,
+      "loss": 0.1891,
+      "step": 3190
+    },
+    {
+      "epoch": 0.07024765040817726,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.20025e-05,
+      "loss": 0.1836,
+      "step": 3200
+    },
+    {
+      "epoch": 0.07046717431570282,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1977500000000002e-05,
+      "loss": 0.2071,
+      "step": 3210
+    },
+    {
+      "epoch": 0.07068669822322837,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.19525e-05,
+      "loss": 0.2144,
+      "step": 3220
+    },
+    {
+      "epoch": 0.07090622213075393,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.19275e-05,
+      "loss": 0.1881,
+      "step": 3230
+    },
+    {
+      "epoch": 0.07112574603827948,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.1902500000000002e-05,
+      "loss": 0.1751,
+      "step": 3240
+    },
+    {
+      "epoch": 0.07134526994580503,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1877500000000001e-05,
+      "loss": 0.1844,
+      "step": 3250
+    },
+    {
+      "epoch": 0.07156479385333059,
+      "grad_norm": 1.0,
+      "learning_rate": 1.18525e-05,
+      "loss": 0.177,
+      "step": 3260
+    },
+    {
+      "epoch": 0.07178431776085614,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.18275e-05,
+      "loss": 0.1831,
+      "step": 3270
+    },
+    {
+      "epoch": 0.0720038416683817,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.1802500000000001e-05,
+      "loss": 0.1793,
+      "step": 3280
+    },
+    {
+      "epoch": 0.07222336557590725,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1777500000000002e-05,
+      "loss": 0.1824,
+      "step": 3290
+    },
+    {
+      "epoch": 0.0724428894834328,
+      "grad_norm": 0.875,
+      "learning_rate": 1.17525e-05,
+      "loss": 0.2064,
+      "step": 3300
+    },
+    {
+      "epoch": 0.07266241339095836,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.17275e-05,
+      "loss": 0.1824,
+      "step": 3310
+    },
+    {
+      "epoch": 0.07288193729848391,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1702500000000002e-05,
+      "loss": 0.1792,
+      "step": 3320
+    },
+    {
+      "epoch": 0.07310146120600947,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.1677500000000001e-05,
+      "loss": 0.1882,
+      "step": 3330
+    },
+    {
+      "epoch": 0.07332098511353502,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.16525e-05,
+      "loss": 0.1945,
+      "step": 3340
+    },
+    {
+      "epoch": 0.07354050902106057,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.16275e-05,
+      "loss": 0.2059,
+      "step": 3350
+    },
+    {
+      "epoch": 0.07376003292858613,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1602500000000001e-05,
+      "loss": 0.2108,
+      "step": 3360
+    },
+    {
+      "epoch": 0.07397955683611168,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1577500000000002e-05,
+      "loss": 0.1833,
+      "step": 3370
+    },
+    {
+      "epoch": 0.07419908074363724,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.1552500000000002e-05,
+      "loss": 0.1719,
+      "step": 3380
+    },
+    {
+      "epoch": 0.07441860465116279,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1527500000000001e-05,
+      "loss": 0.191,
+      "step": 3390
+    },
+    {
+      "epoch": 0.07463812855868834,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.15025e-05,
+      "loss": 0.1933,
+      "step": 3400
+    },
+    {
+      "epoch": 0.0748576524662139,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.1477500000000001e-05,
+      "loss": 0.2048,
+      "step": 3410
+    },
+    {
+      "epoch": 0.07507717637373945,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1452500000000003e-05,
+      "loss": 0.1669,
+      "step": 3420
+    },
+    {
+      "epoch": 0.075296700281265,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.14275e-05,
+      "loss": 0.1684,
+      "step": 3430
+    },
+    {
+      "epoch": 0.07551622418879056,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1402500000000001e-05,
+      "loss": 0.1744,
+      "step": 3440
+    },
+    {
+      "epoch": 0.07573574809631611,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.13775e-05,
+      "loss": 0.1822,
+      "step": 3450
+    },
+    {
+      "epoch": 0.07595527200384167,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1352500000000002e-05,
+      "loss": 0.1808,
+      "step": 3460
+    },
+    {
+      "epoch": 0.07617479591136722,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.13275e-05,
+      "loss": 0.1872,
+      "step": 3470
+    },
+    {
+      "epoch": 0.07639431981889278,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.13025e-05,
+      "loss": 0.192,
+      "step": 3480
+    },
+    {
+      "epoch": 0.07661384372641833,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1277500000000002e-05,
+      "loss": 0.219,
+      "step": 3490
+    },
+    {
+      "epoch": 0.07683336763394388,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.1252500000000001e-05,
+      "loss": 0.2029,
+      "step": 3500
+    },
+    {
+      "epoch": 0.07705289154146944,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.12275e-05,
+      "loss": 0.1787,
+      "step": 3510
+    },
+    {
+      "epoch": 0.07727241544899499,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.12025e-05,
+      "loss": 0.1613,
+      "step": 3520
+    },
+    {
+      "epoch": 0.07749193935652054,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.1177500000000001e-05,
+      "loss": 0.1823,
+      "step": 3530
+    },
+    {
+      "epoch": 0.0777114632640461,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1152500000000002e-05,
+      "loss": 0.1823,
+      "step": 3540
+    },
+    {
+      "epoch": 0.07793098717157165,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.11275e-05,
+      "loss": 0.194,
+      "step": 3550
+    },
+    {
+      "epoch": 0.0781505110790972,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.1102500000000001e-05,
+      "loss": 0.202,
+      "step": 3560
+    },
+    {
+      "epoch": 0.07837003498662276,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.10775e-05,
+      "loss": 0.1599,
+      "step": 3570
+    },
+    {
+      "epoch": 0.07858955889414831,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.1052500000000001e-05,
+      "loss": 0.183,
+      "step": 3580
+    },
+    {
+      "epoch": 0.07880908280167387,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1027499999999999e-05,
+      "loss": 0.1697,
+      "step": 3590
+    },
+    {
+      "epoch": 0.07902860670919942,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.10025e-05,
+      "loss": 0.1872,
+      "step": 3600
+    },
+    {
+      "epoch": 0.07924813061672498,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.0977500000000001e-05,
+      "loss": 0.2008,
+      "step": 3610
+    },
+    {
+      "epoch": 0.07946765452425053,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.0952500000000002e-05,
+      "loss": 0.1857,
+      "step": 3620
+    },
+    {
+      "epoch": 0.07968717843177608,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.0927500000000002e-05,
+      "loss": 0.2013,
+      "step": 3630
+    },
+    {
+      "epoch": 0.07990670233930164,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0902500000000001e-05,
+      "loss": 0.1996,
+      "step": 3640
+    },
+    {
+      "epoch": 0.08012622624682719,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.08775e-05,
+      "loss": 0.1866,
+      "step": 3650
+    },
+    {
+      "epoch": 0.08034575015435275,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.0852500000000002e-05,
+      "loss": 0.1854,
+      "step": 3660
+    },
+    {
+      "epoch": 0.0805652740618783,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0827500000000003e-05,
+      "loss": 0.2041,
+      "step": 3670
+    },
+    {
+      "epoch": 0.08078479796940385,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.08025e-05,
+      "loss": 0.2103,
+      "step": 3680
+    },
+    {
+      "epoch": 0.08100432187692941,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0777500000000001e-05,
+      "loss": 0.1689,
+      "step": 3690
+    },
+    {
+      "epoch": 0.08122384578445496,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.07525e-05,
+      "loss": 0.2075,
+      "step": 3700
+    },
+    {
+      "epoch": 0.08144336969198052,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.0727500000000002e-05,
+      "loss": 0.1958,
+      "step": 3710
+    },
+    {
+      "epoch": 0.08166289359950607,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.07025e-05,
+      "loss": 0.2133,
+      "step": 3720
+    },
+    {
+      "epoch": 0.08188241750703162,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.06775e-05,
+      "loss": 0.2199,
+      "step": 3730
+    },
+    {
+      "epoch": 0.08210194141455718,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0652500000000002e-05,
+      "loss": 0.1816,
+      "step": 3740
+    },
+    {
+      "epoch": 0.08232146532208273,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0627500000000001e-05,
+      "loss": 0.1618,
+      "step": 3750
+    },
+    {
+      "epoch": 0.08254098922960829,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.06025e-05,
+      "loss": 0.2193,
+      "step": 3760
+    },
+    {
+      "epoch": 0.08276051313713384,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.05775e-05,
+      "loss": 0.1626,
+      "step": 3770
+    },
+    {
+      "epoch": 0.0829800370446594,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.0552500000000001e-05,
+      "loss": 0.1873,
+      "step": 3780
+    },
+    {
+      "epoch": 0.08319956095218495,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.0527500000000002e-05,
+      "loss": 0.2129,
+      "step": 3790
+    },
+    {
+      "epoch": 0.0834190848597105,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.05025e-05,
+      "loss": 0.1906,
+      "step": 3800
+    },
+    {
+      "epoch": 0.08363860876723606,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0477500000000001e-05,
+      "loss": 0.1711,
+      "step": 3810
+    },
+    {
+      "epoch": 0.08385813267476161,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.04525e-05,
+      "loss": 0.1813,
+      "step": 3820
+    },
+    {
+      "epoch": 0.08407765658228716,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0427500000000001e-05,
+      "loss": 0.1996,
+      "step": 3830
+    },
+    {
+      "epoch": 0.08429718048981272,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0402499999999999e-05,
+      "loss": 0.1536,
+      "step": 3840
+    },
+    {
+      "epoch": 0.08451670439733827,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.03775e-05,
+      "loss": 0.1617,
+      "step": 3850
+    },
+    {
+      "epoch": 0.08473622830486383,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.0352500000000001e-05,
+      "loss": 0.1874,
+      "step": 3860
+    },
+    {
+      "epoch": 0.08495575221238938,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.03275e-05,
+      "loss": 0.2133,
+      "step": 3870
+    },
+    {
+      "epoch": 0.08517527611991493,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.0302500000000002e-05,
+      "loss": 0.1796,
+      "step": 3880
+    },
+    {
+      "epoch": 0.08539480002744049,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.02775e-05,
+      "loss": 0.1698,
+      "step": 3890
+    },
+    {
+      "epoch": 0.08561432393496604,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.02525e-05,
+      "loss": 0.1938,
+      "step": 3900
+    },
+    {
+      "epoch": 0.0858338478424916,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0227500000000002e-05,
+      "loss": 0.1816,
+      "step": 3910
+    },
+    {
+      "epoch": 0.08605337175001715,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0202500000000003e-05,
+      "loss": 0.1677,
+      "step": 3920
+    },
+    {
+      "epoch": 0.0862728956575427,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.01775e-05,
+      "loss": 0.1956,
+      "step": 3930
+    },
+    {
+      "epoch": 0.08649241956506826,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.0152500000000001e-05,
+      "loss": 0.1769,
+      "step": 3940
+    },
+    {
+      "epoch": 0.08671194347259381,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0127500000000001e-05,
+      "loss": 0.2014,
+      "step": 3950
+    },
+    {
+      "epoch": 0.08693146738011936,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0102500000000002e-05,
+      "loss": 0.2199,
+      "step": 3960
+    },
+    {
+      "epoch": 0.08715099128764492,
+      "grad_norm": 0.75,
+      "learning_rate": 1.00775e-05,
+      "loss": 0.1815,
+      "step": 3970
+    },
+    {
+      "epoch": 0.08737051519517047,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.00525e-05,
+      "loss": 0.1797,
+      "step": 3980
+    },
+    {
+      "epoch": 0.08759003910269603,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.0027500000000002e-05,
+      "loss": 0.2142,
+      "step": 3990
+    },
+    {
+      "epoch": 0.08780956301022158,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0002500000000001e-05,
+      "loss": 0.2015,
+      "step": 4000
+    },
+    {
+      "epoch": 0.08802908691774713,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.9775e-06,
+      "loss": 0.1653,
+      "step": 4010
+    },
+    {
+      "epoch": 0.08824861082527269,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.9525e-06,
+      "loss": 0.1784,
+      "step": 4020
+    },
+    {
+      "epoch": 0.08846813473279824,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.927500000000001e-06,
+      "loss": 0.1915,
+      "step": 4030
+    },
+    {
+      "epoch": 0.0886876586403238,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.9025e-06,
+      "loss": 0.1997,
+      "step": 4040
+    },
+    {
+      "epoch": 0.08890718254784935,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.877500000000002e-06,
+      "loss": 0.1816,
+      "step": 4050
+    },
+    {
+      "epoch": 0.0891267064553749,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.852500000000001e-06,
+      "loss": 0.1954,
+      "step": 4060
+    },
+    {
+      "epoch": 0.08934623036290046,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.8275e-06,
+      "loss": 0.1532,
+      "step": 4070
+    },
+    {
+      "epoch": 0.08956575427042601,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.8025e-06,
+      "loss": 0.2117,
+      "step": 4080
+    },
+    {
+      "epoch": 0.08978527817795157,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.7775e-06,
+      "loss": 0.1957,
+      "step": 4090
+    },
+    {
+      "epoch": 0.09000480208547712,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.7525e-06,
+      "loss": 0.1683,
+      "step": 4100
+    },
+    {
+      "epoch": 0.09022432599300267,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.727500000000001e-06,
+      "loss": 0.202,
+      "step": 4110
+    },
+    {
+      "epoch": 0.09044384990052823,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.7025e-06,
+      "loss": 0.1806,
+      "step": 4120
+    },
+    {
+      "epoch": 0.09066337380805378,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.6775e-06,
+      "loss": 0.1941,
+      "step": 4130
+    },
+    {
+      "epoch": 0.09088289771557934,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.652500000000001e-06,
+      "loss": 0.1824,
+      "step": 4140
+    },
+    {
+      "epoch": 0.09110242162310489,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.6275e-06,
+      "loss": 0.1489,
+      "step": 4150
+    },
+    {
+      "epoch": 0.09132194553063044,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.602500000000002e-06,
+      "loss": 0.1578,
+      "step": 4160
+    },
+    {
+      "epoch": 0.091541469438156,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.577500000000001e-06,
+      "loss": 0.1837,
+      "step": 4170
+    },
+    {
+      "epoch": 0.09176099334568155,
+      "grad_norm": 0.75,
+      "learning_rate": 9.5525e-06,
+      "loss": 0.1943,
+      "step": 4180
+    },
+    {
+      "epoch": 0.0919805172532071,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.5275e-06,
+      "loss": 0.1938,
+      "step": 4190
+    },
+    {
+      "epoch": 0.09220004116073266,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.502500000000001e-06,
+      "loss": 0.1722,
+      "step": 4200
+    },
+    {
+      "epoch": 0.09241956506825821,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.4775e-06,
+      "loss": 0.1893,
+      "step": 4210
+    },
+    {
+      "epoch": 0.09263908897578377,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.452500000000001e-06,
+      "loss": 0.1843,
+      "step": 4220
+    },
+    {
+      "epoch": 0.09285861288330932,
+      "grad_norm": 0.75,
+      "learning_rate": 9.4275e-06,
+      "loss": 0.2091,
+      "step": 4230
+    },
+    {
+      "epoch": 0.09307813679083488,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.402500000000002e-06,
+      "loss": 0.1744,
+      "step": 4240
+    },
+    {
+      "epoch": 0.09329766069836043,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.377500000000001e-06,
+      "loss": 0.1934,
+      "step": 4250
+    },
+    {
+      "epoch": 0.09351718460588598,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.3525e-06,
+      "loss": 0.1982,
+      "step": 4260
+    },
+    {
+      "epoch": 0.09373670851341154,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.3275e-06,
+      "loss": 0.1705,
+      "step": 4270
+    },
+    {
+      "epoch": 0.09395623242093709,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.302500000000001e-06,
+      "loss": 0.1701,
+      "step": 4280
+    },
+    {
+      "epoch": 0.09417575632846265,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.2775e-06,
+      "loss": 0.1956,
+      "step": 4290
+    },
+    {
+      "epoch": 0.0943952802359882,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.252500000000002e-06,
+      "loss": 0.1792,
+      "step": 4300
+    },
+    {
+      "epoch": 0.09461480414351375,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.227500000000001e-06,
+      "loss": 0.1744,
+      "step": 4310
+    },
+    {
+      "epoch": 0.09483432805103931,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.2025e-06,
+      "loss": 0.2185,
+      "step": 4320
+    },
+    {
+      "epoch": 0.09505385195856486,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.1775e-06,
+      "loss": 0.1655,
+      "step": 4330
+    },
+    {
+      "epoch": 0.09527337586609042,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.152500000000001e-06,
+      "loss": 0.2083,
+      "step": 4340
+    },
+    {
+      "epoch": 0.09549289977361597,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.1275e-06,
+      "loss": 0.2039,
+      "step": 4350
+    },
+    {
+      "epoch": 0.09571242368114152,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.102500000000001e-06,
+      "loss": 0.1923,
+      "step": 4360
+    },
+    {
+      "epoch": 0.09593194758866708,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.0775e-06,
+      "loss": 0.2013,
+      "step": 4370
+    },
+    {
+      "epoch": 0.09615147149619263,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.0525e-06,
+      "loss": 0.1856,
+      "step": 4380
+    },
+    {
+      "epoch": 0.09637099540371818,
+      "grad_norm": 0.75,
+      "learning_rate": 9.027500000000001e-06,
+      "loss": 0.2029,
+      "step": 4390
+    },
+    {
+      "epoch": 0.09659051931124374,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.0025e-06,
+      "loss": 0.1753,
+      "step": 4400
+    },
+    {
+      "epoch": 0.09681004321876929,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.977500000000002e-06,
+      "loss": 0.1767,
+      "step": 4410
+    },
+    {
+      "epoch": 0.09702956712629485,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.952500000000001e-06,
+      "loss": 0.1727,
+      "step": 4420
+    },
+    {
+      "epoch": 0.0972490910338204,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.9275e-06,
+      "loss": 0.1802,
+      "step": 4430
+    },
+    {
+      "epoch": 0.09746861494134595,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.9025e-06,
+      "loss": 0.217,
+      "step": 4440
+    },
+    {
+      "epoch": 0.09768813884887151,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.877500000000001e-06,
+      "loss": 0.1644,
+      "step": 4450
+    },
+    {
+      "epoch": 0.09790766275639706,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.8525e-06,
+      "loss": 0.195,
+      "step": 4460
+    },
+    {
+      "epoch": 0.09812718666392262,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.827500000000001e-06,
+      "loss": 0.1872,
+      "step": 4470
+    },
+    {
+      "epoch": 0.09834671057144817,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.802500000000001e-06,
+      "loss": 0.196,
+      "step": 4480
+    },
+    {
+      "epoch": 0.09856623447897372,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.7775e-06,
+      "loss": 0.186,
+      "step": 4490
+    },
+    {
+      "epoch": 0.09878575838649928,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.7525e-06,
+      "loss": 0.1871,
+      "step": 4500
+    },
+    {
+      "epoch": 0.09900528229402483,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.7275e-06,
+      "loss": 0.189,
+      "step": 4510
+    },
+    {
+      "epoch": 0.09922480620155039,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.7025e-06,
+      "loss": 0.1596,
+      "step": 4520
+    },
+    {
+      "epoch": 0.09944433010907594,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.677500000000001e-06,
+      "loss": 0.1781,
+      "step": 4530
+    },
+    {
+      "epoch": 0.0996638540166015,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.6525e-06,
+      "loss": 0.1879,
+      "step": 4540
+    },
+    {
+      "epoch": 0.09988337792412705,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.627500000000002e-06,
+      "loss": 0.2002,
+      "step": 4550
+    },
+    {
+      "epoch": 0.1001029018316526,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.602500000000001e-06,
+      "loss": 0.2004,
+      "step": 4560
+    },
+    {
+      "epoch": 0.10032242573917816,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.5775e-06,
+      "loss": 0.1844,
+      "step": 4570
+    },
+    {
+      "epoch": 0.10054194964670371,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.5525e-06,
+      "loss": 0.1701,
+      "step": 4580
+    },
+    {
+      "epoch": 0.10076147355422926,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.527500000000001e-06,
+      "loss": 0.2176,
+      "step": 4590
+    },
+    {
+      "epoch": 0.10098099746175482,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.5025e-06,
+      "loss": 0.1736,
+      "step": 4600
+    },
+    {
+      "epoch": 0.10120052136928037,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.477500000000001e-06,
+      "loss": 0.1813,
+      "step": 4610
+    },
+    {
+      "epoch": 0.10142004527680593,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.4525e-06,
+      "loss": 0.1878,
+      "step": 4620
+    },
+    {
+      "epoch": 0.10163956918433148,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.4275e-06,
+      "loss": 0.1796,
+      "step": 4630
+    },
+    {
+      "epoch": 0.10185909309185703,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.402500000000001e-06,
+      "loss": 0.1702,
+      "step": 4640
+    },
+    {
+      "epoch": 0.10207861699938259,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.3775e-06,
+      "loss": 0.1689,
+      "step": 4650
+    },
+    {
+      "epoch": 0.10229814090690814,
+      "grad_norm": 0.86328125,
+      "learning_rate": 8.352500000000002e-06,
+      "loss": 0.1922,
+      "step": 4660
+    },
+    {
+      "epoch": 0.1025176648144337,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.327500000000001e-06,
+      "loss": 0.1813,
+      "step": 4670
+    },
+    {
+      "epoch": 0.10273718872195925,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.3025e-06,
+      "loss": 0.1486,
+      "step": 4680
+    },
+    {
+      "epoch": 0.1029567126294848,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.2775e-06,
+      "loss": 0.1606,
+      "step": 4690
+    },
+    {
+      "epoch": 0.10317623653701036,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.252500000000001e-06,
+      "loss": 0.2014,
+      "step": 4700
+    },
+    {
+      "epoch": 0.10339576044453591,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.2275e-06,
+      "loss": 0.1757,
+      "step": 4710
+    },
+    {
+      "epoch": 0.10361528435206147,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.202500000000002e-06,
+      "loss": 0.1895,
+      "step": 4720
+    },
+    {
+      "epoch": 0.10383480825958702,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.177500000000001e-06,
+      "loss": 0.1753,
+      "step": 4730
+    },
+    {
+      "epoch": 0.10405433216711257,
+      "grad_norm": 0.75,
+      "learning_rate": 8.1525e-06,
+      "loss": 0.1705,
+      "step": 4740
+    },
+    {
+      "epoch": 0.10427385607463813,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.1275e-06,
+      "loss": 0.217,
+      "step": 4750
+    },
+    {
+      "epoch": 0.10449337998216368,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.1025e-06,
+      "loss": 0.2155,
+      "step": 4760
+    },
+    {
+      "epoch": 0.10471290388968924,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.0775e-06,
+      "loss": 0.1972,
+      "step": 4770
+    },
+    {
+      "epoch": 0.10493242779721479,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.052500000000001e-06,
+      "loss": 0.1927,
+      "step": 4780
+    },
+    {
+      "epoch": 0.10515195170474034,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.0275e-06,
+      "loss": 0.2067,
+      "step": 4790
+    },
+    {
+      "epoch": 0.1053714756122659,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.0025e-06,
+      "loss": 0.1802,
+      "step": 4800
+    },
+    {
+      "epoch": 0.10559099951979145,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.9775e-06,
+      "loss": 0.1808,
+      "step": 4810
+    },
+    {
+      "epoch": 0.105810523427317,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.9525e-06,
+      "loss": 0.1641,
+      "step": 4820
+    },
+    {
+      "epoch": 0.10603004733484256,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.9275e-06,
+      "loss": 0.175,
+      "step": 4830
+    },
+    {
+      "epoch": 0.10624957124236811,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.902500000000001e-06,
+      "loss": 0.1984,
+      "step": 4840
+    },
+    {
+      "epoch": 0.10646909514989367,
+      "grad_norm": 1.0,
+      "learning_rate": 7.8775e-06,
+      "loss": 0.182,
+      "step": 4850
+    },
+    {
+      "epoch": 0.10668861905741922,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.852500000000001e-06,
+      "loss": 0.1785,
+      "step": 4860
+    },
+    {
+      "epoch": 0.10690814296494477,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.827500000000001e-06,
+      "loss": 0.1917,
+      "step": 4870
+    },
+    {
+      "epoch": 0.10712766687247033,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.8025e-06,
+      "loss": 0.1747,
+      "step": 4880
+    },
+    {
+      "epoch": 0.10734719077999588,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.777500000000001e-06,
+      "loss": 0.1806,
+      "step": 4890
+    },
+    {
+      "epoch": 0.10756671468752144,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.7525e-06,
+      "loss": 0.1672,
+      "step": 4900
+    },
+    {
+      "epoch": 0.10778623859504699,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.727500000000002e-06,
+      "loss": 0.2223,
+      "step": 4910
+    },
+    {
+      "epoch": 0.10800576250257254,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.702500000000001e-06,
+      "loss": 0.1946,
+      "step": 4920
+    },
+    {
+      "epoch": 0.1082252864100981,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.6775e-06,
+      "loss": 0.1682,
+      "step": 4930
+    },
+    {
+      "epoch": 0.10844481031762365,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.6525e-06,
+      "loss": 0.1494,
+      "step": 4940
+    },
+    {
+      "epoch": 0.1086643342251492,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.627500000000001e-06,
+      "loss": 0.1856,
+      "step": 4950
+    },
+    {
+      "epoch": 0.10888385813267476,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.6025000000000005e-06,
+      "loss": 0.1922,
+      "step": 4960
+    },
+    {
+      "epoch": 0.10910338204020031,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.577500000000001e-06,
+      "loss": 0.1919,
+      "step": 4970
+    },
+    {
+      "epoch": 0.10932290594772587,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.5525e-06,
+      "loss": 0.1758,
+      "step": 4980
+    },
+    {
+      "epoch": 0.10954242985525142,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.527500000000001e-06,
+      "loss": 0.2053,
+      "step": 4990
+    },
+    {
+      "epoch": 0.10976195376277698,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.502500000000001e-06,
+      "loss": 0.1742,
+      "step": 5000
+    },
+    {
+      "epoch": 0.10998147767030253,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.477500000000001e-06,
+      "loss": 0.2089,
+      "step": 5010
+    },
+    {
+      "epoch": 0.11020100157782808,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.4525e-06,
+      "loss": 0.1788,
+      "step": 5020
+    },
+    {
+      "epoch": 0.11042052548535364,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.4275000000000005e-06,
+      "loss": 0.1981,
+      "step": 5030
+    },
+    {
+      "epoch": 0.11064004939287919,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.4025e-06,
+      "loss": 0.1708,
+      "step": 5040
+    },
+    {
+      "epoch": 0.11085957330040475,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.377500000000001e-06,
+      "loss": 0.1621,
+      "step": 5050
+    },
+    {
+      "epoch": 0.1110790972079303,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.3525e-06,
+      "loss": 0.1805,
+      "step": 5060
+    },
+    {
+      "epoch": 0.11129862111545585,
+      "grad_norm": 0.8046875,
+      "learning_rate": 7.3275000000000006e-06,
+      "loss": 0.201,
+      "step": 5070
+    },
+    {
+      "epoch": 0.11151814502298141,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.3025e-06,
+      "loss": 0.2034,
+      "step": 5080
+    },
+    {
+      "epoch": 0.11173766893050696,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.277500000000001e-06,
+      "loss": 0.1858,
+      "step": 5090
+    },
+    {
+      "epoch": 0.11195719283803252,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.2525000000000004e-06,
+      "loss": 0.2008,
+      "step": 5100
+    },
+    {
+      "epoch": 0.11217671674555807,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.227500000000001e-06,
+      "loss": 0.1954,
+      "step": 5110
+    },
+    {
+      "epoch": 0.11239624065308362,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.2025e-06,
+      "loss": 0.1638,
+      "step": 5120
+    },
+    {
+      "epoch": 0.11261576456060918,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.1775e-06,
+      "loss": 0.1845,
+      "step": 5130
+    },
+    {
+      "epoch": 0.11283528846813473,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.152500000000001e-06,
+      "loss": 0.1953,
+      "step": 5140
+    },
+    {
+      "epoch": 0.11305481237566029,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.127500000000001e-06,
+      "loss": 0.1937,
+      "step": 5150
+    },
+    {
+      "epoch": 0.11327433628318584,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.102500000000001e-06,
+      "loss": 0.1736,
+      "step": 5160
+    },
+    {
+      "epoch": 0.1134938601907114,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.0775000000000004e-06,
+      "loss": 0.2043,
+      "step": 5170
+    },
+    {
+      "epoch": 0.11371338409823695,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.052500000000001e-06,
+      "loss": 0.1758,
+      "step": 5180
+    },
+    {
+      "epoch": 0.1139329080057625,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.0275e-06,
+      "loss": 0.1895,
+      "step": 5190
+    },
+    {
+      "epoch": 0.11415243191328805,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.002500000000001e-06,
+      "loss": 0.2119,
+      "step": 5200
+    },
+    {
+      "epoch": 0.11437195582081361,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.9775000000000005e-06,
+      "loss": 0.1839,
+      "step": 5210
+    },
+    {
+      "epoch": 0.11459147972833916,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.952500000000001e-06,
+      "loss": 0.186,
+      "step": 5220
+    },
+    {
+      "epoch": 0.11481100363586472,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.9275e-06,
+      "loss": 0.1887,
+      "step": 5230
+    },
+    {
+      "epoch": 0.11503052754339027,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.902500000000001e-06,
+      "loss": 0.1916,
+      "step": 5240
+    },
+    {
+      "epoch": 0.11525005145091582,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.877500000000001e-06,
+      "loss": 0.1755,
+      "step": 5250
+    },
+    {
+      "epoch": 0.11546957535844138,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.852500000000001e-06,
+      "loss": 0.2005,
+      "step": 5260
+    },
+    {
+      "epoch": 0.11568909926596693,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.8275e-06,
+      "loss": 0.229,
+      "step": 5270
+    },
+    {
+      "epoch": 0.11590862317349249,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.8025000000000005e-06,
+      "loss": 0.1747,
+      "step": 5280
+    },
+    {
+      "epoch": 0.11612814708101804,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.7775e-06,
+      "loss": 0.1735,
+      "step": 5290
+    },
+    {
+      "epoch": 0.1163476709885436,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.752500000000001e-06,
+      "loss": 0.1971,
+      "step": 5300
+    },
+    {
+      "epoch": 0.11656719489606915,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.7275e-06,
+      "loss": 0.1885,
+      "step": 5310
+    },
+    {
+      "epoch": 0.1167867188035947,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.702500000000001e-06,
+      "loss": 0.1801,
+      "step": 5320
+    },
+    {
+      "epoch": 0.11700624271112026,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.6775e-06,
+      "loss": 0.2179,
+      "step": 5330
+    },
+    {
+      "epoch": 0.11722576661864581,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.6525e-06,
+      "loss": 0.1607,
+      "step": 5340
+    },
+    {
+      "epoch": 0.11744529052617136,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.6275e-06,
+      "loss": 0.1931,
+      "step": 5350
+    },
+    {
+      "epoch": 0.11766481443369692,
+      "grad_norm": 0.85546875,
+      "learning_rate": 6.602500000000001e-06,
+      "loss": 0.2009,
+      "step": 5360
+    },
+    {
+      "epoch": 0.11788433834122247,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.5775e-06,
+      "loss": 0.1551,
+      "step": 5370
+    },
+    {
+      "epoch": 0.11810386224874803,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.5525e-06,
+      "loss": 0.1775,
+      "step": 5380
+    },
+    {
+      "epoch": 0.11832338615627358,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.5275000000000015e-06,
+      "loss": 0.1673,
+      "step": 5390
+    },
+    {
+      "epoch": 0.11854291006379913,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.502500000000001e-06,
+      "loss": 0.1785,
+      "step": 5400
+    },
+    {
+      "epoch": 0.11876243397132469,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.477500000000001e-06,
+      "loss": 0.1902,
+      "step": 5410
+    },
+    {
+      "epoch": 0.11898195787885024,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.4525000000000005e-06,
+      "loss": 0.1794,
+      "step": 5420
+    },
+    {
+      "epoch": 0.1192014817863758,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.427500000000001e-06,
+      "loss": 0.1954,
+      "step": 5430
+    },
+    {
+      "epoch": 0.11942100569390135,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.4025e-06,
+      "loss": 0.21,
+      "step": 5440
+    },
+    {
+      "epoch": 0.1196405296014269,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.377500000000001e-06,
+      "loss": 0.1887,
+      "step": 5450
+    },
+    {
+      "epoch": 0.11986005350895246,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.352500000000001e-06,
+      "loss": 0.1928,
+      "step": 5460
+    },
+    {
+      "epoch": 0.12007957741647801,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.327500000000001e-06,
+      "loss": 0.2024,
+      "step": 5470
+    },
+    {
+      "epoch": 0.12029910132400357,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.3025e-06,
+      "loss": 0.1855,
+      "step": 5480
+    },
+    {
+      "epoch": 0.12051862523152912,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.2775000000000005e-06,
+      "loss": 0.1693,
+      "step": 5490
+    },
+    {
+      "epoch": 0.12073814913905467,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.2525e-06,
+      "loss": 0.1777,
+      "step": 5500
+    },
+    {
+      "epoch": 0.12095767304658023,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.227500000000001e-06,
+      "loss": 0.221,
+      "step": 5510
+    },
+    {
+      "epoch": 0.12117719695410578,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.2025e-06,
+      "loss": 0.1692,
+      "step": 5520
+    },
+    {
+      "epoch": 0.12139672086163134,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.1775000000000006e-06,
+      "loss": 0.192,
+      "step": 5530
+    },
+    {
+      "epoch": 0.12161624476915689,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.1525e-06,
+      "loss": 0.1882,
+      "step": 5540
+    },
+    {
+      "epoch": 0.12183576867668244,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.127500000000001e-06,
+      "loss": 0.2051,
+      "step": 5550
+    },
+    {
+      "epoch": 0.122055292584208,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.1025000000000004e-06,
+      "loss": 0.2132,
+      "step": 5560
+    },
+    {
+      "epoch": 0.12227481649173355,
+      "grad_norm": 0.76953125,
+      "learning_rate": 6.077500000000001e-06,
+      "loss": 0.1776,
+      "step": 5570
+    },
+    {
+      "epoch": 0.1224943403992591,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.0525e-06,
+      "loss": 0.2029,
+      "step": 5580
+    },
+    {
+      "epoch": 0.12271386430678466,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.0275e-06,
+      "loss": 0.209,
+      "step": 5590
+    },
+    {
+      "epoch": 0.12293338821431021,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.0025e-06,
+      "loss": 0.1967,
+      "step": 5600
+    },
+    {
+      "epoch": 0.12315291212183577,
+      "grad_norm": 1.0,
+      "learning_rate": 5.977500000000001e-06,
+      "loss": 0.2383,
+      "step": 5610
+    },
+    {
+      "epoch": 0.12337243602936132,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.9525e-06,
+      "loss": 0.163,
+      "step": 5620
+    },
+    {
+      "epoch": 0.12359195993688687,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.9275e-06,
+      "loss": 0.196,
+      "step": 5630
+    },
+    {
+      "epoch": 0.12381148384441243,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.902500000000001e-06,
+      "loss": 0.2017,
+      "step": 5640
+    },
+    {
+      "epoch": 0.12403100775193798,
+      "grad_norm": 0.82421875,
+      "learning_rate": 5.8775e-06,
+      "loss": 0.1859,
+      "step": 5650
+    },
+    {
+      "epoch": 0.12425053165946354,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.852500000000001e-06,
+      "loss": 0.2002,
+      "step": 5660
+    },
+    {
+      "epoch": 0.12447005556698909,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.8275000000000005e-06,
+      "loss": 0.1784,
+      "step": 5670
+    },
+    {
+      "epoch": 0.12468957947451464,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.802500000000001e-06,
+      "loss": 0.1809,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1249091033820402,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.7775e-06,
+      "loss": 0.1663,
+      "step": 5690
+    },
+    {
+      "epoch": 0.12512862728956575,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.752500000000001e-06,
+      "loss": 0.173,
+      "step": 5700
+    },
+    {
+      "epoch": 0.1253481511970913,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.727500000000001e-06,
+      "loss": 0.1743,
+      "step": 5710
+    },
+    {
+      "epoch": 0.12556767510461686,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.702500000000001e-06,
+      "loss": 0.2364,
+      "step": 5720
+    },
+    {
+      "epoch": 0.12578719901214241,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.6775e-06,
+      "loss": 0.212,
+      "step": 5730
+    },
+    {
+      "epoch": 0.12600672291966797,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.6525000000000005e-06,
+      "loss": 0.1757,
+      "step": 5740
+    },
+    {
+      "epoch": 0.12622624682719352,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.6275e-06,
+      "loss": 0.1868,
+      "step": 5750
+    },
+    {
+      "epoch": 0.12644577073471908,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.602500000000001e-06,
+      "loss": 0.1901,
+      "step": 5760
+    },
+    {
+      "epoch": 0.12666529464224463,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.5775e-06,
+      "loss": 0.2302,
+      "step": 5770
+    },
+    {
+      "epoch": 0.12688481854977018,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.552500000000001e-06,
+      "loss": 0.1639,
+      "step": 5780
+    },
+    {
+      "epoch": 0.12710434245729574,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.5275e-06,
+      "loss": 0.1778,
+      "step": 5790
+    },
+    {
+      "epoch": 0.1273238663648213,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.5025e-06,
+      "loss": 0.1727,
+      "step": 5800
+    },
+    {
+      "epoch": 0.12754339027234685,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.4775e-06,
+      "loss": 0.1731,
+      "step": 5810
+    },
+    {
+      "epoch": 0.1277629141798724,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.452500000000001e-06,
+      "loss": 0.1762,
+      "step": 5820
+    },
+    {
+      "epoch": 0.12798243808739795,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.4275e-06,
+      "loss": 0.1874,
+      "step": 5830
+    },
+    {
+      "epoch": 0.1282019619949235,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.4025e-06,
+      "loss": 0.199,
+      "step": 5840
+    },
+    {
+      "epoch": 0.12842148590244906,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.3775e-06,
+      "loss": 0.1935,
+      "step": 5850
+    },
+    {
+      "epoch": 0.12864100980997462,
+      "grad_norm": 0.875,
+      "learning_rate": 5.352500000000001e-06,
+      "loss": 0.1862,
+      "step": 5860
+    },
+    {
+      "epoch": 0.12886053371750017,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.3275e-06,
+      "loss": 0.1693,
+      "step": 5870
+    },
+    {
+      "epoch": 0.12908005762502572,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.3025000000000005e-06,
+      "loss": 0.1972,
+      "step": 5880
+    },
+    {
+      "epoch": 0.12929958153255128,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.277500000000001e-06,
+      "loss": 0.1742,
+      "step": 5890
+    },
+    {
+      "epoch": 0.12951910544007683,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.2525e-06,
+      "loss": 0.1765,
+      "step": 5900
+    },
+    {
+      "epoch": 0.12973862934760239,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.227500000000001e-06,
+      "loss": 0.1713,
+      "step": 5910
+    },
+    {
+      "epoch": 0.12995815325512794,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.202500000000001e-06,
+      "loss": 0.1647,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1301776771626535,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.177500000000001e-06,
+      "loss": 0.1795,
+      "step": 5930
+    },
+    {
+      "epoch": 0.13039720107017905,
+      "grad_norm": 0.71875,
+      "learning_rate": 5.1525e-06,
+      "loss": 0.2107,
+      "step": 5940
+    },
+    {
+      "epoch": 0.1306167249777046,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.1275000000000005e-06,
+      "loss": 0.1919,
+      "step": 5950
+    },
+    {
+      "epoch": 0.13083624888523016,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.1025e-06,
+      "loss": 0.1755,
+      "step": 5960
+    },
+    {
+      "epoch": 0.1310557727927557,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.077500000000001e-06,
+      "loss": 0.1673,
+      "step": 5970
+    },
+    {
+      "epoch": 0.13127529670028126,
+      "grad_norm": 0.88671875,
+      "learning_rate": 5.0525e-06,
+      "loss": 0.2152,
+      "step": 5980
+    },
+    {
+      "epoch": 0.13149482060780682,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.0275000000000006e-06,
+      "loss": 0.2161,
+      "step": 5990
+    },
+    {
+      "epoch": 0.13171434451533237,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.0025e-06,
+      "loss": 0.1849,
+      "step": 6000
+    },
+    {
+      "epoch": 0.13193386842285793,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.977500000000001e-06,
+      "loss": 0.1786,
+      "step": 6010
+    },
+    {
+      "epoch": 0.13215339233038348,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.9525000000000004e-06,
+      "loss": 0.1818,
+      "step": 6020
+    },
+    {
+      "epoch": 0.13237291623790903,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.927500000000001e-06,
+      "loss": 0.182,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1325924401454346,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.902500000000001e-06,
+      "loss": 0.1778,
+      "step": 6040
+    },
+    {
+      "epoch": 0.13281196405296014,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.8775e-06,
+      "loss": 0.165,
+      "step": 6050
+    },
+    {
+      "epoch": 0.1330314879604857,
+      "grad_norm": 0.875,
+      "learning_rate": 4.8525000000000006e-06,
+      "loss": 0.2036,
+      "step": 6060
+    },
+    {
+      "epoch": 0.13325101186801125,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.827500000000001e-06,
+      "loss": 0.1749,
+      "step": 6070
+    },
+    {
+      "epoch": 0.1334705357755368,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.8025e-06,
+      "loss": 0.1979,
+      "step": 6080
+    },
+    {
+      "epoch": 0.13369005968306236,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.7775e-06,
+      "loss": 0.1883,
+      "step": 6090
+    },
+    {
+      "epoch": 0.1339095835905879,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.752500000000001e-06,
+      "loss": 0.1742,
+      "step": 6100
+    },
+    {
+      "epoch": 0.13412910749811346,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.7275e-06,
+      "loss": 0.1704,
+      "step": 6110
+    },
+    {
+      "epoch": 0.13434863140563902,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.7025e-06,
+      "loss": 0.1963,
+      "step": 6120
+    },
+    {
+      "epoch": 0.13456815531316457,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.6775000000000005e-06,
+      "loss": 0.1935,
+      "step": 6130
+    },
+    {
+      "epoch": 0.13478767922069013,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.652500000000001e-06,
+      "loss": 0.1756,
+      "step": 6140
+    },
+    {
+      "epoch": 0.13500720312821568,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.6275e-06,
+      "loss": 0.1922,
+      "step": 6150
+    },
+    {
+      "epoch": 0.13522672703574123,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.6025e-06,
+      "loss": 0.2005,
+      "step": 6160
+    },
+    {
+      "epoch": 0.1354462509432668,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.577500000000001e-06,
+      "loss": 0.184,
+      "step": 6170
+    },
+    {
+      "epoch": 0.13566577485079234,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.5525e-06,
+      "loss": 0.1865,
+      "step": 6180
+    },
+    {
+      "epoch": 0.1358852987583179,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.5275e-06,
+      "loss": 0.201,
+      "step": 6190
+    },
+    {
+      "epoch": 0.13610482266584345,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.5025000000000005e-06,
+      "loss": 0.2129,
+      "step": 6200
+    },
+    {
+      "epoch": 0.136324346573369,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.4775e-06,
+      "loss": 0.2103,
+      "step": 6210
+    },
+    {
+      "epoch": 0.13654387048089456,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.4525e-06,
+      "loss": 0.1933,
+      "step": 6220
+    },
+    {
+      "epoch": 0.1367633943884201,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.4275e-06,
+      "loss": 0.1837,
+      "step": 6230
+    },
+    {
+      "epoch": 0.13698291829594567,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.4025e-06,
+      "loss": 0.1696,
+      "step": 6240
+    },
+    {
+      "epoch": 0.13720244220347122,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.3775e-06,
+      "loss": 0.1885,
+      "step": 6250
+    },
+    {
+      "epoch": 0.13742196611099677,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.3525e-06,
+      "loss": 0.175,
+      "step": 6260
+    },
+    {
+      "epoch": 0.13764149001852233,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.3275000000000005e-06,
+      "loss": 0.1905,
+      "step": 6270
+    },
+    {
+      "epoch": 0.13786101392604788,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.302500000000001e-06,
+      "loss": 0.1854,
+      "step": 6280
+    },
+    {
+      "epoch": 0.13808053783357344,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.2775e-06,
+      "loss": 0.2116,
+      "step": 6290
+    },
+    {
+      "epoch": 0.138300061741099,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.2525e-06,
+      "loss": 0.184,
+      "step": 6300
+    },
+    {
+      "epoch": 0.13851958564862454,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.227500000000001e-06,
+      "loss": 0.1831,
+      "step": 6310
+    },
+    {
+      "epoch": 0.1387391095561501,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.202500000000001e-06,
+      "loss": 0.1765,
+      "step": 6320
+    },
+    {
+      "epoch": 0.13895863346367565,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.1775e-06,
+      "loss": 0.1742,
+      "step": 6330
+    },
+    {
+      "epoch": 0.1391781573712012,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.1525000000000005e-06,
+      "loss": 0.2031,
+      "step": 6340
+    },
+    {
+      "epoch": 0.13939768127872676,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.127500000000001e-06,
+      "loss": 0.1976,
+      "step": 6350
+    },
+    {
+      "epoch": 0.1396172051862523,
+      "grad_norm": 0.625,
+      "learning_rate": 4.1025e-06,
+      "loss": 0.1701,
+      "step": 6360
+    },
+    {
+      "epoch": 0.13983672909377787,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.0775e-06,
+      "loss": 0.1827,
+      "step": 6370
+    },
+    {
+      "epoch": 0.14005625300130342,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.052500000000001e-06,
+      "loss": 0.1767,
+      "step": 6380
+    },
+    {
+      "epoch": 0.14027577690882898,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.0275e-06,
+      "loss": 0.1869,
+      "step": 6390
+    },
+    {
+      "epoch": 0.14049530081635453,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.0025e-06,
+      "loss": 0.1954,
+      "step": 6400
+    },
+    {
+      "epoch": 0.14071482472388008,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.9775000000000005e-06,
+      "loss": 0.1762,
+      "step": 6410
+    },
+    {
+      "epoch": 0.14093434863140564,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.9525e-06,
+      "loss": 0.1865,
+      "step": 6420
+    },
+    {
+      "epoch": 0.1411538725389312,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.9275e-06,
+      "loss": 0.1816,
+      "step": 6430
+    },
+    {
+      "epoch": 0.14137339644645674,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.9025e-06,
+      "loss": 0.1757,
+      "step": 6440
+    },
+    {
+      "epoch": 0.1415929203539823,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.8775000000000006e-06,
+      "loss": 0.1836,
+      "step": 6450
+    },
+    {
+      "epoch": 0.14181244426150785,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.8525e-06,
+      "loss": 0.1953,
+      "step": 6460
+    },
+    {
+      "epoch": 0.1420319681690334,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.8275e-06,
+      "loss": 0.1605,
+      "step": 6470
+    },
+    {
+      "epoch": 0.14225149207655896,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.8025e-06,
+      "loss": 0.1865,
+      "step": 6480
+    },
+    {
+      "epoch": 0.14247101598408451,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.7775000000000003e-06,
+      "loss": 0.1746,
+      "step": 6490
+    },
+    {
+      "epoch": 0.14269053989161007,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.7525e-06,
+      "loss": 0.1572,
+      "step": 6500
+    },
+    {
+      "epoch": 0.14291006379913562,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.7275000000000007e-06,
+      "loss": 0.1942,
+      "step": 6510
+    },
+    {
+      "epoch": 0.14312958770666118,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.7025000000000005e-06,
+      "loss": 0.1841,
+      "step": 6520
+    },
+    {
+      "epoch": 0.14334911161418673,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.6775000000000004e-06,
+      "loss": 0.1964,
+      "step": 6530
+    },
+    {
+      "epoch": 0.14356863552171228,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.6525000000000006e-06,
+      "loss": 0.198,
+      "step": 6540
+    },
+    {
+      "epoch": 0.14378815942923784,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.6275000000000004e-06,
+      "loss": 0.1773,
+      "step": 6550
+    },
+    {
+      "epoch": 0.1440076833367634,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.6025000000000002e-06,
+      "loss": 0.1699,
+      "step": 6560
+    },
+    {
+      "epoch": 0.14422720724428895,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.5775000000000005e-06,
+      "loss": 0.2117,
+      "step": 6570
+    },
+    {
+      "epoch": 0.1444467311518145,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.5525000000000003e-06,
+      "loss": 0.1783,
+      "step": 6580
+    },
+    {
+      "epoch": 0.14466625505934005,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.5275000000000005e-06,
+      "loss": 0.1608,
+      "step": 6590
+    },
+    {
+      "epoch": 0.1448857789668656,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.5025000000000003e-06,
+      "loss": 0.1933,
+      "step": 6600
+    },
+    {
+      "epoch": 0.14510530287439116,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.4775e-06,
+      "loss": 0.2031,
+      "step": 6610
+    },
+    {
+      "epoch": 0.14532482678191672,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.4525000000000004e-06,
+      "loss": 0.188,
+      "step": 6620
+    },
+    {
+      "epoch": 0.14554435068944227,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.4275000000000002e-06,
+      "loss": 0.1767,
+      "step": 6630
+    },
+    {
+      "epoch": 0.14576387459696782,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.4025000000000005e-06,
+      "loss": 0.1888,
+      "step": 6640
+    },
+    {
+      "epoch": 0.14598339850449338,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.3775000000000003e-06,
+      "loss": 0.1918,
+      "step": 6650
+    },
+    {
+      "epoch": 0.14620292241201893,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.3525e-06,
+      "loss": 0.167,
+      "step": 6660
+    },
+    {
+      "epoch": 0.14642244631954449,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.3275000000000003e-06,
+      "loss": 0.1635,
+      "step": 6670
+    },
+    {
+      "epoch": 0.14664197022707004,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.3025e-06,
+      "loss": 0.2107,
+      "step": 6680
+    },
+    {
+      "epoch": 0.1468614941345956,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.2775e-06,
+      "loss": 0.1872,
+      "step": 6690
+    },
+    {
+      "epoch": 0.14708101804212115,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.2525e-06,
+      "loss": 0.1627,
+      "step": 6700
+    },
+    {
+      "epoch": 0.1473005419496467,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.2275e-06,
+      "loss": 0.1499,
+      "step": 6710
+    },
+    {
+      "epoch": 0.14752006585717226,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.2025000000000003e-06,
+      "loss": 0.1921,
+      "step": 6720
+    },
+    {
+      "epoch": 0.1477395897646978,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.1775e-06,
+      "loss": 0.1811,
+      "step": 6730
+    },
+    {
+      "epoch": 0.14795911367222336,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.1525e-06,
+      "loss": 0.176,
+      "step": 6740
+    },
+    {
+      "epoch": 0.14817863757974892,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.1275e-06,
+      "loss": 0.2066,
+      "step": 6750
+    },
+    {
+      "epoch": 0.14839816148727447,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.1025000000000004e-06,
+      "loss": 0.1424,
+      "step": 6760
+    },
+    {
+      "epoch": 0.14861768539480003,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.0775000000000006e-06,
+      "loss": 0.1997,
+      "step": 6770
+    },
+    {
+      "epoch": 0.14883720930232558,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.0525000000000004e-06,
+      "loss": 0.1976,
+      "step": 6780
+    },
+    {
+      "epoch": 0.14905673320985113,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.0275000000000002e-06,
+      "loss": 0.1596,
+      "step": 6790
+    },
+    {
+      "epoch": 0.1492762571173767,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.0025000000000005e-06,
+      "loss": 0.1694,
+      "step": 6800
+    },
+    {
+      "epoch": 0.14949578102490224,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.9775000000000003e-06,
+      "loss": 0.1774,
+      "step": 6810
+    },
+    {
+      "epoch": 0.1497153049324278,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.9525000000000005e-06,
+      "loss": 0.1849,
+      "step": 6820
+    },
+    {
+      "epoch": 0.14993482883995335,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.9275000000000003e-06,
+      "loss": 0.2215,
+      "step": 6830
+    },
+    {
+      "epoch": 0.1501543527474789,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.9025e-06,
+      "loss": 0.1916,
+      "step": 6840
+    },
+    {
+      "epoch": 0.15037387665500446,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.8775000000000004e-06,
+      "loss": 0.185,
+      "step": 6850
+    },
+    {
+      "epoch": 0.15059340056253,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.8525000000000002e-06,
+      "loss": 0.1826,
+      "step": 6860
+    },
+    {
+      "epoch": 0.15081292447005556,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.8275e-06,
+      "loss": 0.1935,
+      "step": 6870
+    },
+    {
+      "epoch": 0.15103244837758112,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.8025000000000003e-06,
+      "loss": 0.1683,
+      "step": 6880
+    },
+    {
+      "epoch": 0.15125197228510667,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.7775e-06,
+      "loss": 0.2083,
+      "step": 6890
+    },
+    {
+      "epoch": 0.15147149619263223,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.7525000000000003e-06,
+      "loss": 0.1656,
+      "step": 6900
+    },
+    {
+      "epoch": 0.15169102010015778,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.7275e-06,
+      "loss": 0.1748,
+      "step": 6910
+    },
+    {
+      "epoch": 0.15191054400768333,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.7025e-06,
+      "loss": 0.2087,
+      "step": 6920
+    },
+    {
+      "epoch": 0.1521300679152089,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.6775e-06,
+      "loss": 0.1721,
+      "step": 6930
+    },
+    {
+      "epoch": 0.15234959182273444,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.6525e-06,
+      "loss": 0.2098,
+      "step": 6940
+    },
+    {
+      "epoch": 0.15256911573026,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.6275000000000003e-06,
+      "loss": 0.1765,
+      "step": 6950
+    },
+    {
+      "epoch": 0.15278863963778555,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.6025e-06,
+      "loss": 0.1839,
+      "step": 6960
+    },
+    {
+      "epoch": 0.1530081635453111,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.5775e-06,
+      "loss": 0.1648,
+      "step": 6970
+    },
+    {
+      "epoch": 0.15322768745283666,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5525e-06,
+      "loss": 0.1808,
+      "step": 6980
+    },
+    {
+      "epoch": 0.1534472113603622,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.5275e-06,
+      "loss": 0.1903,
+      "step": 6990
+    },
+    {
+      "epoch": 0.15366673526788777,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.5024999999999998e-06,
+      "loss": 0.1867,
+      "step": 7000
+    },
+    {
+      "epoch": 0.15388625917541332,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.4775e-06,
+      "loss": 0.1942,
+      "step": 7010
+    },
+    {
+      "epoch": 0.15410578308293887,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.4525000000000002e-06,
+      "loss": 0.1753,
+      "step": 7020
+    },
+    {
+      "epoch": 0.15432530699046443,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.4275e-06,
+      "loss": 0.1916,
+      "step": 7030
+    },
+    {
+      "epoch": 0.15454483089798998,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.4025000000000003e-06,
+      "loss": 0.1735,
+      "step": 7040
+    },
+    {
+      "epoch": 0.15476435480551554,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.3775e-06,
+      "loss": 0.1675,
+      "step": 7050
+    },
+    {
+      "epoch": 0.1549838787130411,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.3525e-06,
+      "loss": 0.176,
+      "step": 7060
+    },
+    {
+      "epoch": 0.15520340262056664,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.3275e-06,
+      "loss": 0.1785,
+      "step": 7070
+    },
+    {
+      "epoch": 0.1554229265280922,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.3025000000000004e-06,
+      "loss": 0.1981,
+      "step": 7080
+    },
+    {
+      "epoch": 0.15564245043561775,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.2775000000000002e-06,
+      "loss": 0.2026,
+      "step": 7090
+    },
+    {
+      "epoch": 0.1558619743431433,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.2525e-06,
+      "loss": 0.1676,
+      "step": 7100
+    },
+    {
+      "epoch": 0.15608149825066886,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.2275000000000003e-06,
+      "loss": 0.1657,
+      "step": 7110
+    },
+    {
+      "epoch": 0.1563010221581944,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2025e-06,
+      "loss": 0.1702,
+      "step": 7120
+    },
+    {
+      "epoch": 0.15652054606571997,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.1775000000000003e-06,
+      "loss": 0.1788,
+      "step": 7130
+    },
+    {
+      "epoch": 0.15674006997324552,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.1525e-06,
+      "loss": 0.1713,
+      "step": 7140
+    },
+    {
+      "epoch": 0.15695959388077108,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.1275e-06,
+      "loss": 0.1754,
+      "step": 7150
+    },
+    {
+      "epoch": 0.15717911778829663,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.1025e-06,
+      "loss": 0.1924,
+      "step": 7160
+    },
+    {
+      "epoch": 0.15739864169582218,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.0775e-06,
+      "loss": 0.1997,
+      "step": 7170
+    },
+    {
+      "epoch": 0.15761816560334774,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.0525000000000003e-06,
+      "loss": 0.1917,
+      "step": 7180
+    },
+    {
+      "epoch": 0.1578376895108733,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.0275000000000005e-06,
+      "loss": 0.2014,
+      "step": 7190
+    },
+    {
+      "epoch": 0.15805721341839885,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0025000000000003e-06,
+      "loss": 0.1756,
+      "step": 7200
+    },
+    {
+      "epoch": 0.1582767373259244,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9775e-06,
+      "loss": 0.1767,
+      "step": 7210
+    },
+    {
+      "epoch": 0.15849626123344995,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9525000000000004e-06,
+      "loss": 0.1863,
+      "step": 7220
+    },
+    {
+      "epoch": 0.1587157851409755,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9275e-06,
+      "loss": 0.2036,
+      "step": 7230
+    },
+    {
+      "epoch": 0.15893530904850106,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9025000000000002e-06,
+      "loss": 0.1922,
+      "step": 7240
+    },
+    {
+      "epoch": 0.15915483295602662,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.8775000000000002e-06,
+      "loss": 0.1937,
+      "step": 7250
+    },
+    {
+      "epoch": 0.15937435686355217,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8525e-06,
+      "loss": 0.2084,
+      "step": 7260
+    },
+    {
+      "epoch": 0.15959388077107772,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.8275e-06,
+      "loss": 0.2239,
+      "step": 7270
+    },
+    {
+      "epoch": 0.15981340467860328,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8025000000000001e-06,
+      "loss": 0.1826,
+      "step": 7280
+    },
+    {
+      "epoch": 0.16003292858612883,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.7775000000000001e-06,
+      "loss": 0.1847,
+      "step": 7290
+    },
+    {
+      "epoch": 0.16025245249365438,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7525e-06,
+      "loss": 0.2061,
+      "step": 7300
+    },
+    {
+      "epoch": 0.16047197640117994,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7275e-06,
+      "loss": 0.1872,
+      "step": 7310
+    },
+    {
+      "epoch": 0.1606915003087055,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7025000000000002e-06,
+      "loss": 0.1826,
+      "step": 7320
+    },
+    {
+      "epoch": 0.16091102421623105,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.6775000000000002e-06,
+      "loss": 0.1821,
+      "step": 7330
+    },
+    {
+      "epoch": 0.1611305481237566,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6525000000000003e-06,
+      "loss": 0.1842,
+      "step": 7340
+    },
+    {
+      "epoch": 0.16135007203128215,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6275e-06,
+      "loss": 0.1754,
+      "step": 7350
+    },
+    {
+      "epoch": 0.1615695959388077,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.6025000000000001e-06,
+      "loss": 0.1928,
+      "step": 7360
+    },
+    {
+      "epoch": 0.16178911984633326,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.5775000000000001e-06,
+      "loss": 0.1871,
+      "step": 7370
+    },
+    {
+      "epoch": 0.16200864375385882,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5525000000000002e-06,
+      "loss": 0.2064,
+      "step": 7380
+    },
+    {
+      "epoch": 0.16222816766138437,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.5275000000000002e-06,
+      "loss": 0.2008,
+      "step": 7390
+    },
+    {
+      "epoch": 0.16244769156890992,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5025e-06,
+      "loss": 0.1788,
+      "step": 7400
+    },
+    {
+      "epoch": 0.16266721547643548,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.4775e-06,
+      "loss": 0.1762,
+      "step": 7410
+    },
+    {
+      "epoch": 0.16288673938396103,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4525e-06,
+      "loss": 0.1807,
+      "step": 7420
+    },
+    {
+      "epoch": 0.1631062632914866,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4275e-06,
+      "loss": 0.2052,
+      "step": 7430
+    },
+    {
+      "epoch": 0.16332578719901214,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4025000000000003e-06,
+      "loss": 0.1669,
+      "step": 7440
+    },
+    {
+      "epoch": 0.1635453111065377,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3775000000000002e-06,
+      "loss": 0.1858,
+      "step": 7450
+    },
+    {
+      "epoch": 0.16376483501406325,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3525000000000002e-06,
+      "loss": 0.1636,
+      "step": 7460
+    },
+    {
+      "epoch": 0.1639843589215888,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.3275000000000002e-06,
+      "loss": 0.1912,
+      "step": 7470
+    },
+    {
+      "epoch": 0.16420388282911436,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3025000000000002e-06,
+      "loss": 0.2127,
+      "step": 7480
+    },
+    {
+      "epoch": 0.1644234067366399,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2775e-06,
+      "loss": 0.1856,
+      "step": 7490
+    },
+    {
+      "epoch": 0.16464293064416546,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2525e-06,
+      "loss": 0.1888,
+      "step": 7500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4314080978858947e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}