diff --git "a/checkpoint-18500/trainer_state.json" "b/checkpoint-18500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-18500/trainer_state.json"
@@ -0,0 +1,12984 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5647820246672365,
+  "eval_steps": 500,
+  "global_step": 18500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003052875809012089,
+      "grad_norm": 20.0,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 2.2504,
+      "step": 10
+    },
+    {
+      "epoch": 0.0006105751618024178,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 2.305,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009158627427036268,
+      "grad_norm": 13.75,
+      "learning_rate": 2.9e-06,
+      "loss": 2.1805,
+      "step": 30
+    },
+    {
+      "epoch": 0.0012211503236048357,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 2.045,
+      "step": 40
+    },
+    {
+      "epoch": 0.0015264379045060448,
+      "grad_norm": 7.5,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.8589,
+      "step": 50
+    },
+    {
+      "epoch": 0.0018317254854072536,
+      "grad_norm": 4.96875,
+      "learning_rate": 5.9e-06,
+      "loss": 1.7373,
+      "step": 60
+    },
+    {
+      "epoch": 0.0021370130663084627,
+      "grad_norm": 4.75,
+      "learning_rate": 6.9e-06,
+      "loss": 1.5472,
+      "step": 70
+    },
+    {
+      "epoch": 0.0024423006472096714,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.9e-06,
+      "loss": 1.4499,
+      "step": 80
+    },
+    {
+      "epoch": 0.0027475882281108805,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 1.2686,
+      "step": 90
+    },
+    {
+      "epoch": 0.0030528758090120896,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.9e-06,
+      "loss": 1.2467,
+      "step": 100
+    },
+    {
+      "epoch": 0.003358163389913298,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.997243998040178e-06,
+      "loss": 1.1575,
+      "step": 110
+    },
+    {
+      "epoch": 0.0036634509708145073,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.994181773640373e-06,
+      "loss": 1.1596,
+      "step": 120
+    },
+    {
+      "epoch": 0.003968738551715716,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.99111954924057e-06,
+      "loss": 1.1043,
+      "step": 130
+    },
+    {
+      "epoch": 0.0042740261326169255,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.988057324840765e-06,
+      "loss": 1.1881,
+      "step": 140
+    },
+    {
+      "epoch": 0.004579313713518134,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.98499510044096e-06,
+      "loss": 1.0619,
+      "step": 150
+    },
+    {
+      "epoch": 0.004884601294419343,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.981932876041157e-06,
+      "loss": 1.0942,
+      "step": 160
+    },
+    {
+      "epoch": 0.005189888875320552,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.978870651641354e-06,
+      "loss": 1.0805,
+      "step": 170
+    },
+    {
+      "epoch": 0.005495176456221761,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.975808427241549e-06,
+      "loss": 1.0048,
+      "step": 180
+    },
+    {
+      "epoch": 0.00580046403712297,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.972746202841746e-06,
+      "loss": 1.1054,
+      "step": 190
+    },
+    {
+      "epoch": 0.006105751618024179,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.969683978441941e-06,
+      "loss": 1.0129,
+      "step": 200
+    },
+    {
+      "epoch": 0.006411039198925388,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.966621754042138e-06,
+      "loss": 1.0408,
+      "step": 210
+    },
+    {
+      "epoch": 0.006716326779826596,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.963559529642333e-06,
+      "loss": 0.9895,
+      "step": 220
+    },
+    {
+      "epoch": 0.007021614360727806,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.96049730524253e-06,
+      "loss": 1.0581,
+      "step": 230
+    },
+    {
+      "epoch": 0.007326901941629015,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.957435080842725e-06,
+      "loss": 0.9868,
+      "step": 240
+    },
+    {
+      "epoch": 0.007632189522530223,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.954372856442922e-06,
+      "loss": 1.0124,
+      "step": 250
+    },
+    {
+      "epoch": 0.007937477103431432,
+      "grad_norm": 3.5,
+      "learning_rate": 9.951310632043117e-06,
+      "loss": 1.0133,
+      "step": 260
+    },
+    {
+      "epoch": 0.00824276468433264,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.948248407643312e-06,
+      "loss": 1.0321,
+      "step": 270
+    },
+    {
+      "epoch": 0.008548052265233851,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.945186183243509e-06,
+      "loss": 1.0384,
+      "step": 280
+    },
+    {
+      "epoch": 0.00885333984613506,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.942123958843704e-06,
+      "loss": 1.0244,
+      "step": 290
+    },
+    {
+      "epoch": 0.009158627427036268,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.939061734443901e-06,
+      "loss": 0.9838,
+      "step": 300
+    },
+    {
+      "epoch": 0.009463915007937477,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.935999510044096e-06,
+      "loss": 1.0058,
+      "step": 310
+    },
+    {
+      "epoch": 0.009769202588838686,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.932937285644293e-06,
+      "loss": 0.9537,
+      "step": 320
+    },
+    {
+      "epoch": 0.010074490169739894,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.929875061244488e-06,
+      "loss": 0.9686,
+      "step": 330
+    },
+    {
+      "epoch": 0.010379777750641105,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.926812836844685e-06,
+      "loss": 0.9911,
+      "step": 340
+    },
+    {
+      "epoch": 0.010685065331542313,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.923750612444882e-06,
+      "loss": 0.9323,
+      "step": 350
+    },
+    {
+      "epoch": 0.010990352912443522,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.920688388045077e-06,
+      "loss": 0.9615,
+      "step": 360
+    },
+    {
+      "epoch": 0.01129564049334473,
+      "grad_norm": 3.5,
+      "learning_rate": 9.917626163645274e-06,
+      "loss": 0.9387,
+      "step": 370
+    },
+    {
+      "epoch": 0.01160092807424594,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.914563939245469e-06,
+      "loss": 0.9742,
+      "step": 380
+    },
+    {
+      "epoch": 0.011906215655147148,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.911501714845664e-06,
+      "loss": 0.9568,
+      "step": 390
+    },
+    {
+      "epoch": 0.012211503236048358,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.908439490445861e-06,
+      "loss": 0.9859,
+      "step": 400
+    },
+    {
+      "epoch": 0.012516790816949567,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.905377266046056e-06,
+      "loss": 0.9257,
+      "step": 410
+    },
+    {
+      "epoch": 0.012822078397850776,
+      "grad_norm": 3.125,
+      "learning_rate": 9.902315041646253e-06,
+      "loss": 0.9626,
+      "step": 420
+    },
+    {
+      "epoch": 0.013127365978751984,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.899252817246448e-06,
+      "loss": 0.9794,
+      "step": 430
+    },
+    {
+      "epoch": 0.013432653559653193,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.896190592846645e-06,
+      "loss": 0.9789,
+      "step": 440
+    },
+    {
+      "epoch": 0.013737941140554401,
+      "grad_norm": 3.375,
+      "learning_rate": 9.89312836844684e-06,
+      "loss": 0.9644,
+      "step": 450
+    },
+    {
+      "epoch": 0.014043228721455612,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.890066144047037e-06,
+      "loss": 0.9711,
+      "step": 460
+    },
+    {
+      "epoch": 0.01434851630235682,
+      "grad_norm": 3.5,
+      "learning_rate": 9.887003919647232e-06,
+      "loss": 1.0036,
+      "step": 470
+    },
+    {
+      "epoch": 0.01465380388325803,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.883941695247429e-06,
+      "loss": 1.0357,
+      "step": 480
+    },
+    {
+      "epoch": 0.014959091464159238,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.880879470847624e-06,
+      "loss": 1.1038,
+      "step": 490
+    },
+    {
+      "epoch": 0.015264379045060446,
+      "grad_norm": 3.25,
+      "learning_rate": 9.877817246447821e-06,
+      "loss": 0.9671,
+      "step": 500
+    },
+    {
+      "epoch": 0.015569666625961655,
+      "grad_norm": 3.125,
+      "learning_rate": 9.874755022048016e-06,
+      "loss": 0.9675,
+      "step": 510
+    },
+    {
+      "epoch": 0.015874954206862864,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.871692797648213e-06,
+      "loss": 0.9595,
+      "step": 520
+    },
+    {
+      "epoch": 0.016180241787764074,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.868630573248408e-06,
+      "loss": 1.0033,
+      "step": 530
+    },
+    {
+      "epoch": 0.01648552936866528,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.865568348848605e-06,
+      "loss": 0.9772,
+      "step": 540
+    },
+    {
+      "epoch": 0.01679081694956649,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.8625061244488e-06,
+      "loss": 0.9659,
+      "step": 550
+    },
+    {
+      "epoch": 0.017096104530467702,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.859443900048995e-06,
+      "loss": 0.9197,
+      "step": 560
+    },
+    {
+      "epoch": 0.01740139211136891,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.856381675649192e-06,
+      "loss": 0.9123,
+      "step": 570
+    },
+    {
+      "epoch": 0.01770667969227012,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.853319451249387e-06,
+      "loss": 0.9013,
+      "step": 580
+    },
+    {
+      "epoch": 0.018011967273171326,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.850257226849584e-06,
+      "loss": 0.9523,
+      "step": 590
+    },
+    {
+      "epoch": 0.018317254854072536,
+      "grad_norm": 3.125,
+      "learning_rate": 9.84719500244978e-06,
+      "loss": 1.0012,
+      "step": 600
+    },
+    {
+      "epoch": 0.018622542434973747,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.844132778049976e-06,
+      "loss": 0.9694,
+      "step": 610
+    },
+    {
+      "epoch": 0.018927830015874954,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.841070553650173e-06,
+      "loss": 1.0199,
+      "step": 620
+    },
+    {
+      "epoch": 0.019233117596776164,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.838008329250368e-06,
+      "loss": 0.927,
+      "step": 630
+    },
+    {
+      "epoch": 0.01953840517767737,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.834946104850565e-06,
+      "loss": 0.9334,
+      "step": 640
+    },
+    {
+      "epoch": 0.01984369275857858,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.83188388045076e-06,
+      "loss": 0.9607,
+      "step": 650
+    },
+    {
+      "epoch": 0.02014898033947979,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.828821656050957e-06,
+      "loss": 0.9754,
+      "step": 660
+    },
+    {
+      "epoch": 0.020454267920381,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.825759431651152e-06,
+      "loss": 0.9787,
+      "step": 670
+    },
+    {
+      "epoch": 0.02075955550128221,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.822697207251347e-06,
+      "loss": 0.9622,
+      "step": 680
+    },
+    {
+      "epoch": 0.021064843082183416,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.819634982851544e-06,
+      "loss": 0.9649,
+      "step": 690
+    },
+    {
+      "epoch": 0.021370130663084626,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.81657275845174e-06,
+      "loss": 0.9502,
+      "step": 700
+    },
+    {
+      "epoch": 0.021675418243985833,
+      "grad_norm": 4.0,
+      "learning_rate": 9.813510534051936e-06,
+      "loss": 0.9401,
+      "step": 710
+    },
+    {
+      "epoch": 0.021980705824887044,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.810448309652131e-06,
+      "loss": 0.9389,
+      "step": 720
+    },
+    {
+      "epoch": 0.022285993405788254,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.807386085252328e-06,
+      "loss": 0.9608,
+      "step": 730
+    },
+    {
+      "epoch": 0.02259128098668946,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.804323860852523e-06,
+      "loss": 0.9209,
+      "step": 740
+    },
+    {
+      "epoch": 0.02289656856759067,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.80126163645272e-06,
+      "loss": 0.9926,
+      "step": 750
+    },
+    {
+      "epoch": 0.02320185614849188,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.798199412052915e-06,
+      "loss": 0.9543,
+      "step": 760
+    },
+    {
+      "epoch": 0.02350714372939309,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.795137187653112e-06,
+      "loss": 0.9256,
+      "step": 770
+    },
+    {
+      "epoch": 0.023812431310294296,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.79207496325331e-06,
+      "loss": 0.953,
+      "step": 780
+    },
+    {
+      "epoch": 0.024117718891195506,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.789012738853504e-06,
+      "loss": 0.9684,
+      "step": 790
+    },
+    {
+      "epoch": 0.024423006472096716,
+      "grad_norm": 3.5,
+      "learning_rate": 9.785950514453701e-06,
+      "loss": 1.0046,
+      "step": 800
+    },
+    {
+      "epoch": 0.024728294052997923,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.782888290053896e-06,
+      "loss": 1.0192,
+      "step": 810
+    },
+    {
+      "epoch": 0.025033581633899134,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.779826065654091e-06,
+      "loss": 0.9067,
+      "step": 820
+    },
+    {
+      "epoch": 0.02533886921480034,
+      "grad_norm": 3.25,
+      "learning_rate": 9.776763841254288e-06,
+      "loss": 0.9454,
+      "step": 830
+    },
+    {
+      "epoch": 0.02564415679570155,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.773701616854484e-06,
+      "loss": 0.9162,
+      "step": 840
+    },
+    {
+      "epoch": 0.02594944437660276,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.770639392454679e-06,
+      "loss": 0.9216,
+      "step": 850
+    },
+    {
+      "epoch": 0.02625473195750397,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.767577168054876e-06,
+      "loss": 0.9246,
+      "step": 860
+    },
+    {
+      "epoch": 0.02656001953840518,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.764514943655072e-06,
+      "loss": 0.9081,
+      "step": 870
+    },
+    {
+      "epoch": 0.026865307119306386,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.761452719255268e-06,
+      "loss": 0.9376,
+      "step": 880
+    },
+    {
+      "epoch": 0.027170594700207596,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.758390494855464e-06,
+      "loss": 0.8979,
+      "step": 890
+    },
+    {
+      "epoch": 0.027475882281108803,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.75532827045566e-06,
+      "loss": 0.9662,
+      "step": 900
+    },
+    {
+      "epoch": 0.027781169862010013,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.752266046055856e-06,
+      "loss": 0.9206,
+      "step": 910
+    },
+    {
+      "epoch": 0.028086457442911224,
+      "grad_norm": 3.5,
+      "learning_rate": 9.749203821656052e-06,
+      "loss": 0.9269,
+      "step": 920
+    },
+    {
+      "epoch": 0.02839174502381243,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.746141597256248e-06,
+      "loss": 0.9187,
+      "step": 930
+    },
+    {
+      "epoch": 0.02869703260471364,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.743079372856444e-06,
+      "loss": 0.9733,
+      "step": 940
+    },
+    {
+      "epoch": 0.029002320185614848,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.74001714845664e-06,
+      "loss": 0.9724,
+      "step": 950
+    },
+    {
+      "epoch": 0.02930760776651606,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.736954924056836e-06,
+      "loss": 0.9247,
+      "step": 960
+    },
+    {
+      "epoch": 0.02961289534741727,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.73389269965703e-06,
+      "loss": 0.9503,
+      "step": 970
+    },
+    {
+      "epoch": 0.029918182928318476,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.730830475257228e-06,
+      "loss": 0.9766,
+      "step": 980
+    },
+    {
+      "epoch": 0.030223470509219686,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.727768250857423e-06,
+      "loss": 0.895,
+      "step": 990
+    },
+    {
+      "epoch": 0.030528758090120893,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.72470602645762e-06,
+      "loss": 0.9779,
+      "step": 1000
+    },
+    {
+      "epoch": 0.030834045671022103,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.721643802057815e-06,
+      "loss": 0.9002,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03113933325192331,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.718581577658012e-06,
+      "loss": 0.8777,
+      "step": 1020
+    },
+    {
+      "epoch": 0.031444620832824524,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.715519353258207e-06,
+      "loss": 0.9866,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03174990841372573,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.712457128858404e-06,
+      "loss": 0.9512,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03205519599462694,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.7093949044586e-06,
+      "loss": 0.8489,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03236048357552815,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.706332680058796e-06,
+      "loss": 0.9427,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03266577115642936,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.703270455658992e-06,
+      "loss": 0.8966,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03297105873733056,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.700208231259188e-06,
+      "loss": 0.942,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03327634631823177,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.697146006859384e-06,
+      "loss": 0.9155,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03358163389913298,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.69408378245958e-06,
+      "loss": 0.8871,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03388692148003419,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.691021558059775e-06,
+      "loss": 0.9495,
+      "step": 1110
+    },
+    {
+      "epoch": 0.034192209060935404,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.687959333659972e-06,
+      "loss": 0.953,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03449749664183661,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.684897109260167e-06,
+      "loss": 0.8765,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03480278422273782,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.681834884860364e-06,
+      "loss": 0.9421,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03510807180363903,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.678772660460559e-06,
+      "loss": 0.9437,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03541335938454024,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.675710436060756e-06,
+      "loss": 0.8336,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03571864696544145,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.67264821166095e-06,
+      "loss": 0.8529,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03602393454634265,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.669585987261148e-06,
+      "loss": 0.8623,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03632922212724386,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.666523762861343e-06,
+      "loss": 0.9721,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03663450970814507,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.66346153846154e-06,
+      "loss": 0.8867,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03693979728904628,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.660399314061735e-06,
+      "loss": 0.915,
+      "step": 1210
+    },
+    {
+      "epoch": 0.037245084869947494,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.657337089661932e-06,
+      "loss": 0.8983,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0375503724508487,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.654274865262127e-06,
+      "loss": 0.9718,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03785566003174991,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.651212640862324e-06,
+      "loss": 0.9019,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03816094761265112,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.648150416462519e-06,
+      "loss": 0.8962,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03846623519355233,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.645088192062714e-06,
+      "loss": 0.8857,
+      "step": 1260
+    },
+    {
+      "epoch": 0.03877152277445353,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.642025967662911e-06,
+      "loss": 0.8991,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03907681035535474,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.638963743263106e-06,
+      "loss": 0.9119,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03938209793625595,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.635901518863303e-06,
+      "loss": 0.9231,
+      "step": 1290
+    },
+    {
+      "epoch": 0.03968738551715716,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.632839294463498e-06,
+      "loss": 0.95,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03999267309805837,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.629777070063695e-06,
+      "loss": 0.9255,
+      "step": 1310
+    },
+    {
+      "epoch": 0.04029796067895958,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.626714845663892e-06,
+      "loss": 0.8951,
+      "step": 1320
+    },
+    {
+      "epoch": 0.04060324825986079,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.623652621264087e-06,
+      "loss": 0.8579,
+      "step": 1330
+    },
+    {
+      "epoch": 0.040908535840762,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.620590396864284e-06,
+      "loss": 0.9482,
+      "step": 1340
+    },
+    {
+      "epoch": 0.04121382342166321,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.617528172464479e-06,
+      "loss": 0.8819,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04151911100256442,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.614465948064676e-06,
+      "loss": 0.9442,
+      "step": 1360
+    },
+    {
+      "epoch": 0.04182439858346562,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.611403723664871e-06,
+      "loss": 0.9546,
+      "step": 1370
+    },
+    {
+      "epoch": 0.04212968616436683,
+      "grad_norm": 3.25,
+      "learning_rate": 9.608341499265068e-06,
+      "loss": 0.9114,
+      "step": 1380
+    },
+    {
+      "epoch": 0.04243497374526804,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.605279274865263e-06,
+      "loss": 0.8957,
+      "step": 1390
+    },
+    {
+      "epoch": 0.04274026132616925,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.602217050465458e-06,
+      "loss": 0.9542,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04304554890707046,
+      "grad_norm": 2.875,
+      "learning_rate": 9.599154826065655e-06,
+      "loss": 0.8686,
+      "step": 1410
+    },
+    {
+      "epoch": 0.04335083648797167,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.59609260166585e-06,
+      "loss": 0.9227,
+      "step": 1420
+    },
+    {
+      "epoch": 0.04365612406887288,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.593030377266047e-06,
+      "loss": 0.9245,
+      "step": 1430
+    },
+    {
+      "epoch": 0.04396141164977409,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.589968152866242e-06,
+      "loss": 0.8734,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0442666992306753,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.586905928466439e-06,
+      "loss": 0.9252,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04457198681157651,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.583843704066634e-06,
+      "loss": 0.8329,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04487727439247771,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.580781479666831e-06,
+      "loss": 0.8568,
+      "step": 1470
+    },
+    {
+      "epoch": 0.04518256197337892,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.577719255267028e-06,
+      "loss": 0.9022,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04548784955428013,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.574657030867223e-06,
+      "loss": 0.9385,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04579313713518134,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.57159480646742e-06,
+      "loss": 0.9028,
+      "step": 1500
+    },
+    {
+      "epoch": 0.046098424716082546,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.568532582067615e-06,
+      "loss": 0.855,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04640371229698376,
+      "grad_norm": 3.25,
+      "learning_rate": 9.56547035766781e-06,
+      "loss": 0.8944,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04670899987788497,
+      "grad_norm": 3.125,
+      "learning_rate": 9.562408133268007e-06,
+      "loss": 0.9477,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04701428745878618,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.559345908868202e-06,
+      "loss": 0.9202,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04731957503968739,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.556283684468397e-06,
+      "loss": 0.9337,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04762486262058859,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.553221460068594e-06,
+      "loss": 0.9324,
+      "step": 1560
+    },
+    {
+      "epoch": 0.0479301502014898,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.550159235668791e-06,
+      "loss": 0.9401,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04823543778239101,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.547097011268986e-06,
+      "loss": 0.8255,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04854072536329222,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.544034786869183e-06,
+      "loss": 0.9377,
+      "step": 1590
+    },
+    {
+      "epoch": 0.04884601294419343,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.540972562469378e-06,
+      "loss": 0.9112,
+      "step": 1600
+    },
+    {
+      "epoch": 0.049151300525094636,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.537910338069575e-06,
+      "loss": 0.9106,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04945658810599585,
+      "grad_norm": 3.625,
+      "learning_rate": 9.53484811366977e-06,
+      "loss": 0.9319,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04976187568689706,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.531785889269967e-06,
+      "loss": 0.8611,
+      "step": 1630
+    },
+    {
+      "epoch": 0.05006716326779827,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.528723664870162e-06,
+      "loss": 0.8842,
+      "step": 1640
+    },
+    {
+      "epoch": 0.05037245084869948,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.525661440470359e-06,
+      "loss": 0.929,
+      "step": 1650
+    },
+    {
+      "epoch": 0.05067773842960068,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.522599216070554e-06,
+      "loss": 0.8643,
+      "step": 1660
+    },
+    {
+      "epoch": 0.05098302601050189,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.519536991670751e-06,
+      "loss": 0.9192,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0512883135914031,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.516474767270946e-06,
+      "loss": 0.8826,
+      "step": 1680
+    },
+    {
+      "epoch": 0.05159360117230431,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.513412542871141e-06,
+      "loss": 0.8677,
+      "step": 1690
+    },
+    {
+      "epoch": 0.05189888875320552,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.510350318471338e-06,
+      "loss": 0.9244,
+      "step": 1700
+    },
+    {
+      "epoch": 0.052204176334106726,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.507288094071533e-06,
+      "loss": 0.9288,
+      "step": 1710
+    },
+    {
+      "epoch": 0.05250946391500794,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.50422586967173e-06,
+      "loss": 0.9067,
+      "step": 1720
+    },
+    {
+      "epoch": 0.05281475149590915,
+      "grad_norm": 3.375,
+      "learning_rate": 9.501163645271925e-06,
+      "loss": 0.8933,
+      "step": 1730
+    },
+    {
+      "epoch": 0.05312003907681036,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.498101420872122e-06,
+      "loss": 0.9053,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05342532665771156,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.49503919647232e-06,
+      "loss": 0.9014,
+      "step": 1750
+    },
+    {
+      "epoch": 0.05373061423861277,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.491976972072514e-06,
+      "loss": 0.8948,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05403590181951398,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.488914747672711e-06,
+      "loss": 0.8829,
+      "step": 1770
+    },
+    {
+      "epoch": 0.05434118940041519,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.485852523272906e-06,
+      "loss": 0.9375,
+      "step": 1780
+    },
+    {
+      "epoch": 0.0546464769813164,
+      "grad_norm": 3.25,
+      "learning_rate": 9.482790298873103e-06,
+      "loss": 0.9466,
+      "step": 1790
+    },
+    {
+      "epoch": 0.054951764562217606,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.479728074473298e-06,
+      "loss": 0.9079,
+      "step": 1800
+    },
+    {
+      "epoch": 0.055257052143118816,
+      "grad_norm": 3.375,
+      "learning_rate": 9.476665850073493e-06,
+      "loss": 0.8759,
+      "step": 1810
+    },
+    {
+      "epoch": 0.05556233972402003,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.47360362567369e-06,
+      "loss": 0.8918,
+      "step": 1820
+    },
+    {
+      "epoch": 0.05586762730492124,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.470541401273886e-06,
+      "loss": 0.9727,
+      "step": 1830
+    },
+    {
+      "epoch": 0.05617291488582245,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.467479176874082e-06,
+      "loss": 0.8599,
+      "step": 1840
+    },
+    {
+      "epoch": 0.05647820246672365,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.464416952474278e-06,
+      "loss": 0.9295,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05678349004762486,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.461354728074474e-06,
+      "loss": 0.853,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05708877762852607,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.45829250367467e-06,
+      "loss": 0.9022,
+      "step": 1870
+    },
+    {
+      "epoch": 0.05739406520942728,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.455230279274866e-06,
+      "loss": 0.9311,
+      "step": 1880
+    },
+    {
+      "epoch": 0.05769935279032849,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.452168054875062e-06,
+      "loss": 0.8912,
+      "step": 1890
+    },
+    {
+      "epoch": 0.058004640371229696,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.449105830475258e-06,
+      "loss": 0.8658,
+      "step": 1900
+    },
+    {
+      "epoch": 0.058309927952130906,
+      "grad_norm": 3.25,
+      "learning_rate": 9.446043606075454e-06,
+      "loss": 0.8551,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05861521553303212,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.44298138167565e-06,
+      "loss": 0.9328,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05892050311393333,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.439919157275846e-06,
+      "loss": 0.8999,
+      "step": 1930
+    },
+    {
+      "epoch": 0.05922579069483454,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.436856932876042e-06,
+      "loss": 0.8599,
+      "step": 1940
+    },
+    {
+      "epoch": 0.05953107827573574,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.433794708476238e-06,
+      "loss": 0.8852,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05983636585663695,
+      "grad_norm": 3.125,
+      "learning_rate": 9.430732484076434e-06,
+      "loss": 0.9264,
+      "step": 1960
+    },
+    {
+      "epoch": 0.06014165343753816,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.42767025967663e-06,
+      "loss": 0.8967,
+      "step": 1970
+    },
+    {
+      "epoch": 0.06044694101843937,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.424608035276825e-06,
+      "loss": 0.8914,
+      "step": 1980
+    },
+    {
+      "epoch": 0.060752228599340576,
+      "grad_norm": 3.25,
+      "learning_rate": 9.421545810877022e-06,
+      "loss": 0.8219,
+      "step": 1990
+    },
+    {
+      "epoch": 0.061057516180241786,
+      "grad_norm": 3.375,
+      "learning_rate": 9.418483586477217e-06,
+      "loss": 0.865,
+      "step": 2000
+    },
+    {
+      "epoch": 0.061362803761142996,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.415421362077414e-06,
+      "loss": 0.9355,
+      "step": 2010
+    },
+    {
+      "epoch": 0.06166809134204421,
+      "grad_norm": 2.875,
+      "learning_rate": 9.41235913767761e-06,
+      "loss": 0.8997,
+      "step": 2020
+    },
+    {
+      "epoch": 0.06197337892294542,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.409296913277806e-06,
+      "loss": 0.9249,
+      "step": 2030
+    },
+    {
+      "epoch": 0.06227866650384662,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.406234688878002e-06,
+      "loss": 0.9141,
+      "step": 2040
+    },
+    {
+      "epoch": 0.06258395408474783,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.403172464478198e-06,
+      "loss": 0.8814,
+      "step": 2050
+    },
+    {
+      "epoch": 0.06288924166564905,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.400110240078394e-06,
+      "loss": 0.9306,
+      "step": 2060
+    },
+    {
+      "epoch": 0.06319452924655025,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.39704801567859e-06,
+      "loss": 0.9188,
+      "step": 2070
+    },
+    {
+      "epoch": 0.06349981682745146,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.393985791278786e-06,
+      "loss": 0.9032,
+      "step": 2080
+    },
+    {
+      "epoch": 0.06380510440835267,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.390923566878982e-06,
+      "loss": 0.8616,
+      "step": 2090
+    },
+    {
+      "epoch": 0.06411039198925388,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.387861342479177e-06,
+      "loss": 0.8625,
+      "step": 2100
+    },
+    {
+      "epoch": 0.06441567957015508,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.384799118079374e-06,
+      "loss": 0.8442,
+      "step": 2110
+    },
+    {
+      "epoch": 0.0647209671510563,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.381736893679569e-06,
+      "loss": 0.9003,
+      "step": 2120
+    },
+    {
+      "epoch": 0.0650262547319575,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.378674669279766e-06,
+      "loss": 0.8282,
+      "step": 2130
+    },
+    {
+      "epoch": 0.06533154231285872,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.37561244487996e-06,
+      "loss": 0.8626,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06563682989375992,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.372550220480158e-06,
+      "loss": 0.8251,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06594211747466112,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.369487996080353e-06,
+      "loss": 0.7694,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06624740505556234,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.36642577168055e-06,
+      "loss": 0.8319,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06655269263646355,
+      "grad_norm": 3.25,
+      "learning_rate": 9.363363547280747e-06,
+      "loss": 0.8894,
+      "step": 2180
+    },
+    {
+      "epoch": 0.06685798021736476,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.360301322880942e-06,
+      "loss": 0.8905,
+      "step": 2190
+    },
+    {
+      "epoch": 0.06716326779826597,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.357239098481139e-06,
+      "loss": 0.9242,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06746855537916717,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.354176874081334e-06,
+      "loss": 0.8712,
+      "step": 2210
+    },
+    {
+      "epoch": 0.06777384296006839,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.351114649681529e-06,
+      "loss": 0.9268,
+      "step": 2220
+    },
+    {
+      "epoch": 0.06807913054096959,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.348052425281726e-06,
+      "loss": 0.8769,
+      "step": 2230
+    },
+    {
+      "epoch": 0.06838441812187081,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.344990200881921e-06,
+      "loss": 0.8741,
+      "step": 2240
+    },
+    {
+      "epoch": 0.06868970570277201,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.341927976482118e-06,
+      "loss": 0.8645,
+      "step": 2250
+    },
+    {
+      "epoch": 0.06899499328367321,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.338865752082313e-06,
+      "loss": 0.9076,
+      "step": 2260
+    },
+    {
+      "epoch": 0.06930028086457443,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.33580352768251e-06,
+      "loss": 0.8846,
+      "step": 2270
+    },
+    {
+      "epoch": 0.06960556844547564,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.332741303282705e-06,
+      "loss": 0.8677,
+      "step": 2280
+    },
+    {
+      "epoch": 0.06991085602637685,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.329679078882902e-06,
+      "loss": 0.851,
+      "step": 2290
+    },
+    {
+      "epoch": 0.07021614360727806,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.326616854483097e-06,
+      "loss": 0.8506,
+      "step": 2300
+    },
+    {
+      "epoch": 0.07052143118817926,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.323554630083294e-06,
+      "loss": 0.9108,
+      "step": 2310
+    },
+    {
+      "epoch": 0.07082671876908048,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.320492405683489e-06,
+      "loss": 0.9065,
+      "step": 2320
+    },
+    {
+      "epoch": 0.07113200634998168,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.317430181283686e-06,
+      "loss": 0.9092,
+      "step": 2330
+    },
+    {
+      "epoch": 0.0714372939308829,
+      "grad_norm": 3.25,
+      "learning_rate": 9.314367956883881e-06,
+      "loss": 0.9072,
+      "step": 2340
+    },
+    {
+      "epoch": 0.0717425815117841,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.311305732484078e-06,
+      "loss": 0.9022,
+      "step": 2350
+    },
+    {
+      "epoch": 0.0720478690926853,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.308243508084273e-06,
+      "loss": 0.903,
+      "step": 2360
+    },
+    {
+      "epoch": 0.07235315667358652,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.30518128368447e-06,
+      "loss": 0.8631,
+      "step": 2370
+    },
+    {
+      "epoch": 0.07265844425448773,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.302119059284665e-06,
+      "loss": 0.8207,
+      "step": 2380
+    },
+    {
+      "epoch": 0.07296373183538894,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.29905683488486e-06,
+      "loss": 0.9078,
+      "step": 2390
+    },
+    {
+      "epoch": 0.07326901941629015,
+      "grad_norm": 3.375,
+      "learning_rate": 9.295994610485057e-06,
+      "loss": 0.823,
+      "step": 2400
+    },
+    {
+      "epoch": 0.07357430699719135,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.292932386085252e-06,
+      "loss": 0.8617,
+      "step": 2410
+    },
+    {
+      "epoch": 0.07387959457809257,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.289870161685449e-06,
+      "loss": 0.9061,
+      "step": 2420
+    },
+    {
+      "epoch": 0.07418488215899377,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.286807937285644e-06,
+      "loss": 0.9113,
+      "step": 2430
+    },
+    {
+      "epoch": 0.07449016973989499,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.283745712885841e-06,
+      "loss": 0.9166,
+      "step": 2440
+    },
+    {
+      "epoch": 0.07479545732079619,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.280683488486038e-06,
+      "loss": 0.8694,
+      "step": 2450
+    },
+    {
+      "epoch": 0.0751007449016974,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.277621264086233e-06,
+      "loss": 0.9137,
+      "step": 2460
+    },
+    {
+      "epoch": 0.07540603248259861,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.27455903968643e-06,
+      "loss": 0.8935,
+      "step": 2470
+    },
+    {
+      "epoch": 0.07571132006349982,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.271496815286625e-06,
+      "loss": 0.9103,
+      "step": 2480
+    },
+    {
+      "epoch": 0.07601660764440103,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.268434590886822e-06,
+      "loss": 0.8328,
+      "step": 2490
+    },
+    {
+      "epoch": 0.07632189522530224,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.265372366487017e-06,
+      "loss": 0.9323,
+      "step": 2500
+    },
+    {
+      "epoch": 0.07662718280620344,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.262310142087212e-06,
+      "loss": 0.8654,
+      "step": 2510
+    },
+    {
+      "epoch": 0.07693247038710466,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.259247917687409e-06,
+      "loss": 0.9103,
+      "step": 2520
+    },
+    {
+      "epoch": 0.07723775796800586,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.256185693287604e-06,
+      "loss": 0.8987,
+      "step": 2530
+    },
+    {
+      "epoch": 0.07754304554890706,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.253123468887801e-06,
+      "loss": 0.8538,
+      "step": 2540
+    },
+    {
+      "epoch": 0.07784833312980828,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.250061244487996e-06,
+      "loss": 0.847,
+      "step": 2550
+    },
+    {
+      "epoch": 0.07815362071070948,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.246999020088193e-06,
+      "loss": 0.8614,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0784589082916107,
+      "grad_norm": 2.875,
+      "learning_rate": 9.243936795688388e-06,
+      "loss": 0.8367,
+      "step": 2570
+    },
+    {
+      "epoch": 0.0787641958725119,
+      "grad_norm": 3.5,
+      "learning_rate": 9.240874571288585e-06,
+      "loss": 0.789,
+      "step": 2580
+    },
+    {
+      "epoch": 0.07906948345341311,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.23781234688878e-06,
+      "loss": 0.8573,
+      "step": 2590
+    },
+    {
+      "epoch": 0.07937477103431433,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.234750122488977e-06,
+      "loss": 0.9036,
+      "step": 2600
+    },
+    {
+      "epoch": 0.07968005861521553,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.231687898089172e-06,
+      "loss": 0.8478,
+      "step": 2610
+    },
+    {
+      "epoch": 0.07998534619611675,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.228625673689369e-06,
+      "loss": 0.913,
+      "step": 2620
+    },
+    {
+      "epoch": 0.08029063377701795,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.225563449289566e-06,
+      "loss": 0.8799,
+      "step": 2630
+    },
+    {
+      "epoch": 0.08059592135791915,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.222501224889761e-06,
+      "loss": 0.9194,
+      "step": 2640
+    },
+    {
+      "epoch": 0.08090120893882037,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.219439000489956e-06,
+      "loss": 0.8215,
+      "step": 2650
+    },
+    {
+      "epoch": 0.08120649651972157,
+      "grad_norm": 3.5,
+      "learning_rate": 9.216376776090153e-06,
+      "loss": 0.9236,
+      "step": 2660
+    },
+    {
+      "epoch": 0.08151178410062279,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.213314551690348e-06,
+      "loss": 0.8724,
+      "step": 2670
+    },
+    {
+      "epoch": 0.081817071681524,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.210252327290543e-06,
+      "loss": 0.823,
+      "step": 2680
+    },
+    {
+      "epoch": 0.0821223592624252,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.20719010289074e-06,
+      "loss": 0.8519,
+      "step": 2690
+    },
+    {
+      "epoch": 0.08242764684332642,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.204127878490935e-06,
+      "loss": 0.8627,
+      "step": 2700
+    },
+    {
+      "epoch": 0.08273293442422762,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.201065654091132e-06,
+      "loss": 0.8806,
+      "step": 2710
+    },
+    {
+      "epoch": 0.08303822200512884,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.198003429691329e-06,
+      "loss": 0.8249,
+      "step": 2720
+    },
+    {
+      "epoch": 0.08334350958603004,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.194941205291524e-06,
+      "loss": 0.8761,
+      "step": 2730
+    },
+    {
+      "epoch": 0.08364879716693124,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.191878980891721e-06,
+      "loss": 0.9413,
+      "step": 2740
+    },
+    {
+      "epoch": 0.08395408474783246,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.188816756491916e-06,
+      "loss": 0.8763,
+      "step": 2750
+    },
+    {
+      "epoch": 0.08425937232873366,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.185754532092113e-06,
+      "loss": 0.8503,
+      "step": 2760
+    },
+    {
+      "epoch": 0.08456465990963488,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.182692307692308e-06,
+      "loss": 0.8561,
+      "step": 2770
+    },
+    {
+      "epoch": 0.08486994749053609,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.179630083292505e-06,
+      "loss": 0.8507,
+      "step": 2780
+    },
+    {
+      "epoch": 0.08517523507143729,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.1765678588927e-06,
+      "loss": 0.8607,
+      "step": 2790
+    },
+    {
+      "epoch": 0.0854805226523385,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.173505634492895e-06,
+      "loss": 0.8737,
+      "step": 2800
+    },
+    {
+      "epoch": 0.08578581023323971,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.170443410093092e-06,
+      "loss": 0.9171,
+      "step": 2810
+    },
+    {
+      "epoch": 0.08609109781414093,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.167381185693288e-06,
+      "loss": 0.891,
+      "step": 2820
+    },
+    {
+      "epoch": 0.08639638539504213,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.164318961293484e-06,
+      "loss": 0.8798,
+      "step": 2830
+    },
+    {
+      "epoch": 0.08670167297594333,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.16125673689368e-06,
+      "loss": 0.8738,
+      "step": 2840
+    },
+    {
+      "epoch": 0.08700696055684455,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.158194512493876e-06,
+      "loss": 0.9028,
+      "step": 2850
+    },
+    {
+      "epoch": 0.08731224813774575,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.155132288094072e-06,
+      "loss": 0.8294,
+      "step": 2860
+    },
+    {
+      "epoch": 0.08761753571864697,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.152070063694268e-06,
+      "loss": 0.9248,
+      "step": 2870
+    },
+    {
+      "epoch": 0.08792282329954818,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.149007839294465e-06,
+      "loss": 0.9019,
+      "step": 2880
+    },
+    {
+      "epoch": 0.08822811088044938,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.14594561489466e-06,
+      "loss": 0.8378,
+      "step": 2890
+    },
+    {
+      "epoch": 0.0885333984613506,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.142883390494857e-06,
+      "loss": 0.8593,
+      "step": 2900
+    },
+    {
+      "epoch": 0.0888386860422518,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.139821166095052e-06,
+      "loss": 0.8188,
+      "step": 2910
+    },
+    {
+      "epoch": 0.08914397362315302,
+      "grad_norm": 3.5,
+      "learning_rate": 9.13675894169525e-06,
+      "loss": 0.8724,
+      "step": 2920
+    },
+    {
+      "epoch": 0.08944926120405422,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.133696717295444e-06,
+      "loss": 0.8126,
+      "step": 2930
+    },
+    {
+      "epoch": 0.08975454878495542,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.13063449289564e-06,
+      "loss": 0.9146,
+      "step": 2940
+    },
+    {
+      "epoch": 0.09005983636585664,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.127572268495836e-06,
+      "loss": 0.8715,
+      "step": 2950
+    },
+    {
+      "epoch": 0.09036512394675784,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.124510044096032e-06,
+      "loss": 0.8304,
+      "step": 2960
+    },
+    {
+      "epoch": 0.09067041152765906,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.121447819696228e-06,
+      "loss": 0.8383,
+      "step": 2970
+    },
+    {
+      "epoch": 0.09097569910856027,
+      "grad_norm": 3.125,
+      "learning_rate": 9.118385595296424e-06,
+      "loss": 0.8253,
+      "step": 2980
+    },
+    {
+      "epoch": 0.09128098668946147,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.11532337089662e-06,
+      "loss": 0.8607,
+      "step": 2990
+    },
+    {
+      "epoch": 0.09158627427036269,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.112261146496816e-06,
+      "loss": 0.8829,
+      "step": 3000
+    },
+    {
+      "epoch": 0.09189156185126389,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.109198922097012e-06,
+      "loss": 0.8869,
+      "step": 3010
+    },
+    {
+      "epoch": 0.09219684943216509,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.106136697697208e-06,
+      "loss": 0.8921,
+      "step": 3020
+    },
+    {
+      "epoch": 0.09250213701306631,
+      "grad_norm": 3.0,
+      "learning_rate": 9.103074473297404e-06,
+      "loss": 0.8421,
+      "step": 3030
+    },
+    {
+      "epoch": 0.09280742459396751,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.1000122488976e-06,
+      "loss": 0.8599,
+      "step": 3040
+    },
+    {
+      "epoch": 0.09311271217486873,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.096950024497796e-06,
+      "loss": 0.9195,
+      "step": 3050
+    },
+    {
+      "epoch": 0.09341799975576993,
+      "grad_norm": 3.25,
+      "learning_rate": 9.093887800097992e-06,
+      "loss": 0.8408,
+      "step": 3060
+    },
+    {
+      "epoch": 0.09372328733667114,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.090825575698188e-06,
+      "loss": 0.8219,
+      "step": 3070
+    },
+    {
+      "epoch": 0.09402857491757236,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.087763351298384e-06,
+      "loss": 0.9182,
+      "step": 3080
+    },
+    {
+      "epoch": 0.09433386249847356,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.084701126898579e-06,
+      "loss": 0.8481,
+      "step": 3090
+    },
+    {
+      "epoch": 0.09463915007937478,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.081638902498776e-06,
+      "loss": 0.8348,
+      "step": 3100
+    },
+    {
+      "epoch": 0.09494443766027598,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.07857667809897e-06,
+      "loss": 0.8967,
+      "step": 3110
+    },
+    {
+      "epoch": 0.09524972524117718,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.075514453699168e-06,
+      "loss": 0.8335,
+      "step": 3120
+    },
+    {
+      "epoch": 0.0955550128220784,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.072452229299363e-06,
+      "loss": 0.8171,
+      "step": 3130
+    },
+    {
+      "epoch": 0.0958603004029796,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.06939000489956e-06,
+      "loss": 0.8856,
+      "step": 3140
+    },
+    {
+      "epoch": 0.09616558798388082,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.066327780499757e-06,
+      "loss": 0.8764,
+      "step": 3150
+    },
+    {
+      "epoch": 0.09647087556478202,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.063265556099952e-06,
+      "loss": 0.8428,
+      "step": 3160
+    },
+    {
+      "epoch": 0.09677616314568323,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.060203331700149e-06,
+      "loss": 0.8285,
+      "step": 3170
+    },
+    {
+      "epoch": 0.09708145072658445,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.057141107300344e-06,
+      "loss": 0.8661,
+      "step": 3180
+    },
+    {
+      "epoch": 0.09738673830748565,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.05407888290054e-06,
+      "loss": 0.8261,
+      "step": 3190
+    },
+    {
+      "epoch": 0.09769202588838687,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.051016658500736e-06,
+      "loss": 0.885,
+      "step": 3200
+    },
+    {
+      "epoch": 0.09799731346928807,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.047954434100933e-06,
+      "loss": 0.8675,
+      "step": 3210
+    },
+    {
+      "epoch": 0.09830260105018927,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.044892209701128e-06,
+      "loss": 0.8615,
+      "step": 3220
+    },
+    {
+      "epoch": 0.09860788863109049,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.041829985301323e-06,
+      "loss": 0.8702,
+      "step": 3230
+    },
+    {
+      "epoch": 0.0989131762119917,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.03876776090152e-06,
+      "loss": 0.8303,
+      "step": 3240
+    },
+    {
+      "epoch": 0.09921846379289291,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.035705536501715e-06,
+      "loss": 0.8382,
+      "step": 3250
+    },
+    {
+      "epoch": 0.09952375137379411,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.032643312101912e-06,
+      "loss": 0.8252,
+      "step": 3260
+    },
+    {
+      "epoch": 0.09982903895469532,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.029581087702107e-06,
+      "loss": 0.8512,
+      "step": 3270
+    },
+    {
+      "epoch": 0.10013432653559654,
+      "grad_norm": 3.25,
+      "learning_rate": 9.026518863302304e-06,
+      "loss": 0.8535,
+      "step": 3280
+    },
+    {
+      "epoch": 0.10043961411649774,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.023456638902499e-06,
+      "loss": 0.8823,
+      "step": 3290
+    },
+    {
+      "epoch": 0.10074490169739896,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.020394414502696e-06,
+      "loss": 0.8796,
+      "step": 3300
+    },
+    {
+      "epoch": 0.10105018927830016,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.017332190102891e-06,
+      "loss": 0.8727,
+      "step": 3310
+    },
+    {
+      "epoch": 0.10135547685920136,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.014269965703088e-06,
+      "loss": 0.834,
+      "step": 3320
+    },
+    {
+      "epoch": 0.10166076444010258,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.011207741303285e-06,
+      "loss": 0.9174,
+      "step": 3330
+    },
+    {
+      "epoch": 0.10196605202100378,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.00814551690348e-06,
+      "loss": 0.8577,
+      "step": 3340
+    },
+    {
+      "epoch": 0.102271339601905,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.005083292503675e-06,
+      "loss": 0.861,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1025766271828062,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.002021068103872e-06,
+      "loss": 0.8742,
+      "step": 3360
+    },
+    {
+      "epoch": 0.10288191476370741,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.998958843704067e-06,
+      "loss": 0.812,
+      "step": 3370
+    },
+    {
+      "epoch": 0.10318720234460863,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.995896619304262e-06,
+      "loss": 0.8485,
+      "step": 3380
+    },
+    {
+      "epoch": 0.10349248992550983,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.992834394904459e-06,
+      "loss": 0.8712,
+      "step": 3390
+    },
+    {
+      "epoch": 0.10379777750641105,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.989772170504654e-06,
+      "loss": 0.8392,
+      "step": 3400
+    },
+    {
+      "epoch": 0.10410306508731225,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.986709946104851e-06,
+      "loss": 0.8936,
+      "step": 3410
+    },
+    {
+      "epoch": 0.10440835266821345,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.983647721705048e-06,
+      "loss": 0.9101,
+      "step": 3420
+    },
+    {
+      "epoch": 0.10471364024911467,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.980585497305243e-06,
+      "loss": 0.9279,
+      "step": 3430
+    },
+    {
+      "epoch": 0.10501892783001587,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.97752327290544e-06,
+      "loss": 0.866,
+      "step": 3440
+    },
+    {
+      "epoch": 0.10532421541091709,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.974461048505635e-06,
+      "loss": 0.8797,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1056295029918183,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.971398824105832e-06,
+      "loss": 0.8692,
+      "step": 3460
+    },
+    {
+      "epoch": 0.1059347905727195,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.968336599706027e-06,
+      "loss": 0.869,
+      "step": 3470
+    },
+    {
+      "epoch": 0.10624007815362072,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.965274375306224e-06,
+      "loss": 0.8379,
+      "step": 3480
+    },
+    {
+      "epoch": 0.10654536573452192,
+      "grad_norm": 3.125,
+      "learning_rate": 8.962212150906419e-06,
+      "loss": 0.8567,
+      "step": 3490
+    },
+    {
+      "epoch": 0.10685065331542312,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.959149926506616e-06,
+      "loss": 0.894,
+      "step": 3500
+    },
+    {
+      "epoch": 0.10715594089632434,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.956087702106811e-06,
+      "loss": 0.8947,
+      "step": 3510
+    },
+    {
+      "epoch": 0.10746122847722554,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.953025477707006e-06,
+      "loss": 0.8774,
+      "step": 3520
+    },
+    {
+      "epoch": 0.10776651605812676,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.949963253307203e-06,
+      "loss": 0.8303,
+      "step": 3530
+    },
+    {
+      "epoch": 0.10807180363902796,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.946901028907398e-06,
+      "loss": 0.8159,
+      "step": 3540
+    },
+    {
+      "epoch": 0.10837709121992917,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.943838804507595e-06,
+      "loss": 0.8316,
+      "step": 3550
+    },
+    {
+      "epoch": 0.10868237880083038,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.94077658010779e-06,
+      "loss": 0.8864,
+      "step": 3560
+    },
+    {
+      "epoch": 0.10898766638173159,
+      "grad_norm": 3.125,
+      "learning_rate": 8.937714355707987e-06,
+      "loss": 0.8515,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1092929539626328,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.934652131308184e-06,
+      "loss": 0.9144,
+      "step": 3580
+    },
+    {
+      "epoch": 0.10959824154353401,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.931589906908379e-06,
+      "loss": 0.8294,
+      "step": 3590
+    },
+    {
+      "epoch": 0.10990352912443521,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.928527682508576e-06,
+      "loss": 0.8606,
+      "step": 3600
+    },
+    {
+      "epoch": 0.11020881670533643,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.925465458108771e-06,
+      "loss": 0.8209,
+      "step": 3610
+    },
+    {
+      "epoch": 0.11051410428623763,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.922403233708968e-06,
+      "loss": 0.8359,
+      "step": 3620
+    },
+    {
+      "epoch": 0.11081939186713885,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.919341009309163e-06,
+      "loss": 0.8809,
+      "step": 3630
+    },
+    {
+      "epoch": 0.11112467944804005,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.916278784909358e-06,
+      "loss": 0.7898,
+      "step": 3640
+    },
+    {
+      "epoch": 0.11142996702894126,
+      "grad_norm": 3.0,
+      "learning_rate": 8.913216560509555e-06,
+      "loss": 0.8964,
+      "step": 3650
+    },
+    {
+      "epoch": 0.11173525460984247,
+      "grad_norm": 3.125,
+      "learning_rate": 8.91015433610975e-06,
+      "loss": 0.8355,
+      "step": 3660
+    },
+    {
+      "epoch": 0.11204054219074368,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.907092111709947e-06,
+      "loss": 0.9046,
+      "step": 3670
+    },
+    {
+      "epoch": 0.1123458297716449,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.904029887310142e-06,
+      "loss": 0.8837,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1126511173525461,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.900967662910339e-06,
+      "loss": 0.863,
+      "step": 3690
+    },
+    {
+      "epoch": 0.1129564049334473,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.897905438510534e-06,
+      "loss": 0.8188,
+      "step": 3700
+    },
+    {
+      "epoch": 0.11326169251434852,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.894843214110731e-06,
+      "loss": 0.8824,
+      "step": 3710
+    },
+    {
+      "epoch": 0.11356698009524972,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.891780989710926e-06,
+      "loss": 0.8311,
+      "step": 3720
+    },
+    {
+      "epoch": 0.11387226767615094,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.888718765311123e-06,
+      "loss": 0.8203,
+      "step": 3730
+    },
+    {
+      "epoch": 0.11417755525705214,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.885656540911318e-06,
+      "loss": 0.8409,
+      "step": 3740
+    },
+    {
+      "epoch": 0.11448284283795335,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.882594316511515e-06,
+      "loss": 0.8569,
+      "step": 3750
+    },
+    {
+      "epoch": 0.11478813041885456,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.879532092111712e-06,
+      "loss": 0.8293,
+      "step": 3760
+    },
+    {
+      "epoch": 0.11509341799975577,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.876469867711907e-06,
+      "loss": 0.8263,
+      "step": 3770
+    },
+    {
+      "epoch": 0.11539870558065699,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.873407643312102e-06,
+      "loss": 0.882,
+      "step": 3780
+    },
+    {
+      "epoch": 0.11570399316155819,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.8703454189123e-06,
+      "loss": 0.837,
+      "step": 3790
+    },
+    {
+      "epoch": 0.11600928074245939,
+      "grad_norm": 3.515625,
+      "learning_rate": 8.867283194512494e-06,
+      "loss": 0.8697,
+      "step": 3800
+    },
+    {
+      "epoch": 0.11631456832336061,
+      "grad_norm": 3.0,
+      "learning_rate": 8.86422097011269e-06,
+      "loss": 0.8559,
+      "step": 3810
+    },
+    {
+      "epoch": 0.11661985590426181,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.861158745712886e-06,
+      "loss": 0.9272,
+      "step": 3820
+    },
+    {
+      "epoch": 0.11692514348516303,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.858096521313082e-06,
+      "loss": 0.8549,
+      "step": 3830
+    },
+    {
+      "epoch": 0.11723043106606423,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.855034296913278e-06,
+      "loss": 0.8147,
+      "step": 3840
+    },
+    {
+      "epoch": 0.11753571864696544,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.851972072513475e-06,
+      "loss": 0.8285,
+      "step": 3850
+    },
+    {
+      "epoch": 0.11784100622786665,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.84890984811367e-06,
+      "loss": 0.8705,
+      "step": 3860
+    },
+    {
+      "epoch": 0.11814629380876786,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.845847623713867e-06,
+      "loss": 0.8007,
+      "step": 3870
+    },
+    {
+      "epoch": 0.11845158138966908,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.842785399314062e-06,
+      "loss": 0.8207,
+      "step": 3880
+    },
+    {
+      "epoch": 0.11875686897057028,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.83972317491426e-06,
+      "loss": 0.8006,
+      "step": 3890
+    },
+    {
+      "epoch": 0.11906215655147148,
+      "grad_norm": 3.375,
+      "learning_rate": 8.836660950514454e-06,
+      "loss": 0.8612,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1193674441323727,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.833598726114651e-06,
+      "loss": 0.8499,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1196727317132739,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.830536501714846e-06,
+      "loss": 0.8096,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1199780192941751,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.827474277315042e-06,
+      "loss": 0.8464,
+      "step": 3930
+    },
+    {
+      "epoch": 0.12028330687507632,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.824412052915238e-06,
+      "loss": 0.8199,
+      "step": 3940
+    },
+    {
+      "epoch": 0.12058859445597753,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.821349828515434e-06,
+      "loss": 0.8148,
+      "step": 3950
+    },
+    {
+      "epoch": 0.12089388203687874,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.81828760411563e-06,
+      "loss": 0.8264,
+      "step": 3960
+    },
+    {
+      "epoch": 0.12119916961777995,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.815225379715826e-06,
+      "loss": 0.8713,
+      "step": 3970
+    },
+    {
+      "epoch": 0.12150445719868115,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.812163155316022e-06,
+      "loss": 0.9008,
+      "step": 3980
+    },
+    {
+      "epoch": 0.12180974477958237,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.809100930916218e-06,
+      "loss": 0.8192,
+      "step": 3990
+    },
+    {
+      "epoch": 0.12211503236048357,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.806038706516414e-06,
+      "loss": 0.866,
+      "step": 4000
+    },
+    {
+      "epoch": 0.12242031994138479,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.80297648211661e-06,
+      "loss": 0.8421,
+      "step": 4010
+    },
+    {
+      "epoch": 0.12272560752228599,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.799914257716806e-06,
+      "loss": 0.8178,
+      "step": 4020
+    },
+    {
+      "epoch": 0.1230308951031872,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.796852033317003e-06,
+      "loss": 0.9061,
+      "step": 4030
+    },
+    {
+      "epoch": 0.12333618268408841,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.793789808917198e-06,
+      "loss": 0.8167,
+      "step": 4040
+    },
+    {
+      "epoch": 0.12364147026498962,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.790727584517395e-06,
+      "loss": 0.8309,
+      "step": 4050
+    },
+    {
+      "epoch": 0.12394675784589083,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.78766536011759e-06,
+      "loss": 0.8796,
+      "step": 4060
+    },
+    {
+      "epoch": 0.12425204542679204,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.784603135717786e-06,
+      "loss": 0.8347,
+      "step": 4070
+    },
+    {
+      "epoch": 0.12455733300769324,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.781540911317982e-06,
+      "loss": 0.8531,
+      "step": 4080
+    },
+    {
+      "epoch": 0.12486262058859446,
+      "grad_norm": 3.375,
+      "learning_rate": 8.778478686918178e-06,
+      "loss": 0.7842,
+      "step": 4090
+    },
+    {
+      "epoch": 0.12516790816949566,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.775416462518373e-06,
+      "loss": 0.8242,
+      "step": 4100
+    },
+    {
+      "epoch": 0.12547319575039687,
+      "grad_norm": 3.375,
+      "learning_rate": 8.77235423811857e-06,
+      "loss": 0.8885,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1257784833312981,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.769292013718767e-06,
+      "loss": 0.842,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1260837709121993,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.766229789318962e-06,
+      "loss": 0.7795,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1263890584931005,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.763167564919159e-06,
+      "loss": 0.8791,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1266943460740017,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.760105340519354e-06,
+      "loss": 0.8365,
+      "step": 4150
+    },
+    {
+      "epoch": 0.1269996336549029,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.75704311611955e-06,
+      "loss": 0.8807,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1273049212358041,
+      "grad_norm": 3.25,
+      "learning_rate": 8.753980891719746e-06,
+      "loss": 0.8387,
+      "step": 4170
+    },
+    {
+      "epoch": 0.12761020881670534,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.750918667319943e-06,
+      "loss": 0.8679,
+      "step": 4180
+    },
+    {
+      "epoch": 0.12791549639760655,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.747856442920138e-06,
+      "loss": 0.7803,
+      "step": 4190
+    },
+    {
+      "epoch": 0.12822078397850775,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.744794218520335e-06,
+      "loss": 0.8789,
+      "step": 4200
+    },
+    {
+      "epoch": 0.12852607155940896,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.74173199412053e-06,
+      "loss": 0.8865,
+      "step": 4210
+    },
+    {
+      "epoch": 0.12883135914031016,
+      "grad_norm": 3.25,
+      "learning_rate": 8.738669769720725e-06,
+      "loss": 0.8886,
+      "step": 4220
+    },
+    {
+      "epoch": 0.1291366467212114,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.735607545320922e-06,
+      "loss": 0.8801,
+      "step": 4230
+    },
+    {
+      "epoch": 0.1294419343021126,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.732545320921117e-06,
+      "loss": 0.9391,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1297472218830138,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.729483096521314e-06,
+      "loss": 0.85,
+      "step": 4250
+    },
+    {
+      "epoch": 0.130052509463915,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.726420872121509e-06,
+      "loss": 0.7823,
+      "step": 4260
+    },
+    {
+      "epoch": 0.1303577970448162,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.723358647721706e-06,
+      "loss": 0.8654,
+      "step": 4270
+    },
+    {
+      "epoch": 0.13066308462571743,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.720296423321903e-06,
+      "loss": 0.8737,
+      "step": 4280
+    },
+    {
+      "epoch": 0.13096837220661864,
+      "grad_norm": 3.0,
+      "learning_rate": 8.717234198922098e-06,
+      "loss": 0.8922,
+      "step": 4290
+    },
+    {
+      "epoch": 0.13127365978751984,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.714171974522295e-06,
+      "loss": 0.875,
+      "step": 4300
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 3.0,
+      "learning_rate": 8.71110975012249e-06,
+      "loss": 0.8183,
+      "step": 4310
+    },
+    {
+      "epoch": 0.13188423494932225,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.708047525722687e-06,
+      "loss": 0.8126,
+      "step": 4320
+    },
+    {
+      "epoch": 0.13218952253022348,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.704985301322882e-06,
+      "loss": 0.8259,
+      "step": 4330
+    },
+    {
+      "epoch": 0.13249481011112468,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.701923076923079e-06,
+      "loss": 0.8148,
+      "step": 4340
+    },
+    {
+      "epoch": 0.1328000976920259,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.698860852523274e-06,
+      "loss": 0.8838,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1331053852729271,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.695798628123469e-06,
+      "loss": 0.8401,
+      "step": 4360
+    },
+    {
+      "epoch": 0.1334106728538283,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.692736403723666e-06,
+      "loss": 0.9413,
+      "step": 4370
+    },
+    {
+      "epoch": 0.13371596043472952,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.689674179323861e-06,
+      "loss": 0.8483,
+      "step": 4380
+    },
+    {
+      "epoch": 0.13402124801563073,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.686611954924058e-06,
+      "loss": 0.8461,
+      "step": 4390
+    },
+    {
+      "epoch": 0.13432653559653193,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.683549730524253e-06,
+      "loss": 0.8363,
+      "step": 4400
+    },
+    {
+      "epoch": 0.13463182317743314,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.68048750612445e-06,
+      "loss": 0.872,
+      "step": 4410
+    },
+    {
+      "epoch": 0.13493711075833434,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.677425281724645e-06,
+      "loss": 0.862,
+      "step": 4420
+    },
+    {
+      "epoch": 0.13524239833923557,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.674363057324842e-06,
+      "loss": 0.7503,
+      "step": 4430
+    },
+    {
+      "epoch": 0.13554768592013677,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.671300832925037e-06,
+      "loss": 0.8175,
+      "step": 4440
+    },
+    {
+      "epoch": 0.13585297350103798,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.668238608525234e-06,
+      "loss": 0.8313,
+      "step": 4450
+    },
+    {
+      "epoch": 0.13615826108193918,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.66517638412543e-06,
+      "loss": 0.8072,
+      "step": 4460
+    },
+    {
+      "epoch": 0.13646354866284038,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.662114159725626e-06,
+      "loss": 0.8037,
+      "step": 4470
+    },
+    {
+      "epoch": 0.13676883624374161,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.659051935325821e-06,
+      "loss": 0.8671,
+      "step": 4480
+    },
+    {
+      "epoch": 0.13707412382464282,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.655989710926018e-06,
+      "loss": 0.8378,
+      "step": 4490
+    },
+    {
+      "epoch": 0.13737941140554402,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.652927486526213e-06,
+      "loss": 0.8328,
+      "step": 4500
+    },
+    {
+      "epoch": 0.13768469898644523,
+      "grad_norm": 3.375,
+      "learning_rate": 8.649865262126408e-06,
+      "loss": 0.8883,
+      "step": 4510
+    },
+    {
+      "epoch": 0.13798998656734643,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.646803037726605e-06,
+      "loss": 0.8949,
+      "step": 4520
+    },
+    {
+      "epoch": 0.13829527414824766,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.6437408133268e-06,
+      "loss": 0.9181,
+      "step": 4530
+    },
+    {
+      "epoch": 0.13860056172914886,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.640678588926997e-06,
+      "loss": 0.8328,
+      "step": 4540
+    },
+    {
+      "epoch": 0.13890584931005007,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.637616364527194e-06,
+      "loss": 0.8801,
+      "step": 4550
+    },
+    {
+      "epoch": 0.13921113689095127,
+      "grad_norm": 3.25,
+      "learning_rate": 8.634554140127389e-06,
+      "loss": 0.8368,
+      "step": 4560
+    },
+    {
+      "epoch": 0.13951642447185247,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.631491915727586e-06,
+      "loss": 0.8359,
+      "step": 4570
+    },
+    {
+      "epoch": 0.1398217120527537,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.628429691327781e-06,
+      "loss": 0.828,
+      "step": 4580
+    },
+    {
+      "epoch": 0.1401269996336549,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.625367466927978e-06,
+      "loss": 0.8531,
+      "step": 4590
+    },
+    {
+      "epoch": 0.1404322872145561,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.622305242528173e-06,
+      "loss": 0.8511,
+      "step": 4600
+    },
+    {
+      "epoch": 0.14073757479545732,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.61924301812837e-06,
+      "loss": 0.8667,
+      "step": 4610
+    },
+    {
+      "epoch": 0.14104286237635852,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.616180793728565e-06,
+      "loss": 0.8258,
+      "step": 4620
+    },
+    {
+      "epoch": 0.14134814995725975,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.613118569328762e-06,
+      "loss": 0.8434,
+      "step": 4630
+    },
+    {
+      "epoch": 0.14165343753816095,
+      "grad_norm": 3.796875,
+      "learning_rate": 8.610056344928957e-06,
+      "loss": 0.8189,
+      "step": 4640
+    },
+    {
+      "epoch": 0.14195872511906216,
+      "grad_norm": 3.25,
+      "learning_rate": 8.606994120529152e-06,
+      "loss": 0.8532,
+      "step": 4650
+    },
+    {
+      "epoch": 0.14226401269996336,
+      "grad_norm": 3.25,
+      "learning_rate": 8.603931896129349e-06,
+      "loss": 0.813,
+      "step": 4660
+    },
+    {
+      "epoch": 0.14256930028086456,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.600869671729544e-06,
+      "loss": 0.8077,
+      "step": 4670
+    },
+    {
+      "epoch": 0.1428745878617658,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.597807447329741e-06,
+      "loss": 0.8236,
+      "step": 4680
+    },
+    {
+      "epoch": 0.143179875442667,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.594745222929936e-06,
+      "loss": 0.8691,
+      "step": 4690
+    },
+    {
+      "epoch": 0.1434851630235682,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.591682998530133e-06,
+      "loss": 0.8619,
+      "step": 4700
+    },
+    {
+      "epoch": 0.1437904506044694,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.588620774130328e-06,
+      "loss": 0.841,
+      "step": 4710
+    },
+    {
+      "epoch": 0.1440957381853706,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.585558549730525e-06,
+      "loss": 0.7607,
+      "step": 4720
+    },
+    {
+      "epoch": 0.14440102576627184,
+      "grad_norm": 3.0,
+      "learning_rate": 8.582496325330722e-06,
+      "loss": 0.8268,
+      "step": 4730
+    },
+    {
+      "epoch": 0.14470631334717304,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.579434100930917e-06,
+      "loss": 0.8011,
+      "step": 4740
+    },
+    {
+      "epoch": 0.14501160092807425,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.576371876531114e-06,
+      "loss": 0.8422,
+      "step": 4750
+    },
+    {
+      "epoch": 0.14531688850897545,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.57330965213131e-06,
+      "loss": 0.8528,
+      "step": 4760
+    },
+    {
+      "epoch": 0.14562217608987665,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.570247427731504e-06,
+      "loss": 0.9041,
+      "step": 4770
+    },
+    {
+      "epoch": 0.14592746367077788,
+      "grad_norm": 3.0,
+      "learning_rate": 8.567185203331701e-06,
+      "loss": 0.8455,
+      "step": 4780
+    },
+    {
+      "epoch": 0.1462327512516791,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.564122978931896e-06,
+      "loss": 0.8614,
+      "step": 4790
+    },
+    {
+      "epoch": 0.1465380388325803,
+      "grad_norm": 3.0,
+      "learning_rate": 8.561060754532091e-06,
+      "loss": 0.7656,
+      "step": 4800
+    },
+    {
+      "epoch": 0.1468433264134815,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.557998530132288e-06,
+      "loss": 0.877,
+      "step": 4810
+    },
+    {
+      "epoch": 0.1471486139943827,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.554936305732485e-06,
+      "loss": 0.8231,
+      "step": 4820
+    },
+    {
+      "epoch": 0.14745390157528393,
+      "grad_norm": 3.875,
+      "learning_rate": 8.55187408133268e-06,
+      "loss": 0.8762,
+      "step": 4830
+    },
+    {
+      "epoch": 0.14775918915618513,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.548811856932877e-06,
+      "loss": 0.8296,
+      "step": 4840
+    },
+    {
+      "epoch": 0.14806447673708634,
+      "grad_norm": 3.0,
+      "learning_rate": 8.545749632533072e-06,
+      "loss": 0.8112,
+      "step": 4850
+    },
+    {
+      "epoch": 0.14836976431798754,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.54268740813327e-06,
+      "loss": 0.8193,
+      "step": 4860
+    },
+    {
+      "epoch": 0.14867505189888874,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.539625183733464e-06,
+      "loss": 0.8544,
+      "step": 4870
+    },
+    {
+      "epoch": 0.14898033947978997,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.536562959333661e-06,
+      "loss": 0.7955,
+      "step": 4880
+    },
+    {
+      "epoch": 0.14928562706069118,
+      "grad_norm": 3.890625,
+      "learning_rate": 8.533500734933856e-06,
+      "loss": 0.8793,
+      "step": 4890
+    },
+    {
+      "epoch": 0.14959091464159238,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.530438510534053e-06,
+      "loss": 0.8307,
+      "step": 4900
+    },
+    {
+      "epoch": 0.14989620222249359,
+      "grad_norm": 3.5,
+      "learning_rate": 8.527376286134248e-06,
+      "loss": 0.8702,
+      "step": 4910
+    },
+    {
+      "epoch": 0.1502014898033948,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.524314061734445e-06,
+      "loss": 0.8459,
+      "step": 4920
+    },
+    {
+      "epoch": 0.15050677738429602,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.52125183733464e-06,
+      "loss": 0.833,
+      "step": 4930
+    },
+    {
+      "epoch": 0.15081206496519722,
+      "grad_norm": 6.90625,
+      "learning_rate": 8.518189612934836e-06,
+      "loss": 0.7619,
+      "step": 4940
+    },
+    {
+      "epoch": 0.15111735254609843,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.515127388535032e-06,
+      "loss": 0.83,
+      "step": 4950
+    },
+    {
+      "epoch": 0.15142264012699963,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.512065164135228e-06,
+      "loss": 0.8667,
+      "step": 4960
+    },
+    {
+      "epoch": 0.15172792770790083,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.509002939735424e-06,
+      "loss": 0.8509,
+      "step": 4970
+    },
+    {
+      "epoch": 0.15203321528880206,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.50594071533562e-06,
+      "loss": 0.8448,
+      "step": 4980
+    },
+    {
+      "epoch": 0.15233850286970327,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.502878490935816e-06,
+      "loss": 0.8176,
+      "step": 4990
+    },
+    {
+      "epoch": 0.15264379045060447,
+      "grad_norm": 3.25,
+      "learning_rate": 8.499816266536013e-06,
+      "loss": 0.8212,
+      "step": 5000
+    },
+    {
+      "epoch": 0.15294907803150568,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.496754042136208e-06,
+      "loss": 0.8305,
+      "step": 5010
+    },
+    {
+      "epoch": 0.15325436561240688,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.493691817736405e-06,
+      "loss": 0.8465,
+      "step": 5020
+    },
+    {
+      "epoch": 0.1535596531933081,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.4906295933366e-06,
+      "loss": 0.7885,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1538649407742093,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.487567368936797e-06,
+      "loss": 0.8678,
+      "step": 5040
+    },
+    {
+      "epoch": 0.15417022835511052,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.484505144536992e-06,
+      "loss": 0.8054,
+      "step": 5050
+    },
+    {
+      "epoch": 0.15447551593601172,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.481442920137188e-06,
+      "loss": 0.8662,
+      "step": 5060
+    },
+    {
+      "epoch": 0.15478080351691292,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.478380695737384e-06,
+      "loss": 0.8629,
+      "step": 5070
+    },
+    {
+      "epoch": 0.15508609109781413,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.47531847133758e-06,
+      "loss": 0.8115,
+      "step": 5080
+    },
+    {
+      "epoch": 0.15539137867871536,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.472256246937777e-06,
+      "loss": 0.8543,
+      "step": 5090
+    },
+    {
+      "epoch": 0.15569666625961656,
+      "grad_norm": 2.984375,
+      "learning_rate": 8.469194022537972e-06,
+      "loss": 0.8217,
+      "step": 5100
+    },
+    {
+      "epoch": 0.15600195384051777,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.466131798138169e-06,
+      "loss": 0.8998,
+      "step": 5110
+    },
+    {
+      "epoch": 0.15630724142141897,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.463069573738364e-06,
+      "loss": 0.8696,
+      "step": 5120
+    },
+    {
+      "epoch": 0.15661252900232017,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.46000734933856e-06,
+      "loss": 0.7999,
+      "step": 5130
+    },
+    {
+      "epoch": 0.1569178165832214,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.456945124938756e-06,
+      "loss": 0.8899,
+      "step": 5140
+    },
+    {
+      "epoch": 0.1572231041641226,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.453882900538953e-06,
+      "loss": 0.7956,
+      "step": 5150
+    },
+    {
+      "epoch": 0.1575283917450238,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.45082067613915e-06,
+      "loss": 0.7898,
+      "step": 5160
+    },
+    {
+      "epoch": 0.157833679325925,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.447758451739345e-06,
+      "loss": 0.8035,
+      "step": 5170
+    },
+    {
+      "epoch": 0.15813896690682622,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.44469622733954e-06,
+      "loss": 0.8107,
+      "step": 5180
+    },
+    {
+      "epoch": 0.15844425448772745,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.441634002939737e-06,
+      "loss": 0.8778,
+      "step": 5190
+    },
+    {
+      "epoch": 0.15874954206862865,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.438571778539932e-06,
+      "loss": 0.8188,
+      "step": 5200
+    },
+    {
+      "epoch": 0.15905482964952986,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.435509554140129e-06,
+      "loss": 0.8019,
+      "step": 5210
+    },
+    {
+      "epoch": 0.15936011723043106,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.432447329740324e-06,
+      "loss": 0.8478,
+      "step": 5220
+    },
+    {
+      "epoch": 0.15966540481133226,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.429385105340519e-06,
+      "loss": 0.8322,
+      "step": 5230
+    },
+    {
+      "epoch": 0.1599706923922335,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.426322880940716e-06,
+      "loss": 0.838,
+      "step": 5240
+    },
+    {
+      "epoch": 0.1602759799731347,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.423260656540913e-06,
+      "loss": 0.8104,
+      "step": 5250
+    },
+    {
+      "epoch": 0.1605812675540359,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.420198432141108e-06,
+      "loss": 0.8242,
+      "step": 5260
+    },
+    {
+      "epoch": 0.1608865551349371,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.417136207741305e-06,
+      "loss": 0.7569,
+      "step": 5270
+    },
+    {
+      "epoch": 0.1611918427158383,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.4140739833415e-06,
+      "loss": 0.8559,
+      "step": 5280
+    },
+    {
+      "epoch": 0.16149713029673954,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.411011758941697e-06,
+      "loss": 0.8196,
+      "step": 5290
+    },
+    {
+      "epoch": 0.16180241787764074,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.407949534541892e-06,
+      "loss": 0.8803,
+      "step": 5300
+    },
+    {
+      "epoch": 0.16210770545854195,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.404887310142089e-06,
+      "loss": 0.8442,
+      "step": 5310
+    },
+    {
+      "epoch": 0.16241299303944315,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.401825085742284e-06,
+      "loss": 0.8702,
+      "step": 5320
+    },
+    {
+      "epoch": 0.16271828062034435,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.39876286134248e-06,
+      "loss": 0.8403,
+      "step": 5330
+    },
+    {
+      "epoch": 0.16302356820124558,
+      "grad_norm": 3.5,
+      "learning_rate": 8.395700636942676e-06,
+      "loss": 0.8352,
+      "step": 5340
+    },
+    {
+      "epoch": 0.1633288557821468,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.392638412542871e-06,
+      "loss": 0.8529,
+      "step": 5350
+    },
+    {
+      "epoch": 0.163634143363048,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.389576188143068e-06,
+      "loss": 0.8976,
+      "step": 5360
+    },
+    {
+      "epoch": 0.1639394309439492,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.386513963743263e-06,
+      "loss": 0.8582,
+      "step": 5370
+    },
+    {
+      "epoch": 0.1642447185248504,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.38345173934346e-06,
+      "loss": 0.8355,
+      "step": 5380
+    },
+    {
+      "epoch": 0.16455000610575163,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.380389514943655e-06,
+      "loss": 0.8416,
+      "step": 5390
+    },
+    {
+      "epoch": 0.16485529368665283,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.377327290543852e-06,
+      "loss": 0.8848,
+      "step": 5400
+    },
+    {
+      "epoch": 0.16516058126755404,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.374265066144047e-06,
+      "loss": 0.8344,
+      "step": 5410
+    },
+    {
+      "epoch": 0.16546586884845524,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.371202841744244e-06,
+      "loss": 0.857,
+      "step": 5420
+    },
+    {
+      "epoch": 0.16577115642935644,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.36814061734444e-06,
+      "loss": 0.8208,
+      "step": 5430
+    },
+    {
+      "epoch": 0.16607644401025767,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.365078392944636e-06,
+      "loss": 0.8116,
+      "step": 5440
+    },
+    {
+      "epoch": 0.16638173159115888,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.362016168544833e-06,
+      "loss": 0.8937,
+      "step": 5450
+    },
+    {
+      "epoch": 0.16668701917206008,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.358953944145028e-06,
+      "loss": 0.8168,
+      "step": 5460
+    },
+    {
+      "epoch": 0.16699230675296128,
+      "grad_norm": 3.25,
+      "learning_rate": 8.355891719745223e-06,
+      "loss": 0.8256,
+      "step": 5470
+    },
+    {
+      "epoch": 0.1672975943338625,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.35282949534542e-06,
+      "loss": 0.8078,
+      "step": 5480
+    },
+    {
+      "epoch": 0.16760288191476372,
+      "grad_norm": 3.484375,
+      "learning_rate": 8.349767270945615e-06,
+      "loss": 0.8632,
+      "step": 5490
+    },
+    {
+      "epoch": 0.16790816949566492,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.346705046545812e-06,
+      "loss": 0.8476,
+      "step": 5500
+    },
+    {
+      "epoch": 0.16821345707656613,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.343642822146007e-06,
+      "loss": 0.8308,
+      "step": 5510
+    },
+    {
+      "epoch": 0.16851874465746733,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.340580597746204e-06,
+      "loss": 0.8398,
+      "step": 5520
+    },
+    {
+      "epoch": 0.16882403223836853,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.337518373346399e-06,
+      "loss": 0.7811,
+      "step": 5530
+    },
+    {
+      "epoch": 0.16912931981926976,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.334456148946596e-06,
+      "loss": 0.8219,
+      "step": 5540
+    },
+    {
+      "epoch": 0.16943460740017097,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.331393924546791e-06,
+      "loss": 0.839,
+      "step": 5550
+    },
+    {
+      "epoch": 0.16973989498107217,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.328331700146988e-06,
+      "loss": 0.9018,
+      "step": 5560
+    },
+    {
+      "epoch": 0.17004518256197337,
+      "grad_norm": 3.375,
+      "learning_rate": 8.325269475747183e-06,
+      "loss": 0.8486,
+      "step": 5570
+    },
+    {
+      "epoch": 0.17035047014287458,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.32220725134738e-06,
+      "loss": 0.8133,
+      "step": 5580
+    },
+    {
+      "epoch": 0.1706557577237758,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.319145026947575e-06,
+      "loss": 0.8222,
+      "step": 5590
+    },
+    {
+      "epoch": 0.170961045304677,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.316082802547772e-06,
+      "loss": 0.8412,
+      "step": 5600
+    },
+    {
+      "epoch": 0.17126633288557822,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.313020578147967e-06,
+      "loss": 0.8716,
+      "step": 5610
+    },
+    {
+      "epoch": 0.17157162046647942,
+      "grad_norm": 3.625,
+      "learning_rate": 8.309958353748164e-06,
+      "loss": 0.8482,
+      "step": 5620
+    },
+    {
+      "epoch": 0.17187690804738062,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.306896129348359e-06,
+      "loss": 0.8837,
+      "step": 5630
+    },
+    {
+      "epoch": 0.17218219562828185,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.303833904948554e-06,
+      "loss": 0.8674,
+      "step": 5640
+    },
+    {
+      "epoch": 0.17248748320918306,
+      "grad_norm": 2.75,
+      "learning_rate": 8.300771680548751e-06,
+      "loss": 0.8305,
+      "step": 5650
+    },
+    {
+      "epoch": 0.17279277079008426,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.297709456148946e-06,
+      "loss": 0.8038,
+      "step": 5660
+    },
+    {
+      "epoch": 0.17309805837098546,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.294647231749143e-06,
+      "loss": 0.9063,
+      "step": 5670
+    },
+    {
+      "epoch": 0.17340334595188667,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.291585007349338e-06,
+      "loss": 0.8551,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1737086335327879,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.288522782949535e-06,
+      "loss": 0.8584,
+      "step": 5690
+    },
+    {
+      "epoch": 0.1740139211136891,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.285460558549732e-06,
+      "loss": 0.8583,
+      "step": 5700
+    },
+    {
+      "epoch": 0.1743192086945903,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.282398334149927e-06,
+      "loss": 0.8009,
+      "step": 5710
+    },
+    {
+      "epoch": 0.1746244962754915,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.279336109750124e-06,
+      "loss": 0.8494,
+      "step": 5720
+    },
+    {
+      "epoch": 0.1749297838563927,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.27627388535032e-06,
+      "loss": 0.7956,
+      "step": 5730
+    },
+    {
+      "epoch": 0.17523507143729394,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.273211660950516e-06,
+      "loss": 0.819,
+      "step": 5740
+    },
+    {
+      "epoch": 0.17554035901819515,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.270149436550711e-06,
+      "loss": 0.8153,
+      "step": 5750
+    },
+    {
+      "epoch": 0.17584564659909635,
+      "grad_norm": 3.125,
+      "learning_rate": 8.267087212150906e-06,
+      "loss": 0.8401,
+      "step": 5760
+    },
+    {
+      "epoch": 0.17615093417999755,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.264024987751103e-06,
+      "loss": 0.8243,
+      "step": 5770
+    },
+    {
+      "epoch": 0.17645622176089876,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.260962763351298e-06,
+      "loss": 0.8227,
+      "step": 5780
+    },
+    {
+      "epoch": 0.1767615093418,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.257900538951495e-06,
+      "loss": 0.7952,
+      "step": 5790
+    },
+    {
+      "epoch": 0.1770667969227012,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.25483831455169e-06,
+      "loss": 0.8757,
+      "step": 5800
+    },
+    {
+      "epoch": 0.1773720845036024,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.251776090151887e-06,
+      "loss": 0.8438,
+      "step": 5810
+    },
+    {
+      "epoch": 0.1776773720845036,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.248713865752082e-06,
+      "loss": 0.8666,
+      "step": 5820
+    },
+    {
+      "epoch": 0.1779826596654048,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.24565164135228e-06,
+      "loss": 0.8138,
+      "step": 5830
+    },
+    {
+      "epoch": 0.17828794724630603,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.242589416952474e-06,
+      "loss": 0.8169,
+      "step": 5840
+    },
+    {
+      "epoch": 0.17859323482720724,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.239527192552671e-06,
+      "loss": 0.8517,
+      "step": 5850
+    },
+    {
+      "epoch": 0.17889852240810844,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.236464968152868e-06,
+      "loss": 0.7971,
+      "step": 5860
+    },
+    {
+      "epoch": 0.17920380998900964,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.233402743753063e-06,
+      "loss": 0.8146,
+      "step": 5870
+    },
+    {
+      "epoch": 0.17950909756991085,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.23034051935326e-06,
+      "loss": 0.8552,
+      "step": 5880
+    },
+    {
+      "epoch": 0.17981438515081208,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.227278294953455e-06,
+      "loss": 0.7949,
+      "step": 5890
+    },
+    {
+      "epoch": 0.18011967273171328,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.22421607055365e-06,
+      "loss": 0.7831,
+      "step": 5900
+    },
+    {
+      "epoch": 0.18042496031261449,
+      "grad_norm": 3.75,
+      "learning_rate": 8.221153846153847e-06,
+      "loss": 0.7743,
+      "step": 5910
+    },
+    {
+      "epoch": 0.1807302478935157,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.218091621754042e-06,
+      "loss": 0.8134,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1810355354744169,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.215029397354238e-06,
+      "loss": 0.8612,
+      "step": 5930
+    },
+    {
+      "epoch": 0.18134082305531812,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.211967172954434e-06,
+      "loss": 0.8879,
+      "step": 5940
+    },
+    {
+      "epoch": 0.18164611063621933,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.208904948554631e-06,
+      "loss": 0.8286,
+      "step": 5950
+    },
+    {
+      "epoch": 0.18195139821712053,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.205842724154826e-06,
+      "loss": 0.8511,
+      "step": 5960
+    },
+    {
+      "epoch": 0.18225668579802173,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.202780499755023e-06,
+      "loss": 0.8701,
+      "step": 5970
+    },
+    {
+      "epoch": 0.18256197337892294,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.199718275355218e-06,
+      "loss": 0.8062,
+      "step": 5980
+    },
+    {
+      "epoch": 0.18286726095982414,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.196656050955415e-06,
+      "loss": 0.7979,
+      "step": 5990
+    },
+    {
+      "epoch": 0.18317254854072537,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.19359382655561e-06,
+      "loss": 0.9005,
+      "step": 6000
+    },
+    {
+      "epoch": 0.18347783612162658,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.190531602155807e-06,
+      "loss": 0.8626,
+      "step": 6010
+    },
+    {
+      "epoch": 0.18378312370252778,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.187469377756002e-06,
+      "loss": 0.8634,
+      "step": 6020
+    },
+    {
+      "epoch": 0.18408841128342898,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.1844071533562e-06,
+      "loss": 0.8072,
+      "step": 6030
+    },
+    {
+      "epoch": 0.18439369886433019,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.181344928956394e-06,
+      "loss": 0.8139,
+      "step": 6040
+    },
+    {
+      "epoch": 0.18469898644523142,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.17828270455659e-06,
+      "loss": 0.8409,
+      "step": 6050
+    },
+    {
+      "epoch": 0.18500427402613262,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.175220480156786e-06,
+      "loss": 0.8614,
+      "step": 6060
+    },
+    {
+      "epoch": 0.18530956160703382,
+      "grad_norm": 3.375,
+      "learning_rate": 8.172158255756982e-06,
+      "loss": 0.8034,
+      "step": 6070
+    },
+    {
+      "epoch": 0.18561484918793503,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.169096031357179e-06,
+      "loss": 0.8116,
+      "step": 6080
+    },
+    {
+      "epoch": 0.18592013676883623,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.166033806957374e-06,
+      "loss": 0.8411,
+      "step": 6090
+    },
+    {
+      "epoch": 0.18622542434973746,
+      "grad_norm": 3.25,
+      "learning_rate": 8.16297158255757e-06,
+      "loss": 0.8413,
+      "step": 6100
+    },
+    {
+      "epoch": 0.18653071193063867,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.159909358157766e-06,
+      "loss": 0.8783,
+      "step": 6110
+    },
+    {
+      "epoch": 0.18683599951153987,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.156847133757963e-06,
+      "loss": 0.8543,
+      "step": 6120
+    },
+    {
+      "epoch": 0.18714128709244107,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.15378490935816e-06,
+      "loss": 0.8139,
+      "step": 6130
+    },
+    {
+      "epoch": 0.18744657467334228,
+      "grad_norm": 3.125,
+      "learning_rate": 8.150722684958355e-06,
+      "loss": 0.878,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1877518622542435,
+      "grad_norm": 3.921875,
+      "learning_rate": 8.147660460558551e-06,
+      "loss": 0.8905,
+      "step": 6150
+    },
+    {
+      "epoch": 0.1880571498351447,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.144598236158747e-06,
+      "loss": 0.8198,
+      "step": 6160
+    },
+    {
+      "epoch": 0.1883624374160459,
+      "grad_norm": 3.0,
+      "learning_rate": 8.141536011758943e-06,
+      "loss": 0.8806,
+      "step": 6170
+    },
+    {
+      "epoch": 0.18866772499694712,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.138473787359139e-06,
+      "loss": 0.832,
+      "step": 6180
+    },
+    {
+      "epoch": 0.18897301257784832,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.135411562959334e-06,
+      "loss": 0.803,
+      "step": 6190
+    },
+    {
+      "epoch": 0.18927830015874955,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.13234933855953e-06,
+      "loss": 0.8545,
+      "step": 6200
+    },
+    {
+      "epoch": 0.18958358773965076,
+      "grad_norm": 3.90625,
+      "learning_rate": 8.129287114159726e-06,
+      "loss": 0.8532,
+      "step": 6210
+    },
+    {
+      "epoch": 0.18988887532055196,
+      "grad_norm": 3.125,
+      "learning_rate": 8.126224889759923e-06,
+      "loss": 0.8227,
+      "step": 6220
+    },
+    {
+      "epoch": 0.19019416290145316,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.123162665360118e-06,
+      "loss": 0.8073,
+      "step": 6230
+    },
+    {
+      "epoch": 0.19049945048235437,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.120100440960315e-06,
+      "loss": 0.8115,
+      "step": 6240
+    },
+    {
+      "epoch": 0.1908047380632556,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.11703821656051e-06,
+      "loss": 0.8334,
+      "step": 6250
+    },
+    {
+      "epoch": 0.1911100256441568,
+      "grad_norm": 3.5,
+      "learning_rate": 8.113975992160707e-06,
+      "loss": 0.8335,
+      "step": 6260
+    },
+    {
+      "epoch": 0.191415313225058,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.110913767760902e-06,
+      "loss": 0.8394,
+      "step": 6270
+    },
+    {
+      "epoch": 0.1917206008059592,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.107851543361099e-06,
+      "loss": 0.8761,
+      "step": 6280
+    },
+    {
+      "epoch": 0.1920258883868604,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.104789318961294e-06,
+      "loss": 0.7833,
+      "step": 6290
+    },
+    {
+      "epoch": 0.19233117596776164,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.10172709456149e-06,
+      "loss": 0.8318,
+      "step": 6300
+    },
+    {
+      "epoch": 0.19263646354866285,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.098664870161686e-06,
+      "loss": 0.8299,
+      "step": 6310
+    },
+    {
+      "epoch": 0.19294175112956405,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.095602645761883e-06,
+      "loss": 0.847,
+      "step": 6320
+    },
+    {
+      "epoch": 0.19324703871046525,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.092540421362078e-06,
+      "loss": 0.8291,
+      "step": 6330
+    },
+    {
+      "epoch": 0.19355232629136646,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.089478196962273e-06,
+      "loss": 0.8335,
+      "step": 6340
+    },
+    {
+      "epoch": 0.1938576138722677,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.08641597256247e-06,
+      "loss": 0.7829,
+      "step": 6350
+    },
+    {
+      "epoch": 0.1941629014531689,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.083353748162665e-06,
+      "loss": 0.8651,
+      "step": 6360
+    },
+    {
+      "epoch": 0.1944681890340701,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.080291523762862e-06,
+      "loss": 0.829,
+      "step": 6370
+    },
+    {
+      "epoch": 0.1947734766149713,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.077229299363057e-06,
+      "loss": 0.8417,
+      "step": 6380
+    },
+    {
+      "epoch": 0.1950787641958725,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.074167074963254e-06,
+      "loss": 0.7968,
+      "step": 6390
+    },
+    {
+      "epoch": 0.19538405177677373,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.07110485056345e-06,
+      "loss": 0.8532,
+      "step": 6400
+    },
+    {
+      "epoch": 0.19568933935767494,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.068042626163646e-06,
+      "loss": 0.8633,
+      "step": 6410
+    },
+    {
+      "epoch": 0.19599462693857614,
+      "grad_norm": 3.25,
+      "learning_rate": 8.064980401763843e-06,
+      "loss": 0.9114,
+      "step": 6420
+    },
+    {
+      "epoch": 0.19629991451947734,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.061918177364038e-06,
+      "loss": 0.8119,
+      "step": 6430
+    },
+    {
+      "epoch": 0.19660520210037855,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.058855952964235e-06,
+      "loss": 0.8713,
+      "step": 6440
+    },
+    {
+      "epoch": 0.19691048968127978,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.05579372856443e-06,
+      "loss": 0.896,
+      "step": 6450
+    },
+    {
+      "epoch": 0.19721577726218098,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.052731504164627e-06,
+      "loss": 0.8249,
+      "step": 6460
+    },
+    {
+      "epoch": 0.19752106484308218,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.049669279764822e-06,
+      "loss": 0.8518,
+      "step": 6470
+    },
+    {
+      "epoch": 0.1978263524239834,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.046607055365017e-06,
+      "loss": 0.8126,
+      "step": 6480
+    },
+    {
+      "epoch": 0.1981316400048846,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.043544830965214e-06,
+      "loss": 0.8619,
+      "step": 6490
+    },
+    {
+      "epoch": 0.19843692758578582,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.040482606565409e-06,
+      "loss": 0.8303,
+      "step": 6500
+    },
+    {
+      "epoch": 0.19874221516668703,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.037420382165606e-06,
+      "loss": 0.8775,
+      "step": 6510
+    },
+    {
+      "epoch": 0.19904750274758823,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.034358157765801e-06,
+      "loss": 0.8679,
+      "step": 6520
+    },
+    {
+      "epoch": 0.19935279032848943,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.031295933365998e-06,
+      "loss": 0.8671,
+      "step": 6530
+    },
+    {
+      "epoch": 0.19965807790939064,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.028233708966193e-06,
+      "loss": 0.8421,
+      "step": 6540
+    },
+    {
+      "epoch": 0.19996336549029187,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.02517148456639e-06,
+      "loss": 0.8671,
+      "step": 6550
+    },
+    {
+      "epoch": 0.20026865307119307,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.022109260166587e-06,
+      "loss": 0.8089,
+      "step": 6560
+    },
+    {
+      "epoch": 0.20057394065209427,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.019047035766782e-06,
+      "loss": 0.8101,
+      "step": 6570
+    },
+    {
+      "epoch": 0.20087922823299548,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.015984811366979e-06,
+      "loss": 0.8582,
+      "step": 6580
+    },
+    {
+      "epoch": 0.20118451581389668,
+      "grad_norm": 4.46875,
+      "learning_rate": 8.012922586967174e-06,
+      "loss": 0.7999,
+      "step": 6590
+    },
+    {
+      "epoch": 0.2014898033947979,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.009860362567369e-06,
+      "loss": 0.879,
+      "step": 6600
+    },
+    {
+      "epoch": 0.20179509097569912,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.006798138167566e-06,
+      "loss": 0.8153,
+      "step": 6610
+    },
+    {
+      "epoch": 0.20210037855660032,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.003735913767761e-06,
+      "loss": 0.8281,
+      "step": 6620
+    },
+    {
+      "epoch": 0.20240566613750152,
+      "grad_norm": 3.125,
+      "learning_rate": 8.000673689367956e-06,
+      "loss": 0.8353,
+      "step": 6630
+    },
+    {
+      "epoch": 0.20271095371840273,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.997611464968153e-06,
+      "loss": 0.8689,
+      "step": 6640
+    },
+    {
+      "epoch": 0.20301624129930396,
+      "grad_norm": 3.5,
+      "learning_rate": 7.99454924056835e-06,
+      "loss": 0.8147,
+      "step": 6650
+    },
+    {
+      "epoch": 0.20332152888020516,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.991487016168545e-06,
+      "loss": 0.898,
+      "step": 6660
+    },
+    {
+      "epoch": 0.20362681646110636,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.988424791768742e-06,
+      "loss": 0.8585,
+      "step": 6670
+    },
+    {
+      "epoch": 0.20393210404200757,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.985362567368937e-06,
+      "loss": 0.8079,
+      "step": 6680
+    },
+    {
+      "epoch": 0.20423739162290877,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.982300342969134e-06,
+      "loss": 0.8093,
+      "step": 6690
+    },
+    {
+      "epoch": 0.20454267920381,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.979238118569329e-06,
+      "loss": 0.8259,
+      "step": 6700
+    },
+    {
+      "epoch": 0.2048479667847112,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.976175894169526e-06,
+      "loss": 0.8208,
+      "step": 6710
+    },
+    {
+      "epoch": 0.2051532543656124,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.973113669769721e-06,
+      "loss": 0.8106,
+      "step": 6720
+    },
+    {
+      "epoch": 0.2054585419465136,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.970051445369918e-06,
+      "loss": 0.8249,
+      "step": 6730
+    },
+    {
+      "epoch": 0.20576382952741482,
+      "grad_norm": 3.0,
+      "learning_rate": 7.966989220970113e-06,
+      "loss": 0.824,
+      "step": 6740
+    },
+    {
+      "epoch": 0.20606911710831605,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.96392699657031e-06,
+      "loss": 0.7986,
+      "step": 6750
+    },
+    {
+      "epoch": 0.20637440468921725,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.960864772170505e-06,
+      "loss": 0.8268,
+      "step": 6760
+    },
+    {
+      "epoch": 0.20667969227011845,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.9578025477707e-06,
+      "loss": 0.8432,
+      "step": 6770
+    },
+    {
+      "epoch": 0.20698497985101966,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.954740323370897e-06,
+      "loss": 0.8111,
+      "step": 6780
+    },
+    {
+      "epoch": 0.20729026743192086,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.951678098971092e-06,
+      "loss": 0.8093,
+      "step": 6790
+    },
+    {
+      "epoch": 0.2075955550128221,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.94861587457129e-06,
+      "loss": 0.8566,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2079008425937233,
+      "grad_norm": 20.75,
+      "learning_rate": 7.945553650171484e-06,
+      "loss": 0.8145,
+      "step": 6810
+    },
+    {
+      "epoch": 0.2082061301746245,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.942491425771681e-06,
+      "loss": 0.829,
+      "step": 6820
+    },
+    {
+      "epoch": 0.2085114177555257,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.939429201371878e-06,
+      "loss": 0.8954,
+      "step": 6830
+    },
+    {
+      "epoch": 0.2088167053364269,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.936366976972073e-06,
+      "loss": 0.8227,
+      "step": 6840
+    },
+    {
+      "epoch": 0.20912199291732814,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.93330475257227e-06,
+      "loss": 0.8717,
+      "step": 6850
+    },
+    {
+      "epoch": 0.20942728049822934,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.930242528172465e-06,
+      "loss": 0.8311,
+      "step": 6860
+    },
+    {
+      "epoch": 0.20973256807913054,
+      "grad_norm": 3.375,
+      "learning_rate": 7.927180303772662e-06,
+      "loss": 0.848,
+      "step": 6870
+    },
+    {
+      "epoch": 0.21003785566003175,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.924118079372857e-06,
+      "loss": 0.8474,
+      "step": 6880
+    },
+    {
+      "epoch": 0.21034314324093295,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.921055854973052e-06,
+      "loss": 0.8126,
+      "step": 6890
+    },
+    {
+      "epoch": 0.21064843082183418,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.91799363057325e-06,
+      "loss": 0.8462,
+      "step": 6900
+    },
+    {
+      "epoch": 0.21095371840273539,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.914931406173444e-06,
+      "loss": 0.8501,
+      "step": 6910
+    },
+    {
+      "epoch": 0.2112590059836366,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.911869181773641e-06,
+      "loss": 0.805,
+      "step": 6920
+    },
+    {
+      "epoch": 0.2115642935645378,
+      "grad_norm": 3.375,
+      "learning_rate": 7.908806957373836e-06,
+      "loss": 0.8156,
+      "step": 6930
+    },
+    {
+      "epoch": 0.211869581145439,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.905744732974033e-06,
+      "loss": 0.8136,
+      "step": 6940
+    },
+    {
+      "epoch": 0.2121748687263402,
+      "grad_norm": 6.28125,
+      "learning_rate": 7.902682508574228e-06,
+      "loss": 0.8602,
+      "step": 6950
+    },
+    {
+      "epoch": 0.21248015630724143,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.899620284174425e-06,
+      "loss": 0.8422,
+      "step": 6960
+    },
+    {
+      "epoch": 0.21278544388814263,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.89655805977462e-06,
+      "loss": 0.8191,
+      "step": 6970
+    },
+    {
+      "epoch": 0.21309073146904384,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.893495835374817e-06,
+      "loss": 0.8145,
+      "step": 6980
+    },
+    {
+      "epoch": 0.21339601904994504,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.890433610975012e-06,
+      "loss": 0.7883,
+      "step": 6990
+    },
+    {
+      "epoch": 0.21370130663084624,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.88737138657521e-06,
+      "loss": 0.7845,
+      "step": 7000
+    },
+    {
+      "epoch": 0.21400659421174748,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.884309162175406e-06,
+      "loss": 0.8529,
+      "step": 7010
+    },
+    {
+      "epoch": 0.21431188179264868,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.881246937775601e-06,
+      "loss": 0.8243,
+      "step": 7020
+    },
+    {
+      "epoch": 0.21461716937354988,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.878184713375796e-06,
+      "loss": 0.8707,
+      "step": 7030
+    },
+    {
+      "epoch": 0.21492245695445109,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.875122488975993e-06,
+      "loss": 0.809,
+      "step": 7040
+    },
+    {
+      "epoch": 0.2152277445353523,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.872060264576188e-06,
+      "loss": 0.8324,
+      "step": 7050
+    },
+    {
+      "epoch": 0.21553303211625352,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.868998040176384e-06,
+      "loss": 0.8632,
+      "step": 7060
+    },
+    {
+      "epoch": 0.21583831969715472,
+      "grad_norm": 3.25,
+      "learning_rate": 7.86593581577658e-06,
+      "loss": 0.8427,
+      "step": 7070
+    },
+    {
+      "epoch": 0.21614360727805593,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.862873591376776e-06,
+      "loss": 0.8281,
+      "step": 7080
+    },
+    {
+      "epoch": 0.21644889485895713,
+      "grad_norm": 3.375,
+      "learning_rate": 7.859811366976973e-06,
+      "loss": 0.8278,
+      "step": 7090
+    },
+    {
+      "epoch": 0.21675418243985833,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.85674914257717e-06,
+      "loss": 0.8555,
+      "step": 7100
+    },
+    {
+      "epoch": 0.21705947002075957,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.853686918177365e-06,
+      "loss": 0.8345,
+      "step": 7110
+    },
+    {
+      "epoch": 0.21736475760166077,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.850624693777561e-06,
+      "loss": 0.8644,
+      "step": 7120
+    },
+    {
+      "epoch": 0.21767004518256197,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.847562469377757e-06,
+      "loss": 0.8593,
+      "step": 7130
+    },
+    {
+      "epoch": 0.21797533276346318,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.844500244977953e-06,
+      "loss": 0.8943,
+      "step": 7140
+    },
+    {
+      "epoch": 0.21828062034436438,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.841438020578149e-06,
+      "loss": 0.8249,
+      "step": 7150
+    },
+    {
+      "epoch": 0.2185859079252656,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.838375796178345e-06,
+      "loss": 0.8249,
+      "step": 7160
+    },
+    {
+      "epoch": 0.2188911955061668,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.83531357177854e-06,
+      "loss": 0.8136,
+      "step": 7170
+    },
+    {
+      "epoch": 0.21919648308706802,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.832251347378736e-06,
+      "loss": 0.8405,
+      "step": 7180
+    },
+    {
+      "epoch": 0.21950177066796922,
+      "grad_norm": 3.5,
+      "learning_rate": 7.829189122978933e-06,
+      "loss": 0.8192,
+      "step": 7190
+    },
+    {
+      "epoch": 0.21980705824887042,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.826126898579128e-06,
+      "loss": 0.9015,
+      "step": 7200
+    },
+    {
+      "epoch": 0.22011234582977166,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.823064674179325e-06,
+      "loss": 0.872,
+      "step": 7210
+    },
+    {
+      "epoch": 0.22041763341067286,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.82000244977952e-06,
+      "loss": 0.8135,
+      "step": 7220
+    },
+    {
+      "epoch": 0.22072292099157406,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.816940225379717e-06,
+      "loss": 0.861,
+      "step": 7230
+    },
+    {
+      "epoch": 0.22102820857247527,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.813878000979912e-06,
+      "loss": 0.8554,
+      "step": 7240
+    },
+    {
+      "epoch": 0.22133349615337647,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.810815776580109e-06,
+      "loss": 0.8493,
+      "step": 7250
+    },
+    {
+      "epoch": 0.2216387837342777,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.807753552180305e-06,
+      "loss": 0.8021,
+      "step": 7260
+    },
+    {
+      "epoch": 0.2219440713151789,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.8046913277805e-06,
+      "loss": 0.7537,
+      "step": 7270
+    },
+    {
+      "epoch": 0.2222493588960801,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.801629103380697e-06,
+      "loss": 0.8442,
+      "step": 7280
+    },
+    {
+      "epoch": 0.2225546464769813,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.798566878980893e-06,
+      "loss": 0.8003,
+      "step": 7290
+    },
+    {
+      "epoch": 0.2228599340578825,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.795504654581088e-06,
+      "loss": 0.7773,
+      "step": 7300
+    },
+    {
+      "epoch": 0.22316522163878375,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.792442430181285e-06,
+      "loss": 0.8394,
+      "step": 7310
+    },
+    {
+      "epoch": 0.22347050921968495,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.78938020578148e-06,
+      "loss": 0.8326,
+      "step": 7320
+    },
+    {
+      "epoch": 0.22377579680058615,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.786317981381677e-06,
+      "loss": 0.7966,
+      "step": 7330
+    },
+    {
+      "epoch": 0.22408108438148736,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.783255756981872e-06,
+      "loss": 0.8087,
+      "step": 7340
+    },
+    {
+      "epoch": 0.22438637196238856,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.780193532582069e-06,
+      "loss": 0.8599,
+      "step": 7350
+    },
+    {
+      "epoch": 0.2246916595432898,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.777131308182264e-06,
+      "loss": 0.8453,
+      "step": 7360
+    },
+    {
+      "epoch": 0.224996947124191,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.77406908378246e-06,
+      "loss": 0.8496,
+      "step": 7370
+    },
+    {
+      "epoch": 0.2253022347050922,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.771006859382656e-06,
+      "loss": 0.8037,
+      "step": 7380
+    },
+    {
+      "epoch": 0.2256075222859934,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.767944634982853e-06,
+      "loss": 0.8555,
+      "step": 7390
+    },
+    {
+      "epoch": 0.2259128098668946,
+      "grad_norm": 3.0,
+      "learning_rate": 7.764882410583048e-06,
+      "loss": 0.8354,
+      "step": 7400
+    },
+    {
+      "epoch": 0.22621809744779584,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.761820186183245e-06,
+      "loss": 0.808,
+      "step": 7410
+    },
+    {
+      "epoch": 0.22652338502869704,
+      "grad_norm": 3.125,
+      "learning_rate": 7.75875796178344e-06,
+      "loss": 0.8486,
+      "step": 7420
+    },
+    {
+      "epoch": 0.22682867260959824,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.755695737383637e-06,
+      "loss": 0.8231,
+      "step": 7430
+    },
+    {
+      "epoch": 0.22713396019049945,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.752633512983832e-06,
+      "loss": 0.9148,
+      "step": 7440
+    },
+    {
+      "epoch": 0.22743924777140065,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.749571288584029e-06,
+      "loss": 0.8151,
+      "step": 7450
+    },
+    {
+      "epoch": 0.22774453535230188,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.746509064184224e-06,
+      "loss": 0.8185,
+      "step": 7460
+    },
+    {
+      "epoch": 0.22804982293320308,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.743446839784419e-06,
+      "loss": 0.818,
+      "step": 7470
+    },
+    {
+      "epoch": 0.2283551105141043,
+      "grad_norm": 3.0,
+      "learning_rate": 7.740384615384616e-06,
+      "loss": 0.8283,
+      "step": 7480
+    },
+    {
+      "epoch": 0.2286603980950055,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.737322390984811e-06,
+      "loss": 0.8029,
+      "step": 7490
+    },
+    {
+      "epoch": 0.2289656856759067,
+      "grad_norm": 3.5,
+      "learning_rate": 7.734260166585008e-06,
+      "loss": 0.8554,
+      "step": 7500
+    },
+    {
+      "epoch": 0.22927097325680793,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.731197942185203e-06,
+      "loss": 0.8377,
+      "step": 7510
+    },
+    {
+      "epoch": 0.22957626083770913,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.7281357177854e-06,
+      "loss": 0.8213,
+      "step": 7520
+    },
+    {
+      "epoch": 0.22988154841861033,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.725073493385597e-06,
+      "loss": 0.7862,
+      "step": 7530
+    },
+    {
+      "epoch": 0.23018683599951154,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.722011268985792e-06,
+      "loss": 0.8505,
+      "step": 7540
+    },
+    {
+      "epoch": 0.23049212358041274,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.718949044585989e-06,
+      "loss": 0.7678,
+      "step": 7550
+    },
+    {
+      "epoch": 0.23079741116131397,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.715886820186184e-06,
+      "loss": 0.7816,
+      "step": 7560
+    },
+    {
+      "epoch": 0.23110269874221517,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.71282459578638e-06,
+      "loss": 0.8208,
+      "step": 7570
+    },
+    {
+      "epoch": 0.23140798632311638,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.709762371386576e-06,
+      "loss": 0.7823,
+      "step": 7580
+    },
+    {
+      "epoch": 0.23171327390401758,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.706700146986771e-06,
+      "loss": 0.7939,
+      "step": 7590
+    },
+    {
+      "epoch": 0.23201856148491878,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.703637922586968e-06,
+      "loss": 0.88,
+      "step": 7600
+    },
+    {
+      "epoch": 0.23232384906582002,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.700575698187163e-06,
+      "loss": 0.8258,
+      "step": 7610
+    },
+    {
+      "epoch": 0.23262913664672122,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.69751347378736e-06,
+      "loss": 0.7725,
+      "step": 7620
+    },
+    {
+      "epoch": 0.23293442422762242,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.694451249387555e-06,
+      "loss": 0.7965,
+      "step": 7630
+    },
+    {
+      "epoch": 0.23323971180852363,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.691389024987752e-06,
+      "loss": 0.8473,
+      "step": 7640
+    },
+    {
+      "epoch": 0.23354499938942483,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.688326800587947e-06,
+      "loss": 0.8343,
+      "step": 7650
+    },
+    {
+      "epoch": 0.23385028697032606,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.685264576188144e-06,
+      "loss": 0.8577,
+      "step": 7660
+    },
+    {
+      "epoch": 0.23415557455122726,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.682202351788339e-06,
+      "loss": 0.8582,
+      "step": 7670
+    },
+    {
+      "epoch": 0.23446086213212847,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.679140127388536e-06,
+      "loss": 0.8245,
+      "step": 7680
+    },
+    {
+      "epoch": 0.23476614971302967,
+      "grad_norm": 3.25,
+      "learning_rate": 7.676077902988731e-06,
+      "loss": 0.8073,
+      "step": 7690
+    },
+    {
+      "epoch": 0.23507143729393087,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.673015678588928e-06,
+      "loss": 0.821,
+      "step": 7700
+    },
+    {
+      "epoch": 0.2353767248748321,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.669953454189125e-06,
+      "loss": 0.8017,
+      "step": 7710
+    },
+    {
+      "epoch": 0.2356820124557333,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.66689122978932e-06,
+      "loss": 0.8858,
+      "step": 7720
+    },
+    {
+      "epoch": 0.2359873000366345,
+      "grad_norm": 3.125,
+      "learning_rate": 7.663829005389515e-06,
+      "loss": 0.8543,
+      "step": 7730
+    },
+    {
+      "epoch": 0.23629258761753572,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.660766780989712e-06,
+      "loss": 0.8128,
+      "step": 7740
+    },
+    {
+      "epoch": 0.23659787519843692,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.657704556589907e-06,
+      "loss": 0.8777,
+      "step": 7750
+    },
+    {
+      "epoch": 0.23690316277933815,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.654642332190102e-06,
+      "loss": 0.8239,
+      "step": 7760
+    },
+    {
+      "epoch": 0.23720845036023935,
+      "grad_norm": 3.375,
+      "learning_rate": 7.6515801077903e-06,
+      "loss": 0.8183,
+      "step": 7770
+    },
+    {
+      "epoch": 0.23751373794114056,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.648517883390494e-06,
+      "loss": 0.8071,
+      "step": 7780
+    },
+    {
+      "epoch": 0.23781902552204176,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.645455658990691e-06,
+      "loss": 0.8491,
+      "step": 7790
+    },
+    {
+      "epoch": 0.23812431310294296,
+      "grad_norm": 3.25,
+      "learning_rate": 7.642393434590888e-06,
+      "loss": 0.7804,
+      "step": 7800
+    },
+    {
+      "epoch": 0.2384296006838442,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.639331210191083e-06,
+      "loss": 0.8171,
+      "step": 7810
+    },
+    {
+      "epoch": 0.2387348882647454,
+      "grad_norm": 3.125,
+      "learning_rate": 7.63626898579128e-06,
+      "loss": 0.852,
+      "step": 7820
+    },
+    {
+      "epoch": 0.2390401758456466,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.633206761391475e-06,
+      "loss": 0.8258,
+      "step": 7830
+    },
+    {
+      "epoch": 0.2393454634265478,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.630144536991672e-06,
+      "loss": 0.8349,
+      "step": 7840
+    },
+    {
+      "epoch": 0.239650751007449,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.627082312591867e-06,
+      "loss": 0.8103,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2399560385883502,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.624020088192063e-06,
+      "loss": 0.8746,
+      "step": 7860
+    },
+    {
+      "epoch": 0.24026132616925144,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.620957863792258e-06,
+      "loss": 0.8233,
+      "step": 7870
+    },
+    {
+      "epoch": 0.24056661375015265,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.617895639392455e-06,
+      "loss": 0.7978,
+      "step": 7880
+    },
+    {
+      "epoch": 0.24087190133105385,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.614833414992652e-06,
+      "loss": 0.8494,
+      "step": 7890
+    },
+    {
+      "epoch": 0.24117718891195505,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.611771190592847e-06,
+      "loss": 0.8057,
+      "step": 7900
+    },
+    {
+      "epoch": 0.24148247649285626,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.608708966193043e-06,
+      "loss": 0.813,
+      "step": 7910
+    },
+    {
+      "epoch": 0.2417877640737575,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.605646741793239e-06,
+      "loss": 0.8151,
+      "step": 7920
+    },
+    {
+      "epoch": 0.2420930516546587,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.602584517393435e-06,
+      "loss": 0.7815,
+      "step": 7930
+    },
+    {
+      "epoch": 0.2423983392355599,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.5995222929936304e-06,
+      "loss": 0.8345,
+      "step": 7940
+    },
+    {
+      "epoch": 0.2427036268164611,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.596460068593827e-06,
+      "loss": 0.8557,
+      "step": 7950
+    },
+    {
+      "epoch": 0.2430089143973623,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.593397844194023e-06,
+      "loss": 0.8126,
+      "step": 7960
+    },
+    {
+      "epoch": 0.24331420197826353,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.590335619794219e-06,
+      "loss": 0.8575,
+      "step": 7970
+    },
+    {
+      "epoch": 0.24361948955916474,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.587273395394415e-06,
+      "loss": 0.903,
+      "step": 7980
+    },
+    {
+      "epoch": 0.24392477714006594,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.584211170994611e-06,
+      "loss": 0.8225,
+      "step": 7990
+    },
+    {
+      "epoch": 0.24423006472096714,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.581148946594807e-06,
+      "loss": 0.8445,
+      "step": 8000
+    },
+    {
+      "epoch": 0.24453535230186835,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.5780867221950025e-06,
+      "loss": 0.8014,
+      "step": 8010
+    },
+    {
+      "epoch": 0.24484063988276958,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.575024497795199e-06,
+      "loss": 0.7856,
+      "step": 8020
+    },
+    {
+      "epoch": 0.24514592746367078,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.5719622733953945e-06,
+      "loss": 0.8498,
+      "step": 8030
+    },
+    {
+      "epoch": 0.24545121504457199,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.568900048995591e-06,
+      "loss": 0.8521,
+      "step": 8040
+    },
+    {
+      "epoch": 0.2457565026254732,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.565837824595787e-06,
+      "loss": 0.88,
+      "step": 8050
+    },
+    {
+      "epoch": 0.2460617902063744,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.5627756001959825e-06,
+      "loss": 0.8977,
+      "step": 8060
+    },
+    {
+      "epoch": 0.24636707778727562,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.559713375796179e-06,
+      "loss": 0.8306,
+      "step": 8070
+    },
+    {
+      "epoch": 0.24667236536817683,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.5566511513963745e-06,
+      "loss": 0.7879,
+      "step": 8080
+    },
+    {
+      "epoch": 0.24697765294907803,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.553588926996571e-06,
+      "loss": 0.8388,
+      "step": 8090
+    },
+    {
+      "epoch": 0.24728294052997923,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.5505267025967665e-06,
+      "loss": 0.7985,
+      "step": 8100
+    },
+    {
+      "epoch": 0.24758822811088044,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.547464478196963e-06,
+      "loss": 0.8013,
+      "step": 8110
+    },
+    {
+      "epoch": 0.24789351569178167,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.5444022537971585e-06,
+      "loss": 0.7343,
+      "step": 8120
+    },
+    {
+      "epoch": 0.24819880327268287,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.5413400293973545e-06,
+      "loss": 0.8703,
+      "step": 8130
+    },
+    {
+      "epoch": 0.24850409085358408,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.538277804997551e-06,
+      "loss": 0.8286,
+      "step": 8140
+    },
+    {
+      "epoch": 0.24880937843448528,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.5352155805977466e-06,
+      "loss": 0.8325,
+      "step": 8150
+    },
+    {
+      "epoch": 0.24911466601538648,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.532153356197943e-06,
+      "loss": 0.8602,
+      "step": 8160
+    },
+    {
+      "epoch": 0.2494199535962877,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.5290911317981386e-06,
+      "loss": 0.8225,
+      "step": 8170
+    },
+    {
+      "epoch": 0.24972524117718892,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.526028907398335e-06,
+      "loss": 0.7998,
+      "step": 8180
+    },
+    {
+      "epoch": 0.2500305287580901,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.5229666829985306e-06,
+      "loss": 0.8203,
+      "step": 8190
+    },
+    {
+      "epoch": 0.2503358163389913,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.519904458598727e-06,
+      "loss": 0.8486,
+      "step": 8200
+    },
+    {
+      "epoch": 0.2506411039198925,
+      "grad_norm": 3.5,
+      "learning_rate": 7.516842234198923e-06,
+      "loss": 0.7761,
+      "step": 8210
+    },
+    {
+      "epoch": 0.25094639150079373,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.513780009799119e-06,
+      "loss": 0.807,
+      "step": 8220
+    },
+    {
+      "epoch": 0.25125167908169493,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.5107177853993154e-06,
+      "loss": 0.7882,
+      "step": 8230
+    },
+    {
+      "epoch": 0.2515569666625962,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.507655560999511e-06,
+      "loss": 0.7888,
+      "step": 8240
+    },
+    {
+      "epoch": 0.2518622542434974,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.504593336599707e-06,
+      "loss": 0.7875,
+      "step": 8250
+    },
+    {
+      "epoch": 0.2521675418243986,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.501531112199903e-06,
+      "loss": 0.8001,
+      "step": 8260
+    },
+    {
+      "epoch": 0.2524728294052998,
+      "grad_norm": 3.375,
+      "learning_rate": 7.498468887800099e-06,
+      "loss": 0.8156,
+      "step": 8270
+    },
+    {
+      "epoch": 0.252778116986201,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.495406663400295e-06,
+      "loss": 0.7733,
+      "step": 8280
+    },
+    {
+      "epoch": 0.2530834045671022,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.492344439000491e-06,
+      "loss": 0.7973,
+      "step": 8290
+    },
+    {
+      "epoch": 0.2533886921480034,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.489282214600686e-06,
+      "loss": 0.8297,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2536939797289046,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.486219990200883e-06,
+      "loss": 0.8687,
+      "step": 8310
+    },
+    {
+      "epoch": 0.2539992673098058,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.483157765801079e-06,
+      "loss": 0.7949,
+      "step": 8320
+    },
+    {
+      "epoch": 0.254304554890707,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.480095541401275e-06,
+      "loss": 0.856,
+      "step": 8330
+    },
+    {
+      "epoch": 0.2546098424716082,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.477033317001471e-06,
+      "loss": 0.8052,
+      "step": 8340
+    },
+    {
+      "epoch": 0.2549151300525095,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.473971092601666e-06,
+      "loss": 0.8701,
+      "step": 8350
+    },
+    {
+      "epoch": 0.2552204176334107,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.470908868201863e-06,
+      "loss": 0.8078,
+      "step": 8360
+    },
+    {
+      "epoch": 0.2555257052143119,
+      "grad_norm": 3.125,
+      "learning_rate": 7.467846643802058e-06,
+      "loss": 0.808,
+      "step": 8370
+    },
+    {
+      "epoch": 0.2558309927952131,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.464784419402255e-06,
+      "loss": 0.8324,
+      "step": 8380
+    },
+    {
+      "epoch": 0.2561362803761143,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.46172219500245e-06,
+      "loss": 0.819,
+      "step": 8390
+    },
+    {
+      "epoch": 0.2564415679570155,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.458659970602647e-06,
+      "loss": 0.9034,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2567468555379167,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.455597746202843e-06,
+      "loss": 0.8152,
+      "step": 8410
+    },
+    {
+      "epoch": 0.2570521431188179,
+      "grad_norm": 3.375,
+      "learning_rate": 7.452535521803038e-06,
+      "loss": 0.7887,
+      "step": 8420
+    },
+    {
+      "epoch": 0.2573574306997191,
+      "grad_norm": 3.125,
+      "learning_rate": 7.449473297403235e-06,
+      "loss": 0.8372,
+      "step": 8430
+    },
+    {
+      "epoch": 0.2576627182806203,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.44641107300343e-06,
+      "loss": 0.8607,
+      "step": 8440
+    },
+    {
+      "epoch": 0.2579680058615216,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.443348848603627e-06,
+      "loss": 0.8194,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2582732934424228,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.440286624203822e-06,
+      "loss": 0.8395,
+      "step": 8460
+    },
+    {
+      "epoch": 0.258578581023324,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.437224399804019e-06,
+      "loss": 0.7842,
+      "step": 8470
+    },
+    {
+      "epoch": 0.2588838686042252,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.434162175404214e-06,
+      "loss": 0.8193,
+      "step": 8480
+    },
+    {
+      "epoch": 0.2591891561851264,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.43109995100441e-06,
+      "loss": 0.8374,
+      "step": 8490
+    },
+    {
+      "epoch": 0.2594944437660276,
+      "grad_norm": 3.375,
+      "learning_rate": 7.428037726604607e-06,
+      "loss": 0.7953,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2597997313469288,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.424975502204802e-06,
+      "loss": 0.8287,
+      "step": 8510
+    },
+    {
+      "epoch": 0.26010501892783,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.421913277804999e-06,
+      "loss": 0.8455,
+      "step": 8520
+    },
+    {
+      "epoch": 0.2604103065087312,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.418851053405194e-06,
+      "loss": 0.8436,
+      "step": 8530
+    },
+    {
+      "epoch": 0.2607155940896324,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.41578882900539e-06,
+      "loss": 0.87,
+      "step": 8540
+    },
+    {
+      "epoch": 0.26102088167053367,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.412726604605586e-06,
+      "loss": 0.8353,
+      "step": 8550
+    },
+    {
+      "epoch": 0.26132616925143487,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.409664380205782e-06,
+      "loss": 0.8019,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2616314568323361,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.406602155805978e-06,
+      "loss": 0.7785,
+      "step": 8570
+    },
+    {
+      "epoch": 0.2619367444132373,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.403539931406174e-06,
+      "loss": 0.8503,
+      "step": 8580
+    },
+    {
+      "epoch": 0.2622420319941385,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.400477707006371e-06,
+      "loss": 0.8273,
+      "step": 8590
+    },
+    {
+      "epoch": 0.2625473195750397,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.397415482606566e-06,
+      "loss": 0.7977,
+      "step": 8600
+    },
+    {
+      "epoch": 0.2628526071559409,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.394353258206762e-06,
+      "loss": 0.7769,
+      "step": 8610
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.391291033806958e-06,
+      "loss": 0.8848,
+      "step": 8620
+    },
+    {
+      "epoch": 0.2634631823177433,
+      "grad_norm": 2.75,
+      "learning_rate": 7.388228809407154e-06,
+      "loss": 0.834,
+      "step": 8630
+    },
+    {
+      "epoch": 0.2637684698986445,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.385166585007349e-06,
+      "loss": 0.8197,
+      "step": 8640
+    },
+    {
+      "epoch": 0.26407375747954576,
+      "grad_norm": 3.5,
+      "learning_rate": 7.382104360607546e-06,
+      "loss": 0.8061,
+      "step": 8650
+    },
+    {
+      "epoch": 0.26437904506044696,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.379042136207743e-06,
+      "loss": 0.8402,
+      "step": 8660
+    },
+    {
+      "epoch": 0.26468433264134816,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.375979911807938e-06,
+      "loss": 0.7699,
+      "step": 8670
+    },
+    {
+      "epoch": 0.26498962022224937,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.372917687408134e-06,
+      "loss": 0.8466,
+      "step": 8680
+    },
+    {
+      "epoch": 0.26529490780315057,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.36985546300833e-06,
+      "loss": 0.7881,
+      "step": 8690
+    },
+    {
+      "epoch": 0.2656001953840518,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.366793238608526e-06,
+      "loss": 0.8248,
+      "step": 8700
+    },
+    {
+      "epoch": 0.265905482964953,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.363731014208721e-06,
+      "loss": 0.8352,
+      "step": 8710
+    },
+    {
+      "epoch": 0.2662107705458542,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.360668789808918e-06,
+      "loss": 0.8329,
+      "step": 8720
+    },
+    {
+      "epoch": 0.2665160581267554,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.357606565409113e-06,
+      "loss": 0.8749,
+      "step": 8730
+    },
+    {
+      "epoch": 0.2668213457076566,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.35454434100931e-06,
+      "loss": 0.8186,
+      "step": 8740
+    },
+    {
+      "epoch": 0.26712663328855785,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.351482116609506e-06,
+      "loss": 0.8189,
+      "step": 8750
+    },
+    {
+      "epoch": 0.26743192086945905,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.348419892209702e-06,
+      "loss": 0.8298,
+      "step": 8760
+    },
+    {
+      "epoch": 0.26773720845036025,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.345357667809898e-06,
+      "loss": 0.8247,
+      "step": 8770
+    },
+    {
+      "epoch": 0.26804249603126146,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.342295443410093e-06,
+      "loss": 0.8145,
+      "step": 8780
+    },
+    {
+      "epoch": 0.26834778361216266,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.33923321901029e-06,
+      "loss": 0.8404,
+      "step": 8790
+    },
+    {
+      "epoch": 0.26865307119306386,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.336170994610485e-06,
+      "loss": 0.8542,
+      "step": 8800
+    },
+    {
+      "epoch": 0.26895835877396507,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.333108770210682e-06,
+      "loss": 0.7829,
+      "step": 8810
+    },
+    {
+      "epoch": 0.26926364635486627,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.330046545810877e-06,
+      "loss": 0.77,
+      "step": 8820
+    },
+    {
+      "epoch": 0.2695689339357675,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.326984321411073e-06,
+      "loss": 0.7954,
+      "step": 8830
+    },
+    {
+      "epoch": 0.2698742215166687,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.32392209701127e-06,
+      "loss": 0.8518,
+      "step": 8840
+    },
+    {
+      "epoch": 0.27017950909756994,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.320859872611465e-06,
+      "loss": 0.8396,
+      "step": 8850
+    },
+    {
+      "epoch": 0.27048479667847114,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.317797648211662e-06,
+      "loss": 0.8262,
+      "step": 8860
+    },
+    {
+      "epoch": 0.27079008425937234,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.314735423811857e-06,
+      "loss": 0.8287,
+      "step": 8870
+    },
+    {
+      "epoch": 0.27109537184027355,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.311673199412054e-06,
+      "loss": 0.8185,
+      "step": 8880
+    },
+    {
+      "epoch": 0.27140065942117475,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.308610975012249e-06,
+      "loss": 0.8544,
+      "step": 8890
+    },
+    {
+      "epoch": 0.27170594700207595,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.305548750612445e-06,
+      "loss": 0.8233,
+      "step": 8900
+    },
+    {
+      "epoch": 0.27201123458297716,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.302486526212641e-06,
+      "loss": 0.8117,
+      "step": 8910
+    },
+    {
+      "epoch": 0.27231652216387836,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.299424301812837e-06,
+      "loss": 0.8026,
+      "step": 8920
+    },
+    {
+      "epoch": 0.27262180974477956,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.296362077413034e-06,
+      "loss": 0.8152,
+      "step": 8930
+    },
+    {
+      "epoch": 0.27292709732568077,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.293299853013229e-06,
+      "loss": 0.8012,
+      "step": 8940
+    },
+    {
+      "epoch": 0.273232384906582,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.290237628613426e-06,
+      "loss": 0.8278,
+      "step": 8950
+    },
+    {
+      "epoch": 0.27353767248748323,
+      "grad_norm": 3.0,
+      "learning_rate": 7.287175404213621e-06,
+      "loss": 0.8479,
+      "step": 8960
+    },
+    {
+      "epoch": 0.27384296006838443,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.284113179813817e-06,
+      "loss": 0.8401,
+      "step": 8970
+    },
+    {
+      "epoch": 0.27414824764928564,
+      "grad_norm": 3.375,
+      "learning_rate": 7.281050955414013e-06,
+      "loss": 0.7839,
+      "step": 8980
+    },
+    {
+      "epoch": 0.27445353523018684,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.277988731014209e-06,
+      "loss": 0.8253,
+      "step": 8990
+    },
+    {
+      "epoch": 0.27475882281108804,
+      "grad_norm": 3.5,
+      "learning_rate": 7.2749265066144045e-06,
+      "loss": 0.8001,
+      "step": 9000
+    },
+    {
+      "epoch": 0.27506411039198925,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.271864282214601e-06,
+      "loss": 0.8094,
+      "step": 9010
+    },
+    {
+      "epoch": 0.27536939797289045,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.268802057814797e-06,
+      "loss": 0.821,
+      "step": 9020
+    },
+    {
+      "epoch": 0.27567468555379165,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.265739833414993e-06,
+      "loss": 0.8385,
+      "step": 9030
+    },
+    {
+      "epoch": 0.27597997313469286,
+      "grad_norm": 3.625,
+      "learning_rate": 7.262677609015189e-06,
+      "loss": 0.8193,
+      "step": 9040
+    },
+    {
+      "epoch": 0.2762852607155941,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.259615384615385e-06,
+      "loss": 0.8128,
+      "step": 9050
+    },
+    {
+      "epoch": 0.2765905482964953,
+      "grad_norm": 4.09375,
+      "learning_rate": 7.256553160215581e-06,
+      "loss": 0.826,
+      "step": 9060
+    },
+    {
+      "epoch": 0.2768958358773965,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.2534909358157765e-06,
+      "loss": 0.7977,
+      "step": 9070
+    },
+    {
+      "epoch": 0.2772011234582977,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.250428711415973e-06,
+      "loss": 0.7855,
+      "step": 9080
+    },
+    {
+      "epoch": 0.27750641103919893,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.2473664870161685e-06,
+      "loss": 0.8493,
+      "step": 9090
+    },
+    {
+      "epoch": 0.27781169862010013,
+      "grad_norm": 3.375,
+      "learning_rate": 7.244304262616365e-06,
+      "loss": 0.8626,
+      "step": 9100
+    },
+    {
+      "epoch": 0.27811698620100134,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.241242038216561e-06,
+      "loss": 0.7983,
+      "step": 9110
+    },
+    {
+      "epoch": 0.27842227378190254,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.2381798138167565e-06,
+      "loss": 0.8308,
+      "step": 9120
+    },
+    {
+      "epoch": 0.27872756136280374,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.235117589416953e-06,
+      "loss": 0.8057,
+      "step": 9130
+    },
+    {
+      "epoch": 0.27903284894370495,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.2320553650171485e-06,
+      "loss": 0.8174,
+      "step": 9140
+    },
+    {
+      "epoch": 0.2793381365246062,
+      "grad_norm": 3.5,
+      "learning_rate": 7.228993140617345e-06,
+      "loss": 0.7601,
+      "step": 9150
+    },
+    {
+      "epoch": 0.2796434241055074,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.2259309162175406e-06,
+      "loss": 0.7838,
+      "step": 9160
+    },
+    {
+      "epoch": 0.2799487116864086,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.222868691817737e-06,
+      "loss": 0.8083,
+      "step": 9170
+    },
+    {
+      "epoch": 0.2802539992673098,
+      "grad_norm": 3.125,
+      "learning_rate": 7.2198064674179326e-06,
+      "loss": 0.8295,
+      "step": 9180
+    },
+    {
+      "epoch": 0.280559286848211,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.216744243018129e-06,
+      "loss": 0.834,
+      "step": 9190
+    },
+    {
+      "epoch": 0.2808645744291122,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.2136820186183254e-06,
+      "loss": 0.8238,
+      "step": 9200
+    },
+    {
+      "epoch": 0.2811698620100134,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.210619794218521e-06,
+      "loss": 0.8362,
+      "step": 9210
+    },
+    {
+      "epoch": 0.28147514959091463,
+      "grad_norm": 3.25,
+      "learning_rate": 7.2075575698187174e-06,
+      "loss": 0.8316,
+      "step": 9220
+    },
+    {
+      "epoch": 0.28178043717181583,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.204495345418913e-06,
+      "loss": 0.8276,
+      "step": 9230
+    },
+    {
+      "epoch": 0.28208572475271704,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.2014331210191095e-06,
+      "loss": 0.7921,
+      "step": 9240
+    },
+    {
+      "epoch": 0.28239101233361824,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.198370896619305e-06,
+      "loss": 0.8462,
+      "step": 9250
+    },
+    {
+      "epoch": 0.2826962999145195,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.195308672219501e-06,
+      "loss": 0.8073,
+      "step": 9260
+    },
+    {
+      "epoch": 0.2830015874954207,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.192246447819697e-06,
+      "loss": 0.8301,
+      "step": 9270
+    },
+    {
+      "epoch": 0.2833068750763219,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.189184223419893e-06,
+      "loss": 0.8141,
+      "step": 9280
+    },
+    {
+      "epoch": 0.2836121626572231,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.1861219990200895e-06,
+      "loss": 0.8217,
+      "step": 9290
+    },
+    {
+      "epoch": 0.2839174502381243,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.183059774620285e-06,
+      "loss": 0.8455,
+      "step": 9300
+    },
+    {
+      "epoch": 0.2842227378190255,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.179997550220481e-06,
+      "loss": 0.8221,
+      "step": 9310
+    },
+    {
+      "epoch": 0.2845280253999267,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.176935325820677e-06,
+      "loss": 0.803,
+      "step": 9320
+    },
+    {
+      "epoch": 0.2848333129808279,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.173873101420873e-06,
+      "loss": 0.7936,
+      "step": 9330
+    },
+    {
+      "epoch": 0.2851386005617291,
+      "grad_norm": 3.25,
+      "learning_rate": 7.170810877021069e-06,
+      "loss": 0.8125,
+      "step": 9340
+    },
+    {
+      "epoch": 0.28544388814263033,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.167748652621265e-06,
+      "loss": 0.8198,
+      "step": 9350
+    },
+    {
+      "epoch": 0.2857491757235316,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.1646864282214615e-06,
+      "loss": 0.7978,
+      "step": 9360
+    },
+    {
+      "epoch": 0.2860544633044328,
+      "grad_norm": 3.375,
+      "learning_rate": 7.161624203821657e-06,
+      "loss": 0.844,
+      "step": 9370
+    },
+    {
+      "epoch": 0.286359750885334,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.158561979421853e-06,
+      "loss": 0.881,
+      "step": 9380
+    },
+    {
+      "epoch": 0.2866650384662352,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.155499755022049e-06,
+      "loss": 0.8149,
+      "step": 9390
+    },
+    {
+      "epoch": 0.2869703260471364,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.152437530622245e-06,
+      "loss": 0.8364,
+      "step": 9400
+    },
+    {
+      "epoch": 0.2872756136280376,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.14937530622244e-06,
+      "loss": 0.8384,
+      "step": 9410
+    },
+    {
+      "epoch": 0.2875809012089388,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.146313081822637e-06,
+      "loss": 0.835,
+      "step": 9420
+    },
+    {
+      "epoch": 0.28788618878984,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.143250857422832e-06,
+      "loss": 0.776,
+      "step": 9430
+    },
+    {
+      "epoch": 0.2881914763707412,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.140188633023029e-06,
+      "loss": 0.8483,
+      "step": 9440
+    },
+    {
+      "epoch": 0.2884967639516424,
+      "grad_norm": 3.0,
+      "learning_rate": 7.137126408623225e-06,
+      "loss": 0.7865,
+      "step": 9450
+    },
+    {
+      "epoch": 0.2888020515325437,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.134064184223421e-06,
+      "loss": 0.8316,
+      "step": 9460
+    },
+    {
+      "epoch": 0.2891073391134449,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.131001959823617e-06,
+      "loss": 0.7941,
+      "step": 9470
+    },
+    {
+      "epoch": 0.2894126266943461,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.127939735423812e-06,
+      "loss": 0.8714,
+      "step": 9480
+    },
+    {
+      "epoch": 0.2897179142752473,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.124877511024009e-06,
+      "loss": 0.8557,
+      "step": 9490
+    },
+    {
+      "epoch": 0.2900232018561485,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.121815286624204e-06,
+      "loss": 0.8201,
+      "step": 9500
+    },
+    {
+      "epoch": 0.2903284894370497,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.118753062224401e-06,
+      "loss": 0.797,
+      "step": 9510
+    },
+    {
+      "epoch": 0.2906337770179509,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.115690837824596e-06,
+      "loss": 0.8161,
+      "step": 9520
+    },
+    {
+      "epoch": 0.2909390645988521,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.112628613424793e-06,
+      "loss": 0.8276,
+      "step": 9530
+    },
+    {
+      "epoch": 0.2912443521797533,
+      "grad_norm": 3.96875,
+      "learning_rate": 7.109566389024989e-06,
+      "loss": 0.827,
+      "step": 9540
+    },
+    {
+      "epoch": 0.2915496397606545,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.106504164625184e-06,
+      "loss": 0.8269,
+      "step": 9550
+    },
+    {
+      "epoch": 0.29185492734155577,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.103441940225381e-06,
+      "loss": 0.8233,
+      "step": 9560
+    },
+    {
+      "epoch": 0.292160214922457,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.100379715825576e-06,
+      "loss": 0.8426,
+      "step": 9570
+    },
+    {
+      "epoch": 0.2924655025033582,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.097317491425773e-06,
+      "loss": 0.7728,
+      "step": 9580
+    },
+    {
+      "epoch": 0.2927707900842594,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.094255267025968e-06,
+      "loss": 0.8332,
+      "step": 9590
+    },
+    {
+      "epoch": 0.2930760776651606,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.091193042626164e-06,
+      "loss": 0.8314,
+      "step": 9600
+    },
+    {
+      "epoch": 0.2933813652460618,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.08813081822636e-06,
+      "loss": 0.7699,
+      "step": 9610
+    },
+    {
+      "epoch": 0.293686652826963,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.085068593826556e-06,
+      "loss": 0.8422,
+      "step": 9620
+    },
+    {
+      "epoch": 0.2939919404078642,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.082006369426753e-06,
+      "loss": 0.8182,
+      "step": 9630
+    },
+    {
+      "epoch": 0.2942972279887654,
+      "grad_norm": 3.5,
+      "learning_rate": 7.078944145026948e-06,
+      "loss": 0.7995,
+      "step": 9640
+    },
+    {
+      "epoch": 0.2946025155696666,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.075881920627145e-06,
+      "loss": 0.7858,
+      "step": 9650
+    },
+    {
+      "epoch": 0.29490780315056786,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.07281969622734e-06,
+      "loss": 0.8004,
+      "step": 9660
+    },
+    {
+      "epoch": 0.29521309073146906,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.069757471827536e-06,
+      "loss": 0.7442,
+      "step": 9670
+    },
+    {
+      "epoch": 0.29551837831237027,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.066695247427732e-06,
+      "loss": 0.8123,
+      "step": 9680
+    },
+    {
+      "epoch": 0.29582366589327147,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.063633023027928e-06,
+      "loss": 0.8055,
+      "step": 9690
+    },
+    {
+      "epoch": 0.2961289534741727,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.060570798628123e-06,
+      "loss": 0.8117,
+      "step": 9700
+    },
+    {
+      "epoch": 0.2964342410550739,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.05750857422832e-06,
+      "loss": 0.8392,
+      "step": 9710
+    },
+    {
+      "epoch": 0.2967395286359751,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.054446349828517e-06,
+      "loss": 0.8117,
+      "step": 9720
+    },
+    {
+      "epoch": 0.2970448162168763,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.051384125428712e-06,
+      "loss": 0.8079,
+      "step": 9730
+    },
+    {
+      "epoch": 0.2973501037977775,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.048321901028908e-06,
+      "loss": 0.872,
+      "step": 9740
+    },
+    {
+      "epoch": 0.2976553913786787,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.045259676629104e-06,
+      "loss": 0.8732,
+      "step": 9750
+    },
+    {
+      "epoch": 0.29796067895957995,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.0421974522293e-06,
+      "loss": 0.8212,
+      "step": 9760
+    },
+    {
+      "epoch": 0.29826596654048115,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.039135227829495e-06,
+      "loss": 0.8139,
+      "step": 9770
+    },
+    {
+      "epoch": 0.29857125412138236,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.036073003429692e-06,
+      "loss": 0.8529,
+      "step": 9780
+    },
+    {
+      "epoch": 0.29887654170228356,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.033010779029887e-06,
+      "loss": 0.8768,
+      "step": 9790
+    },
+    {
+      "epoch": 0.29918182928318476,
+      "grad_norm": 3.125,
+      "learning_rate": 7.029948554630084e-06,
+      "loss": 0.8417,
+      "step": 9800
+    },
+    {
+      "epoch": 0.29948711686408597,
+      "grad_norm": 3.25,
+      "learning_rate": 7.02688633023028e-06,
+      "loss": 0.8575,
+      "step": 9810
+    },
+    {
+      "epoch": 0.29979240444498717,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.023824105830476e-06,
+      "loss": 0.8556,
+      "step": 9820
+    },
+    {
+      "epoch": 0.3000976920258884,
+      "grad_norm": 3.375,
+      "learning_rate": 7.020761881430672e-06,
+      "loss": 0.7929,
+      "step": 9830
+    },
+    {
+      "epoch": 0.3004029796067896,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.017699657030867e-06,
+      "loss": 0.7796,
+      "step": 9840
+    },
+    {
+      "epoch": 0.3007082671876908,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.014637432631064e-06,
+      "loss": 0.8651,
+      "step": 9850
+    },
+    {
+      "epoch": 0.30101355476859204,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.011575208231259e-06,
+      "loss": 0.7877,
+      "step": 9860
+    },
+    {
+      "epoch": 0.30131884234949324,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.008512983831456e-06,
+      "loss": 0.8319,
+      "step": 9870
+    },
+    {
+      "epoch": 0.30162412993039445,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.005450759431651e-06,
+      "loss": 0.8205,
+      "step": 9880
+    },
+    {
+      "epoch": 0.30192941751129565,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.002388535031847e-06,
+      "loss": 0.7775,
+      "step": 9890
+    },
+    {
+      "epoch": 0.30223470509219685,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.999326310632044e-06,
+      "loss": 0.823,
+      "step": 9900
+    },
+    {
+      "epoch": 0.30253999267309806,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.996264086232239e-06,
+      "loss": 0.7695,
+      "step": 9910
+    },
+    {
+      "epoch": 0.30284528025399926,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.993201861832436e-06,
+      "loss": 0.7708,
+      "step": 9920
+    },
+    {
+      "epoch": 0.30315056783490046,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.990139637432631e-06,
+      "loss": 0.8462,
+      "step": 9930
+    },
+    {
+      "epoch": 0.30345585541580167,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.987077413032828e-06,
+      "loss": 0.8299,
+      "step": 9940
+    },
+    {
+      "epoch": 0.30376114299670287,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.984015188633023e-06,
+      "loss": 0.8874,
+      "step": 9950
+    },
+    {
+      "epoch": 0.30406643057760413,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.980952964233219e-06,
+      "loss": 0.8071,
+      "step": 9960
+    },
+    {
+      "epoch": 0.30437171815850533,
+      "grad_norm": 3.375,
+      "learning_rate": 6.977890739833415e-06,
+      "loss": 0.8043,
+      "step": 9970
+    },
+    {
+      "epoch": 0.30467700573940654,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.974828515433611e-06,
+      "loss": 0.7988,
+      "step": 9980
+    },
+    {
+      "epoch": 0.30498229332030774,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.971766291033808e-06,
+      "loss": 0.7691,
+      "step": 9990
+    },
+    {
+      "epoch": 0.30528758090120894,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.968704066634003e-06,
+      "loss": 0.7873,
+      "step": 10000
+    },
+    {
+      "epoch": 0.30559286848211015,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.9656418422342e-06,
+      "loss": 0.7654,
+      "step": 10010
+    },
+    {
+      "epoch": 0.30589815606301135,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.962579617834395e-06,
+      "loss": 0.8471,
+      "step": 10020
+    },
+    {
+      "epoch": 0.30620344364391255,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.959517393434591e-06,
+      "loss": 0.8768,
+      "step": 10030
+    },
+    {
+      "epoch": 0.30650873122481376,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.956455169034787e-06,
+      "loss": 0.7962,
+      "step": 10040
+    },
+    {
+      "epoch": 0.30681401880571496,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.953392944634983e-06,
+      "loss": 0.8399,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3071193063866162,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.9503307202351785e-06,
+      "loss": 0.8201,
+      "step": 10060
+    },
+    {
+      "epoch": 0.3074245939675174,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.947268495835375e-06,
+      "loss": 0.801,
+      "step": 10070
+    },
+    {
+      "epoch": 0.3077298815484186,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.944206271435572e-06,
+      "loss": 0.8102,
+      "step": 10080
+    },
+    {
+      "epoch": 0.30803516912931983,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.941144047035767e-06,
+      "loss": 0.8684,
+      "step": 10090
+    },
+    {
+      "epoch": 0.30834045671022103,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.938081822635963e-06,
+      "loss": 0.8668,
+      "step": 10100
+    },
+    {
+      "epoch": 0.30864574429112224,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.935019598236159e-06,
+      "loss": 0.7737,
+      "step": 10110
+    },
+    {
+      "epoch": 0.30895103187202344,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.931957373836355e-06,
+      "loss": 0.8084,
+      "step": 10120
+    },
+    {
+      "epoch": 0.30925631945292464,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.9288951494365505e-06,
+      "loss": 0.8116,
+      "step": 10130
+    },
+    {
+      "epoch": 0.30956160703382585,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.925832925036747e-06,
+      "loss": 0.8033,
+      "step": 10140
+    },
+    {
+      "epoch": 0.30986689461472705,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.922770700636943e-06,
+      "loss": 0.7994,
+      "step": 10150
+    },
+    {
+      "epoch": 0.31017218219562825,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.919708476237139e-06,
+      "loss": 0.751,
+      "step": 10160
+    },
+    {
+      "epoch": 0.3104774697765295,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.916646251837335e-06,
+      "loss": 0.8214,
+      "step": 10170
+    },
+    {
+      "epoch": 0.3107827573574307,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.9135840274375306e-06,
+      "loss": 0.8045,
+      "step": 10180
+    },
+    {
+      "epoch": 0.3110880449383319,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.9105218030377274e-06,
+      "loss": 0.8273,
+      "step": 10190
+    },
+    {
+      "epoch": 0.3113933325192331,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.907459578637923e-06,
+      "loss": 0.7937,
+      "step": 10200
+    },
+    {
+      "epoch": 0.3116986201001343,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.9043973542381194e-06,
+      "loss": 0.7904,
+      "step": 10210
+    },
+    {
+      "epoch": 0.31200390768103553,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.901335129838315e-06,
+      "loss": 0.8104,
+      "step": 10220
+    },
+    {
+      "epoch": 0.31230919526193673,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.8982729054385115e-06,
+      "loss": 0.8483,
+      "step": 10230
+    },
+    {
+      "epoch": 0.31261448284283794,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.8952106810387075e-06,
+      "loss": 0.8281,
+      "step": 10240
+    },
+    {
+      "epoch": 0.31291977042373914,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.892148456638903e-06,
+      "loss": 0.8548,
+      "step": 10250
+    },
+    {
+      "epoch": 0.31322505800464034,
+      "grad_norm": 3.25,
+      "learning_rate": 6.8890862322390995e-06,
+      "loss": 0.8238,
+      "step": 10260
+    },
+    {
+      "epoch": 0.3135303455855416,
+      "grad_norm": 3.5,
+      "learning_rate": 6.886024007839295e-06,
+      "loss": 0.8228,
+      "step": 10270
+    },
+    {
+      "epoch": 0.3138356331664428,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.8829617834394915e-06,
+      "loss": 0.7715,
+      "step": 10280
+    },
+    {
+      "epoch": 0.314140920747344,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.879899559039687e-06,
+      "loss": 0.8042,
+      "step": 10290
+    },
+    {
+      "epoch": 0.3144462083282452,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.8768373346398835e-06,
+      "loss": 0.8451,
+      "step": 10300
+    },
+    {
+      "epoch": 0.3147514959091464,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.873775110240079e-06,
+      "loss": 0.7302,
+      "step": 10310
+    },
+    {
+      "epoch": 0.3150567834900476,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.870712885840275e-06,
+      "loss": 0.8129,
+      "step": 10320
+    },
+    {
+      "epoch": 0.3153620710709488,
+      "grad_norm": 3.125,
+      "learning_rate": 6.8676506614404715e-06,
+      "loss": 0.837,
+      "step": 10330
+    },
+    {
+      "epoch": 0.31566735865185,
+      "grad_norm": 3.625,
+      "learning_rate": 6.864588437040667e-06,
+      "loss": 0.7757,
+      "step": 10340
+    },
+    {
+      "epoch": 0.31597264623275123,
+      "grad_norm": 3.5,
+      "learning_rate": 6.8615262126408635e-06,
+      "loss": 0.8112,
+      "step": 10350
+    },
+    {
+      "epoch": 0.31627793381365243,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.858463988241059e-06,
+      "loss": 0.8199,
+      "step": 10360
+    },
+    {
+      "epoch": 0.3165832213945537,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.8554017638412555e-06,
+      "loss": 0.8302,
+      "step": 10370
+    },
+    {
+      "epoch": 0.3168885089754549,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.852339539441451e-06,
+      "loss": 0.827,
+      "step": 10380
+    },
+    {
+      "epoch": 0.3171937965563561,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.849277315041647e-06,
+      "loss": 0.8235,
+      "step": 10390
+    },
+    {
+      "epoch": 0.3174990841372573,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.846215090641843e-06,
+      "loss": 0.8118,
+      "step": 10400
+    },
+    {
+      "epoch": 0.3178043717181585,
+      "grad_norm": 3.5,
+      "learning_rate": 6.843152866242039e-06,
+      "loss": 0.8121,
+      "step": 10410
+    },
+    {
+      "epoch": 0.3181096592990597,
+      "grad_norm": 3.375,
+      "learning_rate": 6.8400906418422356e-06,
+      "loss": 0.818,
+      "step": 10420
+    },
+    {
+      "epoch": 0.3184149468799609,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.837028417442431e-06,
+      "loss": 0.8403,
+      "step": 10430
+    },
+    {
+      "epoch": 0.3187202344608621,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.833966193042627e-06,
+      "loss": 0.7398,
+      "step": 10440
+    },
+    {
+      "epoch": 0.3190255220417633,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.830903968642823e-06,
+      "loss": 0.8447,
+      "step": 10450
+    },
+    {
+      "epoch": 0.3193308096226645,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.827841744243019e-06,
+      "loss": 0.8419,
+      "step": 10460
+    },
+    {
+      "epoch": 0.3196360972035658,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.824779519843214e-06,
+      "loss": 0.8374,
+      "step": 10470
+    },
+    {
+      "epoch": 0.319941384784467,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.821717295443411e-06,
+      "loss": 0.7864,
+      "step": 10480
+    },
+    {
+      "epoch": 0.3202466723653682,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.818655071043606e-06,
+      "loss": 0.8043,
+      "step": 10490
+    },
+    {
+      "epoch": 0.3205519599462694,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.815592846643803e-06,
+      "loss": 0.8421,
+      "step": 10500
+    },
+    {
+      "epoch": 0.3208572475271706,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.812530622243999e-06,
+      "loss": 0.7808,
+      "step": 10510
+    },
+    {
+      "epoch": 0.3211625351080718,
+      "grad_norm": 4.21875,
+      "learning_rate": 6.809468397844195e-06,
+      "loss": 0.7633,
+      "step": 10520
+    },
+    {
+      "epoch": 0.321467822688973,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.806406173444391e-06,
+      "loss": 0.8551,
+      "step": 10530
+    },
+    {
+      "epoch": 0.3217731102698742,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.803343949044586e-06,
+      "loss": 0.8482,
+      "step": 10540
+    },
+    {
+      "epoch": 0.3220783978507754,
+      "grad_norm": 3.375,
+      "learning_rate": 6.800281724644783e-06,
+      "loss": 0.8483,
+      "step": 10550
+    },
+    {
+      "epoch": 0.3223836854316766,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.797219500244978e-06,
+      "loss": 0.8436,
+      "step": 10560
+    },
+    {
+      "epoch": 0.3226889730125779,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.794157275845175e-06,
+      "loss": 0.8768,
+      "step": 10570
+    },
+    {
+      "epoch": 0.3229942605934791,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.79109505144537e-06,
+      "loss": 0.8515,
+      "step": 10580
+    },
+    {
+      "epoch": 0.3232995481743803,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.788032827045567e-06,
+      "loss": 0.8485,
+      "step": 10590
+    },
+    {
+      "epoch": 0.3236048357552815,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.784970602645763e-06,
+      "loss": 0.8415,
+      "step": 10600
+    },
+    {
+      "epoch": 0.3239101233361827,
+      "grad_norm": 3.125,
+      "learning_rate": 6.781908378245958e-06,
+      "loss": 0.7835,
+      "step": 10610
+    },
+    {
+      "epoch": 0.3242154109170839,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.778846153846155e-06,
+      "loss": 0.8344,
+      "step": 10620
+    },
+    {
+      "epoch": 0.3245206984979851,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.77578392944635e-06,
+      "loss": 0.8463,
+      "step": 10630
+    },
+    {
+      "epoch": 0.3248259860788863,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.772721705046547e-06,
+      "loss": 0.84,
+      "step": 10640
+    },
+    {
+      "epoch": 0.3251312736597875,
+      "grad_norm": 3.0,
+      "learning_rate": 6.769659480646742e-06,
+      "loss": 0.8009,
+      "step": 10650
+    },
+    {
+      "epoch": 0.3254365612406887,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.766597256246939e-06,
+      "loss": 0.8526,
+      "step": 10660
+    },
+    {
+      "epoch": 0.32574184882158996,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.763535031847134e-06,
+      "loss": 0.8439,
+      "step": 10670
+    },
+    {
+      "epoch": 0.32604713640249117,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.76047280744733e-06,
+      "loss": 0.795,
+      "step": 10680
+    },
+    {
+      "epoch": 0.32635242398339237,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.757410583047527e-06,
+      "loss": 0.8391,
+      "step": 10690
+    },
+    {
+      "epoch": 0.3266577115642936,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.754348358647722e-06,
+      "loss": 0.8028,
+      "step": 10700
+    },
+    {
+      "epoch": 0.3269629991451948,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.751286134247919e-06,
+      "loss": 0.7952,
+      "step": 10710
+    },
+    {
+      "epoch": 0.327268286726096,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.748223909848114e-06,
+      "loss": 0.7595,
+      "step": 10720
+    },
+    {
+      "epoch": 0.3275735743069972,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.74516168544831e-06,
+      "loss": 0.8157,
+      "step": 10730
+    },
+    {
+      "epoch": 0.3278788618878984,
+      "grad_norm": 3.25,
+      "learning_rate": 6.742099461048506e-06,
+      "loss": 0.7824,
+      "step": 10740
+    },
+    {
+      "epoch": 0.3281841494687996,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.739037236648702e-06,
+      "loss": 0.8315,
+      "step": 10750
+    },
+    {
+      "epoch": 0.3284894370497008,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.735975012248897e-06,
+      "loss": 0.806,
+      "step": 10760
+    },
+    {
+      "epoch": 0.32879472463060205,
+      "grad_norm": 3.25,
+      "learning_rate": 6.732912787849094e-06,
+      "loss": 0.885,
+      "step": 10770
+    },
+    {
+      "epoch": 0.32910001221150326,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.729850563449291e-06,
+      "loss": 0.8597,
+      "step": 10780
+    },
+    {
+      "epoch": 0.32940529979240446,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.726788339049486e-06,
+      "loss": 0.8516,
+      "step": 10790
+    },
+    {
+      "epoch": 0.32971058737330566,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.723726114649682e-06,
+      "loss": 0.7836,
+      "step": 10800
+    },
+    {
+      "epoch": 0.33001587495420687,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.720663890249878e-06,
+      "loss": 0.7797,
+      "step": 10810
+    },
+    {
+      "epoch": 0.33032116253510807,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.717601665850074e-06,
+      "loss": 0.8194,
+      "step": 10820
+    },
+    {
+      "epoch": 0.3306264501160093,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.714539441450269e-06,
+      "loss": 0.8037,
+      "step": 10830
+    },
+    {
+      "epoch": 0.3309317376969105,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.711477217050466e-06,
+      "loss": 0.824,
+      "step": 10840
+    },
+    {
+      "epoch": 0.3312370252778117,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.708414992650663e-06,
+      "loss": 0.7873,
+      "step": 10850
+    },
+    {
+      "epoch": 0.3315423128587129,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.705352768250858e-06,
+      "loss": 0.846,
+      "step": 10860
+    },
+    {
+      "epoch": 0.33184760043961414,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.702290543851054e-06,
+      "loss": 0.8145,
+      "step": 10870
+    },
+    {
+      "epoch": 0.33215288802051535,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.69922831945125e-06,
+      "loss": 0.7799,
+      "step": 10880
+    },
+    {
+      "epoch": 0.33245817560141655,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.696166095051446e-06,
+      "loss": 0.8167,
+      "step": 10890
+    },
+    {
+      "epoch": 0.33276346318231775,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.693103870651641e-06,
+      "loss": 0.836,
+      "step": 10900
+    },
+    {
+      "epoch": 0.33306875076321896,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.690041646251838e-06,
+      "loss": 0.875,
+      "step": 10910
+    },
+    {
+      "epoch": 0.33337403834412016,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.686979421852033e-06,
+      "loss": 0.8223,
+      "step": 10920
+    },
+    {
+      "epoch": 0.33367932592502136,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.68391719745223e-06,
+      "loss": 0.7646,
+      "step": 10930
+    },
+    {
+      "epoch": 0.33398461350592257,
+      "grad_norm": 3.375,
+      "learning_rate": 6.680854973052426e-06,
+      "loss": 0.8283,
+      "step": 10940
+    },
+    {
+      "epoch": 0.33428990108682377,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.677792748652622e-06,
+      "loss": 0.8035,
+      "step": 10950
+    },
+    {
+      "epoch": 0.334595188667725,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.674730524252818e-06,
+      "loss": 0.8068,
+      "step": 10960
+    },
+    {
+      "epoch": 0.33490047624862623,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.671668299853013e-06,
+      "loss": 0.7965,
+      "step": 10970
+    },
+    {
+      "epoch": 0.33520576382952744,
+      "grad_norm": 3.0,
+      "learning_rate": 6.66860607545321e-06,
+      "loss": 0.802,
+      "step": 10980
+    },
+    {
+      "epoch": 0.33551105141042864,
+      "grad_norm": 3.125,
+      "learning_rate": 6.665543851053405e-06,
+      "loss": 0.8544,
+      "step": 10990
+    },
+    {
+      "epoch": 0.33581633899132984,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.662481626653602e-06,
+      "loss": 0.8094,
+      "step": 11000
+    },
+    {
+      "epoch": 0.33612162657223105,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.659419402253797e-06,
+      "loss": 0.823,
+      "step": 11010
+    },
+    {
+      "epoch": 0.33642691415313225,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.656357177853993e-06,
+      "loss": 0.8237,
+      "step": 11020
+    },
+    {
+      "epoch": 0.33673220173403345,
+      "grad_norm": 4.125,
+      "learning_rate": 6.65329495345419e-06,
+      "loss": 0.8149,
+      "step": 11030
+    },
+    {
+      "epoch": 0.33703748931493466,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.650232729054385e-06,
+      "loss": 0.8194,
+      "step": 11040
+    },
+    {
+      "epoch": 0.33734277689583586,
+      "grad_norm": 3.125,
+      "learning_rate": 6.647170504654582e-06,
+      "loss": 0.8331,
+      "step": 11050
+    },
+    {
+      "epoch": 0.33764806447673706,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.644108280254777e-06,
+      "loss": 0.7932,
+      "step": 11060
+    },
+    {
+      "epoch": 0.33795335205763827,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.641046055854974e-06,
+      "loss": 0.7719,
+      "step": 11070
+    },
+    {
+      "epoch": 0.3382586396385395,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.637983831455169e-06,
+      "loss": 0.7991,
+      "step": 11080
+    },
+    {
+      "epoch": 0.33856392721944073,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.634921607055365e-06,
+      "loss": 0.8011,
+      "step": 11090
+    },
+    {
+      "epoch": 0.33886921480034193,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.631859382655561e-06,
+      "loss": 0.782,
+      "step": 11100
+    },
+    {
+      "epoch": 0.33917450238124314,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.628797158255757e-06,
+      "loss": 0.8044,
+      "step": 11110
+    },
+    {
+      "epoch": 0.33947978996214434,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.625734933855954e-06,
+      "loss": 0.7653,
+      "step": 11120
+    },
+    {
+      "epoch": 0.33978507754304554,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.622672709456149e-06,
+      "loss": 0.839,
+      "step": 11130
+    },
+    {
+      "epoch": 0.34009036512394675,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.619610485056346e-06,
+      "loss": 0.8268,
+      "step": 11140
+    },
+    {
+      "epoch": 0.34039565270484795,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.616548260656541e-06,
+      "loss": 0.8148,
+      "step": 11150
+    },
+    {
+      "epoch": 0.34070094028574915,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.613486036256737e-06,
+      "loss": 0.844,
+      "step": 11160
+    },
+    {
+      "epoch": 0.34100622786665036,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.610423811856933e-06,
+      "loss": 0.8124,
+      "step": 11170
+    },
+    {
+      "epoch": 0.3413115154475516,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.6073615874571294e-06,
+      "loss": 0.8574,
+      "step": 11180
+    },
+    {
+      "epoch": 0.3416168030284528,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.604299363057325e-06,
+      "loss": 0.8024,
+      "step": 11190
+    },
+    {
+      "epoch": 0.341922090609354,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.6012371386575214e-06,
+      "loss": 0.8912,
+      "step": 11200
+    },
+    {
+      "epoch": 0.3422273781902552,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.5981749142577174e-06,
+      "loss": 0.805,
+      "step": 11210
+    },
+    {
+      "epoch": 0.34253266577115643,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.5951126898579134e-06,
+      "loss": 0.8272,
+      "step": 11220
+    },
+    {
+      "epoch": 0.34283795335205763,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.5920504654581095e-06,
+      "loss": 0.8746,
+      "step": 11230
+    },
+    {
+      "epoch": 0.34314324093295884,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.588988241058305e-06,
+      "loss": 0.8235,
+      "step": 11240
+    },
+    {
+      "epoch": 0.34344852851386004,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.5859260166585015e-06,
+      "loss": 0.8292,
+      "step": 11250
+    },
+    {
+      "epoch": 0.34375381609476124,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.582863792258697e-06,
+      "loss": 0.8233,
+      "step": 11260
+    },
+    {
+      "epoch": 0.34405910367566245,
+      "grad_norm": 3.375,
+      "learning_rate": 6.5798015678588935e-06,
+      "loss": 0.7775,
+      "step": 11270
+    },
+    {
+      "epoch": 0.3443643912565637,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.576739343459089e-06,
+      "loss": 0.8496,
+      "step": 11280
+    },
+    {
+      "epoch": 0.3446696788374649,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.5736771190592855e-06,
+      "loss": 0.8563,
+      "step": 11290
+    },
+    {
+      "epoch": 0.3449749664183661,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.5706148946594815e-06,
+      "loss": 0.8082,
+      "step": 11300
+    },
+    {
+      "epoch": 0.3452802539992673,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.567552670259677e-06,
+      "loss": 0.8689,
+      "step": 11310
+    },
+    {
+      "epoch": 0.3455855415801685,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.5644904458598735e-06,
+      "loss": 0.7987,
+      "step": 11320
+    },
+    {
+      "epoch": 0.3458908291610697,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.561428221460069e-06,
+      "loss": 0.7942,
+      "step": 11330
+    },
+    {
+      "epoch": 0.3461961167419709,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.5583659970602655e-06,
+      "loss": 0.8322,
+      "step": 11340
+    },
+    {
+      "epoch": 0.34650140432287213,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.555303772660461e-06,
+      "loss": 0.7533,
+      "step": 11350
+    },
+    {
+      "epoch": 0.34680669190377333,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.5522415482606575e-06,
+      "loss": 0.8528,
+      "step": 11360
+    },
+    {
+      "epoch": 0.34711197948467454,
+      "grad_norm": 3.25,
+      "learning_rate": 6.549179323860853e-06,
+      "loss": 0.8454,
+      "step": 11370
+    },
+    {
+      "epoch": 0.3474172670655758,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.546117099461049e-06,
+      "loss": 0.8889,
+      "step": 11380
+    },
+    {
+      "epoch": 0.347722554646477,
+      "grad_norm": 3.625,
+      "learning_rate": 6.5430548750612455e-06,
+      "loss": 0.8028,
+      "step": 11390
+    },
+    {
+      "epoch": 0.3480278422273782,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.539992650661441e-06,
+      "loss": 0.7569,
+      "step": 11400
+    },
+    {
+      "epoch": 0.3483331298082794,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.5369304262616376e-06,
+      "loss": 0.8042,
+      "step": 11410
+    },
+    {
+      "epoch": 0.3486384173891806,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.533868201861833e-06,
+      "loss": 0.8036,
+      "step": 11420
+    },
+    {
+      "epoch": 0.3489437049700818,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.5308059774620296e-06,
+      "loss": 0.7891,
+      "step": 11430
+    },
+    {
+      "epoch": 0.349248992550983,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.527743753062225e-06,
+      "loss": 0.8598,
+      "step": 11440
+    },
+    {
+      "epoch": 0.3495542801318842,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.524681528662421e-06,
+      "loss": 0.7881,
+      "step": 11450
+    },
+    {
+      "epoch": 0.3498595677127854,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.521619304262617e-06,
+      "loss": 0.7785,
+      "step": 11460
+    },
+    {
+      "epoch": 0.3501648552936866,
+      "grad_norm": 3.5,
+      "learning_rate": 6.518557079862813e-06,
+      "loss": 0.8083,
+      "step": 11470
+    },
+    {
+      "epoch": 0.3504701428745879,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.51549485546301e-06,
+      "loss": 0.7869,
+      "step": 11480
+    },
+    {
+      "epoch": 0.3507754304554891,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.512432631063205e-06,
+      "loss": 0.7881,
+      "step": 11490
+    },
+    {
+      "epoch": 0.3510807180363903,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.509370406663401e-06,
+      "loss": 0.7968,
+      "step": 11500
+    },
+    {
+      "epoch": 0.3513860056172915,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.506308182263597e-06,
+      "loss": 0.7861,
+      "step": 11510
+    },
+    {
+      "epoch": 0.3516912931981927,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.503245957863793e-06,
+      "loss": 0.735,
+      "step": 11520
+    },
+    {
+      "epoch": 0.3519965807790939,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.500183733463988e-06,
+      "loss": 0.8307,
+      "step": 11530
+    },
+    {
+      "epoch": 0.3523018683599951,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.497121509064185e-06,
+      "loss": 0.7807,
+      "step": 11540
+    },
+    {
+      "epoch": 0.3526071559408963,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.494059284664382e-06,
+      "loss": 0.779,
+      "step": 11550
+    },
+    {
+      "epoch": 0.3529124435217975,
+      "grad_norm": 3.25,
+      "learning_rate": 6.490997060264577e-06,
+      "loss": 0.8683,
+      "step": 11560
+    },
+    {
+      "epoch": 0.3532177311026987,
+      "grad_norm": 3.625,
+      "learning_rate": 6.487934835864773e-06,
+      "loss": 0.8297,
+      "step": 11570
+    },
+    {
+      "epoch": 0.3535230186836,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.484872611464969e-06,
+      "loss": 0.8137,
+      "step": 11580
+    },
+    {
+      "epoch": 0.3538283062645012,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.481810387065165e-06,
+      "loss": 0.784,
+      "step": 11590
+    },
+    {
+      "epoch": 0.3541335938454024,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.47874816266536e-06,
+      "loss": 0.7988,
+      "step": 11600
+    },
+    {
+      "epoch": 0.3544388814263036,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.475685938265557e-06,
+      "loss": 0.7873,
+      "step": 11610
+    },
+    {
+      "epoch": 0.3547441690072048,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.472623713865752e-06,
+      "loss": 0.8799,
+      "step": 11620
+    },
+    {
+      "epoch": 0.355049456588106,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.469561489465949e-06,
+      "loss": 0.8689,
+      "step": 11630
+    },
+    {
+      "epoch": 0.3553547441690072,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.466499265066145e-06,
+      "loss": 0.8712,
+      "step": 11640
+    },
+    {
+      "epoch": 0.3556600317499084,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.463437040666341e-06,
+      "loss": 0.8212,
+      "step": 11650
+    },
+    {
+      "epoch": 0.3559653193308096,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.460374816266537e-06,
+      "loss": 0.82,
+      "step": 11660
+    },
+    {
+      "epoch": 0.3562706069117108,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.457312591866732e-06,
+      "loss": 0.8154,
+      "step": 11670
+    },
+    {
+      "epoch": 0.35657589449261207,
+      "grad_norm": 3.5,
+      "learning_rate": 6.454250367466929e-06,
+      "loss": 0.7952,
+      "step": 11680
+    },
+    {
+      "epoch": 0.35688118207351327,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.451188143067124e-06,
+      "loss": 0.7662,
+      "step": 11690
+    },
+    {
+      "epoch": 0.3571864696544145,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.448125918667321e-06,
+      "loss": 0.8298,
+      "step": 11700
+    },
+    {
+      "epoch": 0.3574917572353157,
+      "grad_norm": 3.984375,
+      "learning_rate": 6.445063694267516e-06,
+      "loss": 0.7964,
+      "step": 11710
+    },
+    {
+      "epoch": 0.3577970448162169,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.442001469867713e-06,
+      "loss": 0.8283,
+      "step": 11720
+    },
+    {
+      "epoch": 0.3581023323971181,
+      "grad_norm": 3.375,
+      "learning_rate": 6.438939245467909e-06,
+      "loss": 0.8392,
+      "step": 11730
+    },
+    {
+      "epoch": 0.3584076199780193,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.435877021068104e-06,
+      "loss": 0.8101,
+      "step": 11740
+    },
+    {
+      "epoch": 0.3587129075589205,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.432814796668301e-06,
+      "loss": 0.7965,
+      "step": 11750
+    },
+    {
+      "epoch": 0.3590181951398217,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.429752572268496e-06,
+      "loss": 0.9061,
+      "step": 11760
+    },
+    {
+      "epoch": 0.3593234827207229,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.426690347868693e-06,
+      "loss": 0.8348,
+      "step": 11770
+    },
+    {
+      "epoch": 0.35962877030162416,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.423628123468888e-06,
+      "loss": 0.8263,
+      "step": 11780
+    },
+    {
+      "epoch": 0.35993405788252536,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.420565899069084e-06,
+      "loss": 0.8358,
+      "step": 11790
+    },
+    {
+      "epoch": 0.36023934546342656,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.41750367466928e-06,
+      "loss": 0.787,
+      "step": 11800
+    },
+    {
+      "epoch": 0.36054463304432777,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.414441450269476e-06,
+      "loss": 0.8258,
+      "step": 11810
+    },
+    {
+      "epoch": 0.36084992062522897,
+      "grad_norm": 3.125,
+      "learning_rate": 6.411379225869673e-06,
+      "loss": 0.8392,
+      "step": 11820
+    },
+    {
+      "epoch": 0.3611552082061302,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.408317001469868e-06,
+      "loss": 0.7953,
+      "step": 11830
+    },
+    {
+      "epoch": 0.3614604957870314,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.405254777070065e-06,
+      "loss": 0.8488,
+      "step": 11840
+    },
+    {
+      "epoch": 0.3617657833679326,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.40219255267026e-06,
+      "loss": 0.8288,
+      "step": 11850
+    },
+    {
+      "epoch": 0.3620710709488338,
+      "grad_norm": 3.984375,
+      "learning_rate": 6.399130328270456e-06,
+      "loss": 0.8342,
+      "step": 11860
+    },
+    {
+      "epoch": 0.362376358529735,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.396068103870652e-06,
+      "loss": 0.8718,
+      "step": 11870
+    },
+    {
+      "epoch": 0.36268164611063625,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.393005879470848e-06,
+      "loss": 0.8366,
+      "step": 11880
+    },
+    {
+      "epoch": 0.36298693369153745,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.389943655071043e-06,
+      "loss": 0.8505,
+      "step": 11890
+    },
+    {
+      "epoch": 0.36329222127243865,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.38688143067124e-06,
+      "loss": 0.8029,
+      "step": 11900
+    },
+    {
+      "epoch": 0.36359750885333986,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.383819206271437e-06,
+      "loss": 0.8644,
+      "step": 11910
+    },
+    {
+      "epoch": 0.36390279643424106,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.380756981871632e-06,
+      "loss": 0.8181,
+      "step": 11920
+    },
+    {
+      "epoch": 0.36420808401514226,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.377694757471828e-06,
+      "loss": 0.7968,
+      "step": 11930
+    },
+    {
+      "epoch": 0.36451337159604347,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.374632533072024e-06,
+      "loss": 0.8009,
+      "step": 11940
+    },
+    {
+      "epoch": 0.36481865917694467,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.37157030867222e-06,
+      "loss": 0.8202,
+      "step": 11950
+    },
+    {
+      "epoch": 0.3651239467578459,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.368508084272415e-06,
+      "loss": 0.8637,
+      "step": 11960
+    },
+    {
+      "epoch": 0.3654292343387471,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.365445859872612e-06,
+      "loss": 0.7956,
+      "step": 11970
+    },
+    {
+      "epoch": 0.3657345219196483,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.362383635472807e-06,
+      "loss": 0.7937,
+      "step": 11980
+    },
+    {
+      "epoch": 0.36603980950054954,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.359321411073004e-06,
+      "loss": 0.8155,
+      "step": 11990
+    },
+    {
+      "epoch": 0.36634509708145074,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.3562591866732e-06,
+      "loss": 0.7984,
+      "step": 12000
+    },
+    {
+      "epoch": 0.36665038466235195,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.353196962273396e-06,
+      "loss": 0.804,
+      "step": 12010
+    },
+    {
+      "epoch": 0.36695567224325315,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.350134737873592e-06,
+      "loss": 0.843,
+      "step": 12020
+    },
+    {
+      "epoch": 0.36726095982415435,
+      "grad_norm": 3.5,
+      "learning_rate": 6.347072513473787e-06,
+      "loss": 0.7907,
+      "step": 12030
+    },
+    {
+      "epoch": 0.36756624740505556,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.344010289073984e-06,
+      "loss": 0.8365,
+      "step": 12040
+    },
+    {
+      "epoch": 0.36787153498595676,
+      "grad_norm": 3.125,
+      "learning_rate": 6.340948064674179e-06,
+      "loss": 0.7879,
+      "step": 12050
+    },
+    {
+      "epoch": 0.36817682256685796,
+      "grad_norm": 3.125,
+      "learning_rate": 6.337885840274376e-06,
+      "loss": 0.8378,
+      "step": 12060
+    },
+    {
+      "epoch": 0.36848211014775917,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.334823615874571e-06,
+      "loss": 0.7982,
+      "step": 12070
+    },
+    {
+      "epoch": 0.36878739772866037,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.331761391474767e-06,
+      "loss": 0.8586,
+      "step": 12080
+    },
+    {
+      "epoch": 0.36909268530956163,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.328699167074964e-06,
+      "loss": 0.746,
+      "step": 12090
+    },
+    {
+      "epoch": 0.36939797289046283,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.325636942675159e-06,
+      "loss": 0.7703,
+      "step": 12100
+    },
+    {
+      "epoch": 0.36970326047136404,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.322574718275356e-06,
+      "loss": 0.8068,
+      "step": 12110
+    },
+    {
+      "epoch": 0.37000854805226524,
+      "grad_norm": 3.75,
+      "learning_rate": 6.319512493875551e-06,
+      "loss": 0.8743,
+      "step": 12120
+    },
+    {
+      "epoch": 0.37031383563316644,
+      "grad_norm": 3.625,
+      "learning_rate": 6.316450269475748e-06,
+      "loss": 0.809,
+      "step": 12130
+    },
+    {
+      "epoch": 0.37061912321406765,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.313388045075943e-06,
+      "loss": 0.8291,
+      "step": 12140
+    },
+    {
+      "epoch": 0.37092441079496885,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.310325820676139e-06,
+      "loss": 0.8083,
+      "step": 12150
+    },
+    {
+      "epoch": 0.37122969837587005,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.307263596276335e-06,
+      "loss": 0.7932,
+      "step": 12160
+    },
+    {
+      "epoch": 0.37153498595677126,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.3042013718765314e-06,
+      "loss": 0.7809,
+      "step": 12170
+    },
+    {
+      "epoch": 0.37184027353767246,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.301139147476728e-06,
+      "loss": 0.8642,
+      "step": 12180
+    },
+    {
+      "epoch": 0.3721455611185737,
+      "grad_norm": 3.125,
+      "learning_rate": 6.2980769230769234e-06,
+      "loss": 0.8152,
+      "step": 12190
+    },
+    {
+      "epoch": 0.3724508486994749,
+      "grad_norm": 3.5,
+      "learning_rate": 6.29501469867712e-06,
+      "loss": 0.7777,
+      "step": 12200
+    },
+    {
+      "epoch": 0.3727561362803761,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.2919524742773154e-06,
+      "loss": 0.8224,
+      "step": 12210
+    },
+    {
+      "epoch": 0.37306142386127733,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.2888902498775115e-06,
+      "loss": 0.7754,
+      "step": 12220
+    },
+    {
+      "epoch": 0.37336671144217853,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.2858280254777075e-06,
+      "loss": 0.7906,
+      "step": 12230
+    },
+    {
+      "epoch": 0.37367199902307974,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.2827658010779035e-06,
+      "loss": 0.8251,
+      "step": 12240
+    },
+    {
+      "epoch": 0.37397728660398094,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.279703576678099e-06,
+      "loss": 0.8077,
+      "step": 12250
+    },
+    {
+      "epoch": 0.37428257418488214,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.2766413522782955e-06,
+      "loss": 0.7886,
+      "step": 12260
+    },
+    {
+      "epoch": 0.37458786176578335,
+      "grad_norm": 3.375,
+      "learning_rate": 6.2735791278784915e-06,
+      "loss": 0.8409,
+      "step": 12270
+    },
+    {
+      "epoch": 0.37489314934668455,
+      "grad_norm": 3.25,
+      "learning_rate": 6.2705169034786875e-06,
+      "loss": 0.8321,
+      "step": 12280
+    },
+    {
+      "epoch": 0.3751984369275858,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.2674546790788835e-06,
+      "loss": 0.8014,
+      "step": 12290
+    },
+    {
+      "epoch": 0.375503724508487,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.2643924546790795e-06,
+      "loss": 0.8066,
+      "step": 12300
+    },
+    {
+      "epoch": 0.3758090120893882,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.2613302302792755e-06,
+      "loss": 0.8348,
+      "step": 12310
+    },
+    {
+      "epoch": 0.3761142996702894,
+      "grad_norm": 3.5,
+      "learning_rate": 6.258268005879471e-06,
+      "loss": 0.8051,
+      "step": 12320
+    },
+    {
+      "epoch": 0.3764195872511906,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.2552057814796675e-06,
+      "loss": 0.8056,
+      "step": 12330
+    },
+    {
+      "epoch": 0.3767248748320918,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.2521435570798635e-06,
+      "loss": 0.8511,
+      "step": 12340
+    },
+    {
+      "epoch": 0.37703016241299303,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.2490813326800595e-06,
+      "loss": 0.8288,
+      "step": 12350
+    },
+    {
+      "epoch": 0.37733544999389423,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.2460191082802555e-06,
+      "loss": 0.8113,
+      "step": 12360
+    },
+    {
+      "epoch": 0.37764073757479544,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.242956883880451e-06,
+      "loss": 0.8235,
+      "step": 12370
+    },
+    {
+      "epoch": 0.37794602515569664,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.2398946594806475e-06,
+      "loss": 0.8328,
+      "step": 12380
+    },
+    {
+      "epoch": 0.3782513127365979,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.236832435080843e-06,
+      "loss": 0.8341,
+      "step": 12390
+    },
+    {
+      "epoch": 0.3785566003174991,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.2337702106810396e-06,
+      "loss": 0.7835,
+      "step": 12400
+    },
+    {
+      "epoch": 0.3788618878984003,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.230707986281235e-06,
+      "loss": 0.8403,
+      "step": 12410
+    },
+    {
+      "epoch": 0.3791671754793015,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.2276457618814316e-06,
+      "loss": 0.7817,
+      "step": 12420
+    },
+    {
+      "epoch": 0.3794724630602027,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.2245835374816276e-06,
+      "loss": 0.8529,
+      "step": 12430
+    },
+    {
+      "epoch": 0.3797777506411039,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.221521313081823e-06,
+      "loss": 0.796,
+      "step": 12440
+    },
+    {
+      "epoch": 0.3800830382220051,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.21845908868202e-06,
+      "loss": 0.8019,
+      "step": 12450
+    },
+    {
+      "epoch": 0.3803883258029063,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.215396864282215e-06,
+      "loss": 0.8205,
+      "step": 12460
+    },
+    {
+      "epoch": 0.3806936133838075,
+      "grad_norm": 3.375,
+      "learning_rate": 6.212334639882412e-06,
+      "loss": 0.8516,
+      "step": 12470
+    },
+    {
+      "epoch": 0.38099890096470873,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.209272415482607e-06,
+      "loss": 0.7754,
+      "step": 12480
+    },
+    {
+      "epoch": 0.38130418854561,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.206210191082804e-06,
+      "loss": 0.7938,
+      "step": 12490
+    },
+    {
+      "epoch": 0.3816094761265112,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.203147966682999e-06,
+      "loss": 0.8072,
+      "step": 12500
+    },
+    {
+      "epoch": 0.3819147637074124,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.200085742283195e-06,
+      "loss": 0.7863,
+      "step": 12510
+    },
+    {
+      "epoch": 0.3822200512883136,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.197023517883392e-06,
+      "loss": 0.7506,
+      "step": 12520
+    },
+    {
+      "epoch": 0.3825253388692148,
+      "grad_norm": 3.71875,
+      "learning_rate": 6.193961293483587e-06,
+      "loss": 0.8502,
+      "step": 12530
+    },
+    {
+      "epoch": 0.382830626450116,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.190899069083784e-06,
+      "loss": 0.7924,
+      "step": 12540
+    },
+    {
+      "epoch": 0.3831359140310172,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.187836844683979e-06,
+      "loss": 0.8296,
+      "step": 12550
+    },
+    {
+      "epoch": 0.3834412016119184,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.184774620284175e-06,
+      "loss": 0.7904,
+      "step": 12560
+    },
+    {
+      "epoch": 0.3837464891928196,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.181712395884371e-06,
+      "loss": 0.7794,
+      "step": 12570
+    },
+    {
+      "epoch": 0.3840517767737208,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.178650171484567e-06,
+      "loss": 0.8373,
+      "step": 12580
+    },
+    {
+      "epoch": 0.3843570643546221,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.175587947084763e-06,
+      "loss": 0.8251,
+      "step": 12590
+    },
+    {
+      "epoch": 0.3846623519355233,
+      "grad_norm": 3.71875,
+      "learning_rate": 6.172525722684959e-06,
+      "loss": 0.8476,
+      "step": 12600
+    },
+    {
+      "epoch": 0.3849676395164245,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.169463498285156e-06,
+      "loss": 0.8025,
+      "step": 12610
+    },
+    {
+      "epoch": 0.3852729270973257,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.166401273885351e-06,
+      "loss": 0.8507,
+      "step": 12620
+    },
+    {
+      "epoch": 0.3855782146782269,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.163339049485547e-06,
+      "loss": 0.8623,
+      "step": 12630
+    },
+    {
+      "epoch": 0.3858835022591281,
+      "grad_norm": 3.125,
+      "learning_rate": 6.160276825085743e-06,
+      "loss": 0.8208,
+      "step": 12640
+    },
+    {
+      "epoch": 0.3861887898400293,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.157214600685939e-06,
+      "loss": 0.8111,
+      "step": 12650
+    },
+    {
+      "epoch": 0.3864940774209305,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.154152376286134e-06,
+      "loss": 0.8069,
+      "step": 12660
+    },
+    {
+      "epoch": 0.3867993650018317,
+      "grad_norm": 3.375,
+      "learning_rate": 6.151090151886331e-06,
+      "loss": 0.8024,
+      "step": 12670
+    },
+    {
+      "epoch": 0.3871046525827329,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.148027927486526e-06,
+      "loss": 0.81,
+      "step": 12680
+    },
+    {
+      "epoch": 0.38740994016363417,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.144965703086723e-06,
+      "loss": 0.8515,
+      "step": 12690
+    },
+    {
+      "epoch": 0.3877152277445354,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.141903478686919e-06,
+      "loss": 0.8091,
+      "step": 12700
+    },
+    {
+      "epoch": 0.3880205153254366,
+      "grad_norm": 3.375,
+      "learning_rate": 6.138841254287115e-06,
+      "loss": 0.8051,
+      "step": 12710
+    },
+    {
+      "epoch": 0.3883258029063378,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.135779029887311e-06,
+      "loss": 0.7844,
+      "step": 12720
+    },
+    {
+      "epoch": 0.388631090487239,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.132716805487506e-06,
+      "loss": 0.7924,
+      "step": 12730
+    },
+    {
+      "epoch": 0.3889363780681402,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.129654581087703e-06,
+      "loss": 0.7895,
+      "step": 12740
+    },
+    {
+      "epoch": 0.3892416656490414,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.126592356687898e-06,
+      "loss": 0.7949,
+      "step": 12750
+    },
+    {
+      "epoch": 0.3895469532299426,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.123530132288095e-06,
+      "loss": 0.7941,
+      "step": 12760
+    },
+    {
+      "epoch": 0.3898522408108438,
+      "grad_norm": 3.953125,
+      "learning_rate": 6.12046790788829e-06,
+      "loss": 0.8862,
+      "step": 12770
+    },
+    {
+      "epoch": 0.390157528391745,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.117405683488487e-06,
+      "loss": 0.8141,
+      "step": 12780
+    },
+    {
+      "epoch": 0.39046281597264626,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.114343459088683e-06,
+      "loss": 0.7841,
+      "step": 12790
+    },
+    {
+      "epoch": 0.39076810355354746,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.111281234688878e-06,
+      "loss": 0.8349,
+      "step": 12800
+    },
+    {
+      "epoch": 0.39107339113444867,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.108219010289075e-06,
+      "loss": 0.8262,
+      "step": 12810
+    },
+    {
+      "epoch": 0.39137867871534987,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.10515678588927e-06,
+      "loss": 0.8873,
+      "step": 12820
+    },
+    {
+      "epoch": 0.3916839662962511,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.102094561489467e-06,
+      "loss": 0.8102,
+      "step": 12830
+    },
+    {
+      "epoch": 0.3919892538771523,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.099032337089662e-06,
+      "loss": 0.765,
+      "step": 12840
+    },
+    {
+      "epoch": 0.3922945414580535,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.095970112689858e-06,
+      "loss": 0.8043,
+      "step": 12850
+    },
+    {
+      "epoch": 0.3925998290389547,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.092907888290054e-06,
+      "loss": 0.7729,
+      "step": 12860
+    },
+    {
+      "epoch": 0.3929051166198559,
+      "grad_norm": 3.5,
+      "learning_rate": 6.08984566389025e-06,
+      "loss": 0.7682,
+      "step": 12870
+    },
+    {
+      "epoch": 0.3932104042007571,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.086783439490447e-06,
+      "loss": 0.8146,
+      "step": 12880
+    },
+    {
+      "epoch": 0.3935156917816583,
+      "grad_norm": 3.375,
+      "learning_rate": 6.083721215090642e-06,
+      "loss": 0.7549,
+      "step": 12890
+    },
+    {
+      "epoch": 0.39382097936255955,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.080658990690839e-06,
+      "loss": 0.7878,
+      "step": 12900
+    },
+    {
+      "epoch": 0.39412626694346076,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.077596766291034e-06,
+      "loss": 0.8726,
+      "step": 12910
+    },
+    {
+      "epoch": 0.39443155452436196,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.07453454189123e-06,
+      "loss": 0.772,
+      "step": 12920
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.071472317491426e-06,
+      "loss": 0.7802,
+      "step": 12930
+    },
+    {
+      "epoch": 0.39504212968616437,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.068410093091622e-06,
+      "loss": 0.8244,
+      "step": 12940
+    },
+    {
+      "epoch": 0.39534741726706557,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.065347868691817e-06,
+      "loss": 0.802,
+      "step": 12950
+    },
+    {
+      "epoch": 0.3956527048479668,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.062285644292014e-06,
+      "loss": 0.8215,
+      "step": 12960
+    },
+    {
+      "epoch": 0.395957992428868,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.059223419892211e-06,
+      "loss": 0.7785,
+      "step": 12970
+    },
+    {
+      "epoch": 0.3962632800097692,
+      "grad_norm": 3.25,
+      "learning_rate": 6.056161195492406e-06,
+      "loss": 0.775,
+      "step": 12980
+    },
+    {
+      "epoch": 0.3965685675906704,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.053098971092602e-06,
+      "loss": 0.7706,
+      "step": 12990
+    },
+    {
+      "epoch": 0.39687385517157164,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.050036746692798e-06,
+      "loss": 0.7802,
+      "step": 13000
+    },
+    {
+      "epoch": 0.39717914275247285,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.046974522292994e-06,
+      "loss": 0.7515,
+      "step": 13010
+    },
+    {
+      "epoch": 0.39748443033337405,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.043912297893189e-06,
+      "loss": 0.7724,
+      "step": 13020
+    },
+    {
+      "epoch": 0.39778971791427525,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.040850073493386e-06,
+      "loss": 0.7763,
+      "step": 13030
+    },
+    {
+      "epoch": 0.39809500549517646,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.037787849093582e-06,
+      "loss": 0.8076,
+      "step": 13040
+    },
+    {
+      "epoch": 0.39840029307607766,
+      "grad_norm": 3.71875,
+      "learning_rate": 6.034725624693778e-06,
+      "loss": 0.7818,
+      "step": 13050
+    },
+    {
+      "epoch": 0.39870558065697886,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.031663400293974e-06,
+      "loss": 0.8288,
+      "step": 13060
+    },
+    {
+      "epoch": 0.39901086823788007,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.02860117589417e-06,
+      "loss": 0.8003,
+      "step": 13070
+    },
+    {
+      "epoch": 0.39931615581878127,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.025538951494366e-06,
+      "loss": 0.8395,
+      "step": 13080
+    },
+    {
+      "epoch": 0.3996214433996825,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.022476727094561e-06,
+      "loss": 0.8069,
+      "step": 13090
+    },
+    {
+      "epoch": 0.39992673098058373,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.019414502694758e-06,
+      "loss": 0.7694,
+      "step": 13100
+    },
+    {
+      "epoch": 0.40023201856148494,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.016352278294953e-06,
+      "loss": 0.8141,
+      "step": 13110
+    },
+    {
+      "epoch": 0.40053730614238614,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.01329005389515e-06,
+      "loss": 0.7649,
+      "step": 13120
+    },
+    {
+      "epoch": 0.40084259372328734,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.010227829495346e-06,
+      "loss": 0.7885,
+      "step": 13130
+    },
+    {
+      "epoch": 0.40114788130418855,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.007165605095541e-06,
+      "loss": 0.8066,
+      "step": 13140
+    },
+    {
+      "epoch": 0.40145316888508975,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.004103380695738e-06,
+      "loss": 0.8195,
+      "step": 13150
+    },
+    {
+      "epoch": 0.40175845646599095,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.0010411562959334e-06,
+      "loss": 0.8055,
+      "step": 13160
+    },
+    {
+      "epoch": 0.40206374404689216,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.99797893189613e-06,
+      "loss": 0.7736,
+      "step": 13170
+    },
+    {
+      "epoch": 0.40236903162779336,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.9949167074963254e-06,
+      "loss": 0.8048,
+      "step": 13180
+    },
+    {
+      "epoch": 0.40267431920869456,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.991854483096522e-06,
+      "loss": 0.7655,
+      "step": 13190
+    },
+    {
+      "epoch": 0.4029796067895958,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.9887922586967174e-06,
+      "loss": 0.7945,
+      "step": 13200
+    },
+    {
+      "epoch": 0.403284894370497,
+      "grad_norm": 3.875,
+      "learning_rate": 5.9857300342969135e-06,
+      "loss": 0.815,
+      "step": 13210
+    },
+    {
+      "epoch": 0.40359018195139823,
+      "grad_norm": 3.140625,
+      "learning_rate": 5.98266780989711e-06,
+      "loss": 0.7823,
+      "step": 13220
+    },
+    {
+      "epoch": 0.40389546953229943,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.9796055854973055e-06,
+      "loss": 0.8172,
+      "step": 13230
+    },
+    {
+      "epoch": 0.40420075711320064,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.976543361097502e-06,
+      "loss": 0.8085,
+      "step": 13240
+    },
+    {
+      "epoch": 0.40450604469410184,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.9734811366976975e-06,
+      "loss": 0.8159,
+      "step": 13250
+    },
+    {
+      "epoch": 0.40481133227500304,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.970418912297894e-06,
+      "loss": 0.8192,
+      "step": 13260
+    },
+    {
+      "epoch": 0.40511661985590425,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.9673566878980895e-06,
+      "loss": 0.7681,
+      "step": 13270
+    },
+    {
+      "epoch": 0.40542190743680545,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.9642944634982855e-06,
+      "loss": 0.7912,
+      "step": 13280
+    },
+    {
+      "epoch": 0.40572719501770665,
+      "grad_norm": 3.625,
+      "learning_rate": 5.9612322390984815e-06,
+      "loss": 0.8208,
+      "step": 13290
+    },
+    {
+      "epoch": 0.4060324825986079,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.9581700146986775e-06,
+      "loss": 0.7851,
+      "step": 13300
+    },
+    {
+      "epoch": 0.4063377701795091,
+      "grad_norm": 3.375,
+      "learning_rate": 5.955107790298874e-06,
+      "loss": 0.7926,
+      "step": 13310
+    },
+    {
+      "epoch": 0.4066430577604103,
+      "grad_norm": 3.625,
+      "learning_rate": 5.9520455658990695e-06,
+      "loss": 0.8309,
+      "step": 13320
+    },
+    {
+      "epoch": 0.4069483453413115,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.9489833414992655e-06,
+      "loss": 0.8164,
+      "step": 13330
+    },
+    {
+      "epoch": 0.4072536329222127,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.9459211170994615e-06,
+      "loss": 0.801,
+      "step": 13340
+    },
+    {
+      "epoch": 0.40755892050311393,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.9428588926996575e-06,
+      "loss": 0.8614,
+      "step": 13350
+    },
+    {
+      "epoch": 0.40786420808401513,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.9397966682998535e-06,
+      "loss": 0.8427,
+      "step": 13360
+    },
+    {
+      "epoch": 0.40816949566491634,
+      "grad_norm": 3.5,
+      "learning_rate": 5.9367344439000495e-06,
+      "loss": 0.7983,
+      "step": 13370
+    },
+    {
+      "epoch": 0.40847478324581754,
+      "grad_norm": 3.75,
+      "learning_rate": 5.933672219500245e-06,
+      "loss": 0.826,
+      "step": 13380
+    },
+    {
+      "epoch": 0.40878007082671874,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.9306099951004416e-06,
+      "loss": 0.8072,
+      "step": 13390
+    },
+    {
+      "epoch": 0.40908535840762,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.9275477707006376e-06,
+      "loss": 0.8262,
+      "step": 13400
+    },
+    {
+      "epoch": 0.4093906459885212,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.9244855463008336e-06,
+      "loss": 0.7809,
+      "step": 13410
+    },
+    {
+      "epoch": 0.4096959335694224,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.9214233219010296e-06,
+      "loss": 0.8063,
+      "step": 13420
+    },
+    {
+      "epoch": 0.4100012211503236,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.918361097501225e-06,
+      "loss": 0.866,
+      "step": 13430
+    },
+    {
+      "epoch": 0.4103065087312248,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.915298873101422e-06,
+      "loss": 0.8192,
+      "step": 13440
+    },
+    {
+      "epoch": 0.410611796312126,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.912236648701617e-06,
+      "loss": 0.7973,
+      "step": 13450
+    },
+    {
+      "epoch": 0.4109170838930272,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.909174424301814e-06,
+      "loss": 0.7826,
+      "step": 13460
+    },
+    {
+      "epoch": 0.4112223714739284,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.906112199902009e-06,
+      "loss": 0.8292,
+      "step": 13470
+    },
+    {
+      "epoch": 0.41152765905482963,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.903049975502206e-06,
+      "loss": 0.789,
+      "step": 13480
+    },
+    {
+      "epoch": 0.41183294663573083,
+      "grad_norm": 3.375,
+      "learning_rate": 5.899987751102402e-06,
+      "loss": 0.8416,
+      "step": 13490
+    },
+    {
+      "epoch": 0.4121382342166321,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.896925526702597e-06,
+      "loss": 0.7881,
+      "step": 13500
+    },
+    {
+      "epoch": 0.4124435217975333,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.893863302302794e-06,
+      "loss": 0.8639,
+      "step": 13510
+    },
+    {
+      "epoch": 0.4127488093784345,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.890801077902989e-06,
+      "loss": 0.7993,
+      "step": 13520
+    },
+    {
+      "epoch": 0.4130540969593357,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.887738853503186e-06,
+      "loss": 0.816,
+      "step": 13530
+    },
+    {
+      "epoch": 0.4133593845402369,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.884676629103381e-06,
+      "loss": 0.8247,
+      "step": 13540
+    },
+    {
+      "epoch": 0.4136646721211381,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.881614404703578e-06,
+      "loss": 0.8107,
+      "step": 13550
+    },
+    {
+      "epoch": 0.4139699597020393,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.878552180303773e-06,
+      "loss": 0.7799,
+      "step": 13560
+    },
+    {
+      "epoch": 0.4142752472829405,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.875489955903969e-06,
+      "loss": 0.7955,
+      "step": 13570
+    },
+    {
+      "epoch": 0.4145805348638417,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.872427731504166e-06,
+      "loss": 0.7847,
+      "step": 13580
+    },
+    {
+      "epoch": 0.4148858224447429,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.869365507104361e-06,
+      "loss": 0.7935,
+      "step": 13590
+    },
+    {
+      "epoch": 0.4151911100256442,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.866303282704558e-06,
+      "loss": 0.8272,
+      "step": 13600
+    },
+    {
+      "epoch": 0.4154963976065454,
+      "grad_norm": 3.5,
+      "learning_rate": 5.863241058304753e-06,
+      "loss": 0.7518,
+      "step": 13610
+    },
+    {
+      "epoch": 0.4158016851874466,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.860178833904949e-06,
+      "loss": 0.7862,
+      "step": 13620
+    },
+    {
+      "epoch": 0.4161069727683478,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.857116609505145e-06,
+      "loss": 0.7875,
+      "step": 13630
+    },
+    {
+      "epoch": 0.416412260349249,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.854054385105341e-06,
+      "loss": 0.7736,
+      "step": 13640
+    },
+    {
+      "epoch": 0.4167175479301502,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.850992160705537e-06,
+      "loss": 0.8059,
+      "step": 13650
+    },
+    {
+      "epoch": 0.4170228355110514,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.847929936305733e-06,
+      "loss": 0.8172,
+      "step": 13660
+    },
+    {
+      "epoch": 0.4173281230919526,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.84486771190593e-06,
+      "loss": 0.7703,
+      "step": 13670
+    },
+    {
+      "epoch": 0.4176334106728538,
+      "grad_norm": 3.5,
+      "learning_rate": 5.841805487506125e-06,
+      "loss": 0.8076,
+      "step": 13680
+    },
+    {
+      "epoch": 0.417938698253755,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.838743263106321e-06,
+      "loss": 0.817,
+      "step": 13690
+    },
+    {
+      "epoch": 0.4182439858346563,
+      "grad_norm": 4.34375,
+      "learning_rate": 5.835681038706517e-06,
+      "loss": 0.7947,
+      "step": 13700
+    },
+    {
+      "epoch": 0.4185492734155575,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.832618814306713e-06,
+      "loss": 0.7722,
+      "step": 13710
+    },
+    {
+      "epoch": 0.4188545609964587,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.829556589906908e-06,
+      "loss": 0.79,
+      "step": 13720
+    },
+    {
+      "epoch": 0.4191598485773599,
+      "grad_norm": 3.140625,
+      "learning_rate": 5.826494365507105e-06,
+      "loss": 0.7809,
+      "step": 13730
+    },
+    {
+      "epoch": 0.4194651361582611,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.823432141107302e-06,
+      "loss": 0.8186,
+      "step": 13740
+    },
+    {
+      "epoch": 0.4197704237391623,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.820369916707497e-06,
+      "loss": 0.8287,
+      "step": 13750
+    },
+    {
+      "epoch": 0.4200757113200635,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.817307692307693e-06,
+      "loss": 0.8436,
+      "step": 13760
+    },
+    {
+      "epoch": 0.4203809989009647,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.814245467907889e-06,
+      "loss": 0.8197,
+      "step": 13770
+    },
+    {
+      "epoch": 0.4206862864818659,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.811183243508085e-06,
+      "loss": 0.7793,
+      "step": 13780
+    },
+    {
+      "epoch": 0.4209915740627671,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.80812101910828e-06,
+      "loss": 0.8534,
+      "step": 13790
+    },
+    {
+      "epoch": 0.42129686164366836,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.805058794708477e-06,
+      "loss": 0.8133,
+      "step": 13800
+    },
+    {
+      "epoch": 0.42160214922456957,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.801996570308672e-06,
+      "loss": 0.7813,
+      "step": 13810
+    },
+    {
+      "epoch": 0.42190743680547077,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.798934345908869e-06,
+      "loss": 0.7782,
+      "step": 13820
+    },
+    {
+      "epoch": 0.422212724386372,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.795872121509065e-06,
+      "loss": 0.8043,
+      "step": 13830
+    },
+    {
+      "epoch": 0.4225180119672732,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.792809897109261e-06,
+      "loss": 0.8363,
+      "step": 13840
+    },
+    {
+      "epoch": 0.4228232995481744,
+      "grad_norm": 3.5,
+      "learning_rate": 5.789747672709457e-06,
+      "loss": 0.8057,
+      "step": 13850
+    },
+    {
+      "epoch": 0.4231285871290756,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.786685448309652e-06,
+      "loss": 0.7953,
+      "step": 13860
+    },
+    {
+      "epoch": 0.4234338747099768,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.783623223909849e-06,
+      "loss": 0.8455,
+      "step": 13870
+    },
+    {
+      "epoch": 0.423739162290878,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.780560999510044e-06,
+      "loss": 0.8561,
+      "step": 13880
+    },
+    {
+      "epoch": 0.4240444498717792,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.777498775110241e-06,
+      "loss": 0.8061,
+      "step": 13890
+    },
+    {
+      "epoch": 0.4243497374526804,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.774436550710436e-06,
+      "loss": 0.8608,
+      "step": 13900
+    },
+    {
+      "epoch": 0.42465502503358166,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.771374326310632e-06,
+      "loss": 0.809,
+      "step": 13910
+    },
+    {
+      "epoch": 0.42496031261448286,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.768312101910829e-06,
+      "loss": 0.8078,
+      "step": 13920
+    },
+    {
+      "epoch": 0.42526560019538406,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.765249877511024e-06,
+      "loss": 0.828,
+      "step": 13930
+    },
+    {
+      "epoch": 0.42557088777628527,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.762187653111221e-06,
+      "loss": 0.8197,
+      "step": 13940
+    },
+    {
+      "epoch": 0.42587617535718647,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.759125428711416e-06,
+      "loss": 0.8584,
+      "step": 13950
+    },
+    {
+      "epoch": 0.4261814629380877,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.756063204311613e-06,
+      "loss": 0.8198,
+      "step": 13960
+    },
+    {
+      "epoch": 0.4264867505189889,
+      "grad_norm": 3.140625,
+      "learning_rate": 5.753000979911808e-06,
+      "loss": 0.8009,
+      "step": 13970
+    },
+    {
+      "epoch": 0.4267920380998901,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.749938755512004e-06,
+      "loss": 0.8217,
+      "step": 13980
+    },
+    {
+      "epoch": 0.4270973256807913,
+      "grad_norm": 3.5,
+      "learning_rate": 5.7468765311122e-06,
+      "loss": 0.8373,
+      "step": 13990
+    },
+    {
+      "epoch": 0.4274026132616925,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.743814306712396e-06,
+      "loss": 0.8285,
+      "step": 14000
+    },
+    {
+      "epoch": 0.42770790084259375,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.740752082312593e-06,
+      "loss": 0.8313,
+      "step": 14010
+    },
+    {
+      "epoch": 0.42801318842349495,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.737689857912788e-06,
+      "loss": 0.7749,
+      "step": 14020
+    },
+    {
+      "epoch": 0.42831847600439615,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.734627633512985e-06,
+      "loss": 0.7747,
+      "step": 14030
+    },
+    {
+      "epoch": 0.42862376358529736,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.73156540911318e-06,
+      "loss": 0.7983,
+      "step": 14040
+    },
+    {
+      "epoch": 0.42892905116619856,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.728503184713376e-06,
+      "loss": 0.8619,
+      "step": 14050
+    },
+    {
+      "epoch": 0.42923433874709976,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.725440960313572e-06,
+      "loss": 0.8478,
+      "step": 14060
+    },
+    {
+      "epoch": 0.42953962632800097,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.722378735913768e-06,
+      "loss": 0.8259,
+      "step": 14070
+    },
+    {
+      "epoch": 0.42984491390890217,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.719316511513963e-06,
+      "loss": 0.8101,
+      "step": 14080
+    },
+    {
+      "epoch": 0.4301502014898034,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.71625428711416e-06,
+      "loss": 0.7715,
+      "step": 14090
+    },
+    {
+      "epoch": 0.4304554890707046,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.713192062714357e-06,
+      "loss": 0.7888,
+      "step": 14100
+    },
+    {
+      "epoch": 0.43076077665160584,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.710129838314552e-06,
+      "loss": 0.8106,
+      "step": 14110
+    },
+    {
+      "epoch": 0.43106606423250704,
+      "grad_norm": 5.0625,
+      "learning_rate": 5.707067613914748e-06,
+      "loss": 0.8589,
+      "step": 14120
+    },
+    {
+      "epoch": 0.43137135181340824,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.704005389514944e-06,
+      "loss": 0.7938,
+      "step": 14130
+    },
+    {
+      "epoch": 0.43167663939430945,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.70094316511514e-06,
+      "loss": 0.8349,
+      "step": 14140
+    },
+    {
+      "epoch": 0.43198192697521065,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.6978809407153354e-06,
+      "loss": 0.8658,
+      "step": 14150
+    },
+    {
+      "epoch": 0.43228721455611185,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.694818716315532e-06,
+      "loss": 0.8018,
+      "step": 14160
+    },
+    {
+      "epoch": 0.43259250213701306,
+      "grad_norm": 3.375,
+      "learning_rate": 5.6917564919157274e-06,
+      "loss": 0.8112,
+      "step": 14170
+    },
+    {
+      "epoch": 0.43289778971791426,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.688694267515924e-06,
+      "loss": 0.8136,
+      "step": 14180
+    },
+    {
+      "epoch": 0.43320307729881546,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.68563204311612e-06,
+      "loss": 0.8269,
+      "step": 14190
+    },
+    {
+      "epoch": 0.43350836487971667,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.6825698187163155e-06,
+      "loss": 0.8003,
+      "step": 14200
+    },
+    {
+      "epoch": 0.4338136524606179,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.679507594316512e-06,
+      "loss": 0.8359,
+      "step": 14210
+    },
+    {
+      "epoch": 0.43411894004151913,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.6764453699167075e-06,
+      "loss": 0.8067,
+      "step": 14220
+    },
+    {
+      "epoch": 0.43442422762242033,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.673383145516904e-06,
+      "loss": 0.7776,
+      "step": 14230
+    },
+    {
+      "epoch": 0.43472951520332154,
+      "grad_norm": 3.375,
+      "learning_rate": 5.6703209211170995e-06,
+      "loss": 0.7729,
+      "step": 14240
+    },
+    {
+      "epoch": 0.43503480278422274,
+      "grad_norm": 3.75,
+      "learning_rate": 5.667258696717296e-06,
+      "loss": 0.7953,
+      "step": 14250
+    },
+    {
+      "epoch": 0.43534009036512394,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.6641964723174915e-06,
+      "loss": 0.8167,
+      "step": 14260
+    },
+    {
+      "epoch": 0.43564537794602515,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.6611342479176875e-06,
+      "loss": 0.8137,
+      "step": 14270
+    },
+    {
+      "epoch": 0.43595066552692635,
+      "grad_norm": 4.375,
+      "learning_rate": 5.658072023517884e-06,
+      "loss": 0.8089,
+      "step": 14280
+    },
+    {
+      "epoch": 0.43625595310782755,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.6550097991180795e-06,
+      "loss": 0.7812,
+      "step": 14290
+    },
+    {
+      "epoch": 0.43656124068872876,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.651947574718276e-06,
+      "loss": 0.8322,
+      "step": 14300
+    },
+    {
+      "epoch": 0.43686652826963,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.6488853503184715e-06,
+      "loss": 0.816,
+      "step": 14310
+    },
+    {
+      "epoch": 0.4371718158505312,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.645823125918668e-06,
+      "loss": 0.8602,
+      "step": 14320
+    },
+    {
+      "epoch": 0.4374771034314324,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.6427609015188635e-06,
+      "loss": 0.846,
+      "step": 14330
+    },
+    {
+      "epoch": 0.4377823910123336,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.6396986771190595e-06,
+      "loss": 0.7881,
+      "step": 14340
+    },
+    {
+      "epoch": 0.43808767859323483,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.6366364527192555e-06,
+      "loss": 0.8688,
+      "step": 14350
+    },
+    {
+      "epoch": 0.43839296617413603,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.6335742283194515e-06,
+      "loss": 0.8214,
+      "step": 14360
+    },
+    {
+      "epoch": 0.43869825375503724,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.630512003919648e-06,
+      "loss": 0.8511,
+      "step": 14370
+    },
+    {
+      "epoch": 0.43900354133593844,
+      "grad_norm": 3.25,
+      "learning_rate": 5.6274497795198435e-06,
+      "loss": 0.8495,
+      "step": 14380
+    },
+    {
+      "epoch": 0.43930882891683964,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.6243875551200396e-06,
+      "loss": 0.7912,
+      "step": 14390
+    },
+    {
+      "epoch": 0.43961411649774085,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.6213253307202356e-06,
+      "loss": 0.8339,
+      "step": 14400
+    },
+    {
+      "epoch": 0.4399194040786421,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.6182631063204316e-06,
+      "loss": 0.797,
+      "step": 14410
+    },
+    {
+      "epoch": 0.4402246916595433,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.6152008819206276e-06,
+      "loss": 0.7699,
+      "step": 14420
+    },
+    {
+      "epoch": 0.4405299792404445,
+      "grad_norm": 3.375,
+      "learning_rate": 5.612138657520824e-06,
+      "loss": 0.8098,
+      "step": 14430
+    },
+    {
+      "epoch": 0.4408352668213457,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.609076433121019e-06,
+      "loss": 0.7904,
+      "step": 14440
+    },
+    {
+      "epoch": 0.4411405544022469,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.606014208721216e-06,
+      "loss": 0.8679,
+      "step": 14450
+    },
+    {
+      "epoch": 0.4414458419831481,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.602951984321412e-06,
+      "loss": 0.7832,
+      "step": 14460
+    },
+    {
+      "epoch": 0.4417511295640493,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.599889759921608e-06,
+      "loss": 0.7492,
+      "step": 14470
+    },
+    {
+      "epoch": 0.44205641714495053,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.596827535521804e-06,
+      "loss": 0.829,
+      "step": 14480
+    },
+    {
+      "epoch": 0.44236170472585173,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.593765311121999e-06,
+      "loss": 0.8475,
+      "step": 14490
+    },
+    {
+      "epoch": 0.44266699230675294,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.590703086722196e-06,
+      "loss": 0.8227,
+      "step": 14500
+    },
+    {
+      "epoch": 0.4429722798876542,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.587640862322391e-06,
+      "loss": 0.8152,
+      "step": 14510
+    },
+    {
+      "epoch": 0.4432775674685554,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.584578637922588e-06,
+      "loss": 0.8666,
+      "step": 14520
+    },
+    {
+      "epoch": 0.4435828550494566,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.581516413522784e-06,
+      "loss": 0.8208,
+      "step": 14530
+    },
+    {
+      "epoch": 0.4438881426303578,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.57845418912298e-06,
+      "loss": 0.7931,
+      "step": 14540
+    },
+    {
+      "epoch": 0.444193430211259,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.575391964723176e-06,
+      "loss": 0.8441,
+      "step": 14550
+    },
+    {
+      "epoch": 0.4444987177921602,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.572329740323371e-06,
+      "loss": 0.8215,
+      "step": 14560
+    },
+    {
+      "epoch": 0.4448040053730614,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.569267515923568e-06,
+      "loss": 0.8443,
+      "step": 14570
+    },
+    {
+      "epoch": 0.4451092929539626,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.566205291523763e-06,
+      "loss": 0.8122,
+      "step": 14580
+    },
+    {
+      "epoch": 0.4454145805348638,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.56314306712396e-06,
+      "loss": 0.8029,
+      "step": 14590
+    },
+    {
+      "epoch": 0.445719868115765,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.560080842724155e-06,
+      "loss": 0.8602,
+      "step": 14600
+    },
+    {
+      "epoch": 0.4460251556966663,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.557018618324352e-06,
+      "loss": 0.8823,
+      "step": 14610
+    },
+    {
+      "epoch": 0.4463304432775675,
+      "grad_norm": 3.75,
+      "learning_rate": 5.553956393924548e-06,
+      "loss": 0.8325,
+      "step": 14620
+    },
+    {
+      "epoch": 0.4466357308584687,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.550894169524743e-06,
+      "loss": 0.7539,
+      "step": 14630
+    },
+    {
+      "epoch": 0.4469410184393699,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.54783194512494e-06,
+      "loss": 0.7852,
+      "step": 14640
+    },
+    {
+      "epoch": 0.4472463060202711,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.544769720725135e-06,
+      "loss": 0.8364,
+      "step": 14650
+    },
+    {
+      "epoch": 0.4475515936011723,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.541707496325332e-06,
+      "loss": 0.8177,
+      "step": 14660
+    },
+    {
+      "epoch": 0.4478568811820735,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.538645271925527e-06,
+      "loss": 0.8022,
+      "step": 14670
+    },
+    {
+      "epoch": 0.4481621687629747,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.535583047525723e-06,
+      "loss": 0.8625,
+      "step": 14680
+    },
+    {
+      "epoch": 0.4484674563438759,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.532520823125919e-06,
+      "loss": 0.8495,
+      "step": 14690
+    },
+    {
+      "epoch": 0.4487727439247771,
+      "grad_norm": 3.875,
+      "learning_rate": 5.529458598726115e-06,
+      "loss": 0.8049,
+      "step": 14700
+    },
+    {
+      "epoch": 0.4490780315056784,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.526396374326312e-06,
+      "loss": 0.7751,
+      "step": 14710
+    },
+    {
+      "epoch": 0.4493833190865796,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.523334149926507e-06,
+      "loss": 0.7476,
+      "step": 14720
+    },
+    {
+      "epoch": 0.4496886066674808,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.520271925526704e-06,
+      "loss": 0.7767,
+      "step": 14730
+    },
+    {
+      "epoch": 0.449993894248382,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.517209701126899e-06,
+      "loss": 0.8044,
+      "step": 14740
+    },
+    {
+      "epoch": 0.4502991818292832,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.514147476727095e-06,
+      "loss": 0.8346,
+      "step": 14750
+    },
+    {
+      "epoch": 0.4506044694101844,
+      "grad_norm": 4.25,
+      "learning_rate": 5.511085252327291e-06,
+      "loss": 0.814,
+      "step": 14760
+    },
+    {
+      "epoch": 0.4509097569910856,
+      "grad_norm": 3.25,
+      "learning_rate": 5.508023027927487e-06,
+      "loss": 0.7762,
+      "step": 14770
+    },
+    {
+      "epoch": 0.4512150445719868,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.504960803527682e-06,
+      "loss": 0.7392,
+      "step": 14780
+    },
+    {
+      "epoch": 0.451520332152888,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.501898579127879e-06,
+      "loss": 0.8041,
+      "step": 14790
+    },
+    {
+      "epoch": 0.4518256197337892,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.498836354728076e-06,
+      "loss": 0.8325,
+      "step": 14800
+    },
+    {
+      "epoch": 0.4521309073146904,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.495774130328271e-06,
+      "loss": 0.8757,
+      "step": 14810
+    },
+    {
+      "epoch": 0.45243619489559167,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.492711905928467e-06,
+      "loss": 0.8403,
+      "step": 14820
+    },
+    {
+      "epoch": 0.4527414824764929,
+      "grad_norm": 3.625,
+      "learning_rate": 5.489649681528663e-06,
+      "loss": 0.8558,
+      "step": 14830
+    },
+    {
+      "epoch": 0.4530467700573941,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.486587457128859e-06,
+      "loss": 0.7421,
+      "step": 14840
+    },
+    {
+      "epoch": 0.4533520576382953,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.483525232729054e-06,
+      "loss": 0.7694,
+      "step": 14850
+    },
+    {
+      "epoch": 0.4536573452191965,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.480463008329251e-06,
+      "loss": 0.8245,
+      "step": 14860
+    },
+    {
+      "epoch": 0.4539626328000977,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.477400783929446e-06,
+      "loss": 0.8021,
+      "step": 14870
+    },
+    {
+      "epoch": 0.4542679203809989,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.474338559529643e-06,
+      "loss": 0.7988,
+      "step": 14880
+    },
+    {
+      "epoch": 0.4545732079619001,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.471276335129839e-06,
+      "loss": 0.7907,
+      "step": 14890
+    },
+    {
+      "epoch": 0.4548784955428013,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.468214110730035e-06,
+      "loss": 0.7741,
+      "step": 14900
+    },
+    {
+      "epoch": 0.4551837831237025,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.465151886330231e-06,
+      "loss": 0.7593,
+      "step": 14910
+    },
+    {
+      "epoch": 0.45548907070460376,
+      "grad_norm": 3.25,
+      "learning_rate": 5.462089661930426e-06,
+      "loss": 0.7845,
+      "step": 14920
+    },
+    {
+      "epoch": 0.45579435828550496,
+      "grad_norm": 3.25,
+      "learning_rate": 5.459027437530623e-06,
+      "loss": 0.7839,
+      "step": 14930
+    },
+    {
+      "epoch": 0.45609964586640617,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.455965213130818e-06,
+      "loss": 0.7793,
+      "step": 14940
+    },
+    {
+      "epoch": 0.45640493344730737,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.452902988731015e-06,
+      "loss": 0.7556,
+      "step": 14950
+    },
+    {
+      "epoch": 0.4567102210282086,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.44984076433121e-06,
+      "loss": 0.8027,
+      "step": 14960
+    },
+    {
+      "epoch": 0.4570155086091098,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.446778539931406e-06,
+      "loss": 0.8338,
+      "step": 14970
+    },
+    {
+      "epoch": 0.457320796190011,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.443716315531603e-06,
+      "loss": 0.7339,
+      "step": 14980
+    },
+    {
+      "epoch": 0.4576260837709122,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.440654091131798e-06,
+      "loss": 0.8288,
+      "step": 14990
+    },
+    {
+      "epoch": 0.4579313713518134,
+      "grad_norm": 3.375,
+      "learning_rate": 5.437591866731995e-06,
+      "loss": 0.8365,
+      "step": 15000
+    },
+    {
+      "epoch": 0.4582366589327146,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.43452964233219e-06,
+      "loss": 0.8167,
+      "step": 15010
+    },
+    {
+      "epoch": 0.45854194651361585,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.431467417932387e-06,
+      "loss": 0.8264,
+      "step": 15020
+    },
+    {
+      "epoch": 0.45884723409451705,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.428405193532582e-06,
+      "loss": 0.7945,
+      "step": 15030
+    },
+    {
+      "epoch": 0.45915252167541826,
+      "grad_norm": 3.75,
+      "learning_rate": 5.425342969132778e-06,
+      "loss": 0.7567,
+      "step": 15040
+    },
+    {
+      "epoch": 0.45945780925631946,
+      "grad_norm": 3.75,
+      "learning_rate": 5.422280744732974e-06,
+      "loss": 0.7768,
+      "step": 15050
+    },
+    {
+      "epoch": 0.45976309683722066,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.41921852033317e-06,
+      "loss": 0.8452,
+      "step": 15060
+    },
+    {
+      "epoch": 0.46006838441812187,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.416156295933367e-06,
+      "loss": 0.7802,
+      "step": 15070
+    },
+    {
+      "epoch": 0.46037367199902307,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.413094071533562e-06,
+      "loss": 0.8327,
+      "step": 15080
+    },
+    {
+      "epoch": 0.4606789595799243,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.410031847133759e-06,
+      "loss": 0.8878,
+      "step": 15090
+    },
+    {
+      "epoch": 0.4609842471608255,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.406969622733954e-06,
+      "loss": 0.8189,
+      "step": 15100
+    },
+    {
+      "epoch": 0.4612895347417267,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.40390739833415e-06,
+      "loss": 0.7732,
+      "step": 15110
+    },
+    {
+      "epoch": 0.46159482232262794,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.400845173934346e-06,
+      "loss": 0.8203,
+      "step": 15120
+    },
+    {
+      "epoch": 0.46190010990352914,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.397782949534542e-06,
+      "loss": 0.8396,
+      "step": 15130
+    },
+    {
+      "epoch": 0.46220539748443035,
+      "grad_norm": 3.375,
+      "learning_rate": 5.394720725134737e-06,
+      "loss": 0.7733,
+      "step": 15140
+    },
+    {
+      "epoch": 0.46251068506533155,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.391658500734934e-06,
+      "loss": 0.8431,
+      "step": 15150
+    },
+    {
+      "epoch": 0.46281597264623275,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.388596276335131e-06,
+      "loss": 0.8433,
+      "step": 15160
+    },
+    {
+      "epoch": 0.46312126022713396,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.385534051935326e-06,
+      "loss": 0.8089,
+      "step": 15170
+    },
+    {
+      "epoch": 0.46342654780803516,
+      "grad_norm": 3.25,
+      "learning_rate": 5.382471827535522e-06,
+      "loss": 0.8317,
+      "step": 15180
+    },
+    {
+      "epoch": 0.46373183538893636,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.379409603135718e-06,
+      "loss": 0.7481,
+      "step": 15190
+    },
+    {
+      "epoch": 0.46403712296983757,
+      "grad_norm": 3.75,
+      "learning_rate": 5.376347378735914e-06,
+      "loss": 0.8199,
+      "step": 15200
+    },
+    {
+      "epoch": 0.46434241055073877,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.3732851543361095e-06,
+      "loss": 0.8184,
+      "step": 15210
+    },
+    {
+      "epoch": 0.46464769813164003,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.370222929936306e-06,
+      "loss": 0.8235,
+      "step": 15220
+    },
+    {
+      "epoch": 0.46495298571254123,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.367160705536502e-06,
+      "loss": 0.7822,
+      "step": 15230
+    },
+    {
+      "epoch": 0.46525827329344244,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.364098481136698e-06,
+      "loss": 0.8034,
+      "step": 15240
+    },
+    {
+      "epoch": 0.46556356087434364,
+      "grad_norm": 4.375,
+      "learning_rate": 5.361036256736894e-06,
+      "loss": 0.7588,
+      "step": 15250
+    },
+    {
+      "epoch": 0.46586884845524484,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.3579740323370895e-06,
+      "loss": 0.815,
+      "step": 15260
+    },
+    {
+      "epoch": 0.46617413603614605,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.354911807937286e-06,
+      "loss": 0.8501,
+      "step": 15270
+    },
+    {
+      "epoch": 0.46647942361704725,
+      "grad_norm": 3.125,
+      "learning_rate": 5.3518495835374815e-06,
+      "loss": 0.8162,
+      "step": 15280
+    },
+    {
+      "epoch": 0.46678471119794845,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.348787359137678e-06,
+      "loss": 0.7943,
+      "step": 15290
+    },
+    {
+      "epoch": 0.46708999877884966,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.3457251347378735e-06,
+      "loss": 0.7674,
+      "step": 15300
+    },
+    {
+      "epoch": 0.46739528635975086,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.34266291033807e-06,
+      "loss": 0.8296,
+      "step": 15310
+    },
+    {
+      "epoch": 0.4677005739406521,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.339600685938266e-06,
+      "loss": 0.8242,
+      "step": 15320
+    },
+    {
+      "epoch": 0.4680058615215533,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.3365384615384615e-06,
+      "loss": 0.7928,
+      "step": 15330
+    },
+    {
+      "epoch": 0.4683111491024545,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.333476237138658e-06,
+      "loss": 0.8642,
+      "step": 15340
+    },
+    {
+      "epoch": 0.46861643668335573,
+      "grad_norm": 3.25,
+      "learning_rate": 5.3304140127388535e-06,
+      "loss": 0.8253,
+      "step": 15350
+    },
+    {
+      "epoch": 0.46892172426425693,
+      "grad_norm": 3.5,
+      "learning_rate": 5.32735178833905e-06,
+      "loss": 0.7704,
+      "step": 15360
+    },
+    {
+      "epoch": 0.46922701184515814,
+      "grad_norm": 3.5,
+      "learning_rate": 5.3242895639392455e-06,
+      "loss": 0.8254,
+      "step": 15370
+    },
+    {
+      "epoch": 0.46953229942605934,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.321227339539442e-06,
+      "loss": 0.7333,
+      "step": 15380
+    },
+    {
+      "epoch": 0.46983758700696054,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.3181651151396376e-06,
+      "loss": 0.757,
+      "step": 15390
+    },
+    {
+      "epoch": 0.47014287458786175,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.3151028907398336e-06,
+      "loss": 0.7607,
+      "step": 15400
+    },
+    {
+      "epoch": 0.47044816216876295,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.31204066634003e-06,
+      "loss": 0.7941,
+      "step": 15410
+    },
+    {
+      "epoch": 0.4707534497496642,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.3089784419402256e-06,
+      "loss": 0.7695,
+      "step": 15420
+    },
+    {
+      "epoch": 0.4710587373305654,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.3059162175404224e-06,
+      "loss": 0.7948,
+      "step": 15430
+    },
+    {
+      "epoch": 0.4713640249114666,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.302853993140618e-06,
+      "loss": 0.8433,
+      "step": 15440
+    },
+    {
+      "epoch": 0.4716693124923678,
+      "grad_norm": 3.25,
+      "learning_rate": 5.2997917687408144e-06,
+      "loss": 0.8227,
+      "step": 15450
+    },
+    {
+      "epoch": 0.471974600073269,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.29672954434101e-06,
+      "loss": 0.8108,
+      "step": 15460
+    },
+    {
+      "epoch": 0.4722798876541702,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.293667319941206e-06,
+      "loss": 0.766,
+      "step": 15470
+    },
+    {
+      "epoch": 0.47258517523507143,
+      "grad_norm": 3.140625,
+      "learning_rate": 5.290605095541402e-06,
+      "loss": 0.8041,
+      "step": 15480
+    },
+    {
+      "epoch": 0.47289046281597263,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.287542871141598e-06,
+      "loss": 0.7894,
+      "step": 15490
+    },
+    {
+      "epoch": 0.47319575039687384,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.2844806467417945e-06,
+      "loss": 0.8022,
+      "step": 15500
+    },
+    {
+      "epoch": 0.47350103797777504,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.28141842234199e-06,
+      "loss": 0.7405,
+      "step": 15510
+    },
+    {
+      "epoch": 0.4738063255586763,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.278356197942186e-06,
+      "loss": 0.8118,
+      "step": 15520
+    },
+    {
+      "epoch": 0.4741116131395775,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.275293973542382e-06,
+      "loss": 0.8178,
+      "step": 15530
+    },
+    {
+      "epoch": 0.4744169007204787,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.272231749142578e-06,
+      "loss": 0.8575,
+      "step": 15540
+    },
+    {
+      "epoch": 0.4747221883013799,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.269169524742773e-06,
+      "loss": 0.8526,
+      "step": 15550
+    },
+    {
+      "epoch": 0.4750274758822811,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.26610730034297e-06,
+      "loss": 0.8072,
+      "step": 15560
+    },
+    {
+      "epoch": 0.4753327634631823,
+      "grad_norm": 3.25,
+      "learning_rate": 5.263045075943165e-06,
+      "loss": 0.788,
+      "step": 15570
+    },
+    {
+      "epoch": 0.4756380510440835,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.259982851543362e-06,
+      "loss": 0.8561,
+      "step": 15580
+    },
+    {
+      "epoch": 0.4759433386249847,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.256920627143558e-06,
+      "loss": 0.8203,
+      "step": 15590
+    },
+    {
+      "epoch": 0.4762486262058859,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.253858402743754e-06,
+      "loss": 0.8078,
+      "step": 15600
+    },
+    {
+      "epoch": 0.47655391378678713,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.25079617834395e-06,
+      "loss": 0.7717,
+      "step": 15610
+    },
+    {
+      "epoch": 0.4768592013676884,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.247733953944145e-06,
+      "loss": 0.8255,
+      "step": 15620
+    },
+    {
+      "epoch": 0.4771644889485896,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.244671729544342e-06,
+      "loss": 0.7858,
+      "step": 15630
+    },
+    {
+      "epoch": 0.4774697765294908,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.241609505144537e-06,
+      "loss": 0.8175,
+      "step": 15640
+    },
+    {
+      "epoch": 0.477775064110392,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.238547280744734e-06,
+      "loss": 0.7723,
+      "step": 15650
+    },
+    {
+      "epoch": 0.4780803516912932,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.235485056344929e-06,
+      "loss": 0.7981,
+      "step": 15660
+    },
+    {
+      "epoch": 0.4783856392721944,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.232422831945126e-06,
+      "loss": 0.7473,
+      "step": 15670
+    },
+    {
+      "epoch": 0.4786909268530956,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.229360607545322e-06,
+      "loss": 0.8068,
+      "step": 15680
+    },
+    {
+      "epoch": 0.4789962144339968,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.226298383145517e-06,
+      "loss": 0.8156,
+      "step": 15690
+    },
+    {
+      "epoch": 0.479301502014898,
+      "grad_norm": 3.25,
+      "learning_rate": 5.223236158745714e-06,
+      "loss": 0.8262,
+      "step": 15700
+    },
+    {
+      "epoch": 0.4796067895957992,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.220173934345909e-06,
+      "loss": 0.7932,
+      "step": 15710
+    },
+    {
+      "epoch": 0.4799120771767004,
+      "grad_norm": 3.375,
+      "learning_rate": 5.217111709946106e-06,
+      "loss": 0.822,
+      "step": 15720
+    },
+    {
+      "epoch": 0.4802173647576017,
+      "grad_norm": 3.625,
+      "learning_rate": 5.214049485546301e-06,
+      "loss": 0.808,
+      "step": 15730
+    },
+    {
+      "epoch": 0.4805226523385029,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.210987261146498e-06,
+      "loss": 0.805,
+      "step": 15740
+    },
+    {
+      "epoch": 0.4808279399194041,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.207925036746693e-06,
+      "loss": 0.822,
+      "step": 15750
+    },
+    {
+      "epoch": 0.4811332275003053,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.204862812346889e-06,
+      "loss": 0.8413,
+      "step": 15760
+    },
+    {
+      "epoch": 0.4814385150812065,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.201800587947086e-06,
+      "loss": 0.7772,
+      "step": 15770
+    },
+    {
+      "epoch": 0.4817438026621077,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.198738363547281e-06,
+      "loss": 0.7844,
+      "step": 15780
+    },
+    {
+      "epoch": 0.4820490902430089,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.195676139147478e-06,
+      "loss": 0.8056,
+      "step": 15790
+    },
+    {
+      "epoch": 0.4823543778239101,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.192613914747673e-06,
+      "loss": 0.8868,
+      "step": 15800
+    },
+    {
+      "epoch": 0.4826596654048113,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.189551690347869e-06,
+      "loss": 0.75,
+      "step": 15810
+    },
+    {
+      "epoch": 0.4829649529857125,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.186489465948065e-06,
+      "loss": 0.7888,
+      "step": 15820
+    },
+    {
+      "epoch": 0.4832702405666138,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.183427241548261e-06,
+      "loss": 0.8241,
+      "step": 15830
+    },
+    {
+      "epoch": 0.483575528147515,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.180365017148456e-06,
+      "loss": 0.826,
+      "step": 15840
+    },
+    {
+      "epoch": 0.4838808157284162,
+      "grad_norm": 3.5,
+      "learning_rate": 5.177302792748653e-06,
+      "loss": 0.8058,
+      "step": 15850
+    },
+    {
+      "epoch": 0.4841861033093174,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.17424056834885e-06,
+      "loss": 0.8281,
+      "step": 15860
+    },
+    {
+      "epoch": 0.4844913908902186,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.171178343949045e-06,
+      "loss": 0.8034,
+      "step": 15870
+    },
+    {
+      "epoch": 0.4847966784711198,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.168116119549241e-06,
+      "loss": 0.8211,
+      "step": 15880
+    },
+    {
+      "epoch": 0.485101966052021,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.165053895149437e-06,
+      "loss": 0.8021,
+      "step": 15890
+    },
+    {
+      "epoch": 0.4854072536329222,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.161991670749633e-06,
+      "loss": 0.8239,
+      "step": 15900
+    },
+    {
+      "epoch": 0.4857125412138234,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.158929446349828e-06,
+      "loss": 0.8075,
+      "step": 15910
+    },
+    {
+      "epoch": 0.4860178287947246,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.155867221950025e-06,
+      "loss": 0.7938,
+      "step": 15920
+    },
+    {
+      "epoch": 0.48632311637562586,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.152804997550222e-06,
+      "loss": 0.8619,
+      "step": 15930
+    },
+    {
+      "epoch": 0.48662840395652707,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.149742773150417e-06,
+      "loss": 0.8155,
+      "step": 15940
+    },
+    {
+      "epoch": 0.48693369153742827,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.146680548750613e-06,
+      "loss": 0.8542,
+      "step": 15950
+    },
+    {
+      "epoch": 0.4872389791183295,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.143618324350809e-06,
+      "loss": 0.8466,
+      "step": 15960
+    },
+    {
+      "epoch": 0.4875442666992307,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.140556099951005e-06,
+      "loss": 0.8071,
+      "step": 15970
+    },
+    {
+      "epoch": 0.4878495542801319,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.1374938755512e-06,
+      "loss": 0.9111,
+      "step": 15980
+    },
+    {
+      "epoch": 0.4881548418610331,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.134431651151397e-06,
+      "loss": 0.8415,
+      "step": 15990
+    },
+    {
+      "epoch": 0.4884601294419343,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.131369426751592e-06,
+      "loss": 0.7873,
+      "step": 16000
+    },
+    {
+      "epoch": 0.4887654170228355,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.128307202351789e-06,
+      "loss": 0.8605,
+      "step": 16010
+    },
+    {
+      "epoch": 0.4890707046037367,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.125244977951985e-06,
+      "loss": 0.8486,
+      "step": 16020
+    },
+    {
+      "epoch": 0.48937599218463795,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.122182753552181e-06,
+      "loss": 0.7967,
+      "step": 16030
+    },
+    {
+      "epoch": 0.48968127976553916,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.119120529152377e-06,
+      "loss": 0.7694,
+      "step": 16040
+    },
+    {
+      "epoch": 0.48998656734644036,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.116058304752572e-06,
+      "loss": 0.7692,
+      "step": 16050
+    },
+    {
+      "epoch": 0.49029185492734156,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.112996080352769e-06,
+      "loss": 0.8435,
+      "step": 16060
+    },
+    {
+      "epoch": 0.49059714250824277,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.109933855952964e-06,
+      "loss": 0.7675,
+      "step": 16070
+    },
+    {
+      "epoch": 0.49090243008914397,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.106871631553161e-06,
+      "loss": 0.8011,
+      "step": 16080
+    },
+    {
+      "epoch": 0.4912077176700452,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.103809407153356e-06,
+      "loss": 0.838,
+      "step": 16090
+    },
+    {
+      "epoch": 0.4915130052509464,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.100747182753552e-06,
+      "loss": 0.7757,
+      "step": 16100
+    },
+    {
+      "epoch": 0.4918182928318476,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.097684958353749e-06,
+      "loss": 0.7805,
+      "step": 16110
+    },
+    {
+      "epoch": 0.4921235804127488,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.094622733953944e-06,
+      "loss": 0.8423,
+      "step": 16120
+    },
+    {
+      "epoch": 0.49242886799365004,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.091560509554141e-06,
+      "loss": 0.8422,
+      "step": 16130
+    },
+    {
+      "epoch": 0.49273415557455125,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.088498285154336e-06,
+      "loss": 0.8251,
+      "step": 16140
+    },
+    {
+      "epoch": 0.49303944315545245,
+      "grad_norm": 3.875,
+      "learning_rate": 5.085436060754533e-06,
+      "loss": 0.8103,
+      "step": 16150
+    },
+    {
+      "epoch": 0.49334473073635365,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.082373836354728e-06,
+      "loss": 0.766,
+      "step": 16160
+    },
+    {
+      "epoch": 0.49365001831725486,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.079311611954924e-06,
+      "loss": 0.7794,
+      "step": 16170
+    },
+    {
+      "epoch": 0.49395530589815606,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.07624938755512e-06,
+      "loss": 0.8016,
+      "step": 16180
+    },
+    {
+      "epoch": 0.49426059347905726,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.073187163155316e-06,
+      "loss": 0.7914,
+      "step": 16190
+    },
+    {
+      "epoch": 0.49456588105995847,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.070124938755513e-06,
+      "loss": 0.8093,
+      "step": 16200
+    },
+    {
+      "epoch": 0.49487116864085967,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.067062714355708e-06,
+      "loss": 0.8191,
+      "step": 16210
+    },
+    {
+      "epoch": 0.4951764562217609,
+      "grad_norm": 3.5,
+      "learning_rate": 5.064000489955905e-06,
+      "loss": 0.8108,
+      "step": 16220
+    },
+    {
+      "epoch": 0.49548174380266213,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.0609382655561e-06,
+      "loss": 0.7865,
+      "step": 16230
+    },
+    {
+      "epoch": 0.49578703138356334,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.057876041156296e-06,
+      "loss": 0.8146,
+      "step": 16240
+    },
+    {
+      "epoch": 0.49609231896446454,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.054813816756492e-06,
+      "loss": 0.7888,
+      "step": 16250
+    },
+    {
+      "epoch": 0.49639760654536574,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.051751592356688e-06,
+      "loss": 0.8306,
+      "step": 16260
+    },
+    {
+      "epoch": 0.49670289412626695,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.0486893679568835e-06,
+      "loss": 0.7989,
+      "step": 16270
+    },
+    {
+      "epoch": 0.49700818170716815,
+      "grad_norm": 3.453125,
+      "learning_rate": 5.04562714355708e-06,
+      "loss": 0.8276,
+      "step": 16280
+    },
+    {
+      "epoch": 0.49731346928806935,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.042564919157276e-06,
+      "loss": 0.8109,
+      "step": 16290
+    },
+    {
+      "epoch": 0.49761875686897056,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.039502694757472e-06,
+      "loss": 0.8103,
+      "step": 16300
+    },
+    {
+      "epoch": 0.49792404444987176,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.036440470357668e-06,
+      "loss": 0.7519,
+      "step": 16310
+    },
+    {
+      "epoch": 0.49822933203077296,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.033378245957864e-06,
+      "loss": 0.7668,
+      "step": 16320
+    },
+    {
+      "epoch": 0.4985346196116742,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.03031602155806e-06,
+      "loss": 0.8585,
+      "step": 16330
+    },
+    {
+      "epoch": 0.4988399071925754,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.0272537971582555e-06,
+      "loss": 0.7893,
+      "step": 16340
+    },
+    {
+      "epoch": 0.49914519477347663,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.024191572758452e-06,
+      "loss": 0.7783,
+      "step": 16350
+    },
+    {
+      "epoch": 0.49945048235437783,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.0211293483586475e-06,
+      "loss": 0.7861,
+      "step": 16360
+    },
+    {
+      "epoch": 0.49975576993527904,
+      "grad_norm": 3.375,
+      "learning_rate": 5.018067123958844e-06,
+      "loss": 0.8385,
+      "step": 16370
+    },
+    {
+      "epoch": 0.5000610575161802,
+      "grad_norm": 4.125,
+      "learning_rate": 5.01500489955904e-06,
+      "loss": 0.7972,
+      "step": 16380
+    },
+    {
+      "epoch": 0.5003663450970814,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.0119426751592356e-06,
+      "loss": 0.8189,
+      "step": 16390
+    },
+    {
+      "epoch": 0.5006716326779826,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.008880450759432e-06,
+      "loss": 0.8147,
+      "step": 16400
+    },
+    {
+      "epoch": 0.5009769202588839,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.0058182263596276e-06,
+      "loss": 0.8237,
+      "step": 16410
+    },
+    {
+      "epoch": 0.501282207839785,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.0027560019598244e-06,
+      "loss": 0.8573,
+      "step": 16420
+    },
+    {
+      "epoch": 0.5015874954206863,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.9996937775600204e-06,
+      "loss": 0.7565,
+      "step": 16430
+    },
+    {
+      "epoch": 0.5018927830015875,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.9966315531602164e-06,
+      "loss": 0.8359,
+      "step": 16440
+    },
+    {
+      "epoch": 0.5021980705824887,
+      "grad_norm": 3.25,
+      "learning_rate": 4.993569328760412e-06,
+      "loss": 0.7653,
+      "step": 16450
+    },
+    {
+      "epoch": 0.5025033581633899,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.990507104360608e-06,
+      "loss": 0.8125,
+      "step": 16460
+    },
+    {
+      "epoch": 0.5028086457442911,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.987444879960804e-06,
+      "loss": 0.7783,
+      "step": 16470
+    },
+    {
+      "epoch": 0.5031139333251924,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.984382655561e-06,
+      "loss": 0.8092,
+      "step": 16480
+    },
+    {
+      "epoch": 0.5034192209060936,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.981320431161196e-06,
+      "loss": 0.7971,
+      "step": 16490
+    },
+    {
+      "epoch": 0.5037245084869948,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.9782582067613925e-06,
+      "loss": 0.8069,
+      "step": 16500
+    },
+    {
+      "epoch": 0.504029796067896,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.9751959823615885e-06,
+      "loss": 0.7935,
+      "step": 16510
+    },
+    {
+      "epoch": 0.5043350836487972,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.972133757961784e-06,
+      "loss": 0.8356,
+      "step": 16520
+    },
+    {
+      "epoch": 0.5046403712296984,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.96907153356198e-06,
+      "loss": 0.8137,
+      "step": 16530
+    },
+    {
+      "epoch": 0.5049456588105996,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.966009309162176e-06,
+      "loss": 0.8367,
+      "step": 16540
+    },
+    {
+      "epoch": 0.5052509463915008,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.962947084762372e-06,
+      "loss": 0.7956,
+      "step": 16550
+    },
+    {
+      "epoch": 0.505556233972402,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.959884860362568e-06,
+      "loss": 0.8704,
+      "step": 16560
+    },
+    {
+      "epoch": 0.5058615215533032,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.956822635962764e-06,
+      "loss": 0.8107,
+      "step": 16570
+    },
+    {
+      "epoch": 0.5061668091342044,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.95376041156296e-06,
+      "loss": 0.7869,
+      "step": 16580
+    },
+    {
+      "epoch": 0.5064720967151056,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.950698187163156e-06,
+      "loss": 0.8104,
+      "step": 16590
+    },
+    {
+      "epoch": 0.5067773842960068,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.947635962763352e-06,
+      "loss": 0.8443,
+      "step": 16600
+    },
+    {
+      "epoch": 0.507082671876908,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.944573738363548e-06,
+      "loss": 0.8053,
+      "step": 16610
+    },
+    {
+      "epoch": 0.5073879594578092,
+      "grad_norm": 3.125,
+      "learning_rate": 4.941511513963744e-06,
+      "loss": 0.8275,
+      "step": 16620
+    },
+    {
+      "epoch": 0.5076932470387104,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.93844928956394e-06,
+      "loss": 0.7848,
+      "step": 16630
+    },
+    {
+      "epoch": 0.5079985346196116,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.935387065164136e-06,
+      "loss": 0.8199,
+      "step": 16640
+    },
+    {
+      "epoch": 0.5083038222005128,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.932324840764332e-06,
+      "loss": 0.7888,
+      "step": 16650
+    },
+    {
+      "epoch": 0.508609109781414,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.929262616364528e-06,
+      "loss": 0.8064,
+      "step": 16660
+    },
+    {
+      "epoch": 0.5089143973623153,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.926200391964724e-06,
+      "loss": 0.7564,
+      "step": 16670
+    },
+    {
+      "epoch": 0.5092196849432165,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.92313816756492e-06,
+      "loss": 0.8545,
+      "step": 16680
+    },
+    {
+      "epoch": 0.5095249725241178,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.920075943165116e-06,
+      "loss": 0.7723,
+      "step": 16690
+    },
+    {
+      "epoch": 0.509830260105019,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.917013718765312e-06,
+      "loss": 0.7822,
+      "step": 16700
+    },
+    {
+      "epoch": 0.5101355476859202,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.913951494365508e-06,
+      "loss": 0.7904,
+      "step": 16710
+    },
+    {
+      "epoch": 0.5104408352668214,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.910889269965704e-06,
+      "loss": 0.7832,
+      "step": 16720
+    },
+    {
+      "epoch": 0.5107461228477226,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.9078270455659e-06,
+      "loss": 0.7702,
+      "step": 16730
+    },
+    {
+      "epoch": 0.5110514104286238,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.904764821166095e-06,
+      "loss": 0.7593,
+      "step": 16740
+    },
+    {
+      "epoch": 0.511356698009525,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.901702596766291e-06,
+      "loss": 0.7595,
+      "step": 16750
+    },
+    {
+      "epoch": 0.5116619855904262,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.898640372366487e-06,
+      "loss": 0.7978,
+      "step": 16760
+    },
+    {
+      "epoch": 0.5119672731713274,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.895578147966684e-06,
+      "loss": 0.7778,
+      "step": 16770
+    },
+    {
+      "epoch": 0.5122725607522286,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.89251592356688e-06,
+      "loss": 0.8443,
+      "step": 16780
+    },
+    {
+      "epoch": 0.5125778483331298,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.889453699167076e-06,
+      "loss": 0.7967,
+      "step": 16790
+    },
+    {
+      "epoch": 0.512883135914031,
+      "grad_norm": 3.5,
+      "learning_rate": 4.886391474767272e-06,
+      "loss": 0.8123,
+      "step": 16800
+    },
+    {
+      "epoch": 0.5131884234949322,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.883329250367467e-06,
+      "loss": 0.7451,
+      "step": 16810
+    },
+    {
+      "epoch": 0.5134937110758334,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.880267025967663e-06,
+      "loss": 0.821,
+      "step": 16820
+    },
+    {
+      "epoch": 0.5137989986567346,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.877204801567859e-06,
+      "loss": 0.7896,
+      "step": 16830
+    },
+    {
+      "epoch": 0.5141042862376358,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.874142577168055e-06,
+      "loss": 0.8424,
+      "step": 16840
+    },
+    {
+      "epoch": 0.514409573818537,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.871080352768252e-06,
+      "loss": 0.8024,
+      "step": 16850
+    },
+    {
+      "epoch": 0.5147148613994382,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.868018128368448e-06,
+      "loss": 0.8015,
+      "step": 16860
+    },
+    {
+      "epoch": 0.5150201489803394,
+      "grad_norm": 3.25,
+      "learning_rate": 4.864955903968643e-06,
+      "loss": 0.7463,
+      "step": 16870
+    },
+    {
+      "epoch": 0.5153254365612406,
+      "grad_norm": 3.375,
+      "learning_rate": 4.861893679568839e-06,
+      "loss": 0.8586,
+      "step": 16880
+    },
+    {
+      "epoch": 0.515630724142142,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.858831455169035e-06,
+      "loss": 0.767,
+      "step": 16890
+    },
+    {
+      "epoch": 0.5159360117230432,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.855769230769231e-06,
+      "loss": 0.7769,
+      "step": 16900
+    },
+    {
+      "epoch": 0.5162412993039444,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.852707006369427e-06,
+      "loss": 0.8051,
+      "step": 16910
+    },
+    {
+      "epoch": 0.5165465868848456,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.849644781969623e-06,
+      "loss": 0.7816,
+      "step": 16920
+    },
+    {
+      "epoch": 0.5168518744657468,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.846582557569819e-06,
+      "loss": 0.8317,
+      "step": 16930
+    },
+    {
+      "epoch": 0.517157162046648,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.843520333170015e-06,
+      "loss": 0.7953,
+      "step": 16940
+    },
+    {
+      "epoch": 0.5174624496275492,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.840458108770211e-06,
+      "loss": 0.7776,
+      "step": 16950
+    },
+    {
+      "epoch": 0.5177677372084504,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.837395884370407e-06,
+      "loss": 0.793,
+      "step": 16960
+    },
+    {
+      "epoch": 0.5180730247893516,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.834333659970603e-06,
+      "loss": 0.8256,
+      "step": 16970
+    },
+    {
+      "epoch": 0.5183783123702528,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.831271435570799e-06,
+      "loss": 0.7701,
+      "step": 16980
+    },
+    {
+      "epoch": 0.518683599951154,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.828209211170995e-06,
+      "loss": 0.7368,
+      "step": 16990
+    },
+    {
+      "epoch": 0.5189888875320552,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.825146986771191e-06,
+      "loss": 0.8092,
+      "step": 17000
+    },
+    {
+      "epoch": 0.5192941751129564,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.822084762371387e-06,
+      "loss": 0.8413,
+      "step": 17010
+    },
+    {
+      "epoch": 0.5195994626938576,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.819022537971583e-06,
+      "loss": 0.865,
+      "step": 17020
+    },
+    {
+      "epoch": 0.5199047502747588,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.815960313571779e-06,
+      "loss": 0.8556,
+      "step": 17030
+    },
+    {
+      "epoch": 0.52021003785566,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.812898089171975e-06,
+      "loss": 0.7862,
+      "step": 17040
+    },
+    {
+      "epoch": 0.5205153254365612,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.809835864772171e-06,
+      "loss": 0.7834,
+      "step": 17050
+    },
+    {
+      "epoch": 0.5208206130174624,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.806773640372367e-06,
+      "loss": 0.7869,
+      "step": 17060
+    },
+    {
+      "epoch": 0.5211259005983636,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.803711415972563e-06,
+      "loss": 0.7864,
+      "step": 17070
+    },
+    {
+      "epoch": 0.5214311881792648,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.800649191572759e-06,
+      "loss": 0.8154,
+      "step": 17080
+    },
+    {
+      "epoch": 0.5217364757601661,
+      "grad_norm": 3.25,
+      "learning_rate": 4.797586967172955e-06,
+      "loss": 0.8159,
+      "step": 17090
+    },
+    {
+      "epoch": 0.5220417633410673,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.79452474277315e-06,
+      "loss": 0.8058,
+      "step": 17100
+    },
+    {
+      "epoch": 0.5223470509219685,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.791462518373346e-06,
+      "loss": 0.797,
+      "step": 17110
+    },
+    {
+      "epoch": 0.5226523385028697,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.788400293973543e-06,
+      "loss": 0.8013,
+      "step": 17120
+    },
+    {
+      "epoch": 0.5229576260837709,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.785338069573739e-06,
+      "loss": 0.8494,
+      "step": 17130
+    },
+    {
+      "epoch": 0.5232629136646721,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.782275845173935e-06,
+      "loss": 0.777,
+      "step": 17140
+    },
+    {
+      "epoch": 0.5235682012455734,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.779213620774131e-06,
+      "loss": 0.8104,
+      "step": 17150
+    },
+    {
+      "epoch": 0.5238734888264746,
+      "grad_norm": 3.375,
+      "learning_rate": 4.776151396374326e-06,
+      "loss": 0.8522,
+      "step": 17160
+    },
+    {
+      "epoch": 0.5241787764073758,
+      "grad_norm": 4.0,
+      "learning_rate": 4.773089171974522e-06,
+      "loss": 0.8592,
+      "step": 17170
+    },
+    {
+      "epoch": 0.524484063988277,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.770026947574718e-06,
+      "loss": 0.8511,
+      "step": 17180
+    },
+    {
+      "epoch": 0.5247893515691782,
+      "grad_norm": 3.0,
+      "learning_rate": 4.766964723174914e-06,
+      "loss": 0.7891,
+      "step": 17190
+    },
+    {
+      "epoch": 0.5250946391500794,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.763902498775111e-06,
+      "loss": 0.7536,
+      "step": 17200
+    },
+    {
+      "epoch": 0.5253999267309806,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.760840274375307e-06,
+      "loss": 0.7374,
+      "step": 17210
+    },
+    {
+      "epoch": 0.5257052143118818,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.757778049975502e-06,
+      "loss": 0.8073,
+      "step": 17220
+    },
+    {
+      "epoch": 0.526010501892783,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.754715825575698e-06,
+      "loss": 0.7773,
+      "step": 17230
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.751653601175894e-06,
+      "loss": 0.8339,
+      "step": 17240
+    },
+    {
+      "epoch": 0.5266210770545854,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.74859137677609e-06,
+      "loss": 0.7578,
+      "step": 17250
+    },
+    {
+      "epoch": 0.5269263646354866,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.745529152376286e-06,
+      "loss": 0.7902,
+      "step": 17260
+    },
+    {
+      "epoch": 0.5272316522163878,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.742466927976482e-06,
+      "loss": 0.8174,
+      "step": 17270
+    },
+    {
+      "epoch": 0.527536939797289,
+      "grad_norm": 3.375,
+      "learning_rate": 4.739404703576678e-06,
+      "loss": 0.8432,
+      "step": 17280
+    },
+    {
+      "epoch": 0.5278422273781903,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.736342479176874e-06,
+      "loss": 0.8362,
+      "step": 17290
+    },
+    {
+      "epoch": 0.5281475149590915,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.73328025477707e-06,
+      "loss": 0.8299,
+      "step": 17300
+    },
+    {
+      "epoch": 0.5284528025399927,
+      "grad_norm": 3.125,
+      "learning_rate": 4.730218030377266e-06,
+      "loss": 0.7856,
+      "step": 17310
+    },
+    {
+      "epoch": 0.5287580901208939,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.727155805977462e-06,
+      "loss": 0.7949,
+      "step": 17320
+    },
+    {
+      "epoch": 0.5290633777017951,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.724093581577658e-06,
+      "loss": 0.8405,
+      "step": 17330
+    },
+    {
+      "epoch": 0.5293686652826963,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.721031357177854e-06,
+      "loss": 0.8187,
+      "step": 17340
+    },
+    {
+      "epoch": 0.5296739528635975,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.71796913277805e-06,
+      "loss": 0.7359,
+      "step": 17350
+    },
+    {
+      "epoch": 0.5299792404444987,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.714906908378246e-06,
+      "loss": 0.7775,
+      "step": 17360
+    },
+    {
+      "epoch": 0.5302845280253999,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.711844683978442e-06,
+      "loss": 0.8274,
+      "step": 17370
+    },
+    {
+      "epoch": 0.5305898156063011,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.708782459578638e-06,
+      "loss": 0.7326,
+      "step": 17380
+    },
+    {
+      "epoch": 0.5308951031872023,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.705720235178834e-06,
+      "loss": 0.7731,
+      "step": 17390
+    },
+    {
+      "epoch": 0.5312003907681035,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.70265801077903e-06,
+      "loss": 0.8301,
+      "step": 17400
+    },
+    {
+      "epoch": 0.5315056783490048,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.6995957863792264e-06,
+      "loss": 0.855,
+      "step": 17410
+    },
+    {
+      "epoch": 0.531810965929906,
+      "grad_norm": 3.5,
+      "learning_rate": 4.6965335619794224e-06,
+      "loss": 0.8298,
+      "step": 17420
+    },
+    {
+      "epoch": 0.5321162535108072,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.6934713375796184e-06,
+      "loss": 0.8643,
+      "step": 17430
+    },
+    {
+      "epoch": 0.5324215410917084,
+      "grad_norm": 3.25,
+      "learning_rate": 4.6904091131798144e-06,
+      "loss": 0.8577,
+      "step": 17440
+    },
+    {
+      "epoch": 0.5327268286726096,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.68734688878001e-06,
+      "loss": 0.8181,
+      "step": 17450
+    },
+    {
+      "epoch": 0.5330321162535108,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.684284664380206e-06,
+      "loss": 0.8383,
+      "step": 17460
+    },
+    {
+      "epoch": 0.533337403834412,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.6812224399804025e-06,
+      "loss": 0.8018,
+      "step": 17470
+    },
+    {
+      "epoch": 0.5336426914153132,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.6781602155805985e-06,
+      "loss": 0.8813,
+      "step": 17480
+    },
+    {
+      "epoch": 0.5339479789962144,
+      "grad_norm": 3.375,
+      "learning_rate": 4.6750979911807945e-06,
+      "loss": 0.7727,
+      "step": 17490
+    },
+    {
+      "epoch": 0.5342532665771157,
+      "grad_norm": 3.75,
+      "learning_rate": 4.6720357667809905e-06,
+      "loss": 0.7393,
+      "step": 17500
+    },
+    {
+      "epoch": 0.5345585541580169,
+      "grad_norm": 3.5,
+      "learning_rate": 4.668973542381186e-06,
+      "loss": 0.8465,
+      "step": 17510
+    },
+    {
+      "epoch": 0.5348638417389181,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.665911317981382e-06,
+      "loss": 0.7744,
+      "step": 17520
+    },
+    {
+      "epoch": 0.5351691293198193,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.662849093581578e-06,
+      "loss": 0.7922,
+      "step": 17530
+    },
+    {
+      "epoch": 0.5354744169007205,
+      "grad_norm": 3.25,
+      "learning_rate": 4.659786869181774e-06,
+      "loss": 0.8045,
+      "step": 17540
+    },
+    {
+      "epoch": 0.5357797044816217,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.65672464478197e-06,
+      "loss": 0.8123,
+      "step": 17550
+    },
+    {
+      "epoch": 0.5360849920625229,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.6536624203821665e-06,
+      "loss": 0.7445,
+      "step": 17560
+    },
+    {
+      "epoch": 0.5363902796434241,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.6506001959823625e-06,
+      "loss": 0.7757,
+      "step": 17570
+    },
+    {
+      "epoch": 0.5366955672243253,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.647537971582558e-06,
+      "loss": 0.7754,
+      "step": 17580
+    },
+    {
+      "epoch": 0.5370008548052265,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.644475747182754e-06,
+      "loss": 0.7999,
+      "step": 17590
+    },
+    {
+      "epoch": 0.5373061423861277,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.64141352278295e-06,
+      "loss": 0.8045,
+      "step": 17600
+    },
+    {
+      "epoch": 0.5376114299670289,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.638351298383146e-06,
+      "loss": 0.8008,
+      "step": 17610
+    },
+    {
+      "epoch": 0.5379167175479301,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.635289073983342e-06,
+      "loss": 0.8155,
+      "step": 17620
+    },
+    {
+      "epoch": 0.5382220051288313,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.632226849583538e-06,
+      "loss": 0.8048,
+      "step": 17630
+    },
+    {
+      "epoch": 0.5385272927097325,
+      "grad_norm": 3.125,
+      "learning_rate": 4.629164625183734e-06,
+      "loss": 0.8001,
+      "step": 17640
+    },
+    {
+      "epoch": 0.5388325802906337,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.62610240078393e-06,
+      "loss": 0.8002,
+      "step": 17650
+    },
+    {
+      "epoch": 0.539137867871535,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.623040176384126e-06,
+      "loss": 0.7839,
+      "step": 17660
+    },
+    {
+      "epoch": 0.5394431554524362,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.619977951984322e-06,
+      "loss": 0.7825,
+      "step": 17670
+    },
+    {
+      "epoch": 0.5397484430333374,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.616915727584518e-06,
+      "loss": 0.8081,
+      "step": 17680
+    },
+    {
+      "epoch": 0.5400537306142386,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.613853503184714e-06,
+      "loss": 0.7514,
+      "step": 17690
+    },
+    {
+      "epoch": 0.5403590181951399,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.61079127878491e-06,
+      "loss": 0.7746,
+      "step": 17700
+    },
+    {
+      "epoch": 0.5406643057760411,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.607729054385106e-06,
+      "loss": 0.7556,
+      "step": 17710
+    },
+    {
+      "epoch": 0.5409695933569423,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.604666829985302e-06,
+      "loss": 0.7751,
+      "step": 17720
+    },
+    {
+      "epoch": 0.5412748809378435,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.601604605585498e-06,
+      "loss": 0.8678,
+      "step": 17730
+    },
+    {
+      "epoch": 0.5415801685187447,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.598542381185694e-06,
+      "loss": 0.7707,
+      "step": 17740
+    },
+    {
+      "epoch": 0.5418854560996459,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.59548015678589e-06,
+      "loss": 0.7985,
+      "step": 17750
+    },
+    {
+      "epoch": 0.5421907436805471,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.592417932386086e-06,
+      "loss": 0.79,
+      "step": 17760
+    },
+    {
+      "epoch": 0.5424960312614483,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.589355707986282e-06,
+      "loss": 0.8133,
+      "step": 17770
+    },
+    {
+      "epoch": 0.5428013188423495,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.586293483586478e-06,
+      "loss": 0.8331,
+      "step": 17780
+    },
+    {
+      "epoch": 0.5431066064232507,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.583231259186674e-06,
+      "loss": 0.769,
+      "step": 17790
+    },
+    {
+      "epoch": 0.5434118940041519,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.580169034786869e-06,
+      "loss": 0.7808,
+      "step": 17800
+    },
+    {
+      "epoch": 0.5437171815850531,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.577106810387065e-06,
+      "loss": 0.7582,
+      "step": 17810
+    },
+    {
+      "epoch": 0.5440224691659543,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.574044585987262e-06,
+      "loss": 0.7686,
+      "step": 17820
+    },
+    {
+      "epoch": 0.5443277567468555,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.570982361587458e-06,
+      "loss": 0.7842,
+      "step": 17830
+    },
+    {
+      "epoch": 0.5446330443277567,
+      "grad_norm": 3.75,
+      "learning_rate": 4.567920137187654e-06,
+      "loss": 0.7815,
+      "step": 17840
+    },
+    {
+      "epoch": 0.5449383319086579,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.56485791278785e-06,
+      "loss": 0.7862,
+      "step": 17850
+    },
+    {
+      "epoch": 0.5452436194895591,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.561795688388046e-06,
+      "loss": 0.8058,
+      "step": 17860
+    },
+    {
+      "epoch": 0.5455489070704603,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.558733463988241e-06,
+      "loss": 0.8098,
+      "step": 17870
+    },
+    {
+      "epoch": 0.5458541946513615,
+      "grad_norm": 3.75,
+      "learning_rate": 4.555671239588437e-06,
+      "loss": 0.8005,
+      "step": 17880
+    },
+    {
+      "epoch": 0.5461594822322627,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.552609015188633e-06,
+      "loss": 0.8209,
+      "step": 17890
+    },
+    {
+      "epoch": 0.546464769813164,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.549546790788829e-06,
+      "loss": 0.7862,
+      "step": 17900
+    },
+    {
+      "epoch": 0.5467700573940653,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.546484566389026e-06,
+      "loss": 0.7619,
+      "step": 17910
+    },
+    {
+      "epoch": 0.5470753449749665,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.543422341989222e-06,
+      "loss": 0.8557,
+      "step": 17920
+    },
+    {
+      "epoch": 0.5473806325558677,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.540360117589417e-06,
+      "loss": 0.8185,
+      "step": 17930
+    },
+    {
+      "epoch": 0.5476859201367689,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.537297893189613e-06,
+      "loss": 0.835,
+      "step": 17940
+    },
+    {
+      "epoch": 0.5479912077176701,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.534235668789809e-06,
+      "loss": 0.7937,
+      "step": 17950
+    },
+    {
+      "epoch": 0.5482964952985713,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.531173444390005e-06,
+      "loss": 0.7967,
+      "step": 17960
+    },
+    {
+      "epoch": 0.5486017828794725,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.528111219990201e-06,
+      "loss": 0.7819,
+      "step": 17970
+    },
+    {
+      "epoch": 0.5489070704603737,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.525048995590397e-06,
+      "loss": 0.7357,
+      "step": 17980
+    },
+    {
+      "epoch": 0.5492123580412749,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.521986771190594e-06,
+      "loss": 0.803,
+      "step": 17990
+    },
+    {
+      "epoch": 0.5495176456221761,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.518924546790789e-06,
+      "loss": 0.7713,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5498229332030773,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.515862322390985e-06,
+      "loss": 0.7718,
+      "step": 18010
+    },
+    {
+      "epoch": 0.5501282207839785,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.512800097991181e-06,
+      "loss": 0.8175,
+      "step": 18020
+    },
+    {
+      "epoch": 0.5504335083648797,
+      "grad_norm": 3.0,
+      "learning_rate": 4.509737873591377e-06,
+      "loss": 0.7526,
+      "step": 18030
+    },
+    {
+      "epoch": 0.5507387959457809,
+      "grad_norm": 3.375,
+      "learning_rate": 4.506675649191573e-06,
+      "loss": 0.7398,
+      "step": 18040
+    },
+    {
+      "epoch": 0.5510440835266821,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.503613424791769e-06,
+      "loss": 0.7746,
+      "step": 18050
+    },
+    {
+      "epoch": 0.5513493711075833,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.500551200391965e-06,
+      "loss": 0.8758,
+      "step": 18060
+    },
+    {
+      "epoch": 0.5516546586884845,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.497488975992161e-06,
+      "loss": 0.8349,
+      "step": 18070
+    },
+    {
+      "epoch": 0.5519599462693857,
+      "grad_norm": 4.125,
+      "learning_rate": 4.494426751592357e-06,
+      "loss": 0.7582,
+      "step": 18080
+    },
+    {
+      "epoch": 0.5522652338502869,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.491364527192553e-06,
+      "loss": 0.8774,
+      "step": 18090
+    },
+    {
+      "epoch": 0.5525705214311882,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.488302302792749e-06,
+      "loss": 0.8023,
+      "step": 18100
+    },
+    {
+      "epoch": 0.5528758090120894,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.485240078392945e-06,
+      "loss": 0.765,
+      "step": 18110
+    },
+    {
+      "epoch": 0.5531810965929906,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.482177853993141e-06,
+      "loss": 0.8201,
+      "step": 18120
+    },
+    {
+      "epoch": 0.5534863841738918,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.479115629593337e-06,
+      "loss": 0.8326,
+      "step": 18130
+    },
+    {
+      "epoch": 0.553791671754793,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.476053405193533e-06,
+      "loss": 0.7838,
+      "step": 18140
+    },
+    {
+      "epoch": 0.5540969593356943,
+      "grad_norm": 3.375,
+      "learning_rate": 4.472991180793729e-06,
+      "loss": 0.8406,
+      "step": 18150
+    },
+    {
+      "epoch": 0.5544022469165955,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.469928956393924e-06,
+      "loss": 0.7691,
+      "step": 18160
+    },
+    {
+      "epoch": 0.5547075344974967,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.466866731994121e-06,
+      "loss": 0.8022,
+      "step": 18170
+    },
+    {
+      "epoch": 0.5550128220783979,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.463804507594317e-06,
+      "loss": 0.8177,
+      "step": 18180
+    },
+    {
+      "epoch": 0.5553181096592991,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.460742283194513e-06,
+      "loss": 0.8007,
+      "step": 18190
+    },
+    {
+      "epoch": 0.5556233972402003,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.457680058794709e-06,
+      "loss": 0.7814,
+      "step": 18200
+    },
+    {
+      "epoch": 0.5559286848211015,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.454617834394905e-06,
+      "loss": 0.8132,
+      "step": 18210
+    },
+    {
+      "epoch": 0.5562339724020027,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.4515556099951e-06,
+      "loss": 0.7809,
+      "step": 18220
+    },
+    {
+      "epoch": 0.5565392599829039,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.448493385595296e-06,
+      "loss": 0.7844,
+      "step": 18230
+    },
+    {
+      "epoch": 0.5568445475638051,
+      "grad_norm": 3.5,
+      "learning_rate": 4.445431161195492e-06,
+      "loss": 0.7941,
+      "step": 18240
+    },
+    {
+      "epoch": 0.5571498351447063,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.442368936795688e-06,
+      "loss": 0.7751,
+      "step": 18250
+    },
+    {
+      "epoch": 0.5574551227256075,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.439306712395885e-06,
+      "loss": 0.7805,
+      "step": 18260
+    },
+    {
+      "epoch": 0.5577604103065087,
+      "grad_norm": 3.375,
+      "learning_rate": 4.436244487996081e-06,
+      "loss": 0.8182,
+      "step": 18270
+    },
+    {
+      "epoch": 0.5580656978874099,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.433182263596277e-06,
+      "loss": 0.8076,
+      "step": 18280
+    },
+    {
+      "epoch": 0.5583709854683111,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.430120039196472e-06,
+      "loss": 0.8208,
+      "step": 18290
+    },
+    {
+      "epoch": 0.5586762730492124,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.427057814796668e-06,
+      "loss": 0.7684,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5589815606301136,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.423995590396864e-06,
+      "loss": 0.815,
+      "step": 18310
+    },
+    {
+      "epoch": 0.5592868482110148,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.42093336599706e-06,
+      "loss": 0.8096,
+      "step": 18320
+    },
+    {
+      "epoch": 0.559592135791916,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.417871141597256e-06,
+      "loss": 0.8373,
+      "step": 18330
+    },
+    {
+      "epoch": 0.5598974233728172,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.414808917197453e-06,
+      "loss": 0.7671,
+      "step": 18340
+    },
+    {
+      "epoch": 0.5602027109537184,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.411746692797648e-06,
+      "loss": 0.783,
+      "step": 18350
+    },
+    {
+      "epoch": 0.5605079985346196,
+      "grad_norm": 3.25,
+      "learning_rate": 4.408684468397844e-06,
+      "loss": 0.8287,
+      "step": 18360
+    },
+    {
+      "epoch": 0.5608132861155208,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.40562224399804e-06,
+      "loss": 0.7778,
+      "step": 18370
+    },
+    {
+      "epoch": 0.561118573696422,
+      "grad_norm": 4.125,
+      "learning_rate": 4.402560019598236e-06,
+      "loss": 0.8401,
+      "step": 18380
+    },
+    {
+      "epoch": 0.5614238612773232,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.399497795198432e-06,
+      "loss": 0.8094,
+      "step": 18390
+    },
+    {
+      "epoch": 0.5617291488582244,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.3964355707986284e-06,
+      "loss": 0.7997,
+      "step": 18400
+    },
+    {
+      "epoch": 0.5620344364391257,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.3933733463988244e-06,
+      "loss": 0.8068,
+      "step": 18410
+    },
+    {
+      "epoch": 0.5623397240200269,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.3903111219990204e-06,
+      "loss": 0.8253,
+      "step": 18420
+    },
+    {
+      "epoch": 0.5626450116009281,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.3872488975992164e-06,
+      "loss": 0.8315,
+      "step": 18430
+    },
+    {
+      "epoch": 0.5629502991818293,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.3841866731994124e-06,
+      "loss": 0.8053,
+      "step": 18440
+    },
+    {
+      "epoch": 0.5632555867627305,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.3811244487996085e-06,
+      "loss": 0.8124,
+      "step": 18450
+    },
+    {
+      "epoch": 0.5635608743436317,
+      "grad_norm": 3.625,
+      "learning_rate": 4.3780622243998045e-06,
+      "loss": 0.7877,
+      "step": 18460
+    },
+    {
+      "epoch": 0.5638661619245329,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.8157,
+      "step": 18470
+    },
+    {
+      "epoch": 0.5641714495054341,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.3719377756001965e-06,
+      "loss": 0.8788,
+      "step": 18480
+    },
+    {
+      "epoch": 0.5644767370863353,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.3688755512003925e-06,
+      "loss": 0.8171,
+      "step": 18490
+    },
+    {
+      "epoch": 0.5647820246672365,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.3658133268005885e-06,
+      "loss": 0.8271,
+      "step": 18500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 32756,
+  "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": 2.1016700400278077e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}