diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14787 @@
+{
+  "best_global_step": 21000,
+  "best_metric": 0.9441189767578706,
+  "best_model_checkpoint": "./saved_best/sst2_ELECTRA-20000/full_text/checkpoint-21000",
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 21000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014285714285714286,
+      "grad_norm": 357206.34375,
+      "learning_rate": 1.9991428571428572e-05,
+      "loss": 0.6967,
+      "step": 10
+    },
+    {
+      "epoch": 0.002857142857142857,
+      "grad_norm": 273965.15625,
+      "learning_rate": 1.9981904761904764e-05,
+      "loss": 0.6837,
+      "step": 20
+    },
+    {
+      "epoch": 0.004285714285714286,
+      "grad_norm": 325222.28125,
+      "learning_rate": 1.9972380952380953e-05,
+      "loss": 0.6183,
+      "step": 30
+    },
+    {
+      "epoch": 0.005714285714285714,
+      "grad_norm": 217390.0,
+      "learning_rate": 1.9962857142857145e-05,
+      "loss": 0.6108,
+      "step": 40
+    },
+    {
+      "epoch": 0.007142857142857143,
+      "grad_norm": 304553.1875,
+      "learning_rate": 1.9953333333333334e-05,
+      "loss": 0.7513,
+      "step": 50
+    },
+    {
+      "epoch": 0.008571428571428572,
+      "grad_norm": 452970.25,
+      "learning_rate": 1.9943809523809527e-05,
+      "loss": 0.869,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 207090.15625,
+      "learning_rate": 1.9934285714285716e-05,
+      "loss": 0.7148,
+      "step": 70
+    },
+    {
+      "epoch": 0.011428571428571429,
+      "grad_norm": 480391.0625,
+      "learning_rate": 1.9924761904761908e-05,
+      "loss": 0.7425,
+      "step": 80
+    },
+    {
+      "epoch": 0.012857142857142857,
+      "grad_norm": 527771.5625,
+      "learning_rate": 1.9915238095238097e-05,
+      "loss": 0.8046,
+      "step": 90
+    },
+    {
+      "epoch": 0.014285714285714285,
+      "grad_norm": 303525.40625,
+      "learning_rate": 1.990571428571429e-05,
+      "loss": 0.6573,
+      "step": 100
+    },
+    {
+      "epoch": 0.015714285714285715,
+      "grad_norm": 290561.75,
+      "learning_rate": 1.9896190476190478e-05,
+      "loss": 0.7066,
+      "step": 110
+    },
+    {
+      "epoch": 0.017142857142857144,
+      "grad_norm": 369162.96875,
+      "learning_rate": 1.9886666666666667e-05,
+      "loss": 0.715,
+      "step": 120
+    },
+    {
+      "epoch": 0.018571428571428572,
+      "grad_norm": 350161.5,
+      "learning_rate": 1.987714285714286e-05,
+      "loss": 0.6096,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 43413988.0,
+      "learning_rate": 1.986761904761905e-05,
+      "loss": 0.599,
+      "step": 140
+    },
+    {
+      "epoch": 0.02142857142857143,
+      "grad_norm": 1415179.25,
+      "learning_rate": 1.985809523809524e-05,
+      "loss": 0.6316,
+      "step": 150
+    },
+    {
+      "epoch": 0.022857142857142857,
+      "grad_norm": 1907422.125,
+      "learning_rate": 1.984857142857143e-05,
+      "loss": 0.5664,
+      "step": 160
+    },
+    {
+      "epoch": 0.024285714285714285,
+      "grad_norm": 1027747.625,
+      "learning_rate": 1.983904761904762e-05,
+      "loss": 0.5913,
+      "step": 170
+    },
+    {
+      "epoch": 0.025714285714285714,
+      "grad_norm": 207440.515625,
+      "learning_rate": 1.982952380952381e-05,
+      "loss": 0.3931,
+      "step": 180
+    },
+    {
+      "epoch": 0.027142857142857142,
+      "grad_norm": 128810.21875,
+      "learning_rate": 1.982e-05,
+      "loss": 0.563,
+      "step": 190
+    },
+    {
+      "epoch": 0.02857142857142857,
+      "grad_norm": 101973.6328125,
+      "learning_rate": 1.9810476190476192e-05,
+      "loss": 0.7755,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 755123.5625,
+      "learning_rate": 1.980095238095238e-05,
+      "loss": 0.4284,
+      "step": 210
+    },
+    {
+      "epoch": 0.03142857142857143,
+      "grad_norm": 531576.75,
+      "learning_rate": 1.9791428571428574e-05,
+      "loss": 0.6093,
+      "step": 220
+    },
+    {
+      "epoch": 0.032857142857142856,
+      "grad_norm": 583141.6875,
+      "learning_rate": 1.9781904761904766e-05,
+      "loss": 0.2275,
+      "step": 230
+    },
+    {
+      "epoch": 0.03428571428571429,
+      "grad_norm": 8570300.0,
+      "learning_rate": 1.9772380952380955e-05,
+      "loss": 0.8347,
+      "step": 240
+    },
+    {
+      "epoch": 0.03571428571428571,
+      "grad_norm": 4085802.25,
+      "learning_rate": 1.9762857142857144e-05,
+      "loss": 0.3431,
+      "step": 250
+    },
+    {
+      "epoch": 0.037142857142857144,
+      "grad_norm": 45669.38671875,
+      "learning_rate": 1.9753333333333336e-05,
+      "loss": 0.9943,
+      "step": 260
+    },
+    {
+      "epoch": 0.03857142857142857,
+      "grad_norm": 57421.49609375,
+      "learning_rate": 1.9743809523809525e-05,
+      "loss": 0.5699,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 30245.845703125,
+      "learning_rate": 1.9734285714285717e-05,
+      "loss": 0.5177,
+      "step": 280
+    },
+    {
+      "epoch": 0.041428571428571426,
+      "grad_norm": 120243.7421875,
+      "learning_rate": 1.9724761904761906e-05,
+      "loss": 0.3983,
+      "step": 290
+    },
+    {
+      "epoch": 0.04285714285714286,
+      "grad_norm": 12103.5107421875,
+      "learning_rate": 1.9715238095238095e-05,
+      "loss": 0.6515,
+      "step": 300
+    },
+    {
+      "epoch": 0.04428571428571428,
+      "grad_norm": 28955.283203125,
+      "learning_rate": 1.9705714285714288e-05,
+      "loss": 0.9749,
+      "step": 310
+    },
+    {
+      "epoch": 0.045714285714285714,
+      "grad_norm": 7034.7275390625,
+      "learning_rate": 1.9696190476190477e-05,
+      "loss": 0.2087,
+      "step": 320
+    },
+    {
+      "epoch": 0.047142857142857146,
+      "grad_norm": 7419.80224609375,
+      "learning_rate": 1.9686666666666666e-05,
+      "loss": 0.1864,
+      "step": 330
+    },
+    {
+      "epoch": 0.04857142857142857,
+      "grad_norm": 2310242.5,
+      "learning_rate": 1.9677142857142858e-05,
+      "loss": 1.0082,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 794568.6875,
+      "learning_rate": 1.966761904761905e-05,
+      "loss": 1.9931,
+      "step": 350
+    },
+    {
+      "epoch": 0.05142857142857143,
+      "grad_norm": 1542447.25,
+      "learning_rate": 1.965809523809524e-05,
+      "loss": 1.5728,
+      "step": 360
+    },
+    {
+      "epoch": 0.05285714285714286,
+      "grad_norm": 2202.64111328125,
+      "learning_rate": 1.964857142857143e-05,
+      "loss": 0.952,
+      "step": 370
+    },
+    {
+      "epoch": 0.054285714285714284,
+      "grad_norm": 30127.619140625,
+      "learning_rate": 1.963904761904762e-05,
+      "loss": 2.6266,
+      "step": 380
+    },
+    {
+      "epoch": 0.055714285714285716,
+      "grad_norm": 1528363.625,
+      "learning_rate": 1.9629523809523813e-05,
+      "loss": 0.6687,
+      "step": 390
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 1231.0552978515625,
+      "learning_rate": 1.9620000000000002e-05,
+      "loss": 1.4041,
+      "step": 400
+    },
+    {
+      "epoch": 0.05857142857142857,
+      "grad_norm": 1543422.125,
+      "learning_rate": 1.9610476190476194e-05,
+      "loss": 0.7019,
+      "step": 410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 827856.5625,
+      "learning_rate": 1.9600952380952383e-05,
+      "loss": 1.1672,
+      "step": 420
+    },
+    {
+      "epoch": 0.06142857142857143,
+      "grad_norm": 839391.9375,
+      "learning_rate": 1.9591428571428572e-05,
+      "loss": 1.1452,
+      "step": 430
+    },
+    {
+      "epoch": 0.06285714285714286,
+      "grad_norm": 3276064.75,
+      "learning_rate": 1.9581904761904764e-05,
+      "loss": 1.3072,
+      "step": 440
+    },
+    {
+      "epoch": 0.06428571428571428,
+      "grad_norm": 192.73764038085938,
+      "learning_rate": 1.9572380952380953e-05,
+      "loss": 1.2794,
+      "step": 450
+    },
+    {
+      "epoch": 0.06571428571428571,
+      "grad_norm": 134.8483123779297,
+      "learning_rate": 1.9562857142857142e-05,
+      "loss": 2.5575,
+      "step": 460
+    },
+    {
+      "epoch": 0.06714285714285714,
+      "grad_norm": 978355.25,
+      "learning_rate": 1.9553333333333335e-05,
+      "loss": 0.9656,
+      "step": 470
+    },
+    {
+      "epoch": 0.06857142857142857,
+      "grad_norm": 78.88859558105469,
+      "learning_rate": 1.9543809523809524e-05,
+      "loss": 0.5232,
+      "step": 480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 79.68238830566406,
+      "learning_rate": 1.9534285714285716e-05,
+      "loss": 0.0001,
+      "step": 490
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 56.468196868896484,
+      "learning_rate": 1.9524761904761905e-05,
+      "loss": 2.0069,
+      "step": 500
+    },
+    {
+      "epoch": 0.07285714285714286,
+      "grad_norm": 24.5145263671875,
+      "learning_rate": 1.9515238095238097e-05,
+      "loss": 1.0956,
+      "step": 510
+    },
+    {
+      "epoch": 0.07428571428571429,
+      "grad_norm": 8801615.0,
+      "learning_rate": 1.9505714285714286e-05,
+      "loss": 2.9341,
+      "step": 520
+    },
+    {
+      "epoch": 0.07571428571428572,
+      "grad_norm": 25558882.0,
+      "learning_rate": 1.949619047619048e-05,
+      "loss": 1.5093,
+      "step": 530
+    },
+    {
+      "epoch": 0.07714285714285714,
+      "grad_norm": 3.5658111572265625,
+      "learning_rate": 1.948666666666667e-05,
+      "loss": 0.6283,
+      "step": 540
+    },
+    {
+      "epoch": 0.07857142857142857,
+      "grad_norm": 2.4725310802459717,
+      "learning_rate": 1.947714285714286e-05,
+      "loss": 3.243,
+      "step": 550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.12267804145813,
+      "learning_rate": 1.946761904761905e-05,
+      "loss": 1.4092,
+      "step": 560
+    },
+    {
+      "epoch": 0.08142857142857143,
+      "grad_norm": 8150.515625,
+      "learning_rate": 1.945809523809524e-05,
+      "loss": 1.9894,
+      "step": 570
+    },
+    {
+      "epoch": 0.08285714285714285,
+      "grad_norm": 11240.0068359375,
+      "learning_rate": 1.944857142857143e-05,
+      "loss": 1.128,
+      "step": 580
+    },
+    {
+      "epoch": 0.08428571428571428,
+      "grad_norm": 1262046.5,
+      "learning_rate": 1.943904761904762e-05,
+      "loss": 2.4431,
+      "step": 590
+    },
+    {
+      "epoch": 0.08571428571428572,
+      "grad_norm": 1295977.25,
+      "learning_rate": 1.942952380952381e-05,
+      "loss": 4.039,
+      "step": 600
+    },
+    {
+      "epoch": 0.08714285714285715,
+      "grad_norm": 0.8773088455200195,
+      "learning_rate": 1.942e-05,
+      "loss": 2.099,
+      "step": 610
+    },
+    {
+      "epoch": 0.08857142857142856,
+      "grad_norm": 1273879.5,
+      "learning_rate": 1.9410476190476193e-05,
+      "loss": 1.4771,
+      "step": 620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1020.5031127929688,
+      "learning_rate": 1.9400952380952382e-05,
+      "loss": 3.4148,
+      "step": 630
+    },
+    {
+      "epoch": 0.09142857142857143,
+      "grad_norm": 0.6120594143867493,
+      "learning_rate": 1.939142857142857e-05,
+      "loss": 3.1082,
+      "step": 640
+    },
+    {
+      "epoch": 0.09285714285714286,
+      "grad_norm": 0.3739018142223358,
+      "learning_rate": 1.9381904761904763e-05,
+      "loss": 1.5167,
+      "step": 650
+    },
+    {
+      "epoch": 0.09428571428571429,
+      "grad_norm": 1219139.375,
+      "learning_rate": 1.9372380952380952e-05,
+      "loss": 0.8132,
+      "step": 660
+    },
+    {
+      "epoch": 0.09571428571428571,
+      "grad_norm": 0.21987225115299225,
+      "learning_rate": 1.9362857142857144e-05,
+      "loss": 1.5299,
+      "step": 670
+    },
+    {
+      "epoch": 0.09714285714285714,
+      "grad_norm": 1384200.0,
+      "learning_rate": 1.9353333333333337e-05,
+      "loss": 1.5968,
+      "step": 680
+    },
+    {
+      "epoch": 0.09857142857142857,
+      "grad_norm": 0.07125857472419739,
+      "learning_rate": 1.9343809523809526e-05,
+      "loss": 1.5798,
+      "step": 690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.114037036895752,
+      "learning_rate": 1.9334285714285718e-05,
+      "loss": 2.313,
+      "step": 700
+    },
+    {
+      "epoch": 0.10142857142857142,
+      "grad_norm": 0.3785915970802307,
+      "learning_rate": 1.9324761904761907e-05,
+      "loss": 2.4355,
+      "step": 710
+    },
+    {
+      "epoch": 0.10285714285714286,
+      "grad_norm": 5890543.0,
+      "learning_rate": 1.9315238095238096e-05,
+      "loss": 1.6286,
+      "step": 720
+    },
+    {
+      "epoch": 0.10428571428571429,
+      "grad_norm": 1362617.75,
+      "learning_rate": 1.9305714285714288e-05,
+      "loss": 1.5431,
+      "step": 730
+    },
+    {
+      "epoch": 0.10571428571428572,
+      "grad_norm": 0.4030442535877228,
+      "learning_rate": 1.9296190476190477e-05,
+      "loss": 2.4325,
+      "step": 740
+    },
+    {
+      "epoch": 0.10714285714285714,
+      "grad_norm": 5006750.5,
+      "learning_rate": 1.928666666666667e-05,
+      "loss": 3.1126,
+      "step": 750
+    },
+    {
+      "epoch": 0.10857142857142857,
+      "grad_norm": 1.071424126625061,
+      "learning_rate": 1.927714285714286e-05,
+      "loss": 1.9652,
+      "step": 760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2596314549446106,
+      "learning_rate": 1.9267619047619047e-05,
+      "loss": 3.2091,
+      "step": 770
+    },
+    {
+      "epoch": 0.11142857142857143,
+      "grad_norm": 0.3768131136894226,
+      "learning_rate": 1.925809523809524e-05,
+      "loss": 3.1588,
+      "step": 780
+    },
+    {
+      "epoch": 0.11285714285714285,
+      "grad_norm": 71880208.0,
+      "learning_rate": 1.924857142857143e-05,
+      "loss": 2.2799,
+      "step": 790
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 0.3107571601867676,
+      "learning_rate": 1.923904761904762e-05,
+      "loss": 0.0,
+      "step": 800
+    },
+    {
+      "epoch": 0.11571428571428571,
+      "grad_norm": 7004024.5,
+      "learning_rate": 1.922952380952381e-05,
+      "loss": 1.5763,
+      "step": 810
+    },
+    {
+      "epoch": 0.11714285714285715,
+      "grad_norm": 0.0899968072772026,
+      "learning_rate": 1.9220000000000002e-05,
+      "loss": 2.0608,
+      "step": 820
+    },
+    {
+      "epoch": 0.11857142857142858,
+      "grad_norm": 0.3030736446380615,
+      "learning_rate": 1.921047619047619e-05,
+      "loss": 3.0442,
+      "step": 830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1347684.125,
+      "learning_rate": 1.9200952380952384e-05,
+      "loss": 1.6234,
+      "step": 840
+    },
+    {
+      "epoch": 0.12142857142857143,
+      "grad_norm": 0.05146798864006996,
+      "learning_rate": 1.9191428571428573e-05,
+      "loss": 0.8441,
+      "step": 850
+    },
+    {
+      "epoch": 0.12285714285714286,
+      "grad_norm": 0.07182995975017548,
+      "learning_rate": 1.9181904761904765e-05,
+      "loss": 0.8328,
+      "step": 860
+    },
+    {
+      "epoch": 0.12428571428571429,
+      "grad_norm": 0.5721035003662109,
+      "learning_rate": 1.9172380952380954e-05,
+      "loss": 2.369,
+      "step": 870
+    },
+    {
+      "epoch": 0.12571428571428572,
+      "grad_norm": 0.40266790986061096,
+      "learning_rate": 1.9162857142857146e-05,
+      "loss": 2.5126,
+      "step": 880
+    },
+    {
+      "epoch": 0.12714285714285714,
+      "grad_norm": 0.1481495052576065,
+      "learning_rate": 1.9153333333333335e-05,
+      "loss": 2.1925,
+      "step": 890
+    },
+    {
+      "epoch": 0.12857142857142856,
+      "grad_norm": 328981280.0,
+      "learning_rate": 1.9143809523809524e-05,
+      "loss": 1.7446,
+      "step": 900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34252291917800903,
+      "learning_rate": 1.9134285714285716e-05,
+      "loss": 0.8323,
+      "step": 910
+    },
+    {
+      "epoch": 0.13142857142857142,
+      "grad_norm": 10.440815925598145,
+      "learning_rate": 1.9124761904761905e-05,
+      "loss": 0.0012,
+      "step": 920
+    },
+    {
+      "epoch": 0.13285714285714287,
+      "grad_norm": 0.3010234534740448,
+      "learning_rate": 1.9115238095238098e-05,
+      "loss": 0.817,
+      "step": 930
+    },
+    {
+      "epoch": 0.13428571428571429,
+      "grad_norm": 1395567.125,
+      "learning_rate": 1.9105714285714287e-05,
+      "loss": 0.8341,
+      "step": 940
+    },
+    {
+      "epoch": 0.1357142857142857,
+      "grad_norm": 0.04209592193365097,
+      "learning_rate": 1.9096190476190476e-05,
+      "loss": 0.8558,
+      "step": 950
+    },
+    {
+      "epoch": 0.13714285714285715,
+      "grad_norm": 0.029842762276530266,
+      "learning_rate": 1.9086666666666668e-05,
+      "loss": 0.0,
+      "step": 960
+    },
+    {
+      "epoch": 0.13857142857142857,
+      "grad_norm": 1373328.25,
+      "learning_rate": 1.9077142857142857e-05,
+      "loss": 1.735,
+      "step": 970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.05820516124367714,
+      "learning_rate": 1.906761904761905e-05,
+      "loss": 0.0,
+      "step": 980
+    },
+    {
+      "epoch": 0.14142857142857143,
+      "grad_norm": 0.24058863520622253,
+      "learning_rate": 1.905809523809524e-05,
+      "loss": 5.055,
+      "step": 990
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.09235672652721405,
+      "learning_rate": 1.904857142857143e-05,
+      "loss": 0.7835,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1442857142857143,
+      "grad_norm": 0.32474470138549805,
+      "learning_rate": 1.9039047619047623e-05,
+      "loss": 1.5934,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1457142857142857,
+      "grad_norm": 1335350.625,
+      "learning_rate": 1.9029523809523812e-05,
+      "loss": 2.1985,
+      "step": 1020
+    },
+    {
+      "epoch": 0.14714285714285713,
+      "grad_norm": 0.04273092374205589,
+      "learning_rate": 1.902e-05,
+      "loss": 3.1129,
+      "step": 1030
+    },
+    {
+      "epoch": 0.14857142857142858,
+      "grad_norm": 0.4427473545074463,
+      "learning_rate": 1.9010476190476193e-05,
+      "loss": 3.0703,
+      "step": 1040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.747932434082031,
+      "learning_rate": 1.9000952380952382e-05,
+      "loss": 1.4166,
+      "step": 1050
+    },
+    {
+      "epoch": 0.15142857142857144,
+      "grad_norm": 0.036276426166296005,
+      "learning_rate": 1.8991428571428574e-05,
+      "loss": 0.756,
+      "step": 1060
+    },
+    {
+      "epoch": 0.15285714285714286,
+      "grad_norm": 0.35469186305999756,
+      "learning_rate": 1.8981904761904763e-05,
+      "loss": 1.4557,
+      "step": 1070
+    },
+    {
+      "epoch": 0.15428571428571428,
+      "grad_norm": 0.09299268573522568,
+      "learning_rate": 1.8972380952380952e-05,
+      "loss": 0.7547,
+      "step": 1080
+    },
+    {
+      "epoch": 0.15571428571428572,
+      "grad_norm": 0.043454334139823914,
+      "learning_rate": 1.8962857142857145e-05,
+      "loss": 0.9009,
+      "step": 1090
+    },
+    {
+      "epoch": 0.15714285714285714,
+      "grad_norm": 0.15980952978134155,
+      "learning_rate": 1.8953333333333334e-05,
+      "loss": 0.7166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.15857142857142856,
+      "grad_norm": 0.14604657888412476,
+      "learning_rate": 1.8943809523809523e-05,
+      "loss": 0.0,
+      "step": 1110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1859690.125,
+      "learning_rate": 1.8934285714285715e-05,
+      "loss": 3.2153,
+      "step": 1120
+    },
+    {
+      "epoch": 0.16142857142857142,
+      "grad_norm": 0.32800203561782837,
+      "learning_rate": 1.8924761904761907e-05,
+      "loss": 0.8639,
+      "step": 1130
+    },
+    {
+      "epoch": 0.16285714285714287,
+      "grad_norm": 3560083.25,
+      "learning_rate": 1.8915238095238096e-05,
+      "loss": 1.6759,
+      "step": 1140
+    },
+    {
+      "epoch": 0.16428571428571428,
+      "grad_norm": 0.033000826835632324,
+      "learning_rate": 1.890571428571429e-05,
+      "loss": 0.0,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1657142857142857,
+      "grad_norm": 2305505.75,
+      "learning_rate": 1.8896190476190478e-05,
+      "loss": 3.3306,
+      "step": 1160
+    },
+    {
+      "epoch": 0.16714285714285715,
+      "grad_norm": 0.4728598892688751,
+      "learning_rate": 1.888666666666667e-05,
+      "loss": 1.6841,
+      "step": 1170
+    },
+    {
+      "epoch": 0.16857142857142857,
+      "grad_norm": 0.07491559535264969,
+      "learning_rate": 1.887714285714286e-05,
+      "loss": 2.3649,
+      "step": 1180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1323724.375,
+      "learning_rate": 1.886761904761905e-05,
+      "loss": 0.7792,
+      "step": 1190
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 0.04032323881983757,
+      "learning_rate": 1.885809523809524e-05,
+      "loss": 1.7077,
+      "step": 1200
+    },
+    {
+      "epoch": 0.17285714285714285,
+      "grad_norm": 0.25903213024139404,
+      "learning_rate": 1.884857142857143e-05,
+      "loss": 2.4008,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1742857142857143,
+      "grad_norm": 0.2927476465702057,
+      "learning_rate": 1.883904761904762e-05,
+      "loss": 0.8384,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1757142857142857,
+      "grad_norm": 373211264.0,
+      "learning_rate": 1.882952380952381e-05,
+      "loss": 1.4814,
+      "step": 1230
+    },
+    {
+      "epoch": 0.17714285714285713,
+      "grad_norm": 0.14299170672893524,
+      "learning_rate": 1.882e-05,
+      "loss": 4.0988,
+      "step": 1240
+    },
+    {
+      "epoch": 0.17857142857142858,
+      "grad_norm": 0.08436620980501175,
+      "learning_rate": 1.8810476190476192e-05,
+      "loss": 0.739,
+      "step": 1250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 62.826168060302734,
+      "learning_rate": 1.880095238095238e-05,
+      "loss": 1.621,
+      "step": 1260
+    },
+    {
+      "epoch": 0.18142857142857144,
+      "grad_norm": 1.0080652236938477,
+      "learning_rate": 1.8791428571428573e-05,
+      "loss": 1.9567,
+      "step": 1270
+    },
+    {
+      "epoch": 0.18285714285714286,
+      "grad_norm": 13.952902793884277,
+      "learning_rate": 1.8781904761904762e-05,
+      "loss": 2.2569,
+      "step": 1280
+    },
+    {
+      "epoch": 0.18428571428571427,
+      "grad_norm": 0.05663137137889862,
+      "learning_rate": 1.8772380952380954e-05,
+      "loss": 1.564,
+      "step": 1290
+    },
+    {
+      "epoch": 0.18571428571428572,
+      "grad_norm": 0.0583365261554718,
+      "learning_rate": 1.8762857142857143e-05,
+      "loss": 0.7827,
+      "step": 1300
+    },
+    {
+      "epoch": 0.18714285714285714,
+      "grad_norm": 16425307.0,
+      "learning_rate": 1.8753333333333336e-05,
+      "loss": 2.3316,
+      "step": 1310
+    },
+    {
+      "epoch": 0.18857142857142858,
+      "grad_norm": 1591508.75,
+      "learning_rate": 1.8743809523809528e-05,
+      "loss": 2.4518,
+      "step": 1320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.05410398170351982,
+      "learning_rate": 1.8734285714285717e-05,
+      "loss": 2.4694,
+      "step": 1330
+    },
+    {
+      "epoch": 0.19142857142857142,
+      "grad_norm": 0.14513859152793884,
+      "learning_rate": 1.8724761904761906e-05,
+      "loss": 0.8967,
+      "step": 1340
+    },
+    {
+      "epoch": 0.19285714285714287,
+      "grad_norm": 2025066.5,
+      "learning_rate": 1.8715238095238098e-05,
+      "loss": 0.8685,
+      "step": 1350
+    },
+    {
+      "epoch": 0.19428571428571428,
+      "grad_norm": 0.07941562682390213,
+      "learning_rate": 1.8705714285714287e-05,
+      "loss": 1.5942,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1957142857142857,
+      "grad_norm": 0.1275363713502884,
+      "learning_rate": 1.8696190476190476e-05,
+      "loss": 0.8094,
+      "step": 1370
+    },
+    {
+      "epoch": 0.19714285714285715,
+      "grad_norm": 0.11744116991758347,
+      "learning_rate": 1.868666666666667e-05,
+      "loss": 4.2713,
+      "step": 1380
+    },
+    {
+      "epoch": 0.19857142857142857,
+      "grad_norm": 0.28333768248558044,
+      "learning_rate": 1.8677142857142857e-05,
+      "loss": 3.2719,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 29538128.0,
+      "learning_rate": 1.866761904761905e-05,
+      "loss": 0.6668,
+      "step": 1400
+    },
+    {
+      "epoch": 0.20142857142857143,
+      "grad_norm": 0.28693240880966187,
+      "learning_rate": 1.865809523809524e-05,
+      "loss": 0.8302,
+      "step": 1410
+    },
+    {
+      "epoch": 0.20285714285714285,
+      "grad_norm": 0.25721368193626404,
+      "learning_rate": 1.8648571428571428e-05,
+      "loss": 0.8161,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2042857142857143,
+      "grad_norm": 0.06433752924203873,
+      "learning_rate": 1.863904761904762e-05,
+      "loss": 0.8245,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2057142857142857,
+      "grad_norm": 0.06798818707466125,
+      "learning_rate": 1.8629523809523812e-05,
+      "loss": 4.2317,
+      "step": 1440
+    },
+    {
+      "epoch": 0.20714285714285716,
+      "grad_norm": 0.16887658834457397,
+      "learning_rate": 1.862e-05,
+      "loss": 2.9026,
+      "step": 1450
+    },
+    {
+      "epoch": 0.20857142857142857,
+      "grad_norm": 12852670.0,
+      "learning_rate": 1.8610476190476194e-05,
+      "loss": 3.7959,
+      "step": 1460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6697117686271667,
+      "learning_rate": 1.8600952380952383e-05,
+      "loss": 1.4401,
+      "step": 1470
+    },
+    {
+      "epoch": 0.21142857142857144,
+      "grad_norm": 0.034754689782857895,
+      "learning_rate": 1.8591428571428575e-05,
+      "loss": 0.8624,
+      "step": 1480
+    },
+    {
+      "epoch": 0.21285714285714286,
+      "grad_norm": 0.08864087611436844,
+      "learning_rate": 1.8581904761904764e-05,
+      "loss": 0.3113,
+      "step": 1490
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 0.07307349145412445,
+      "learning_rate": 1.8572380952380953e-05,
+      "loss": 3.072,
+      "step": 1500
+    },
+    {
+      "epoch": 0.21571428571428572,
+      "grad_norm": 0.027432406321167946,
+      "learning_rate": 1.8562857142857145e-05,
+      "loss": 0.8366,
+      "step": 1510
+    },
+    {
+      "epoch": 0.21714285714285714,
+      "grad_norm": 0.6425167918205261,
+      "learning_rate": 1.8553333333333334e-05,
+      "loss": 1.6603,
+      "step": 1520
+    },
+    {
+      "epoch": 0.21857142857142858,
+      "grad_norm": 0.04340536147356033,
+      "learning_rate": 1.8543809523809526e-05,
+      "loss": 1.5243,
+      "step": 1530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.07766393572092056,
+      "learning_rate": 1.8534285714285715e-05,
+      "loss": 1.0306,
+      "step": 1540
+    },
+    {
+      "epoch": 0.22142857142857142,
+      "grad_norm": 0.0509803369641304,
+      "learning_rate": 1.8524761904761904e-05,
+      "loss": 0.0,
+      "step": 1550
+    },
+    {
+      "epoch": 0.22285714285714286,
+      "grad_norm": 0.1785711795091629,
+      "learning_rate": 1.8515238095238097e-05,
+      "loss": 1.6456,
+      "step": 1560
+    },
+    {
+      "epoch": 0.22428571428571428,
+      "grad_norm": 0.43144822120666504,
+      "learning_rate": 1.8505714285714286e-05,
+      "loss": 0.1305,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2257142857142857,
+      "grad_norm": 0.0535007044672966,
+      "learning_rate": 1.8496190476190478e-05,
+      "loss": 1.5294,
+      "step": 1580
+    },
+    {
+      "epoch": 0.22714285714285715,
+      "grad_norm": 0.03454805165529251,
+      "learning_rate": 1.8486666666666667e-05,
+      "loss": 1.7291,
+      "step": 1590
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.059376366436481476,
+      "learning_rate": 1.847714285714286e-05,
+      "loss": 1.6528,
+      "step": 1600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.03089080937206745,
+      "learning_rate": 1.8467619047619048e-05,
+      "loss": 0.8402,
+      "step": 1610
+    },
+    {
+      "epoch": 0.23142857142857143,
+      "grad_norm": 1512756.375,
+      "learning_rate": 1.845809523809524e-05,
+      "loss": 1.1037,
+      "step": 1620
+    },
+    {
+      "epoch": 0.23285714285714285,
+      "grad_norm": 0.28252115845680237,
+      "learning_rate": 1.844857142857143e-05,
+      "loss": 2.1044,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2342857142857143,
+      "grad_norm": 0.06322910636663437,
+      "learning_rate": 1.8439047619047622e-05,
+      "loss": 0.6468,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2357142857142857,
+      "grad_norm": 0.055058956146240234,
+      "learning_rate": 1.842952380952381e-05,
+      "loss": 0.8566,
+      "step": 1650
+    },
+    {
+      "epoch": 0.23714285714285716,
+      "grad_norm": 2935924.25,
+      "learning_rate": 1.8420000000000003e-05,
+      "loss": 2.4899,
+      "step": 1660
+    },
+    {
+      "epoch": 0.23857142857142857,
+      "grad_norm": 157389216.0,
+      "learning_rate": 1.8410476190476192e-05,
+      "loss": 1.9805,
+      "step": 1670
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0954144075512886,
+      "learning_rate": 1.840095238095238e-05,
+      "loss": 1.6709,
+      "step": 1680
+    },
+    {
+      "epoch": 0.24142857142857144,
+      "grad_norm": 0.17186371982097626,
+      "learning_rate": 1.8391428571428573e-05,
+      "loss": 2.1504,
+      "step": 1690
+    },
+    {
+      "epoch": 0.24285714285714285,
+      "grad_norm": 0.1450832635164261,
+      "learning_rate": 1.8381904761904762e-05,
+      "loss": 0.0,
+      "step": 1700
+    },
+    {
+      "epoch": 0.24428571428571427,
+      "grad_norm": 0.07599328458309174,
+      "learning_rate": 1.8372380952380955e-05,
+      "loss": 0.8149,
+      "step": 1710
+    },
+    {
+      "epoch": 0.24571428571428572,
+      "grad_norm": 0.05278587341308594,
+      "learning_rate": 1.8362857142857144e-05,
+      "loss": 2.4889,
+      "step": 1720
+    },
+    {
+      "epoch": 0.24714285714285714,
+      "grad_norm": 0.15684713423252106,
+      "learning_rate": 1.8353333333333333e-05,
+      "loss": 2.545,
+      "step": 1730
+    },
+    {
+      "epoch": 0.24857142857142858,
+      "grad_norm": 95031248.0,
+      "learning_rate": 1.8343809523809525e-05,
+      "loss": 2.1774,
+      "step": 1740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6461169719696045,
+      "learning_rate": 1.8334285714285714e-05,
+      "loss": 0.0,
+      "step": 1750
+    },
+    {
+      "epoch": 0.25142857142857145,
+      "grad_norm": 1486288.5,
+      "learning_rate": 1.8324761904761906e-05,
+      "loss": 4.6965,
+      "step": 1760
+    },
+    {
+      "epoch": 0.25285714285714284,
+      "grad_norm": 0.25031182169914246,
+      "learning_rate": 1.83152380952381e-05,
+      "loss": 0.0,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2542857142857143,
+      "grad_norm": 0.10704527795314789,
+      "learning_rate": 1.8305714285714288e-05,
+      "loss": 2.4952,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2557142857142857,
+      "grad_norm": 0.041458845138549805,
+      "learning_rate": 1.829619047619048e-05,
+      "loss": 1.556,
+      "step": 1790
+    },
+    {
+      "epoch": 0.2571428571428571,
+      "grad_norm": 0.04927599057555199,
+      "learning_rate": 1.828666666666667e-05,
+      "loss": 1.4157,
+      "step": 1800
+    },
+    {
+      "epoch": 0.25857142857142856,
+      "grad_norm": 0.2001764476299286,
+      "learning_rate": 1.8277142857142858e-05,
+      "loss": 0.8047,
+      "step": 1810
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14734210073947906,
+      "learning_rate": 1.826761904761905e-05,
+      "loss": 0.3979,
+      "step": 1820
+    },
+    {
+      "epoch": 0.26142857142857145,
+      "grad_norm": 1627507.25,
+      "learning_rate": 1.825809523809524e-05,
+      "loss": 0.8763,
+      "step": 1830
+    },
+    {
+      "epoch": 0.26285714285714284,
+      "grad_norm": 0.062433917075395584,
+      "learning_rate": 1.824857142857143e-05,
+      "loss": 0.0,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2642857142857143,
+      "grad_norm": 0.07109157741069794,
+      "learning_rate": 1.823904761904762e-05,
+      "loss": 2.6052,
+      "step": 1850
+    },
+    {
+      "epoch": 0.26571428571428574,
+      "grad_norm": 11132525.0,
+      "learning_rate": 1.822952380952381e-05,
+      "loss": 2.1691,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2671428571428571,
+      "grad_norm": 0.06913779675960541,
+      "learning_rate": 1.8220000000000002e-05,
+      "loss": 1.3963,
+      "step": 1870
+    },
+    {
+      "epoch": 0.26857142857142857,
+      "grad_norm": 130.31219482421875,
+      "learning_rate": 1.821047619047619e-05,
+      "loss": 2.3367,
+      "step": 1880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 221452288.0,
+      "learning_rate": 1.8200952380952383e-05,
+      "loss": 2.4815,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2714285714285714,
+      "grad_norm": 0.33905863761901855,
+      "learning_rate": 1.8191428571428572e-05,
+      "loss": 2.1748,
+      "step": 1900
+    },
+    {
+      "epoch": 0.27285714285714285,
+      "grad_norm": 0.18102413415908813,
+      "learning_rate": 1.8181904761904764e-05,
+      "loss": 2.4733,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2742857142857143,
+      "grad_norm": 0.1635466068983078,
+      "learning_rate": 1.8172380952380953e-05,
+      "loss": 0.0,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2757142857142857,
+      "grad_norm": 420298400.0,
+      "learning_rate": 1.8162857142857146e-05,
+      "loss": 2.6854,
+      "step": 1930
+    },
+    {
+      "epoch": 0.27714285714285714,
+      "grad_norm": 137570688.0,
+      "learning_rate": 1.8153333333333335e-05,
+      "loss": 0.556,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2785714285714286,
+      "grad_norm": 0.36597055196762085,
+      "learning_rate": 1.8143809523809527e-05,
+      "loss": 3.3709,
+      "step": 1950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14056594.0,
+      "learning_rate": 1.8134285714285716e-05,
+      "loss": 0.8047,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2814285714285714,
+      "grad_norm": 0.1454048901796341,
+      "learning_rate": 1.8124761904761908e-05,
+      "loss": 1.5552,
+      "step": 1970
+    },
+    {
+      "epoch": 0.28285714285714286,
+      "grad_norm": 8297593.5,
+      "learning_rate": 1.8115238095238097e-05,
+      "loss": 0.8067,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2842857142857143,
+      "grad_norm": 0.08763445913791656,
+      "learning_rate": 1.8105714285714286e-05,
+      "loss": 1.6985,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 53748192.0,
+      "learning_rate": 1.809619047619048e-05,
+      "loss": 0.815,
+      "step": 2000
+    },
+    {
+      "epoch": 0.28714285714285714,
+      "grad_norm": 0.2859274446964264,
+      "learning_rate": 1.8086666666666667e-05,
+      "loss": 2.4286,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2885714285714286,
+      "grad_norm": 0.17298242449760437,
+      "learning_rate": 1.8077142857142856e-05,
+      "loss": 1.5246,
+      "step": 2020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.03391052782535553,
+      "learning_rate": 1.806761904761905e-05,
+      "loss": 2.183,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2914285714285714,
+      "grad_norm": 0.0398762971162796,
+      "learning_rate": 1.8058095238095238e-05,
+      "loss": 1.9852,
+      "step": 2040
+    },
+    {
+      "epoch": 0.29285714285714287,
+      "grad_norm": 0.0729857012629509,
+      "learning_rate": 1.804857142857143e-05,
+      "loss": 4.1042,
+      "step": 2050
+    },
+    {
+      "epoch": 0.29428571428571426,
+      "grad_norm": 0.2221122682094574,
+      "learning_rate": 1.803904761904762e-05,
+      "loss": 1.5513,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2957142857142857,
+      "grad_norm": 0.029412029311060905,
+      "learning_rate": 1.802952380952381e-05,
+      "loss": 1.5268,
+      "step": 2070
+    },
+    {
+      "epoch": 0.29714285714285715,
+      "grad_norm": 1573598.125,
+      "learning_rate": 1.802e-05,
+      "loss": 3.5076,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2985714285714286,
+      "grad_norm": 0.297464519739151,
+      "learning_rate": 1.8010476190476193e-05,
+      "loss": 2.3512,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6622931957244873,
+      "learning_rate": 1.8000952380952385e-05,
+      "loss": 1.6363,
+      "step": 2100
+    },
+    {
+      "epoch": 0.30142857142857143,
+      "grad_norm": 0.054901257157325745,
+      "learning_rate": 1.7991428571428574e-05,
+      "loss": 0.0,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3028571428571429,
+      "grad_norm": 0.018535833805799484,
+      "learning_rate": 1.7981904761904763e-05,
+      "loss": 1.2835,
+      "step": 2120
+    },
+    {
+      "epoch": 0.30428571428571427,
+      "grad_norm": 0.060639988631010056,
+      "learning_rate": 1.7972380952380955e-05,
+      "loss": 3.9758,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3057142857142857,
+      "grad_norm": 1.1017646789550781,
+      "learning_rate": 1.7962857142857144e-05,
+      "loss": 2.1858,
+      "step": 2140
+    },
+    {
+      "epoch": 0.30714285714285716,
+      "grad_norm": 0.014938214793801308,
+      "learning_rate": 1.7953333333333333e-05,
+      "loss": 1.4247,
+      "step": 2150
+    },
+    {
+      "epoch": 0.30857142857142855,
+      "grad_norm": 0.013980339281260967,
+      "learning_rate": 1.7943809523809525e-05,
+      "loss": 0.8391,
+      "step": 2160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.04963662847876549,
+      "learning_rate": 1.7934285714285714e-05,
+      "loss": 1.7244,
+      "step": 2170
+    },
+    {
+      "epoch": 0.31142857142857144,
+      "grad_norm": 5.18486213684082,
+      "learning_rate": 1.7924761904761907e-05,
+      "loss": 1.6278,
+      "step": 2180
+    },
+    {
+      "epoch": 0.31285714285714283,
+      "grad_norm": 29450932.0,
+      "learning_rate": 1.7915238095238096e-05,
+      "loss": 1.5809,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3142857142857143,
+      "grad_norm": 0.08646556735038757,
+      "learning_rate": 1.7905714285714285e-05,
+      "loss": 0.8065,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3157142857142857,
+      "grad_norm": 3471954.0,
+      "learning_rate": 1.7896190476190477e-05,
+      "loss": 2.3902,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3171428571428571,
+      "grad_norm": 0.19620957970619202,
+      "learning_rate": 1.788666666666667e-05,
+      "loss": 0.8095,
+      "step": 2220
+    },
+    {
+      "epoch": 0.31857142857142856,
+      "grad_norm": 0.06481914967298508,
+      "learning_rate": 1.7877142857142858e-05,
+      "loss": 0.807,
+      "step": 2230
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1556574.0,
+      "learning_rate": 1.786761904761905e-05,
+      "loss": 0.8517,
+      "step": 2240
+    },
+    {
+      "epoch": 0.32142857142857145,
+      "grad_norm": 0.48912906646728516,
+      "learning_rate": 1.785809523809524e-05,
+      "loss": 0.87,
+      "step": 2250
+    },
+    {
+      "epoch": 0.32285714285714284,
+      "grad_norm": 11.912801742553711,
+      "learning_rate": 1.7848571428571432e-05,
+      "loss": 0.0003,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3242857142857143,
+      "grad_norm": 0.7917404174804688,
+      "learning_rate": 1.783904761904762e-05,
+      "loss": 1.5514,
+      "step": 2270
+    },
+    {
+      "epoch": 0.32571428571428573,
+      "grad_norm": 0.022578110918402672,
+      "learning_rate": 1.782952380952381e-05,
+      "loss": 0.8782,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3271428571428571,
+      "grad_norm": 0.037791941314935684,
+      "learning_rate": 1.7820000000000002e-05,
+      "loss": 0.0,
+      "step": 2290
+    },
+    {
+      "epoch": 0.32857142857142857,
+      "grad_norm": 0.03612258657813072,
+      "learning_rate": 1.781047619047619e-05,
+      "loss": 0.8474,
+      "step": 2300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.01866544969379902,
+      "learning_rate": 1.7800952380952383e-05,
+      "loss": 0.0,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3314285714285714,
+      "grad_norm": 1678833.125,
+      "learning_rate": 1.7791428571428572e-05,
+      "loss": 1.7373,
+      "step": 2320
+    },
+    {
+      "epoch": 0.33285714285714285,
+      "grad_norm": 0.135600283741951,
+      "learning_rate": 1.778190476190476e-05,
+      "loss": 1.7049,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3342857142857143,
+      "grad_norm": 0.064423568546772,
+      "learning_rate": 1.7772380952380954e-05,
+      "loss": 0.8397,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3357142857142857,
+      "grad_norm": 0.06785961985588074,
+      "learning_rate": 1.7762857142857143e-05,
+      "loss": 0.8938,
+      "step": 2350
+    },
+    {
+      "epoch": 0.33714285714285713,
+      "grad_norm": 1447406.75,
+      "learning_rate": 1.7753333333333335e-05,
+      "loss": 4.9967,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3385714285714286,
+      "grad_norm": 22610.8671875,
+      "learning_rate": 1.7743809523809524e-05,
+      "loss": 0.7993,
+      "step": 2370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.05351774021983147,
+      "learning_rate": 1.7734285714285716e-05,
+      "loss": 0.6811,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3414285714285714,
+      "grad_norm": 0.06253944337368011,
+      "learning_rate": 1.7724761904761905e-05,
+      "loss": 2.4307,
+      "step": 2390
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.8237541317939758,
+      "learning_rate": 1.7715238095238098e-05,
+      "loss": 1.5178,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3442857142857143,
+      "grad_norm": 0.08923335373401642,
+      "learning_rate": 1.7705714285714287e-05,
+      "loss": 3.2116,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3457142857142857,
+      "grad_norm": 1580250.625,
+      "learning_rate": 1.769619047619048e-05,
+      "loss": 3.202,
+      "step": 2420
+    },
+    {
+      "epoch": 0.34714285714285714,
+      "grad_norm": 0.4089590609073639,
+      "learning_rate": 1.7686666666666668e-05,
+      "loss": 1.4688,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3485714285714286,
+      "grad_norm": 0.7222450375556946,
+      "learning_rate": 1.767714285714286e-05,
+      "loss": 0.8301,
+      "step": 2440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.054450735449790955,
+      "learning_rate": 1.766761904761905e-05,
+      "loss": 1.7625,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3514285714285714,
+      "grad_norm": 0.21793900430202484,
+      "learning_rate": 1.7658095238095238e-05,
+      "loss": 2.4857,
+      "step": 2460
+    },
+    {
+      "epoch": 0.35285714285714287,
+      "grad_norm": 0.07498624920845032,
+      "learning_rate": 1.764857142857143e-05,
+      "loss": 2.5804,
+      "step": 2470
+    },
+    {
+      "epoch": 0.35428571428571426,
+      "grad_norm": 0.16685189306735992,
+      "learning_rate": 1.763904761904762e-05,
+      "loss": 0.7932,
+      "step": 2480
+    },
+    {
+      "epoch": 0.3557142857142857,
+      "grad_norm": 23.8129940032959,
+      "learning_rate": 1.7629523809523812e-05,
+      "loss": 0.8562,
+      "step": 2490
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 4104183.5,
+      "learning_rate": 1.762e-05,
+      "loss": 3.4057,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3585714285714286,
+      "grad_norm": 0.14345857501029968,
+      "learning_rate": 1.761047619047619e-05,
+      "loss": 0.8401,
+      "step": 2510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.23295703530311584,
+      "learning_rate": 1.7600952380952382e-05,
+      "loss": 1.6453,
+      "step": 2520
+    },
+    {
+      "epoch": 0.36142857142857143,
+      "grad_norm": 0.05060197412967682,
+      "learning_rate": 1.759142857142857e-05,
+      "loss": 0.8376,
+      "step": 2530
+    },
+    {
+      "epoch": 0.3628571428571429,
+      "grad_norm": 0.08178196102380753,
+      "learning_rate": 1.7581904761904763e-05,
+      "loss": 0.8834,
+      "step": 2540
+    },
+    {
+      "epoch": 0.36428571428571427,
+      "grad_norm": 0.3259715735912323,
+      "learning_rate": 1.7572380952380956e-05,
+      "loss": 5.8522,
+      "step": 2550
+    },
+    {
+      "epoch": 0.3657142857142857,
+      "grad_norm": 0.17609982192516327,
+      "learning_rate": 1.7562857142857145e-05,
+      "loss": 1.4409,
+      "step": 2560
+    },
+    {
+      "epoch": 0.36714285714285716,
+      "grad_norm": 0.10048504173755646,
+      "learning_rate": 1.7553333333333337e-05,
+      "loss": 0.8062,
+      "step": 2570
+    },
+    {
+      "epoch": 0.36857142857142855,
+      "grad_norm": 63.24231719970703,
+      "learning_rate": 1.7543809523809526e-05,
+      "loss": 0.0,
+      "step": 2580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15639500319957733,
+      "learning_rate": 1.7534285714285715e-05,
+      "loss": 4.2732,
+      "step": 2590
+    },
+    {
+      "epoch": 0.37142857142857144,
+      "grad_norm": 1686949.875,
+      "learning_rate": 1.7524761904761907e-05,
+      "loss": 4.1394,
+      "step": 2600
+    },
+    {
+      "epoch": 0.37285714285714283,
+      "grad_norm": 0.17479872703552246,
+      "learning_rate": 1.7515238095238096e-05,
+      "loss": 2.2588,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3742857142857143,
+      "grad_norm": 0.07670348137617111,
+      "learning_rate": 1.750571428571429e-05,
+      "loss": 0.8299,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3757142857142857,
+      "grad_norm": 2286924.0,
+      "learning_rate": 1.7496190476190477e-05,
+      "loss": 2.4817,
+      "step": 2630
+    },
+    {
+      "epoch": 0.37714285714285717,
+      "grad_norm": 0.2645169496536255,
+      "learning_rate": 1.7486666666666666e-05,
+      "loss": 0.0,
+      "step": 2640
+    },
+    {
+      "epoch": 0.37857142857142856,
+      "grad_norm": 0.053946301341056824,
+      "learning_rate": 1.747714285714286e-05,
+      "loss": 1.5829,
+      "step": 2650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1426443.375,
+      "learning_rate": 1.7467619047619048e-05,
+      "loss": 3.4777,
+      "step": 2660
+    },
+    {
+      "epoch": 0.38142857142857145,
+      "grad_norm": 3341132.0,
+      "learning_rate": 1.745809523809524e-05,
+      "loss": 0.8423,
+      "step": 2670
+    },
+    {
+      "epoch": 0.38285714285714284,
+      "grad_norm": 2664341.5,
+      "learning_rate": 1.744857142857143e-05,
+      "loss": 4.0496,
+      "step": 2680
+    },
+    {
+      "epoch": 0.3842857142857143,
+      "grad_norm": 0.18499571084976196,
+      "learning_rate": 1.743904761904762e-05,
+      "loss": 0.8219,
+      "step": 2690
+    },
+    {
+      "epoch": 0.38571428571428573,
+      "grad_norm": 1484994.625,
+      "learning_rate": 1.742952380952381e-05,
+      "loss": 1.7036,
+      "step": 2700
+    },
+    {
+      "epoch": 0.3871428571428571,
+      "grad_norm": 0.07417990267276764,
+      "learning_rate": 1.7420000000000003e-05,
+      "loss": 1.5078,
+      "step": 2710
+    },
+    {
+      "epoch": 0.38857142857142857,
+      "grad_norm": 0.06702981889247894,
+      "learning_rate": 1.741047619047619e-05,
+      "loss": 2.6348,
+      "step": 2720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15484023094177246,
+      "learning_rate": 1.7400952380952384e-05,
+      "loss": 1.6027,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3914285714285714,
+      "grad_norm": 3.3918752670288086,
+      "learning_rate": 1.7391428571428573e-05,
+      "loss": 2.0755,
+      "step": 2740
+    },
+    {
+      "epoch": 0.39285714285714285,
+      "grad_norm": 0.1758022904396057,
+      "learning_rate": 1.7381904761904765e-05,
+      "loss": 0.8815,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3942857142857143,
+      "grad_norm": 0.13718213140964508,
+      "learning_rate": 1.7372380952380954e-05,
+      "loss": 1.1375,
+      "step": 2760
+    },
+    {
+      "epoch": 0.39571428571428574,
+      "grad_norm": 2860862.25,
+      "learning_rate": 1.7362857142857143e-05,
+      "loss": 2.4797,
+      "step": 2770
+    },
+    {
+      "epoch": 0.39714285714285713,
+      "grad_norm": 0.05986272916197777,
+      "learning_rate": 1.7353333333333335e-05,
+      "loss": 0.8218,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3985714285714286,
+      "grad_norm": 0.30865082144737244,
+      "learning_rate": 1.7343809523809524e-05,
+      "loss": 1.5241,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15577.8974609375,
+      "learning_rate": 1.7334285714285713e-05,
+      "loss": 0.0,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4014285714285714,
+      "grad_norm": 1453266.875,
+      "learning_rate": 1.7324761904761906e-05,
+      "loss": 1.7952,
+      "step": 2810
+    },
+    {
+      "epoch": 0.40285714285714286,
+      "grad_norm": 306321.46875,
+      "learning_rate": 1.7315238095238095e-05,
+      "loss": 3.5621,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4042857142857143,
+      "grad_norm": 0.21034109592437744,
+      "learning_rate": 1.7305714285714287e-05,
+      "loss": 0.7977,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4057142857142857,
+      "grad_norm": 3717422.0,
+      "learning_rate": 1.7296190476190476e-05,
+      "loss": 2.4353,
+      "step": 2840
+    },
+    {
+      "epoch": 0.40714285714285714,
+      "grad_norm": 1575349.5,
+      "learning_rate": 1.7286666666666668e-05,
+      "loss": 2.4967,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4085714285714286,
+      "grad_norm": 0.3163071572780609,
+      "learning_rate": 1.727714285714286e-05,
+      "loss": 0.7461,
+      "step": 2860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1651473045349121,
+      "learning_rate": 1.726761904761905e-05,
+      "loss": 0.0,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4114285714285714,
+      "grad_norm": 0.05035606771707535,
+      "learning_rate": 1.7258095238095242e-05,
+      "loss": 0.0,
+      "step": 2880
+    },
+    {
+      "epoch": 0.41285714285714287,
+      "grad_norm": 0.044252727180719376,
+      "learning_rate": 1.724857142857143e-05,
+      "loss": 0.4125,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4142857142857143,
+      "grad_norm": 0.030622383579611778,
+      "learning_rate": 1.723904761904762e-05,
+      "loss": 0.8443,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4157142857142857,
+      "grad_norm": 0.08758899569511414,
+      "learning_rate": 1.7229523809523812e-05,
+      "loss": 0.8326,
+      "step": 2910
+    },
+    {
+      "epoch": 0.41714285714285715,
+      "grad_norm": 0.03288514167070389,
+      "learning_rate": 1.722e-05,
+      "loss": 0.8929,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4185714285714286,
+      "grad_norm": 14822603.0,
+      "learning_rate": 1.721047619047619e-05,
+      "loss": 2.5097,
+      "step": 2930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1679517.125,
+      "learning_rate": 1.7200952380952382e-05,
+      "loss": 1.8925,
+      "step": 2940
+    },
+    {
+      "epoch": 0.42142857142857143,
+      "grad_norm": 0.18690992891788483,
+      "learning_rate": 1.719142857142857e-05,
+      "loss": 2.9266,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4228571428571429,
+      "grad_norm": 0.21930739283561707,
+      "learning_rate": 1.7181904761904764e-05,
+      "loss": 0.7632,
+      "step": 2960
+    },
+    {
+      "epoch": 0.42428571428571427,
+      "grad_norm": 0.44387927651405334,
+      "learning_rate": 1.7172380952380953e-05,
+      "loss": 0.8896,
+      "step": 2970
+    },
+    {
+      "epoch": 0.4257142857142857,
+      "grad_norm": 0.09059398621320724,
+      "learning_rate": 1.716285714285714e-05,
+      "loss": 3.3224,
+      "step": 2980
+    },
+    {
+      "epoch": 0.42714285714285716,
+      "grad_norm": 1806601.5,
+      "learning_rate": 1.7153333333333334e-05,
+      "loss": 2.3839,
+      "step": 2990
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 2356865.75,
+      "learning_rate": 1.7143809523809526e-05,
+      "loss": 1.024,
+      "step": 3000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.08273036032915115,
+      "learning_rate": 1.7134285714285715e-05,
+      "loss": 0.7487,
+      "step": 3010
+    },
+    {
+      "epoch": 0.43142857142857144,
+      "grad_norm": 1.2428717613220215,
+      "learning_rate": 1.7124761904761908e-05,
+      "loss": 2.3967,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4328571428571429,
+      "grad_norm": 2.2564940452575684,
+      "learning_rate": 1.7115238095238097e-05,
+      "loss": 0.8374,
+      "step": 3030
+    },
+    {
+      "epoch": 0.4342857142857143,
+      "grad_norm": 0.856573760509491,
+      "learning_rate": 1.710571428571429e-05,
+      "loss": 1.4601,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4357142857142857,
+      "grad_norm": 0.047206174582242966,
+      "learning_rate": 1.7096190476190478e-05,
+      "loss": 1.642,
+      "step": 3050
+    },
+    {
+      "epoch": 0.43714285714285717,
+      "grad_norm": 0.03543708473443985,
+      "learning_rate": 1.7086666666666667e-05,
+      "loss": 1.5986,
+      "step": 3060
+    },
+    {
+      "epoch": 0.43857142857142856,
+      "grad_norm": 0.09097450971603394,
+      "learning_rate": 1.707714285714286e-05,
+      "loss": 1.602,
+      "step": 3070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.595640182495117,
+      "learning_rate": 1.7067619047619048e-05,
+      "loss": 2.3394,
+      "step": 3080
+    },
+    {
+      "epoch": 0.44142857142857145,
+      "grad_norm": 0.45068445801734924,
+      "learning_rate": 1.705809523809524e-05,
+      "loss": 0.7684,
+      "step": 3090
+    },
+    {
+      "epoch": 0.44285714285714284,
+      "grad_norm": 2016719.5,
+      "learning_rate": 1.704857142857143e-05,
+      "loss": 2.3262,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4442857142857143,
+      "grad_norm": 0.2602471709251404,
+      "learning_rate": 1.703904761904762e-05,
+      "loss": 2.4262,
+      "step": 3110
+    },
+    {
+      "epoch": 0.44571428571428573,
+      "grad_norm": 2420584.25,
+      "learning_rate": 1.702952380952381e-05,
+      "loss": 1.6075,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4471428571428571,
+      "grad_norm": 0.7045326828956604,
+      "learning_rate": 1.702e-05,
+      "loss": 0.8309,
+      "step": 3130
+    },
+    {
+      "epoch": 0.44857142857142857,
+      "grad_norm": 5.372880935668945,
+      "learning_rate": 1.7010476190476192e-05,
+      "loss": 1.5598,
+      "step": 3140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.09766633808612823,
+      "learning_rate": 1.700095238095238e-05,
+      "loss": 2.3349,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4514285714285714,
+      "grad_norm": 1.2179263830184937,
+      "learning_rate": 1.6991428571428573e-05,
+      "loss": 0.6939,
+      "step": 3160
+    },
+    {
+      "epoch": 0.45285714285714285,
+      "grad_norm": 0.20540456473827362,
+      "learning_rate": 1.6981904761904762e-05,
+      "loss": 3.1762,
+      "step": 3170
+    },
+    {
+      "epoch": 0.4542857142857143,
+      "grad_norm": 0.19691500067710876,
+      "learning_rate": 1.6972380952380955e-05,
+      "loss": 0.7013,
+      "step": 3180
+    },
+    {
+      "epoch": 0.45571428571428574,
+      "grad_norm": 0.0723930299282074,
+      "learning_rate": 1.6962857142857147e-05,
+      "loss": 0.6731,
+      "step": 3190
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.04215807095170021,
+      "learning_rate": 1.6953333333333336e-05,
+      "loss": 0.4963,
+      "step": 3200
+    },
+    {
+      "epoch": 0.4585714285714286,
+      "grad_norm": 6.351660251617432,
+      "learning_rate": 1.6943809523809525e-05,
+      "loss": 0.0,
+      "step": 3210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.049287691712379456,
+      "learning_rate": 1.6934285714285717e-05,
+      "loss": 2.4251,
+      "step": 3220
+    },
+    {
+      "epoch": 0.4614285714285714,
+      "grad_norm": 10735640.0,
+      "learning_rate": 1.6924761904761906e-05,
+      "loss": 3.053,
+      "step": 3230
+    },
+    {
+      "epoch": 0.46285714285714286,
+      "grad_norm": 0.05822628363966942,
+      "learning_rate": 1.6915238095238095e-05,
+      "loss": 1.5764,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4642857142857143,
+      "grad_norm": 0.061946626752614975,
+      "learning_rate": 1.6905714285714287e-05,
+      "loss": 0.7522,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4657142857142857,
+      "grad_norm": 0.8822937607765198,
+      "learning_rate": 1.6896190476190476e-05,
+      "loss": 0.5685,
+      "step": 3260
+    },
+    {
+      "epoch": 0.46714285714285714,
+      "grad_norm": 0.04442232474684715,
+      "learning_rate": 1.688666666666667e-05,
+      "loss": 2.3987,
+      "step": 3270
+    },
+    {
+      "epoch": 0.4685714285714286,
+      "grad_norm": 0.023770179599523544,
+      "learning_rate": 1.6877142857142858e-05,
+      "loss": 0.6281,
+      "step": 3280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2446277141571045,
+      "learning_rate": 1.6867619047619047e-05,
+      "loss": 3.0717,
+      "step": 3290
+    },
+    {
+      "epoch": 0.4714285714285714,
+      "grad_norm": 1786930.625,
+      "learning_rate": 1.685809523809524e-05,
+      "loss": 2.5331,
+      "step": 3300
+    },
+    {
+      "epoch": 0.47285714285714286,
+      "grad_norm": 0.10868965834379196,
+      "learning_rate": 1.684857142857143e-05,
+      "loss": 1.5862,
+      "step": 3310
+    },
+    {
+      "epoch": 0.4742857142857143,
+      "grad_norm": 0.028069553896784782,
+      "learning_rate": 1.683904761904762e-05,
+      "loss": 0.748,
+      "step": 3320
+    },
+    {
+      "epoch": 0.4757142857142857,
+      "grad_norm": 30653904.0,
+      "learning_rate": 1.6829523809523813e-05,
+      "loss": 2.0415,
+      "step": 3330
+    },
+    {
+      "epoch": 0.47714285714285715,
+      "grad_norm": 0.031270284205675125,
+      "learning_rate": 1.682e-05,
+      "loss": 0.0,
+      "step": 3340
+    },
+    {
+      "epoch": 0.4785714285714286,
+      "grad_norm": 2415140.5,
+      "learning_rate": 1.6810476190476194e-05,
+      "loss": 2.2417,
+      "step": 3350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 35874900.0,
+      "learning_rate": 1.6800952380952383e-05,
+      "loss": 3.3652,
+      "step": 3360
+    },
+    {
+      "epoch": 0.48142857142857143,
+      "grad_norm": 0.06744790077209473,
+      "learning_rate": 1.6791428571428572e-05,
+      "loss": 0.6412,
+      "step": 3370
+    },
+    {
+      "epoch": 0.4828571428571429,
+      "grad_norm": 0.13700543344020844,
+      "learning_rate": 1.6781904761904764e-05,
+      "loss": 1.5269,
+      "step": 3380
+    },
+    {
+      "epoch": 0.48428571428571426,
+      "grad_norm": 0.13974149525165558,
+      "learning_rate": 1.6772380952380953e-05,
+      "loss": 0.8696,
+      "step": 3390
+    },
+    {
+      "epoch": 0.4857142857142857,
+      "grad_norm": 0.03642427921295166,
+      "learning_rate": 1.6762857142857145e-05,
+      "loss": 2.508,
+      "step": 3400
+    },
+    {
+      "epoch": 0.48714285714285716,
+      "grad_norm": 0.2967716455459595,
+      "learning_rate": 1.6753333333333334e-05,
+      "loss": 0.7858,
+      "step": 3410
+    },
+    {
+      "epoch": 0.48857142857142855,
+      "grad_norm": 0.2844642698764801,
+      "learning_rate": 1.6743809523809523e-05,
+      "loss": 0.8987,
+      "step": 3420
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.793733835220337,
+      "learning_rate": 1.6734285714285716e-05,
+      "loss": 1.5364,
+      "step": 3430
+    },
+    {
+      "epoch": 0.49142857142857144,
+      "grad_norm": 0.7094473838806152,
+      "learning_rate": 1.6724761904761905e-05,
+      "loss": 0.0,
+      "step": 3440
+    },
+    {
+      "epoch": 0.4928571428571429,
+      "grad_norm": 1889719.0,
+      "learning_rate": 1.6715238095238097e-05,
+      "loss": 1.7665,
+      "step": 3450
+    },
+    {
+      "epoch": 0.4942857142857143,
+      "grad_norm": 1421811.5,
+      "learning_rate": 1.6705714285714286e-05,
+      "loss": 3.4253,
+      "step": 3460
+    },
+    {
+      "epoch": 0.4957142857142857,
+      "grad_norm": 0.28111425042152405,
+      "learning_rate": 1.669619047619048e-05,
+      "loss": 1.6206,
+      "step": 3470
+    },
+    {
+      "epoch": 0.49714285714285716,
+      "grad_norm": 0.7700180411338806,
+      "learning_rate": 1.6686666666666667e-05,
+      "loss": 0.0,
+      "step": 3480
+    },
+    {
+      "epoch": 0.49857142857142855,
+      "grad_norm": 8501649.0,
+      "learning_rate": 1.667714285714286e-05,
+      "loss": 1.4956,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2384973.0,
+      "learning_rate": 1.666761904761905e-05,
+      "loss": 2.4883,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5014285714285714,
+      "grad_norm": 0.15234528481960297,
+      "learning_rate": 1.665809523809524e-05,
+      "loss": 0.7706,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5028571428571429,
+      "grad_norm": 0.0750003308057785,
+      "learning_rate": 1.664857142857143e-05,
+      "loss": 0.7697,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5042857142857143,
+      "grad_norm": 0.1655336320400238,
+      "learning_rate": 1.6639047619047622e-05,
+      "loss": 0.8432,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5057142857142857,
+      "grad_norm": 1.9231886863708496,
+      "learning_rate": 1.662952380952381e-05,
+      "loss": 1.6966,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5071428571428571,
+      "grad_norm": 874642816.0,
+      "learning_rate": 1.662e-05,
+      "loss": 2.8104,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5085714285714286,
+      "grad_norm": 0.058356814086437225,
+      "learning_rate": 1.6610476190476192e-05,
+      "loss": 0.6007,
+      "step": 3560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7702398.5,
+      "learning_rate": 1.660095238095238e-05,
+      "loss": 3.9404,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5114285714285715,
+      "grad_norm": 0.07619433850049973,
+      "learning_rate": 1.6591428571428574e-05,
+      "loss": 3.1428,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5128571428571429,
+      "grad_norm": 0.09882547706365585,
+      "learning_rate": 1.6581904761904763e-05,
+      "loss": 1.6734,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 0.1984606832265854,
+      "learning_rate": 1.657238095238095e-05,
+      "loss": 4.0641,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5157142857142857,
+      "grad_norm": 0.7016361355781555,
+      "learning_rate": 1.6562857142857144e-05,
+      "loss": 0.7612,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5171428571428571,
+      "grad_norm": 0.11243625730276108,
+      "learning_rate": 1.6553333333333333e-05,
+      "loss": 3.1157,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5185714285714286,
+      "grad_norm": 0.7789314389228821,
+      "learning_rate": 1.6543809523809525e-05,
+      "loss": 0.7992,
+      "step": 3630
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.04589006304740906,
+      "learning_rate": 1.6534285714285718e-05,
+      "loss": 2.5127,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5214285714285715,
+      "grad_norm": 0.17558953166007996,
+      "learning_rate": 1.6524761904761907e-05,
+      "loss": 0.7957,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5228571428571429,
+      "grad_norm": 0.421630859375,
+      "learning_rate": 1.65152380952381e-05,
+      "loss": 1.5556,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5242857142857142,
+      "grad_norm": 1.222598671913147,
+      "learning_rate": 1.6505714285714288e-05,
+      "loss": 2.5331,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5257142857142857,
+      "grad_norm": 0.056295666843652725,
+      "learning_rate": 1.6496190476190477e-05,
+      "loss": 1.2278,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5271428571428571,
+      "grad_norm": 0.022797036916017532,
+      "learning_rate": 1.648666666666667e-05,
+      "loss": 0.829,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5285714285714286,
+      "grad_norm": 0.02164800837635994,
+      "learning_rate": 1.6477142857142858e-05,
+      "loss": 2.4497,
+      "step": 3700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.03037365898489952,
+      "learning_rate": 1.646761904761905e-05,
+      "loss": 2.4738,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5314285714285715,
+      "grad_norm": 0.08893661946058273,
+      "learning_rate": 1.645809523809524e-05,
+      "loss": 2.5326,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5328571428571428,
+      "grad_norm": 0.04576045274734497,
+      "learning_rate": 1.644857142857143e-05,
+      "loss": 1.5429,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5342857142857143,
+      "grad_norm": 24673640.0,
+      "learning_rate": 1.643904761904762e-05,
+      "loss": 2.1989,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5357142857142857,
+      "grad_norm": 0.27416226267814636,
+      "learning_rate": 1.642952380952381e-05,
+      "loss": 2.2888,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5371428571428571,
+      "grad_norm": 0.05079071596264839,
+      "learning_rate": 1.6420000000000002e-05,
+      "loss": 0.8744,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5385714285714286,
+      "grad_norm": 3838081.75,
+      "learning_rate": 1.641047619047619e-05,
+      "loss": 0.7605,
+      "step": 3770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.16330742835998535,
+      "learning_rate": 1.6400952380952383e-05,
+      "loss": 1.5943,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5414285714285715,
+      "grad_norm": 0.04954441636800766,
+      "learning_rate": 1.6391428571428572e-05,
+      "loss": 1.6769,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5428571428571428,
+      "grad_norm": 0.06305906176567078,
+      "learning_rate": 1.6381904761904765e-05,
+      "loss": 1.6315,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5442857142857143,
+      "grad_norm": 0.19511088728904724,
+      "learning_rate": 1.6372380952380954e-05,
+      "loss": 0.9165,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5457142857142857,
+      "grad_norm": 0.29016056656837463,
+      "learning_rate": 1.6362857142857146e-05,
+      "loss": 1.6863,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5471428571428572,
+      "grad_norm": 0.15878696739673615,
+      "learning_rate": 1.6353333333333335e-05,
+      "loss": 1.6171,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5485714285714286,
+      "grad_norm": 79111240.0,
+      "learning_rate": 1.6343809523809527e-05,
+      "loss": 1.3804,
+      "step": 3840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1685488224029541,
+      "learning_rate": 1.6334285714285716e-05,
+      "loss": 1.6099,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5514285714285714,
+      "grad_norm": 0.03058214671909809,
+      "learning_rate": 1.6324761904761905e-05,
+      "loss": 0.6697,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5528571428571428,
+      "grad_norm": 0.31740903854370117,
+      "learning_rate": 1.6315238095238097e-05,
+      "loss": 1.7145,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5542857142857143,
+      "grad_norm": 0.0580446831882,
+      "learning_rate": 1.6305714285714286e-05,
+      "loss": 2.1095,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5557142857142857,
+      "grad_norm": 0.03153647854924202,
+      "learning_rate": 1.6296190476190475e-05,
+      "loss": 2.1704,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5571428571428572,
+      "grad_norm": 0.09184670448303223,
+      "learning_rate": 1.6286666666666668e-05,
+      "loss": 2.0447,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5585714285714286,
+      "grad_norm": 0.06404896825551987,
+      "learning_rate": 1.6277142857142857e-05,
+      "loss": 2.4203,
+      "step": 3910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.04509298875927925,
+      "learning_rate": 1.626761904761905e-05,
+      "loss": 1.5814,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5614285714285714,
+      "grad_norm": 0.051130931824445724,
+      "learning_rate": 1.6258095238095238e-05,
+      "loss": 0.8732,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5628571428571428,
+      "grad_norm": 0.041874341666698456,
+      "learning_rate": 1.624857142857143e-05,
+      "loss": 3.0014,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5642857142857143,
+      "grad_norm": 0.3717464208602905,
+      "learning_rate": 1.623904761904762e-05,
+      "loss": 0.8422,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5657142857142857,
+      "grad_norm": 8.081599235534668,
+      "learning_rate": 1.622952380952381e-05,
+      "loss": 0.8162,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5671428571428572,
+      "grad_norm": 0.04418623819947243,
+      "learning_rate": 1.6220000000000004e-05,
+      "loss": 0.0,
+      "step": 3970
+    },
+    {
+      "epoch": 0.5685714285714286,
+      "grad_norm": 0.028846165165305138,
+      "learning_rate": 1.6210476190476193e-05,
+      "loss": 0.0,
+      "step": 3980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1987395.25,
+      "learning_rate": 1.6200952380952382e-05,
+      "loss": 2.2799,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.02784336358308792,
+      "learning_rate": 1.6191428571428574e-05,
+      "loss": 0.0095,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5728571428571428,
+      "grad_norm": 0.0763741135597229,
+      "learning_rate": 1.6181904761904763e-05,
+      "loss": 0.8292,
+      "step": 4010
+    },
+    {
+      "epoch": 0.5742857142857143,
+      "grad_norm": 0.06242310628294945,
+      "learning_rate": 1.6172380952380952e-05,
+      "loss": 3.3072,
+      "step": 4020
+    },
+    {
+      "epoch": 0.5757142857142857,
+      "grad_norm": 0.11338083446025848,
+      "learning_rate": 1.6162857142857144e-05,
+      "loss": 1.5409,
+      "step": 4030
+    },
+    {
+      "epoch": 0.5771428571428572,
+      "grad_norm": 0.021449211984872818,
+      "learning_rate": 1.6153333333333333e-05,
+      "loss": 0.0,
+      "step": 4040
+    },
+    {
+      "epoch": 0.5785714285714286,
+      "grad_norm": 0.08869735151529312,
+      "learning_rate": 1.6143809523809526e-05,
+      "loss": 0.0,
+      "step": 4050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.04914731904864311,
+      "learning_rate": 1.6134285714285715e-05,
+      "loss": 0.0469,
+      "step": 4060
+    },
+    {
+      "epoch": 0.5814285714285714,
+      "grad_norm": 0.005268457345664501,
+      "learning_rate": 1.6124761904761904e-05,
+      "loss": 1.7542,
+      "step": 4070
+    },
+    {
+      "epoch": 0.5828571428571429,
+      "grad_norm": 0.018505113199353218,
+      "learning_rate": 1.6115238095238096e-05,
+      "loss": 0.6863,
+      "step": 4080
+    },
+    {
+      "epoch": 0.5842857142857143,
+      "grad_norm": 0.08119050413370132,
+      "learning_rate": 1.610571428571429e-05,
+      "loss": 0.891,
+      "step": 4090
+    },
+    {
+      "epoch": 0.5857142857142857,
+      "grad_norm": 0.01272314041852951,
+      "learning_rate": 1.6096190476190477e-05,
+      "loss": 0.0,
+      "step": 4100
+    },
+    {
+      "epoch": 0.5871428571428572,
+      "grad_norm": 8027002.5,
+      "learning_rate": 1.608666666666667e-05,
+      "loss": 3.3431,
+      "step": 4110
+    },
+    {
+      "epoch": 0.5885714285714285,
+      "grad_norm": 0.038243573158979416,
+      "learning_rate": 1.607714285714286e-05,
+      "loss": 0.8793,
+      "step": 4120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.031105700880289078,
+      "learning_rate": 1.606761904761905e-05,
+      "loss": 1.2851,
+      "step": 4130
+    },
+    {
+      "epoch": 0.5914285714285714,
+      "grad_norm": 0.42261093854904175,
+      "learning_rate": 1.605809523809524e-05,
+      "loss": 0.0,
+      "step": 4140
+    },
+    {
+      "epoch": 0.5928571428571429,
+      "grad_norm": 0.06418009102344513,
+      "learning_rate": 1.604857142857143e-05,
+      "loss": 1.6324,
+      "step": 4150
+    },
+    {
+      "epoch": 0.5942857142857143,
+      "grad_norm": 0.026129314675927162,
+      "learning_rate": 1.603904761904762e-05,
+      "loss": 0.0,
+      "step": 4160
+    },
+    {
+      "epoch": 0.5957142857142858,
+      "grad_norm": 2833746.0,
+      "learning_rate": 1.602952380952381e-05,
+      "loss": 1.717,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5971428571428572,
+      "grad_norm": 0.02794097363948822,
+      "learning_rate": 1.6020000000000002e-05,
+      "loss": 0.0,
+      "step": 4180
+    },
+    {
+      "epoch": 0.5985714285714285,
+      "grad_norm": 129719920.0,
+      "learning_rate": 1.601047619047619e-05,
+      "loss": 3.9165,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.03892814368009567,
+      "learning_rate": 1.600095238095238e-05,
+      "loss": 0.0,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6014285714285714,
+      "grad_norm": 0.03918730095028877,
+      "learning_rate": 1.5991428571428573e-05,
+      "loss": 1.4544,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6028571428571429,
+      "grad_norm": 1.4248087406158447,
+      "learning_rate": 1.598190476190476e-05,
+      "loss": 2.3866,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6042857142857143,
+      "grad_norm": 0.4004291594028473,
+      "learning_rate": 1.5972380952380954e-05,
+      "loss": 2.9849,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6057142857142858,
+      "grad_norm": 0.3124048709869385,
+      "learning_rate": 1.5962857142857143e-05,
+      "loss": 1.6759,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6071428571428571,
+      "grad_norm": 0.053076185286045074,
+      "learning_rate": 1.5953333333333335e-05,
+      "loss": 2.5873,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6085714285714285,
+      "grad_norm": 102.07270050048828,
+      "learning_rate": 1.5943809523809524e-05,
+      "loss": 0.8144,
+      "step": 4260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.15384997427463531,
+      "learning_rate": 1.5934285714285717e-05,
+      "loss": 1.6372,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6114285714285714,
+      "grad_norm": 1497.6566162109375,
+      "learning_rate": 1.5924761904761906e-05,
+      "loss": 1.6852,
+      "step": 4280
+    },
+    {
+      "epoch": 0.6128571428571429,
+      "grad_norm": 284105024.0,
+      "learning_rate": 1.5915238095238098e-05,
+      "loss": 0.5114,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6142857142857143,
+      "grad_norm": 0.028531407937407494,
+      "learning_rate": 1.5905714285714287e-05,
+      "loss": 1.6759,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6157142857142858,
+      "grad_norm": 0.23290741443634033,
+      "learning_rate": 1.589619047619048e-05,
+      "loss": 1.6183,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6171428571428571,
+      "grad_norm": 0.23219768702983856,
+      "learning_rate": 1.5886666666666668e-05,
+      "loss": 1.7323,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6185714285714285,
+      "grad_norm": 0.5855656266212463,
+      "learning_rate": 1.5877142857142857e-05,
+      "loss": 0.8075,
+      "step": 4330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7736623287200928,
+      "learning_rate": 1.586761904761905e-05,
+      "loss": 1.6709,
+      "step": 4340
+    },
+    {
+      "epoch": 0.6214285714285714,
+      "grad_norm": 0.15940716862678528,
+      "learning_rate": 1.585809523809524e-05,
+      "loss": 2.557,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6228571428571429,
+      "grad_norm": 0.05566216632723808,
+      "learning_rate": 1.584857142857143e-05,
+      "loss": 2.2228,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6242857142857143,
+      "grad_norm": 0.31042152643203735,
+      "learning_rate": 1.583904761904762e-05,
+      "loss": 2.2638,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6257142857142857,
+      "grad_norm": 0.2624918222427368,
+      "learning_rate": 1.582952380952381e-05,
+      "loss": 0.7863,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6271428571428571,
+      "grad_norm": 0.027225211262702942,
+      "learning_rate": 1.582e-05,
+      "loss": 1.5694,
+      "step": 4390
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 0.11011450737714767,
+      "learning_rate": 1.581047619047619e-05,
+      "loss": 0.8092,
+      "step": 4400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1561906933784485,
+      "learning_rate": 1.5800952380952382e-05,
+      "loss": 1.456,
+      "step": 4410
+    },
+    {
+      "epoch": 0.6314285714285715,
+      "grad_norm": 0.011988009326159954,
+      "learning_rate": 1.5791428571428575e-05,
+      "loss": 3.1073,
+      "step": 4420
+    },
+    {
+      "epoch": 0.6328571428571429,
+      "grad_norm": 0.052196986973285675,
+      "learning_rate": 1.5781904761904764e-05,
+      "loss": 0.8104,
+      "step": 4430
+    },
+    {
+      "epoch": 0.6342857142857142,
+      "grad_norm": 4848959.0,
+      "learning_rate": 1.5772380952380956e-05,
+      "loss": 4.0612,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6357142857142857,
+      "grad_norm": 1676278.75,
+      "learning_rate": 1.5762857142857145e-05,
+      "loss": 1.646,
+      "step": 4450
+    },
+    {
+      "epoch": 0.6371428571428571,
+      "grad_norm": 0.1543026715517044,
+      "learning_rate": 1.5753333333333334e-05,
+      "loss": 1.5677,
+      "step": 4460
+    },
+    {
+      "epoch": 0.6385714285714286,
+      "grad_norm": 0.03917285054922104,
+      "learning_rate": 1.5743809523809526e-05,
+      "loss": 0.7225,
+      "step": 4470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11666043.0,
+      "learning_rate": 1.5734285714285715e-05,
+      "loss": 0.7108,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6414285714285715,
+      "grad_norm": 0.08988025039434433,
+      "learning_rate": 1.5724761904761907e-05,
+      "loss": 4.0472,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.08803051710128784,
+      "learning_rate": 1.5715238095238096e-05,
+      "loss": 0.0,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6442857142857142,
+      "grad_norm": 0.06431713700294495,
+      "learning_rate": 1.5705714285714285e-05,
+      "loss": 0.8055,
+      "step": 4510
+    },
+    {
+      "epoch": 0.6457142857142857,
+      "grad_norm": 0.030746005475521088,
+      "learning_rate": 1.5696190476190478e-05,
+      "loss": 2.4603,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6471428571428571,
+      "grad_norm": 111501864.0,
+      "learning_rate": 1.5686666666666667e-05,
+      "loss": 0.7225,
+      "step": 4530
+    },
+    {
+      "epoch": 0.6485714285714286,
+      "grad_norm": 0.07517790794372559,
+      "learning_rate": 1.567714285714286e-05,
+      "loss": 1.6432,
+      "step": 4540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.09191897511482239,
+      "learning_rate": 1.5667619047619048e-05,
+      "loss": 0.8216,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6514285714285715,
+      "grad_norm": 0.050468698143959045,
+      "learning_rate": 1.565809523809524e-05,
+      "loss": 0.8377,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6528571428571428,
+      "grad_norm": 0.05092250555753708,
+      "learning_rate": 1.564857142857143e-05,
+      "loss": 1.6338,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6542857142857142,
+      "grad_norm": 0.45415133237838745,
+      "learning_rate": 1.563904761904762e-05,
+      "loss": 0.7838,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6557142857142857,
+      "grad_norm": 0.023216435685753822,
+      "learning_rate": 1.562952380952381e-05,
+      "loss": 0.0,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6571428571428571,
+      "grad_norm": 0.11792700737714767,
+      "learning_rate": 1.5620000000000003e-05,
+      "loss": 0.5726,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6585714285714286,
+      "grad_norm": 0.016707276925444603,
+      "learning_rate": 1.5610476190476192e-05,
+      "loss": 0.8856,
+      "step": 4610
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.03643623739480972,
+      "learning_rate": 1.5600952380952384e-05,
+      "loss": 0.9279,
+      "step": 4620
+    },
+    {
+      "epoch": 0.6614285714285715,
+      "grad_norm": 0.03741033375263214,
+      "learning_rate": 1.5591428571428573e-05,
+      "loss": 0.8108,
+      "step": 4630
+    },
+    {
+      "epoch": 0.6628571428571428,
+      "grad_norm": 0.009000065736472607,
+      "learning_rate": 1.5581904761904762e-05,
+      "loss": 1.7023,
+      "step": 4640
+    },
+    {
+      "epoch": 0.6642857142857143,
+      "grad_norm": 0.05466756969690323,
+      "learning_rate": 1.5572380952380954e-05,
+      "loss": 2.6071,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6657142857142857,
+      "grad_norm": 0.06815799325704575,
+      "learning_rate": 1.5562857142857143e-05,
+      "loss": 0.0,
+      "step": 4660
+    },
+    {
+      "epoch": 0.6671428571428571,
+      "grad_norm": 0.015344436280429363,
+      "learning_rate": 1.5553333333333332e-05,
+      "loss": 2.3252,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6685714285714286,
+      "grad_norm": 0.007389787118881941,
+      "learning_rate": 1.5543809523809525e-05,
+      "loss": 1.3197,
+      "step": 4680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29478850960731506,
+      "learning_rate": 1.5534285714285714e-05,
+      "loss": 3.2803,
+      "step": 4690
+    },
+    {
+      "epoch": 0.6714285714285714,
+      "grad_norm": 0.015877464786171913,
+      "learning_rate": 1.5524761904761906e-05,
+      "loss": 1.7069,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6728571428571428,
+      "grad_norm": 1.0021435022354126,
+      "learning_rate": 1.5515238095238095e-05,
+      "loss": 2.4961,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6742857142857143,
+      "grad_norm": 0.22391639649868011,
+      "learning_rate": 1.5505714285714287e-05,
+      "loss": 2.2959,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6757142857142857,
+      "grad_norm": 0.3877425789833069,
+      "learning_rate": 1.549619047619048e-05,
+      "loss": 2.0702,
+      "step": 4730
+    },
+    {
+      "epoch": 0.6771428571428572,
+      "grad_norm": 0.06968878209590912,
+      "learning_rate": 1.548666666666667e-05,
+      "loss": 1.7748,
+      "step": 4740
+    },
+    {
+      "epoch": 0.6785714285714286,
+      "grad_norm": 0.10345824807882309,
+      "learning_rate": 1.547714285714286e-05,
+      "loss": 1.1181,
+      "step": 4750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3385922908782959,
+      "learning_rate": 1.546761904761905e-05,
+      "loss": 0.7649,
+      "step": 4760
+    },
+    {
+      "epoch": 0.6814285714285714,
+      "grad_norm": 12207911.0,
+      "learning_rate": 1.545809523809524e-05,
+      "loss": 3.3313,
+      "step": 4770
+    },
+    {
+      "epoch": 0.6828571428571428,
+      "grad_norm": 0.035624485462903976,
+      "learning_rate": 1.544857142857143e-05,
+      "loss": 0.0,
+      "step": 4780
+    },
+    {
+      "epoch": 0.6842857142857143,
+      "grad_norm": 0.049205634742975235,
+      "learning_rate": 1.543904761904762e-05,
+      "loss": 1.7713,
+      "step": 4790
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 6889844.5,
+      "learning_rate": 1.542952380952381e-05,
+      "loss": 2.2819,
+      "step": 4800
+    },
+    {
+      "epoch": 0.6871428571428572,
+      "grad_norm": 2614956.75,
+      "learning_rate": 1.542e-05,
+      "loss": 2.5574,
+      "step": 4810
+    },
+    {
+      "epoch": 0.6885714285714286,
+      "grad_norm": 0.3002869188785553,
+      "learning_rate": 1.541047619047619e-05,
+      "loss": 1.6936,
+      "step": 4820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0481896810233593,
+      "learning_rate": 1.5400952380952383e-05,
+      "loss": 1.9114,
+      "step": 4830
+    },
+    {
+      "epoch": 0.6914285714285714,
+      "grad_norm": 0.061788108199834824,
+      "learning_rate": 1.539142857142857e-05,
+      "loss": 1.6621,
+      "step": 4840
+    },
+    {
+      "epoch": 0.6928571428571428,
+      "grad_norm": 0.0743480920791626,
+      "learning_rate": 1.5381904761904764e-05,
+      "loss": 0.8806,
+      "step": 4850
+    },
+    {
+      "epoch": 0.6942857142857143,
+      "grad_norm": 0.039855096489191055,
+      "learning_rate": 1.5372380952380953e-05,
+      "loss": 1.6578,
+      "step": 4860
+    },
+    {
+      "epoch": 0.6957142857142857,
+      "grad_norm": 14652494.0,
+      "learning_rate": 1.5362857142857145e-05,
+      "loss": 1.5516,
+      "step": 4870
+    },
+    {
+      "epoch": 0.6971428571428572,
+      "grad_norm": 0.433768630027771,
+      "learning_rate": 1.5353333333333334e-05,
+      "loss": 2.2015,
+      "step": 4880
+    },
+    {
+      "epoch": 0.6985714285714286,
+      "grad_norm": 0.02782445400953293,
+      "learning_rate": 1.5343809523809527e-05,
+      "loss": 0.5954,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.01976465992629528,
+      "learning_rate": 1.5334285714285716e-05,
+      "loss": 0.0,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7014285714285714,
+      "grad_norm": 1936457.875,
+      "learning_rate": 1.5324761904761908e-05,
+      "loss": 2.678,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7028571428571428,
+      "grad_norm": 0.20556390285491943,
+      "learning_rate": 1.5315238095238097e-05,
+      "loss": 3.2382,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7042857142857143,
+      "grad_norm": 0.14977577328681946,
+      "learning_rate": 1.5305714285714286e-05,
+      "loss": 0.527,
+      "step": 4930
+    },
+    {
+      "epoch": 0.7057142857142857,
+      "grad_norm": 4.969018459320068,
+      "learning_rate": 1.5296190476190478e-05,
+      "loss": 0.7928,
+      "step": 4940
+    },
+    {
+      "epoch": 0.7071428571428572,
+      "grad_norm": 3442071.0,
+      "learning_rate": 1.5286666666666667e-05,
+      "loss": 2.4703,
+      "step": 4950
+    },
+    {
+      "epoch": 0.7085714285714285,
+      "grad_norm": 0.18574975430965424,
+      "learning_rate": 1.527714285714286e-05,
+      "loss": 0.0002,
+      "step": 4960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7152432203292847,
+      "learning_rate": 1.526761904761905e-05,
+      "loss": 3.9767,
+      "step": 4970
+    },
+    {
+      "epoch": 0.7114285714285714,
+      "grad_norm": 0.27751776576042175,
+      "learning_rate": 1.5258095238095239e-05,
+      "loss": 0.0,
+      "step": 4980
+    },
+    {
+      "epoch": 0.7128571428571429,
+      "grad_norm": 0.039866797626018524,
+      "learning_rate": 1.524857142857143e-05,
+      "loss": 0.0,
+      "step": 4990
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.03406675532460213,
+      "learning_rate": 1.523904761904762e-05,
+      "loss": 2.1839,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7157142857142857,
+      "grad_norm": 1495503.125,
+      "learning_rate": 1.5229523809523811e-05,
+      "loss": 2.4481,
+      "step": 5010
+    },
+    {
+      "epoch": 0.7171428571428572,
+      "grad_norm": 0.20187106728553772,
+      "learning_rate": 1.5220000000000002e-05,
+      "loss": 1.6172,
+      "step": 5020
+    },
+    {
+      "epoch": 0.7185714285714285,
+      "grad_norm": 2.8645334243774414,
+      "learning_rate": 1.521047619047619e-05,
+      "loss": 0.0,
+      "step": 5030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8273412585258484,
+      "learning_rate": 1.5200952380952383e-05,
+      "loss": 0.7804,
+      "step": 5040
+    },
+    {
+      "epoch": 0.7214285714285714,
+      "grad_norm": 0.025248244404792786,
+      "learning_rate": 1.5191428571428572e-05,
+      "loss": 2.6018,
+      "step": 5050
+    },
+    {
+      "epoch": 0.7228571428571429,
+      "grad_norm": 0.023136647418141365,
+      "learning_rate": 1.5181904761904763e-05,
+      "loss": 0.0,
+      "step": 5060
+    },
+    {
+      "epoch": 0.7242857142857143,
+      "grad_norm": 0.02263324148952961,
+      "learning_rate": 1.5172380952380953e-05,
+      "loss": 0.0,
+      "step": 5070
+    },
+    {
+      "epoch": 0.7257142857142858,
+      "grad_norm": 0.03192922845482826,
+      "learning_rate": 1.5162857142857144e-05,
+      "loss": 0.0,
+      "step": 5080
+    },
+    {
+      "epoch": 0.7271428571428571,
+      "grad_norm": 0.031805649399757385,
+      "learning_rate": 1.5153333333333336e-05,
+      "loss": 3.0023,
+      "step": 5090
+    },
+    {
+      "epoch": 0.7285714285714285,
+      "grad_norm": 0.07224005460739136,
+      "learning_rate": 1.5143809523809525e-05,
+      "loss": 0.8982,
+      "step": 5100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.02881191298365593,
+      "learning_rate": 1.5134285714285714e-05,
+      "loss": 2.4058,
+      "step": 5110
+    },
+    {
+      "epoch": 0.7314285714285714,
+      "grad_norm": 0.042384617030620575,
+      "learning_rate": 1.5124761904761906e-05,
+      "loss": 2.1204,
+      "step": 5120
+    },
+    {
+      "epoch": 0.7328571428571429,
+      "grad_norm": 0.030775081366300583,
+      "learning_rate": 1.5115238095238095e-05,
+      "loss": 0.7534,
+      "step": 5130
+    },
+    {
+      "epoch": 0.7342857142857143,
+      "grad_norm": 0.010330983437597752,
+      "learning_rate": 1.5105714285714288e-05,
+      "loss": 0.0,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7357142857142858,
+      "grad_norm": 0.024451762437820435,
+      "learning_rate": 1.5096190476190478e-05,
+      "loss": 1.6723,
+      "step": 5150
+    },
+    {
+      "epoch": 0.7371428571428571,
+      "grad_norm": 0.028540106490254402,
+      "learning_rate": 1.5086666666666667e-05,
+      "loss": 0.9306,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7385714285714285,
+      "grad_norm": 0.38035863637924194,
+      "learning_rate": 1.507714285714286e-05,
+      "loss": 0.0,
+      "step": 5170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.09059204906225204,
+      "learning_rate": 1.5067619047619049e-05,
+      "loss": 0.0,
+      "step": 5180
+    },
+    {
+      "epoch": 0.7414285714285714,
+      "grad_norm": 0.01671271026134491,
+      "learning_rate": 1.5058095238095238e-05,
+      "loss": 1.7543,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 0.04799080267548561,
+      "learning_rate": 1.504857142857143e-05,
+      "loss": 0.7536,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7442857142857143,
+      "grad_norm": 0.0492214635014534,
+      "learning_rate": 1.503904761904762e-05,
+      "loss": 1.3643,
+      "step": 5210
+    },
+    {
+      "epoch": 0.7457142857142857,
+      "grad_norm": 0.014559627510607243,
+      "learning_rate": 1.5029523809523811e-05,
+      "loss": 0.6892,
+      "step": 5220
+    },
+    {
+      "epoch": 0.7471428571428571,
+      "grad_norm": 0.03892429172992706,
+      "learning_rate": 1.5020000000000002e-05,
+      "loss": 0.0,
+      "step": 5230
+    },
+    {
+      "epoch": 0.7485714285714286,
+      "grad_norm": 1764341.125,
+      "learning_rate": 1.501047619047619e-05,
+      "loss": 2.5398,
+      "step": 5240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4894507.0,
+      "learning_rate": 1.5000952380952383e-05,
+      "loss": 1.7224,
+      "step": 5250
+    },
+    {
+      "epoch": 0.7514285714285714,
+      "grad_norm": 0.03135019913315773,
+      "learning_rate": 1.4991428571428572e-05,
+      "loss": 2.661,
+      "step": 5260
+    },
+    {
+      "epoch": 0.7528571428571429,
+      "grad_norm": 2519961.5,
+      "learning_rate": 1.4981904761904764e-05,
+      "loss": 3.0068,
+      "step": 5270
+    },
+    {
+      "epoch": 0.7542857142857143,
+      "grad_norm": 3.765960454940796,
+      "learning_rate": 1.4972380952380953e-05,
+      "loss": 0.0,
+      "step": 5280
+    },
+    {
+      "epoch": 0.7557142857142857,
+      "grad_norm": 0.0468648299574852,
+      "learning_rate": 1.4962857142857144e-05,
+      "loss": 2.4229,
+      "step": 5290
+    },
+    {
+      "epoch": 0.7571428571428571,
+      "grad_norm": 0.040297310799360275,
+      "learning_rate": 1.4953333333333335e-05,
+      "loss": 0.0,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7585714285714286,
+      "grad_norm": 0.1936325579881668,
+      "learning_rate": 1.4943809523809525e-05,
+      "loss": 2.5875,
+      "step": 5310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.58046954870224,
+      "learning_rate": 1.4934285714285714e-05,
+      "loss": 1.5567,
+      "step": 5320
+    },
+    {
+      "epoch": 0.7614285714285715,
+      "grad_norm": 0.2597913146018982,
+      "learning_rate": 1.4924761904761907e-05,
+      "loss": 0.8234,
+      "step": 5330
+    },
+    {
+      "epoch": 0.7628571428571429,
+      "grad_norm": 2260657.25,
+      "learning_rate": 1.4915238095238096e-05,
+      "loss": 1.683,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7642857142857142,
+      "grad_norm": 5.085236549377441,
+      "learning_rate": 1.4905714285714288e-05,
+      "loss": 3.128,
+      "step": 5350
+    },
+    {
+      "epoch": 0.7657142857142857,
+      "grad_norm": 0.19006921350955963,
+      "learning_rate": 1.4896190476190477e-05,
+      "loss": 0.7481,
+      "step": 5360
+    },
+    {
+      "epoch": 0.7671428571428571,
+      "grad_norm": 0.038023192435503006,
+      "learning_rate": 1.4886666666666668e-05,
+      "loss": 0.0,
+      "step": 5370
+    },
+    {
+      "epoch": 0.7685714285714286,
+      "grad_norm": 0.4512747526168823,
+      "learning_rate": 1.4877142857142858e-05,
+      "loss": 3.4095,
+      "step": 5380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3925737.0,
+      "learning_rate": 1.4867619047619049e-05,
+      "loss": 2.4748,
+      "step": 5390
+    },
+    {
+      "epoch": 0.7714285714285715,
+      "grad_norm": 0.2081717550754547,
+      "learning_rate": 1.485809523809524e-05,
+      "loss": 0.7847,
+      "step": 5400
+    },
+    {
+      "epoch": 0.7728571428571429,
+      "grad_norm": 8309592.5,
+      "learning_rate": 1.484857142857143e-05,
+      "loss": 1.6296,
+      "step": 5410
+    },
+    {
+      "epoch": 0.7742857142857142,
+      "grad_norm": 0.018125707283616066,
+      "learning_rate": 1.4839047619047619e-05,
+      "loss": 0.7555,
+      "step": 5420
+    },
+    {
+      "epoch": 0.7757142857142857,
+      "grad_norm": 3352556.25,
+      "learning_rate": 1.4829523809523811e-05,
+      "loss": 2.6838,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7771428571428571,
+      "grad_norm": 4618625.0,
+      "learning_rate": 1.482e-05,
+      "loss": 1.7289,
+      "step": 5440
+    },
+    {
+      "epoch": 0.7785714285714286,
+      "grad_norm": 0.06682883948087692,
+      "learning_rate": 1.4810476190476191e-05,
+      "loss": 0.8177,
+      "step": 5450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.021823491901159286,
+      "learning_rate": 1.4800952380952382e-05,
+      "loss": 2.405,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7814285714285715,
+      "grad_norm": 11929681.0,
+      "learning_rate": 1.4791428571428572e-05,
+      "loss": 1.6453,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7828571428571428,
+      "grad_norm": 0.20002047717571259,
+      "learning_rate": 1.4781904761904765e-05,
+      "loss": 0.7483,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7842857142857143,
+      "grad_norm": 0.04473010078072548,
+      "learning_rate": 1.4772380952380954e-05,
+      "loss": 0.0,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 5089199.0,
+      "learning_rate": 1.4762857142857143e-05,
+      "loss": 2.4129,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7871428571428571,
+      "grad_norm": 0.22969965636730194,
+      "learning_rate": 1.4753333333333335e-05,
+      "loss": 1.7155,
+      "step": 5510
+    },
+    {
+      "epoch": 0.7885714285714286,
+      "grad_norm": 0.09568771719932556,
+      "learning_rate": 1.4743809523809524e-05,
+      "loss": 0.7618,
+      "step": 5520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1725606471300125,
+      "learning_rate": 1.4734285714285716e-05,
+      "loss": 1.8068,
+      "step": 5530
+    },
+    {
+      "epoch": 0.7914285714285715,
+      "grad_norm": 1.1473311185836792,
+      "learning_rate": 1.4724761904761907e-05,
+      "loss": 1.5703,
+      "step": 5540
+    },
+    {
+      "epoch": 0.7928571428571428,
+      "grad_norm": 8073361.5,
+      "learning_rate": 1.4715238095238096e-05,
+      "loss": 3.8896,
+      "step": 5550
+    },
+    {
+      "epoch": 0.7942857142857143,
+      "grad_norm": 0.03440004587173462,
+      "learning_rate": 1.4705714285714288e-05,
+      "loss": 3.9302,
+      "step": 5560
+    },
+    {
+      "epoch": 0.7957142857142857,
+      "grad_norm": 0.29978862404823303,
+      "learning_rate": 1.4696190476190477e-05,
+      "loss": 0.8179,
+      "step": 5570
+    },
+    {
+      "epoch": 0.7971428571428572,
+      "grad_norm": 0.09151644259691238,
+      "learning_rate": 1.4686666666666666e-05,
+      "loss": 0.4076,
+      "step": 5580
+    },
+    {
+      "epoch": 0.7985714285714286,
+      "grad_norm": 30.097061157226562,
+      "learning_rate": 1.4677142857142858e-05,
+      "loss": 1.6611,
+      "step": 5590
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.030335601419210434,
+      "learning_rate": 1.4667619047619049e-05,
+      "loss": 0.8214,
+      "step": 5600
+    },
+    {
+      "epoch": 0.8014285714285714,
+      "grad_norm": 0.025141309946775436,
+      "learning_rate": 1.465809523809524e-05,
+      "loss": 0.865,
+      "step": 5610
+    },
+    {
+      "epoch": 0.8028571428571428,
+      "grad_norm": 0.10917967557907104,
+      "learning_rate": 1.464857142857143e-05,
+      "loss": 0.8827,
+      "step": 5620
+    },
+    {
+      "epoch": 0.8042857142857143,
+      "grad_norm": 0.04122152179479599,
+      "learning_rate": 1.463904761904762e-05,
+      "loss": 0.8521,
+      "step": 5630
+    },
+    {
+      "epoch": 0.8057142857142857,
+      "grad_norm": 0.1017841324210167,
+      "learning_rate": 1.4629523809523812e-05,
+      "loss": 1.5681,
+      "step": 5640
+    },
+    {
+      "epoch": 0.8071428571428572,
+      "grad_norm": 6142818.0,
+      "learning_rate": 1.462e-05,
+      "loss": 2.2192,
+      "step": 5650
+    },
+    {
+      "epoch": 0.8085714285714286,
+      "grad_norm": 0.06244712695479393,
+      "learning_rate": 1.4610476190476193e-05,
+      "loss": 2.3877,
+      "step": 5660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15657942.0,
+      "learning_rate": 1.4600952380952382e-05,
+      "loss": 2.3634,
+      "step": 5670
+    },
+    {
+      "epoch": 0.8114285714285714,
+      "grad_norm": 0.0383012518286705,
+      "learning_rate": 1.4591428571428573e-05,
+      "loss": 0.8709,
+      "step": 5680
+    },
+    {
+      "epoch": 0.8128571428571428,
+      "grad_norm": 0.05412408709526062,
+      "learning_rate": 1.4581904761904763e-05,
+      "loss": 3.151,
+      "step": 5690
+    },
+    {
+      "epoch": 0.8142857142857143,
+      "grad_norm": 2078394.375,
+      "learning_rate": 1.4572380952380954e-05,
+      "loss": 3.2697,
+      "step": 5700
+    },
+    {
+      "epoch": 0.8157142857142857,
+      "grad_norm": 2149895.5,
+      "learning_rate": 1.4562857142857143e-05,
+      "loss": 2.9575,
+      "step": 5710
+    },
+    {
+      "epoch": 0.8171428571428572,
+      "grad_norm": 2821806.25,
+      "learning_rate": 1.4553333333333335e-05,
+      "loss": 3.0106,
+      "step": 5720
+    },
+    {
+      "epoch": 0.8185714285714286,
+      "grad_norm": 0.30308884382247925,
+      "learning_rate": 1.4543809523809524e-05,
+      "loss": 0.0,
+      "step": 5730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.08973406255245209,
+      "learning_rate": 1.4534285714285716e-05,
+      "loss": 0.0,
+      "step": 5740
+    },
+    {
+      "epoch": 0.8214285714285714,
+      "grad_norm": 0.0339081734418869,
+      "learning_rate": 1.4524761904761905e-05,
+      "loss": 1.726,
+      "step": 5750
+    },
+    {
+      "epoch": 0.8228571428571428,
+      "grad_norm": 1604376.875,
+      "learning_rate": 1.4515238095238096e-05,
+      "loss": 1.8279,
+      "step": 5760
+    },
+    {
+      "epoch": 0.8242857142857143,
+      "grad_norm": 0.09628063440322876,
+      "learning_rate": 1.4505714285714287e-05,
+      "loss": 1.713,
+      "step": 5770
+    },
+    {
+      "epoch": 0.8257142857142857,
+      "grad_norm": 0.023135002702474594,
+      "learning_rate": 1.4496190476190477e-05,
+      "loss": 0.0,
+      "step": 5780
+    },
+    {
+      "epoch": 0.8271428571428572,
+      "grad_norm": 0.017433326691389084,
+      "learning_rate": 1.448666666666667e-05,
+      "loss": 0.7461,
+      "step": 5790
+    },
+    {
+      "epoch": 0.8285714285714286,
+      "grad_norm": 0.3067975342273712,
+      "learning_rate": 1.4477142857142859e-05,
+      "loss": 0.0,
+      "step": 5800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.018409695476293564,
+      "learning_rate": 1.4467619047619048e-05,
+      "loss": 0.0,
+      "step": 5810
+    },
+    {
+      "epoch": 0.8314285714285714,
+      "grad_norm": 0.0625164806842804,
+      "learning_rate": 1.445809523809524e-05,
+      "loss": 0.9618,
+      "step": 5820
+    },
+    {
+      "epoch": 0.8328571428571429,
+      "grad_norm": 0.05618647113442421,
+      "learning_rate": 1.4448571428571429e-05,
+      "loss": 0.0,
+      "step": 5830
+    },
+    {
+      "epoch": 0.8342857142857143,
+      "grad_norm": 0.047780610620975494,
+      "learning_rate": 1.443904761904762e-05,
+      "loss": 1.7724,
+      "step": 5840
+    },
+    {
+      "epoch": 0.8357142857142857,
+      "grad_norm": 0.06281869113445282,
+      "learning_rate": 1.442952380952381e-05,
+      "loss": 0.0,
+      "step": 5850
+    },
+    {
+      "epoch": 0.8371428571428572,
+      "grad_norm": 0.07008826732635498,
+      "learning_rate": 1.4420000000000001e-05,
+      "loss": 0.9328,
+      "step": 5860
+    },
+    {
+      "epoch": 0.8385714285714285,
+      "grad_norm": 0.03894483670592308,
+      "learning_rate": 1.4410476190476193e-05,
+      "loss": 0.8263,
+      "step": 5870
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.05758967623114586,
+      "learning_rate": 1.4400952380952382e-05,
+      "loss": 0.0,
+      "step": 5880
+    },
+    {
+      "epoch": 0.8414285714285714,
+      "grad_norm": 0.03688669949769974,
+      "learning_rate": 1.4391428571428571e-05,
+      "loss": 0.8443,
+      "step": 5890
+    },
+    {
+      "epoch": 0.8428571428571429,
+      "grad_norm": 0.11691343784332275,
+      "learning_rate": 1.4381904761904763e-05,
+      "loss": 0.0,
+      "step": 5900
+    },
+    {
+      "epoch": 0.8442857142857143,
+      "grad_norm": 0.03900890052318573,
+      "learning_rate": 1.4372380952380952e-05,
+      "loss": 2.4718,
+      "step": 5910
+    },
+    {
+      "epoch": 0.8457142857142858,
+      "grad_norm": 1789943.125,
+      "learning_rate": 1.4362857142857145e-05,
+      "loss": 1.7603,
+      "step": 5920
+    },
+    {
+      "epoch": 0.8471428571428572,
+      "grad_norm": 0.18674631416797638,
+      "learning_rate": 1.4353333333333335e-05,
+      "loss": 1.6927,
+      "step": 5930
+    },
+    {
+      "epoch": 0.8485714285714285,
+      "grad_norm": 0.07008085399866104,
+      "learning_rate": 1.4343809523809524e-05,
+      "loss": 2.4182,
+      "step": 5940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8286402821540833,
+      "learning_rate": 1.4334285714285717e-05,
+      "loss": 1.6454,
+      "step": 5950
+    },
+    {
+      "epoch": 0.8514285714285714,
+      "grad_norm": 0.09325440227985382,
+      "learning_rate": 1.4324761904761906e-05,
+      "loss": 3.2833,
+      "step": 5960
+    },
+    {
+      "epoch": 0.8528571428571429,
+      "grad_norm": 0.11135806888341904,
+      "learning_rate": 1.4315238095238095e-05,
+      "loss": 1.6303,
+      "step": 5970
+    },
+    {
+      "epoch": 0.8542857142857143,
+      "grad_norm": 0.04822812229394913,
+      "learning_rate": 1.4305714285714287e-05,
+      "loss": 1.6242,
+      "step": 5980
+    },
+    {
+      "epoch": 0.8557142857142858,
+      "grad_norm": 0.13590972125530243,
+      "learning_rate": 1.4296190476190478e-05,
+      "loss": 3.1999,
+      "step": 5990
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.579108476638794,
+      "learning_rate": 1.4286666666666668e-05,
+      "loss": 0.7803,
+      "step": 6000
+    },
+    {
+      "epoch": 0.8585714285714285,
+      "grad_norm": 0.09968885034322739,
+      "learning_rate": 1.4277142857142859e-05,
+      "loss": 0.8302,
+      "step": 6010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1327065229415894,
+      "learning_rate": 1.4267619047619048e-05,
+      "loss": 0.7951,
+      "step": 6020
+    },
+    {
+      "epoch": 0.8614285714285714,
+      "grad_norm": 0.05613395944237709,
+      "learning_rate": 1.425809523809524e-05,
+      "loss": 1.684,
+      "step": 6030
+    },
+    {
+      "epoch": 0.8628571428571429,
+      "grad_norm": 12.744712829589844,
+      "learning_rate": 1.4248571428571429e-05,
+      "loss": 0.8333,
+      "step": 6040
+    },
+    {
+      "epoch": 0.8642857142857143,
+      "grad_norm": 5565093.5,
+      "learning_rate": 1.4239047619047621e-05,
+      "loss": 4.2774,
+      "step": 6050
+    },
+    {
+      "epoch": 0.8657142857142858,
+      "grad_norm": 0.38834524154663086,
+      "learning_rate": 1.422952380952381e-05,
+      "loss": 0.8024,
+      "step": 6060
+    },
+    {
+      "epoch": 0.8671428571428571,
+      "grad_norm": 0.3369869887828827,
+      "learning_rate": 1.4220000000000001e-05,
+      "loss": 0.904,
+      "step": 6070
+    },
+    {
+      "epoch": 0.8685714285714285,
+      "grad_norm": 0.35916051268577576,
+      "learning_rate": 1.4210476190476192e-05,
+      "loss": 0.7889,
+      "step": 6080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.04615074023604393,
+      "learning_rate": 1.4200952380952382e-05,
+      "loss": 0.0,
+      "step": 6090
+    },
+    {
+      "epoch": 0.8714285714285714,
+      "grad_norm": 0.043160345405340195,
+      "learning_rate": 1.4191428571428571e-05,
+      "loss": 4.3018,
+      "step": 6100
+    },
+    {
+      "epoch": 0.8728571428571429,
+      "grad_norm": 0.1539774239063263,
+      "learning_rate": 1.4181904761904764e-05,
+      "loss": 0.0,
+      "step": 6110
+    },
+    {
+      "epoch": 0.8742857142857143,
+      "grad_norm": 0.03139315918087959,
+      "learning_rate": 1.4172380952380953e-05,
+      "loss": 0.0,
+      "step": 6120
+    },
+    {
+      "epoch": 0.8757142857142857,
+      "grad_norm": 0.07688730210065842,
+      "learning_rate": 1.4162857142857145e-05,
+      "loss": 0.8926,
+      "step": 6130
+    },
+    {
+      "epoch": 0.8771428571428571,
+      "grad_norm": 0.05393550172448158,
+      "learning_rate": 1.4153333333333334e-05,
+      "loss": 0.4848,
+      "step": 6140
+    },
+    {
+      "epoch": 0.8785714285714286,
+      "grad_norm": 0.05173574388027191,
+      "learning_rate": 1.4143809523809525e-05,
+      "loss": 0.0,
+      "step": 6150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.027865346521139145,
+      "learning_rate": 1.4134285714285715e-05,
+      "loss": 1.7565,
+      "step": 6160
+    },
+    {
+      "epoch": 0.8814285714285715,
+      "grad_norm": 0.4441341757774353,
+      "learning_rate": 1.4124761904761906e-05,
+      "loss": 0.8876,
+      "step": 6170
+    },
+    {
+      "epoch": 0.8828571428571429,
+      "grad_norm": 0.028943132609128952,
+      "learning_rate": 1.4115238095238098e-05,
+      "loss": 0.5953,
+      "step": 6180
+    },
+    {
+      "epoch": 0.8842857142857142,
+      "grad_norm": 0.1134498193860054,
+      "learning_rate": 1.4105714285714287e-05,
+      "loss": 2.4883,
+      "step": 6190
+    },
+    {
+      "epoch": 0.8857142857142857,
+      "grad_norm": 0.046840086579322815,
+      "learning_rate": 1.4096190476190476e-05,
+      "loss": 1.4985,
+      "step": 6200
+    },
+    {
+      "epoch": 0.8871428571428571,
+      "grad_norm": 0.13220427930355072,
+      "learning_rate": 1.4086666666666668e-05,
+      "loss": 1.462,
+      "step": 6210
+    },
+    {
+      "epoch": 0.8885714285714286,
+      "grad_norm": 401825.6875,
+      "learning_rate": 1.4077142857142857e-05,
+      "loss": 0.6985,
+      "step": 6220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2233458.25,
+      "learning_rate": 1.4067619047619048e-05,
+      "loss": 2.6779,
+      "step": 6230
+    },
+    {
+      "epoch": 0.8914285714285715,
+      "grad_norm": 0.17707909643650055,
+      "learning_rate": 1.405809523809524e-05,
+      "loss": 0.8122,
+      "step": 6240
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 2873446.75,
+      "learning_rate": 1.404857142857143e-05,
+      "loss": 2.482,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8942857142857142,
+      "grad_norm": 2447852.0,
+      "learning_rate": 1.4039047619047622e-05,
+      "loss": 3.2592,
+      "step": 6260
+    },
+    {
+      "epoch": 0.8957142857142857,
+      "grad_norm": 0.5254727602005005,
+      "learning_rate": 1.402952380952381e-05,
+      "loss": 0.0,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8971428571428571,
+      "grad_norm": 0.7800436019897461,
+      "learning_rate": 1.402e-05,
+      "loss": 3.2526,
+      "step": 6280
+    },
+    {
+      "epoch": 0.8985714285714286,
+      "grad_norm": 0.0733805000782013,
+      "learning_rate": 1.4010476190476192e-05,
+      "loss": 0.8553,
+      "step": 6290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.03802318871021271,
+      "learning_rate": 1.4000952380952381e-05,
+      "loss": 1.5417,
+      "step": 6300
+    },
+    {
+      "epoch": 0.9014285714285715,
+      "grad_norm": 0.029381372034549713,
+      "learning_rate": 1.3991428571428573e-05,
+      "loss": 1.683,
+      "step": 6310
+    },
+    {
+      "epoch": 0.9028571428571428,
+      "grad_norm": 0.052462879568338394,
+      "learning_rate": 1.3981904761904764e-05,
+      "loss": 0.8425,
+      "step": 6320
+    },
+    {
+      "epoch": 0.9042857142857142,
+      "grad_norm": 0.08147754520177841,
+      "learning_rate": 1.3972380952380953e-05,
+      "loss": 1.5305,
+      "step": 6330
+    },
+    {
+      "epoch": 0.9057142857142857,
+      "grad_norm": 0.08017618209123611,
+      "learning_rate": 1.3962857142857145e-05,
+      "loss": 0.7424,
+      "step": 6340
+    },
+    {
+      "epoch": 0.9071428571428571,
+      "grad_norm": 0.03884577378630638,
+      "learning_rate": 1.3953333333333334e-05,
+      "loss": 0.4332,
+      "step": 6350
+    },
+    {
+      "epoch": 0.9085714285714286,
+      "grad_norm": 0.05725259333848953,
+      "learning_rate": 1.3943809523809523e-05,
+      "loss": 0.8403,
+      "step": 6360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.03881106153130531,
+      "learning_rate": 1.3934285714285715e-05,
+      "loss": 0.7507,
+      "step": 6370
+    },
+    {
+      "epoch": 0.9114285714285715,
+      "grad_norm": 0.06903299689292908,
+      "learning_rate": 1.3924761904761906e-05,
+      "loss": 0.0,
+      "step": 6380
+    },
+    {
+      "epoch": 0.9128571428571428,
+      "grad_norm": 0.25843337178230286,
+      "learning_rate": 1.3915238095238097e-05,
+      "loss": 0.8332,
+      "step": 6390
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.04133506491780281,
+      "learning_rate": 1.3905714285714287e-05,
+      "loss": 1.7546,
+      "step": 6400
+    },
+    {
+      "epoch": 0.9157142857142857,
+      "grad_norm": 0.01921152137219906,
+      "learning_rate": 1.3896190476190476e-05,
+      "loss": 3.2392,
+      "step": 6410
+    },
+    {
+      "epoch": 0.9171428571428571,
+      "grad_norm": 0.29022958874702454,
+      "learning_rate": 1.3886666666666669e-05,
+      "loss": 0.8298,
+      "step": 6420
+    },
+    {
+      "epoch": 0.9185714285714286,
+      "grad_norm": 0.037898045033216476,
+      "learning_rate": 1.3877142857142858e-05,
+      "loss": 0.0,
+      "step": 6430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5045204162597656,
+      "learning_rate": 1.386761904761905e-05,
+      "loss": 0.0,
+      "step": 6440
+    },
+    {
+      "epoch": 0.9214285714285714,
+      "grad_norm": 0.018266407772898674,
+      "learning_rate": 1.3858095238095239e-05,
+      "loss": 0.9452,
+      "step": 6450
+    },
+    {
+      "epoch": 0.9228571428571428,
+      "grad_norm": 0.03972134366631508,
+      "learning_rate": 1.384857142857143e-05,
+      "loss": 0.0,
+      "step": 6460
+    },
+    {
+      "epoch": 0.9242857142857143,
+      "grad_norm": 0.03234800696372986,
+      "learning_rate": 1.383904761904762e-05,
+      "loss": 1.7769,
+      "step": 6470
+    },
+    {
+      "epoch": 0.9257142857142857,
+      "grad_norm": 0.04271892458200455,
+      "learning_rate": 1.3829523809523811e-05,
+      "loss": 0.7913,
+      "step": 6480
+    },
+    {
+      "epoch": 0.9271428571428572,
+      "grad_norm": 0.5514897704124451,
+      "learning_rate": 1.382e-05,
+      "loss": 3.5094,
+      "step": 6490
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 10.10671329498291,
+      "learning_rate": 1.3810476190476192e-05,
+      "loss": 2.4951,
+      "step": 6500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.16368746757507324,
+      "learning_rate": 1.3800952380952381e-05,
+      "loss": 3.0247,
+      "step": 6510
+    },
+    {
+      "epoch": 0.9314285714285714,
+      "grad_norm": 13.292880058288574,
+      "learning_rate": 1.3791428571428573e-05,
+      "loss": 0.0,
+      "step": 6520
+    },
+    {
+      "epoch": 0.9328571428571428,
+      "grad_norm": 0.06766203790903091,
+      "learning_rate": 1.3781904761904762e-05,
+      "loss": 0.8841,
+      "step": 6530
+    },
+    {
+      "epoch": 0.9342857142857143,
+      "grad_norm": 0.9948472380638123,
+      "learning_rate": 1.3772380952380953e-05,
+      "loss": 0.8256,
+      "step": 6540
+    },
+    {
+      "epoch": 0.9357142857142857,
+      "grad_norm": 0.07876208424568176,
+      "learning_rate": 1.3762857142857144e-05,
+      "loss": 1.624,
+      "step": 6550
+    },
+    {
+      "epoch": 0.9371428571428572,
+      "grad_norm": 4687747.5,
+      "learning_rate": 1.3753333333333334e-05,
+      "loss": 2.5279,
+      "step": 6560
+    },
+    {
+      "epoch": 0.9385714285714286,
+      "grad_norm": 0.07252531498670578,
+      "learning_rate": 1.3743809523809527e-05,
+      "loss": 0.8658,
+      "step": 6570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 40964140.0,
+      "learning_rate": 1.3734285714285716e-05,
+      "loss": 2.2721,
+      "step": 6580
+    },
+    {
+      "epoch": 0.9414285714285714,
+      "grad_norm": 0.05478376895189285,
+      "learning_rate": 1.3724761904761905e-05,
+      "loss": 1.1866,
+      "step": 6590
+    },
+    {
+      "epoch": 0.9428571428571428,
+      "grad_norm": 0.48456478118896484,
+      "learning_rate": 1.3715238095238097e-05,
+      "loss": 0.8513,
+      "step": 6600
+    },
+    {
+      "epoch": 0.9442857142857143,
+      "grad_norm": 0.07476714998483658,
+      "learning_rate": 1.3705714285714286e-05,
+      "loss": 0.8624,
+      "step": 6610
+    },
+    {
+      "epoch": 0.9457142857142857,
+      "grad_norm": 0.05129942670464516,
+      "learning_rate": 1.3696190476190477e-05,
+      "loss": 0.0,
+      "step": 6620
+    },
+    {
+      "epoch": 0.9471428571428572,
+      "grad_norm": 0.08390059322118759,
+      "learning_rate": 1.3686666666666669e-05,
+      "loss": 2.2917,
+      "step": 6630
+    },
+    {
+      "epoch": 0.9485714285714286,
+      "grad_norm": 0.6709385514259338,
+      "learning_rate": 1.3677142857142858e-05,
+      "loss": 0.0,
+      "step": 6640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 373780544.0,
+      "learning_rate": 1.366761904761905e-05,
+      "loss": 2.4119,
+      "step": 6650
+    },
+    {
+      "epoch": 0.9514285714285714,
+      "grad_norm": 0.13225433230400085,
+      "learning_rate": 1.3658095238095239e-05,
+      "loss": 1.493,
+      "step": 6660
+    },
+    {
+      "epoch": 0.9528571428571428,
+      "grad_norm": 0.22055622935295105,
+      "learning_rate": 1.3648571428571428e-05,
+      "loss": 2.583,
+      "step": 6670
+    },
+    {
+      "epoch": 0.9542857142857143,
+      "grad_norm": 0.06398754566907883,
+      "learning_rate": 1.363904761904762e-05,
+      "loss": 2.0571,
+      "step": 6680
+    },
+    {
+      "epoch": 0.9557142857142857,
+      "grad_norm": 0.2691834270954132,
+      "learning_rate": 1.3629523809523811e-05,
+      "loss": 0.0,
+      "step": 6690
+    },
+    {
+      "epoch": 0.9571428571428572,
+      "grad_norm": 0.029388394206762314,
+      "learning_rate": 1.3620000000000002e-05,
+      "loss": 3.1291,
+      "step": 6700
+    },
+    {
+      "epoch": 0.9585714285714285,
+      "grad_norm": 31.72698974609375,
+      "learning_rate": 1.3610476190476192e-05,
+      "loss": 2.4497,
+      "step": 6710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.057778529822826385,
+      "learning_rate": 1.3600952380952381e-05,
+      "loss": 0.8618,
+      "step": 6720
+    },
+    {
+      "epoch": 0.9614285714285714,
+      "grad_norm": 0.08452066034078598,
+      "learning_rate": 1.3591428571428574e-05,
+      "loss": 0.8475,
+      "step": 6730
+    },
+    {
+      "epoch": 0.9628571428571429,
+      "grad_norm": 1866970.375,
+      "learning_rate": 1.3581904761904763e-05,
+      "loss": 1.658,
+      "step": 6740
+    },
+    {
+      "epoch": 0.9642857142857143,
+      "grad_norm": 0.09029223769903183,
+      "learning_rate": 1.3572380952380953e-05,
+      "loss": 0.8342,
+      "step": 6750
+    },
+    {
+      "epoch": 0.9657142857142857,
+      "grad_norm": 0.05192052200436592,
+      "learning_rate": 1.3562857142857144e-05,
+      "loss": 0.8517,
+      "step": 6760
+    },
+    {
+      "epoch": 0.9671428571428572,
+      "grad_norm": 0.05285080149769783,
+      "learning_rate": 1.3553333333333335e-05,
+      "loss": 0.8017,
+      "step": 6770
+    },
+    {
+      "epoch": 0.9685714285714285,
+      "grad_norm": 0.03326203674077988,
+      "learning_rate": 1.3543809523809525e-05,
+      "loss": 1.6645,
+      "step": 6780
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5953431.0,
+      "learning_rate": 1.3534285714285716e-05,
+      "loss": 1.6982,
+      "step": 6790
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 0.7066969275474548,
+      "learning_rate": 1.3524761904761905e-05,
+      "loss": 0.816,
+      "step": 6800
+    },
+    {
+      "epoch": 0.9728571428571429,
+      "grad_norm": 5681333.5,
+      "learning_rate": 1.3515238095238097e-05,
+      "loss": 0.8079,
+      "step": 6810
+    },
+    {
+      "epoch": 0.9742857142857143,
+      "grad_norm": 0.052510276436805725,
+      "learning_rate": 1.3505714285714286e-05,
+      "loss": 0.7558,
+      "step": 6820
+    },
+    {
+      "epoch": 0.9757142857142858,
+      "grad_norm": 0.3362160921096802,
+      "learning_rate": 1.3496190476190478e-05,
+      "loss": 1.6614,
+      "step": 6830
+    },
+    {
+      "epoch": 0.9771428571428571,
+      "grad_norm": 0.2505156993865967,
+      "learning_rate": 1.3486666666666667e-05,
+      "loss": 0.8901,
+      "step": 6840
+    },
+    {
+      "epoch": 0.9785714285714285,
+      "grad_norm": 0.09322378784418106,
+      "learning_rate": 1.3477142857142858e-05,
+      "loss": 0.8112,
+      "step": 6850
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4786252.5,
+      "learning_rate": 1.3467619047619049e-05,
+      "loss": 1.8555,
+      "step": 6860
+    },
+    {
+      "epoch": 0.9814285714285714,
+      "grad_norm": 0.02122361771762371,
+      "learning_rate": 1.345809523809524e-05,
+      "loss": 0.8965,
+      "step": 6870
+    },
+    {
+      "epoch": 0.9828571428571429,
+      "grad_norm": 0.04701412096619606,
+      "learning_rate": 1.3448571428571428e-05,
+      "loss": 3.0345,
+      "step": 6880
+    },
+    {
+      "epoch": 0.9842857142857143,
+      "grad_norm": 0.06466910988092422,
+      "learning_rate": 1.343904761904762e-05,
+      "loss": 1.62,
+      "step": 6890
+    },
+    {
+      "epoch": 0.9857142857142858,
+      "grad_norm": 0.04547485336661339,
+      "learning_rate": 1.342952380952381e-05,
+      "loss": 0.7344,
+      "step": 6900
+    },
+    {
+      "epoch": 0.9871428571428571,
+      "grad_norm": 2787963.75,
+      "learning_rate": 1.3420000000000002e-05,
+      "loss": 1.5204,
+      "step": 6910
+    },
+    {
+      "epoch": 0.9885714285714285,
+      "grad_norm": 0.47706320881843567,
+      "learning_rate": 1.3410476190476191e-05,
+      "loss": 3.1118,
+      "step": 6920
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.044258058071136475,
+      "learning_rate": 1.3400952380952382e-05,
+      "loss": 2.4776,
+      "step": 6930
+    },
+    {
+      "epoch": 0.9914285714285714,
+      "grad_norm": 0.11442265659570694,
+      "learning_rate": 1.3391428571428572e-05,
+      "loss": 1.5529,
+      "step": 6940
+    },
+    {
+      "epoch": 0.9928571428571429,
+      "grad_norm": 0.14743246138095856,
+      "learning_rate": 1.3381904761904763e-05,
+      "loss": 0.0,
+      "step": 6950
+    },
+    {
+      "epoch": 0.9942857142857143,
+      "grad_norm": 0.04826438054442406,
+      "learning_rate": 1.3372380952380955e-05,
+      "loss": 0.0,
+      "step": 6960
+    },
+    {
+      "epoch": 0.9957142857142857,
+      "grad_norm": 0.03823941573500633,
+      "learning_rate": 1.3362857142857144e-05,
+      "loss": 1.6068,
+      "step": 6970
+    },
+    {
+      "epoch": 0.9971428571428571,
+      "grad_norm": 0.13535606861114502,
+      "learning_rate": 1.3353333333333333e-05,
+      "loss": 0.8673,
+      "step": 6980
+    },
+    {
+      "epoch": 0.9985714285714286,
+      "grad_norm": 0.09028885513544083,
+      "learning_rate": 1.3343809523809525e-05,
+      "loss": 0.0,
+      "step": 6990
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.031916309148073196,
+      "learning_rate": 1.3334285714285714e-05,
+      "loss": 0.8368,
+      "step": 7000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.933,
+      "eval_f1_macro": 0.9322909432343213,
+      "eval_f1_weighted": 0.9330115481557928,
+      "eval_loss": 1.1717169284820557,
+      "eval_runtime": 185.8073,
+      "eval_samples_per_second": 32.292,
+      "eval_steps_per_second": 16.146,
+      "step": 7000
+    },
+    {
+      "epoch": 1.0014285714285713,
+      "grad_norm": 25939850.0,
+      "learning_rate": 1.3324761904761905e-05,
+      "loss": 1.6311,
+      "step": 7010
+    },
+    {
+      "epoch": 1.002857142857143,
+      "grad_norm": 0.030088942497968674,
+      "learning_rate": 1.3315238095238097e-05,
+      "loss": 0.9511,
+      "step": 7020
+    },
+    {
+      "epoch": 1.0042857142857142,
+      "grad_norm": 3631735.25,
+      "learning_rate": 1.3305714285714286e-05,
+      "loss": 1.6915,
+      "step": 7030
+    },
+    {
+      "epoch": 1.0057142857142858,
+      "grad_norm": 0.04201086610555649,
+      "learning_rate": 1.3296190476190479e-05,
+      "loss": 0.8783,
+      "step": 7040
+    },
+    {
+      "epoch": 1.0071428571428571,
+      "grad_norm": 0.03625119850039482,
+      "learning_rate": 1.3286666666666668e-05,
+      "loss": 0.0,
+      "step": 7050
+    },
+    {
+      "epoch": 1.0085714285714287,
+      "grad_norm": 0.023486418649554253,
+      "learning_rate": 1.3277142857142857e-05,
+      "loss": 0.8329,
+      "step": 7060
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.06448784470558167,
+      "learning_rate": 1.3267619047619049e-05,
+      "loss": 0.0,
+      "step": 7070
+    },
+    {
+      "epoch": 1.0114285714285713,
+      "grad_norm": 0.03243952617049217,
+      "learning_rate": 1.325809523809524e-05,
+      "loss": 0.8519,
+      "step": 7080
+    },
+    {
+      "epoch": 1.012857142857143,
+      "grad_norm": 0.03520049527287483,
+      "learning_rate": 1.324857142857143e-05,
+      "loss": 0.0,
+      "step": 7090
+    },
+    {
+      "epoch": 1.0142857142857142,
+      "grad_norm": 1730741.75,
+      "learning_rate": 1.3239047619047621e-05,
+      "loss": 0.8996,
+      "step": 7100
+    },
+    {
+      "epoch": 1.0157142857142858,
+      "grad_norm": 0.027172576636075974,
+      "learning_rate": 1.322952380952381e-05,
+      "loss": 0.0,
+      "step": 7110
+    },
+    {
+      "epoch": 1.0171428571428571,
+      "grad_norm": 0.051058441400527954,
+      "learning_rate": 1.3220000000000002e-05,
+      "loss": 0.8274,
+      "step": 7120
+    },
+    {
+      "epoch": 1.0185714285714287,
+      "grad_norm": 0.11518155038356781,
+      "learning_rate": 1.3210476190476191e-05,
+      "loss": 0.0,
+      "step": 7130
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.023390529677271843,
+      "learning_rate": 1.3200952380952382e-05,
+      "loss": 0.8415,
+      "step": 7140
+    },
+    {
+      "epoch": 1.0214285714285714,
+      "grad_norm": 0.05566827952861786,
+      "learning_rate": 1.3191428571428572e-05,
+      "loss": 0.0,
+      "step": 7150
+    },
+    {
+      "epoch": 1.022857142857143,
+      "grad_norm": 0.025334279984235764,
+      "learning_rate": 1.3181904761904763e-05,
+      "loss": 0.8686,
+      "step": 7160
+    },
+    {
+      "epoch": 1.0242857142857142,
+      "grad_norm": 0.014513662084937096,
+      "learning_rate": 1.3172380952380954e-05,
+      "loss": 0.0,
+      "step": 7170
+    },
+    {
+      "epoch": 1.0257142857142858,
+      "grad_norm": 0.04419514164328575,
+      "learning_rate": 1.3162857142857144e-05,
+      "loss": 0.8766,
+      "step": 7180
+    },
+    {
+      "epoch": 1.0271428571428571,
+      "grad_norm": 0.01611058972775936,
+      "learning_rate": 1.3153333333333333e-05,
+      "loss": 0.756,
+      "step": 7190
+    },
+    {
+      "epoch": 1.0285714285714285,
+      "grad_norm": 2685416.25,
+      "learning_rate": 1.3143809523809526e-05,
+      "loss": 0.9965,
+      "step": 7200
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.09039904177188873,
+      "learning_rate": 1.3134285714285715e-05,
+      "loss": 0.8409,
+      "step": 7210
+    },
+    {
+      "epoch": 1.0314285714285714,
+      "grad_norm": 0.03149201720952988,
+      "learning_rate": 1.3124761904761907e-05,
+      "loss": 0.7547,
+      "step": 7220
+    },
+    {
+      "epoch": 1.032857142857143,
+      "grad_norm": 0.04580838233232498,
+      "learning_rate": 1.3115238095238096e-05,
+      "loss": 1.7343,
+      "step": 7230
+    },
+    {
+      "epoch": 1.0342857142857143,
+      "grad_norm": 0.025980187579989433,
+      "learning_rate": 1.3105714285714287e-05,
+      "loss": 0.9262,
+      "step": 7240
+    },
+    {
+      "epoch": 1.0357142857142858,
+      "grad_norm": 0.05715817213058472,
+      "learning_rate": 1.3096190476190477e-05,
+      "loss": 0.0,
+      "step": 7250
+    },
+    {
+      "epoch": 1.0371428571428571,
+      "grad_norm": 0.07200311869382858,
+      "learning_rate": 1.3086666666666668e-05,
+      "loss": 1.6685,
+      "step": 7260
+    },
+    {
+      "epoch": 1.0385714285714285,
+      "grad_norm": 0.08002328872680664,
+      "learning_rate": 1.3077142857142857e-05,
+      "loss": 0.8463,
+      "step": 7270
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3134590.5,
+      "learning_rate": 1.306761904761905e-05,
+      "loss": 0.8629,
+      "step": 7280
+    },
+    {
+      "epoch": 1.0414285714285714,
+      "grad_norm": 0.025882644578814507,
+      "learning_rate": 1.3058095238095238e-05,
+      "loss": 1.7126,
+      "step": 7290
+    },
+    {
+      "epoch": 1.042857142857143,
+      "grad_norm": 0.05894533917307854,
+      "learning_rate": 1.304857142857143e-05,
+      "loss": 0.0,
+      "step": 7300
+    },
+    {
+      "epoch": 1.0442857142857143,
+      "grad_norm": 0.033157117664813995,
+      "learning_rate": 1.303904761904762e-05,
+      "loss": 0.0,
+      "step": 7310
+    },
+    {
+      "epoch": 1.0457142857142858,
+      "grad_norm": 0.04596635326743126,
+      "learning_rate": 1.302952380952381e-05,
+      "loss": 0.0,
+      "step": 7320
+    },
+    {
+      "epoch": 1.0471428571428572,
+      "grad_norm": 0.21744632720947266,
+      "learning_rate": 1.302e-05,
+      "loss": 1.7861,
+      "step": 7330
+    },
+    {
+      "epoch": 1.0485714285714285,
+      "grad_norm": 0.09278339147567749,
+      "learning_rate": 1.3010476190476191e-05,
+      "loss": 0.0,
+      "step": 7340
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.02117972820997238,
+      "learning_rate": 1.3000952380952384e-05,
+      "loss": 0.9283,
+      "step": 7350
+    },
+    {
+      "epoch": 1.0514285714285714,
+      "grad_norm": 0.02321944572031498,
+      "learning_rate": 1.2991428571428573e-05,
+      "loss": 0.5161,
+      "step": 7360
+    },
+    {
+      "epoch": 1.052857142857143,
+      "grad_norm": 0.014927205629646778,
+      "learning_rate": 1.2981904761904762e-05,
+      "loss": 0.0,
+      "step": 7370
+    },
+    {
+      "epoch": 1.0542857142857143,
+      "grad_norm": 0.013720797374844551,
+      "learning_rate": 1.2972380952380954e-05,
+      "loss": 0.0,
+      "step": 7380
+    },
+    {
+      "epoch": 1.0557142857142856,
+      "grad_norm": 0.0524919293820858,
+      "learning_rate": 1.2962857142857143e-05,
+      "loss": 0.0,
+      "step": 7390
+    },
+    {
+      "epoch": 1.0571428571428572,
+      "grad_norm": 0.017233923077583313,
+      "learning_rate": 1.2953333333333334e-05,
+      "loss": 0.0,
+      "step": 7400
+    },
+    {
+      "epoch": 1.0585714285714285,
+      "grad_norm": 0.01686735451221466,
+      "learning_rate": 1.2943809523809526e-05,
+      "loss": 0.8657,
+      "step": 7410
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.019288865849375725,
+      "learning_rate": 1.2934285714285715e-05,
+      "loss": 0.0,
+      "step": 7420
+    },
+    {
+      "epoch": 1.0614285714285714,
+      "grad_norm": 2273186.75,
+      "learning_rate": 1.2924761904761907e-05,
+      "loss": 1.8573,
+      "step": 7430
+    },
+    {
+      "epoch": 1.062857142857143,
+      "grad_norm": 0.019001882523298264,
+      "learning_rate": 1.2915238095238096e-05,
+      "loss": 1.8237,
+      "step": 7440
+    },
+    {
+      "epoch": 1.0642857142857143,
+      "grad_norm": 0.08115746080875397,
+      "learning_rate": 1.2905714285714285e-05,
+      "loss": 0.0,
+      "step": 7450
+    },
+    {
+      "epoch": 1.0657142857142856,
+      "grad_norm": 2016966.75,
+      "learning_rate": 1.2896190476190477e-05,
+      "loss": 1.7038,
+      "step": 7460
+    },
+    {
+      "epoch": 1.0671428571428572,
+      "grad_norm": 0.424878865480423,
+      "learning_rate": 1.2886666666666668e-05,
+      "loss": 0.0,
+      "step": 7470
+    },
+    {
+      "epoch": 1.0685714285714285,
+      "grad_norm": 0.02476075477898121,
+      "learning_rate": 1.2877142857142859e-05,
+      "loss": 0.0,
+      "step": 7480
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.06589873135089874,
+      "learning_rate": 1.286761904761905e-05,
+      "loss": 0.0,
+      "step": 7490
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.028531435877084732,
+      "learning_rate": 1.2858095238095238e-05,
+      "loss": 0.0,
+      "step": 7500
+    },
+    {
+      "epoch": 1.072857142857143,
+      "grad_norm": 0.01938190683722496,
+      "learning_rate": 1.284857142857143e-05,
+      "loss": 0.8508,
+      "step": 7510
+    },
+    {
+      "epoch": 1.0742857142857143,
+      "grad_norm": 0.08631902933120728,
+      "learning_rate": 1.283904761904762e-05,
+      "loss": 0.0,
+      "step": 7520
+    },
+    {
+      "epoch": 1.0757142857142856,
+      "grad_norm": 0.017822710797190666,
+      "learning_rate": 1.282952380952381e-05,
+      "loss": 1.7343,
+      "step": 7530
+    },
+    {
+      "epoch": 1.0771428571428572,
+      "grad_norm": 0.07766088098287582,
+      "learning_rate": 1.2820000000000001e-05,
+      "loss": 1.7524,
+      "step": 7540
+    },
+    {
+      "epoch": 1.0785714285714285,
+      "grad_norm": 0.12040656059980392,
+      "learning_rate": 1.2810476190476192e-05,
+      "loss": 2.5261,
+      "step": 7550
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.23868650197982788,
+      "learning_rate": 1.2800952380952382e-05,
+      "loss": 0.844,
+      "step": 7560
+    },
+    {
+      "epoch": 1.0814285714285714,
+      "grad_norm": 0.025973087176680565,
+      "learning_rate": 1.2791428571428573e-05,
+      "loss": 0.0,
+      "step": 7570
+    },
+    {
+      "epoch": 1.0828571428571427,
+      "grad_norm": 0.045755863189697266,
+      "learning_rate": 1.2781904761904762e-05,
+      "loss": 0.0,
+      "step": 7580
+    },
+    {
+      "epoch": 1.0842857142857143,
+      "grad_norm": 0.05611782148480415,
+      "learning_rate": 1.2772380952380954e-05,
+      "loss": 0.7788,
+      "step": 7590
+    },
+    {
+      "epoch": 1.0857142857142856,
+      "grad_norm": 0.020975563675165176,
+      "learning_rate": 1.2762857142857143e-05,
+      "loss": 0.0,
+      "step": 7600
+    },
+    {
+      "epoch": 1.0871428571428572,
+      "grad_norm": 0.02233831211924553,
+      "learning_rate": 1.2753333333333335e-05,
+      "loss": 0.0,
+      "step": 7610
+    },
+    {
+      "epoch": 1.0885714285714285,
+      "grad_norm": 0.4176329970359802,
+      "learning_rate": 1.2743809523809524e-05,
+      "loss": 0.0,
+      "step": 7620
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0269179604947567,
+      "learning_rate": 1.2734285714285715e-05,
+      "loss": 0.858,
+      "step": 7630
+    },
+    {
+      "epoch": 1.0914285714285714,
+      "grad_norm": 0.03285587206482887,
+      "learning_rate": 1.2724761904761906e-05,
+      "loss": 0.8638,
+      "step": 7640
+    },
+    {
+      "epoch": 1.092857142857143,
+      "grad_norm": 0.013301646336913109,
+      "learning_rate": 1.2715238095238096e-05,
+      "loss": 0.0,
+      "step": 7650
+    },
+    {
+      "epoch": 1.0942857142857143,
+      "grad_norm": 0.04487910121679306,
+      "learning_rate": 1.2705714285714285e-05,
+      "loss": 0.9667,
+      "step": 7660
+    },
+    {
+      "epoch": 1.0957142857142856,
+      "grad_norm": 0.22912175953388214,
+      "learning_rate": 1.2696190476190478e-05,
+      "loss": 0.0,
+      "step": 7670
+    },
+    {
+      "epoch": 1.0971428571428572,
+      "grad_norm": 0.028395144268870354,
+      "learning_rate": 1.2686666666666667e-05,
+      "loss": 0.8474,
+      "step": 7680
+    },
+    {
+      "epoch": 1.0985714285714285,
+      "grad_norm": 0.03506438806653023,
+      "learning_rate": 1.2677142857142859e-05,
+      "loss": 0.9481,
+      "step": 7690
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.03143644332885742,
+      "learning_rate": 1.2667619047619048e-05,
+      "loss": 0.8583,
+      "step": 7700
+    },
+    {
+      "epoch": 1.1014285714285714,
+      "grad_norm": 0.013262659311294556,
+      "learning_rate": 1.2658095238095239e-05,
+      "loss": 0.0,
+      "step": 7710
+    },
+    {
+      "epoch": 1.1028571428571428,
+      "grad_norm": 0.014700079336762428,
+      "learning_rate": 1.2648571428571431e-05,
+      "loss": 0.889,
+      "step": 7720
+    },
+    {
+      "epoch": 1.1042857142857143,
+      "grad_norm": 0.12083712220191956,
+      "learning_rate": 1.263904761904762e-05,
+      "loss": 0.0,
+      "step": 7730
+    },
+    {
+      "epoch": 1.1057142857142856,
+      "grad_norm": 0.00990900956094265,
+      "learning_rate": 1.2629523809523812e-05,
+      "loss": 0.0,
+      "step": 7740
+    },
+    {
+      "epoch": 1.1071428571428572,
+      "grad_norm": 0.28362390398979187,
+      "learning_rate": 1.2620000000000001e-05,
+      "loss": 1.6391,
+      "step": 7750
+    },
+    {
+      "epoch": 1.1085714285714285,
+      "grad_norm": 0.33541566133499146,
+      "learning_rate": 1.261047619047619e-05,
+      "loss": 0.0,
+      "step": 7760
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.04129500687122345,
+      "learning_rate": 1.2600952380952382e-05,
+      "loss": 2.4589,
+      "step": 7770
+    },
+    {
+      "epoch": 1.1114285714285714,
+      "grad_norm": 0.09728424996137619,
+      "learning_rate": 1.2591428571428571e-05,
+      "loss": 1.7056,
+      "step": 7780
+    },
+    {
+      "epoch": 1.1128571428571428,
+      "grad_norm": 0.008812558837234974,
+      "learning_rate": 1.2581904761904762e-05,
+      "loss": 0.8989,
+      "step": 7790
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 0.012032631784677505,
+      "learning_rate": 1.2572380952380954e-05,
+      "loss": 0.7424,
+      "step": 7800
+    },
+    {
+      "epoch": 1.1157142857142857,
+      "grad_norm": 0.033953145146369934,
+      "learning_rate": 1.2562857142857143e-05,
+      "loss": 0.7078,
+      "step": 7810
+    },
+    {
+      "epoch": 1.1171428571428572,
+      "grad_norm": 0.01798797957599163,
+      "learning_rate": 1.2553333333333336e-05,
+      "loss": 0.9149,
+      "step": 7820
+    },
+    {
+      "epoch": 1.1185714285714285,
+      "grad_norm": 0.027739539742469788,
+      "learning_rate": 1.2543809523809525e-05,
+      "loss": 0.8968,
+      "step": 7830
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.02431493066251278,
+      "learning_rate": 1.2534285714285714e-05,
+      "loss": 0.7941,
+      "step": 7840
+    },
+    {
+      "epoch": 1.1214285714285714,
+      "grad_norm": 3477903.75,
+      "learning_rate": 1.2524761904761906e-05,
+      "loss": 0.8903,
+      "step": 7850
+    },
+    {
+      "epoch": 1.1228571428571428,
+      "grad_norm": 0.025965692475438118,
+      "learning_rate": 1.2515238095238097e-05,
+      "loss": 0.7517,
+      "step": 7860
+    },
+    {
+      "epoch": 1.1242857142857143,
+      "grad_norm": 0.012062611989676952,
+      "learning_rate": 1.2505714285714287e-05,
+      "loss": 2.4532,
+      "step": 7870
+    },
+    {
+      "epoch": 1.1257142857142857,
+      "grad_norm": 0.05875033512711525,
+      "learning_rate": 1.2496190476190478e-05,
+      "loss": 0.0,
+      "step": 7880
+    },
+    {
+      "epoch": 1.1271428571428572,
+      "grad_norm": 0.12643501162528992,
+      "learning_rate": 1.2486666666666667e-05,
+      "loss": 2.4006,
+      "step": 7890
+    },
+    {
+      "epoch": 1.1285714285714286,
+      "grad_norm": 0.015335256233811378,
+      "learning_rate": 1.247714285714286e-05,
+      "loss": 0.0,
+      "step": 7900
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.07105027139186859,
+      "learning_rate": 1.2467619047619048e-05,
+      "loss": 0.0,
+      "step": 7910
+    },
+    {
+      "epoch": 1.1314285714285715,
+      "grad_norm": 0.03675895184278488,
+      "learning_rate": 1.2458095238095239e-05,
+      "loss": 0.0006,
+      "step": 7920
+    },
+    {
+      "epoch": 1.1328571428571428,
+      "grad_norm": 10.108134269714355,
+      "learning_rate": 1.244857142857143e-05,
+      "loss": 0.7972,
+      "step": 7930
+    },
+    {
+      "epoch": 1.1342857142857143,
+      "grad_norm": 2904994.5,
+      "learning_rate": 1.243904761904762e-05,
+      "loss": 1.7591,
+      "step": 7940
+    },
+    {
+      "epoch": 1.1357142857142857,
+      "grad_norm": 0.06973975151777267,
+      "learning_rate": 1.242952380952381e-05,
+      "loss": 0.0,
+      "step": 7950
+    },
+    {
+      "epoch": 1.1371428571428572,
+      "grad_norm": 0.11097346246242523,
+      "learning_rate": 1.2420000000000001e-05,
+      "loss": 0.8988,
+      "step": 7960
+    },
+    {
+      "epoch": 1.1385714285714286,
+      "grad_norm": 0.009019365534186363,
+      "learning_rate": 1.241047619047619e-05,
+      "loss": 0.8746,
+      "step": 7970
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 0.026287801563739777,
+      "learning_rate": 1.2400952380952383e-05,
+      "loss": 1.7323,
+      "step": 7980
+    },
+    {
+      "epoch": 1.1414285714285715,
+      "grad_norm": 0.05487057939171791,
+      "learning_rate": 1.2391428571428572e-05,
+      "loss": 0.0,
+      "step": 7990
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.02337987907230854,
+      "learning_rate": 1.2381904761904764e-05,
+      "loss": 1.706,
+      "step": 8000
+    },
+    {
+      "epoch": 1.1442857142857144,
+      "grad_norm": 0.03765910863876343,
+      "learning_rate": 1.2372380952380953e-05,
+      "loss": 0.9021,
+      "step": 8010
+    },
+    {
+      "epoch": 1.1457142857142857,
+      "grad_norm": 6137160.5,
+      "learning_rate": 1.2362857142857144e-05,
+      "loss": 0.822,
+      "step": 8020
+    },
+    {
+      "epoch": 1.1471428571428572,
+      "grad_norm": 0.03027947060763836,
+      "learning_rate": 1.2353333333333334e-05,
+      "loss": 0.7918,
+      "step": 8030
+    },
+    {
+      "epoch": 1.1485714285714286,
+      "grad_norm": 0.05116349458694458,
+      "learning_rate": 1.2343809523809525e-05,
+      "loss": 0.8364,
+      "step": 8040
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.03407183662056923,
+      "learning_rate": 1.2334285714285714e-05,
+      "loss": 0.0,
+      "step": 8050
+    },
+    {
+      "epoch": 1.1514285714285715,
+      "grad_norm": 0.022328512743115425,
+      "learning_rate": 1.2324761904761906e-05,
+      "loss": 2.6156,
+      "step": 8060
+    },
+    {
+      "epoch": 1.1528571428571428,
+      "grad_norm": 2150228.0,
+      "learning_rate": 1.2315238095238095e-05,
+      "loss": 0.9354,
+      "step": 8070
+    },
+    {
+      "epoch": 1.1542857142857144,
+      "grad_norm": 0.025652026757597923,
+      "learning_rate": 1.2305714285714287e-05,
+      "loss": 0.0,
+      "step": 8080
+    },
+    {
+      "epoch": 1.1557142857142857,
+      "grad_norm": 0.030916299670934677,
+      "learning_rate": 1.2296190476190476e-05,
+      "loss": 0.8506,
+      "step": 8090
+    },
+    {
+      "epoch": 1.157142857142857,
+      "grad_norm": 0.014385643415153027,
+      "learning_rate": 1.2286666666666667e-05,
+      "loss": 0.0,
+      "step": 8100
+    },
+    {
+      "epoch": 1.1585714285714286,
+      "grad_norm": 0.04859547317028046,
+      "learning_rate": 1.227714285714286e-05,
+      "loss": 0.0,
+      "step": 8110
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0473051592707634,
+      "learning_rate": 1.2267619047619048e-05,
+      "loss": 0.795,
+      "step": 8120
+    },
+    {
+      "epoch": 1.1614285714285715,
+      "grad_norm": 0.04349032789468765,
+      "learning_rate": 1.225809523809524e-05,
+      "loss": 0.0,
+      "step": 8130
+    },
+    {
+      "epoch": 1.1628571428571428,
+      "grad_norm": 0.02459562011063099,
+      "learning_rate": 1.224857142857143e-05,
+      "loss": 0.819,
+      "step": 8140
+    },
+    {
+      "epoch": 1.1642857142857144,
+      "grad_norm": 0.03080209158360958,
+      "learning_rate": 1.2239047619047619e-05,
+      "loss": 1.1574,
+      "step": 8150
+    },
+    {
+      "epoch": 1.1657142857142857,
+      "grad_norm": 0.01987123116850853,
+      "learning_rate": 1.2229523809523811e-05,
+      "loss": 0.0,
+      "step": 8160
+    },
+    {
+      "epoch": 1.1671428571428573,
+      "grad_norm": 0.013408007100224495,
+      "learning_rate": 1.2220000000000002e-05,
+      "loss": 0.8295,
+      "step": 8170
+    },
+    {
+      "epoch": 1.1685714285714286,
+      "grad_norm": 2128650.75,
+      "learning_rate": 1.221047619047619e-05,
+      "loss": 0.8242,
+      "step": 8180
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0937790498137474,
+      "learning_rate": 1.2200952380952383e-05,
+      "loss": 0.0,
+      "step": 8190
+    },
+    {
+      "epoch": 1.1714285714285715,
+      "grad_norm": 0.650391161441803,
+      "learning_rate": 1.2191428571428572e-05,
+      "loss": 0.0,
+      "step": 8200
+    },
+    {
+      "epoch": 1.1728571428571428,
+      "grad_norm": 0.0932248905301094,
+      "learning_rate": 1.2181904761904764e-05,
+      "loss": 0.7684,
+      "step": 8210
+    },
+    {
+      "epoch": 1.1742857142857144,
+      "grad_norm": 0.1041816920042038,
+      "learning_rate": 1.2172380952380953e-05,
+      "loss": 0.0,
+      "step": 8220
+    },
+    {
+      "epoch": 1.1757142857142857,
+      "grad_norm": 2450221.25,
+      "learning_rate": 1.2162857142857142e-05,
+      "loss": 0.8838,
+      "step": 8230
+    },
+    {
+      "epoch": 1.177142857142857,
+      "grad_norm": 0.014858823269605637,
+      "learning_rate": 1.2153333333333334e-05,
+      "loss": 1.6649,
+      "step": 8240
+    },
+    {
+      "epoch": 1.1785714285714286,
+      "grad_norm": 0.23467659950256348,
+      "learning_rate": 1.2143809523809525e-05,
+      "loss": 0.0,
+      "step": 8250
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.044659532606601715,
+      "learning_rate": 1.2134285714285716e-05,
+      "loss": 0.0,
+      "step": 8260
+    },
+    {
+      "epoch": 1.1814285714285715,
+      "grad_norm": 0.06898272782564163,
+      "learning_rate": 1.2124761904761906e-05,
+      "loss": 0.0,
+      "step": 8270
+    },
+    {
+      "epoch": 1.1828571428571428,
+      "grad_norm": 0.006830879021435976,
+      "learning_rate": 1.2115238095238095e-05,
+      "loss": 0.0044,
+      "step": 8280
+    },
+    {
+      "epoch": 1.1842857142857142,
+      "grad_norm": 6.919030666351318,
+      "learning_rate": 1.2105714285714288e-05,
+      "loss": 0.9737,
+      "step": 8290
+    },
+    {
+      "epoch": 1.1857142857142857,
+      "grad_norm": 2252633.25,
+      "learning_rate": 1.2096190476190477e-05,
+      "loss": 0.8837,
+      "step": 8300
+    },
+    {
+      "epoch": 1.187142857142857,
+      "grad_norm": 0.026708241552114487,
+      "learning_rate": 1.2086666666666667e-05,
+      "loss": 0.936,
+      "step": 8310
+    },
+    {
+      "epoch": 1.1885714285714286,
+      "grad_norm": 0.013957766816020012,
+      "learning_rate": 1.2077142857142858e-05,
+      "loss": 0.823,
+      "step": 8320
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.012033933773636818,
+      "learning_rate": 1.2067619047619049e-05,
+      "loss": 0.0,
+      "step": 8330
+    },
+    {
+      "epoch": 1.1914285714285715,
+      "grad_norm": 0.026677537709474564,
+      "learning_rate": 1.205809523809524e-05,
+      "loss": 0.0,
+      "step": 8340
+    },
+    {
+      "epoch": 1.1928571428571428,
+      "grad_norm": 0.009044887498021126,
+      "learning_rate": 1.204857142857143e-05,
+      "loss": 0.0,
+      "step": 8350
+    },
+    {
+      "epoch": 1.1942857142857144,
+      "grad_norm": 0.027003323659300804,
+      "learning_rate": 1.2039047619047619e-05,
+      "loss": 0.8056,
+      "step": 8360
+    },
+    {
+      "epoch": 1.1957142857142857,
+      "grad_norm": 3738021.25,
+      "learning_rate": 1.2029523809523811e-05,
+      "loss": 1.922,
+      "step": 8370
+    },
+    {
+      "epoch": 1.197142857142857,
+      "grad_norm": 0.023733044043183327,
+      "learning_rate": 1.202e-05,
+      "loss": 0.0,
+      "step": 8380
+    },
+    {
+      "epoch": 1.1985714285714286,
+      "grad_norm": 0.01520014088600874,
+      "learning_rate": 1.2010476190476192e-05,
+      "loss": 0.0,
+      "step": 8390
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.007492155767977238,
+      "learning_rate": 1.2000952380952381e-05,
+      "loss": 0.8716,
+      "step": 8400
+    },
+    {
+      "epoch": 1.2014285714285715,
+      "grad_norm": 0.17145775258541107,
+      "learning_rate": 1.1991428571428572e-05,
+      "loss": 0.0,
+      "step": 8410
+    },
+    {
+      "epoch": 1.2028571428571428,
+      "grad_norm": 0.033446650952100754,
+      "learning_rate": 1.1981904761904763e-05,
+      "loss": 0.9318,
+      "step": 8420
+    },
+    {
+      "epoch": 1.2042857142857142,
+      "grad_norm": 5488741.5,
+      "learning_rate": 1.1972380952380953e-05,
+      "loss": 0.8212,
+      "step": 8430
+    },
+    {
+      "epoch": 1.2057142857142857,
+      "grad_norm": 0.012246781028807163,
+      "learning_rate": 1.1962857142857146e-05,
+      "loss": 0.833,
+      "step": 8440
+    },
+    {
+      "epoch": 1.207142857142857,
+      "grad_norm": 0.12609416246414185,
+      "learning_rate": 1.1953333333333335e-05,
+      "loss": 0.0,
+      "step": 8450
+    },
+    {
+      "epoch": 1.2085714285714286,
+      "grad_norm": 0.05155313014984131,
+      "learning_rate": 1.1943809523809524e-05,
+      "loss": 0.0,
+      "step": 8460
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.01732013374567032,
+      "learning_rate": 1.1934285714285716e-05,
+      "loss": 0.0,
+      "step": 8470
+    },
+    {
+      "epoch": 1.2114285714285715,
+      "grad_norm": 0.012313324958086014,
+      "learning_rate": 1.1924761904761905e-05,
+      "loss": 0.968,
+      "step": 8480
+    },
+    {
+      "epoch": 1.2128571428571429,
+      "grad_norm": 16.698265075683594,
+      "learning_rate": 1.1915238095238096e-05,
+      "loss": 0.8923,
+      "step": 8490
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 0.05277903750538826,
+      "learning_rate": 1.1905714285714288e-05,
+      "loss": 0.0,
+      "step": 8500
+    },
+    {
+      "epoch": 1.2157142857142857,
+      "grad_norm": 0.008104232139885426,
+      "learning_rate": 1.1896190476190477e-05,
+      "loss": 0.8776,
+      "step": 8510
+    },
+    {
+      "epoch": 1.217142857142857,
+      "grad_norm": 0.04466225206851959,
+      "learning_rate": 1.188666666666667e-05,
+      "loss": 0.9748,
+      "step": 8520
+    },
+    {
+      "epoch": 1.2185714285714286,
+      "grad_norm": 4705172.0,
+      "learning_rate": 1.1877142857142858e-05,
+      "loss": 0.89,
+      "step": 8530
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.03340664878487587,
+      "learning_rate": 1.1867619047619047e-05,
+      "loss": 0.0,
+      "step": 8540
+    },
+    {
+      "epoch": 1.2214285714285715,
+      "grad_norm": 0.02333148382604122,
+      "learning_rate": 1.185809523809524e-05,
+      "loss": 0.8778,
+      "step": 8550
+    },
+    {
+      "epoch": 1.2228571428571429,
+      "grad_norm": 0.052439361810684204,
+      "learning_rate": 1.184857142857143e-05,
+      "loss": 0.8318,
+      "step": 8560
+    },
+    {
+      "epoch": 1.2242857142857142,
+      "grad_norm": 0.01787891425192356,
+      "learning_rate": 1.183904761904762e-05,
+      "loss": 0.0,
+      "step": 8570
+    },
+    {
+      "epoch": 1.2257142857142858,
+      "grad_norm": 2889095.5,
+      "learning_rate": 1.1829523809523811e-05,
+      "loss": 0.8166,
+      "step": 8580
+    },
+    {
+      "epoch": 1.227142857142857,
+      "grad_norm": 0.03564672917127609,
+      "learning_rate": 1.182e-05,
+      "loss": 2.431,
+      "step": 8590
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 0.057759907096624374,
+      "learning_rate": 1.1810476190476193e-05,
+      "loss": 0.0,
+      "step": 8600
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2917649447917938,
+      "learning_rate": 1.1800952380952382e-05,
+      "loss": 0.7244,
+      "step": 8610
+    },
+    {
+      "epoch": 1.2314285714285713,
+      "grad_norm": 0.03878479078412056,
+      "learning_rate": 1.1791428571428572e-05,
+      "loss": 2.3126,
+      "step": 8620
+    },
+    {
+      "epoch": 1.2328571428571429,
+      "grad_norm": 0.02152009680867195,
+      "learning_rate": 1.1781904761904763e-05,
+      "loss": 0.9751,
+      "step": 8630
+    },
+    {
+      "epoch": 1.2342857142857142,
+      "grad_norm": 0.02633904293179512,
+      "learning_rate": 1.1772380952380954e-05,
+      "loss": 1.7672,
+      "step": 8640
+    },
+    {
+      "epoch": 1.2357142857142858,
+      "grad_norm": 0.5630979537963867,
+      "learning_rate": 1.1762857142857144e-05,
+      "loss": 0.847,
+      "step": 8650
+    },
+    {
+      "epoch": 1.237142857142857,
+      "grad_norm": 0.0721784234046936,
+      "learning_rate": 1.1753333333333335e-05,
+      "loss": 0.8346,
+      "step": 8660
+    },
+    {
+      "epoch": 1.2385714285714287,
+      "grad_norm": 0.020708810538053513,
+      "learning_rate": 1.1743809523809524e-05,
+      "loss": 1.7973,
+      "step": 8670
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.1945323944091797,
+      "learning_rate": 1.1734285714285716e-05,
+      "loss": 0.8997,
+      "step": 8680
+    },
+    {
+      "epoch": 1.2414285714285715,
+      "grad_norm": 0.23876921832561493,
+      "learning_rate": 1.1724761904761905e-05,
+      "loss": 0.8523,
+      "step": 8690
+    },
+    {
+      "epoch": 1.2428571428571429,
+      "grad_norm": 0.02663440816104412,
+      "learning_rate": 1.1715238095238097e-05,
+      "loss": 1.5108,
+      "step": 8700
+    },
+    {
+      "epoch": 1.2442857142857142,
+      "grad_norm": 0.021907707676291466,
+      "learning_rate": 1.1705714285714286e-05,
+      "loss": 1.5268,
+      "step": 8710
+    },
+    {
+      "epoch": 1.2457142857142858,
+      "grad_norm": 0.02269059233367443,
+      "learning_rate": 1.1696190476190477e-05,
+      "loss": 2.3094,
+      "step": 8720
+    },
+    {
+      "epoch": 1.247142857142857,
+      "grad_norm": 0.04022393003106117,
+      "learning_rate": 1.1686666666666668e-05,
+      "loss": 0.8331,
+      "step": 8730
+    },
+    {
+      "epoch": 1.2485714285714287,
+      "grad_norm": 0.029710130766034126,
+      "learning_rate": 1.1677142857142858e-05,
+      "loss": 0.8464,
+      "step": 8740
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.017246633768081665,
+      "learning_rate": 1.1667619047619047e-05,
+      "loss": 1.5864,
+      "step": 8750
+    },
+    {
+      "epoch": 1.2514285714285713,
+      "grad_norm": 0.008614294230937958,
+      "learning_rate": 1.165809523809524e-05,
+      "loss": 1.7972,
+      "step": 8760
+    },
+    {
+      "epoch": 1.252857142857143,
+      "grad_norm": 0.04210745543241501,
+      "learning_rate": 1.1648571428571429e-05,
+      "loss": 0.9188,
+      "step": 8770
+    },
+    {
+      "epoch": 1.2542857142857142,
+      "grad_norm": 3295093.25,
+      "learning_rate": 1.1639047619047621e-05,
+      "loss": 0.8174,
+      "step": 8780
+    },
+    {
+      "epoch": 1.2557142857142858,
+      "grad_norm": 0.027939507737755775,
+      "learning_rate": 1.162952380952381e-05,
+      "loss": 0.0,
+      "step": 8790
+    },
+    {
+      "epoch": 1.2571428571428571,
+      "grad_norm": 0.06812241673469543,
+      "learning_rate": 1.162e-05,
+      "loss": 0.0,
+      "step": 8800
+    },
+    {
+      "epoch": 1.2585714285714285,
+      "grad_norm": 0.07074253261089325,
+      "learning_rate": 1.1610476190476191e-05,
+      "loss": 0.0,
+      "step": 8810
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.020804984495043755,
+      "learning_rate": 1.1600952380952382e-05,
+      "loss": 2.5898,
+      "step": 8820
+    },
+    {
+      "epoch": 1.2614285714285716,
+      "grad_norm": 0.06660498678684235,
+      "learning_rate": 1.1591428571428574e-05,
+      "loss": 0.0,
+      "step": 8830
+    },
+    {
+      "epoch": 1.262857142857143,
+      "grad_norm": 0.010008258745074272,
+      "learning_rate": 1.1581904761904763e-05,
+      "loss": 0.8746,
+      "step": 8840
+    },
+    {
+      "epoch": 1.2642857142857142,
+      "grad_norm": 0.46079233288764954,
+      "learning_rate": 1.1572380952380952e-05,
+      "loss": 1.8013,
+      "step": 8850
+    },
+    {
+      "epoch": 1.2657142857142858,
+      "grad_norm": 0.06187233328819275,
+      "learning_rate": 1.1562857142857144e-05,
+      "loss": 0.0,
+      "step": 8860
+    },
+    {
+      "epoch": 1.2671428571428571,
+      "grad_norm": 0.055006545037031174,
+      "learning_rate": 1.1553333333333333e-05,
+      "loss": 0.0,
+      "step": 8870
+    },
+    {
+      "epoch": 1.2685714285714287,
+      "grad_norm": 0.029951203614473343,
+      "learning_rate": 1.1543809523809524e-05,
+      "loss": 0.0,
+      "step": 8880
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.011264192871749401,
+      "learning_rate": 1.1534285714285716e-05,
+      "loss": 0.0,
+      "step": 8890
+    },
+    {
+      "epoch": 1.2714285714285714,
+      "grad_norm": 0.013191123493015766,
+      "learning_rate": 1.1524761904761905e-05,
+      "loss": 4.5148,
+      "step": 8900
+    },
+    {
+      "epoch": 1.272857142857143,
+      "grad_norm": 0.028085438534617424,
+      "learning_rate": 1.1515238095238098e-05,
+      "loss": 0.9181,
+      "step": 8910
+    },
+    {
+      "epoch": 1.2742857142857142,
+      "grad_norm": 0.026875896379351616,
+      "learning_rate": 1.1505714285714287e-05,
+      "loss": 1.5101,
+      "step": 8920
+    },
+    {
+      "epoch": 1.2757142857142858,
+      "grad_norm": 15013122.0,
+      "learning_rate": 1.1496190476190476e-05,
+      "loss": 3.4408,
+      "step": 8930
+    },
+    {
+      "epoch": 1.2771428571428571,
+      "grad_norm": 0.06057320162653923,
+      "learning_rate": 1.1486666666666668e-05,
+      "loss": 0.0,
+      "step": 8940
+    },
+    {
+      "epoch": 1.2785714285714285,
+      "grad_norm": 0.04739050194621086,
+      "learning_rate": 1.1477142857142859e-05,
+      "loss": 0.0,
+      "step": 8950
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.14422151446342468,
+      "learning_rate": 1.146761904761905e-05,
+      "loss": 1.6029,
+      "step": 8960
+    },
+    {
+      "epoch": 1.2814285714285714,
+      "grad_norm": 0.06465866416692734,
+      "learning_rate": 1.145809523809524e-05,
+      "loss": 2.6486,
+      "step": 8970
+    },
+    {
+      "epoch": 1.282857142857143,
+      "grad_norm": 0.008445010520517826,
+      "learning_rate": 1.1448571428571429e-05,
+      "loss": 0.0,
+      "step": 8980
+    },
+    {
+      "epoch": 1.2842857142857143,
+      "grad_norm": 0.049681611359119415,
+      "learning_rate": 1.1439047619047621e-05,
+      "loss": 0.0,
+      "step": 8990
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.013962008990347385,
+      "learning_rate": 1.142952380952381e-05,
+      "loss": 0.0,
+      "step": 9000
+    },
+    {
+      "epoch": 1.2871428571428571,
+      "grad_norm": 0.02123785950243473,
+      "learning_rate": 1.142e-05,
+      "loss": 1.8728,
+      "step": 9010
+    },
+    {
+      "epoch": 1.2885714285714287,
+      "grad_norm": 0.082945317029953,
+      "learning_rate": 1.1410476190476191e-05,
+      "loss": 0.0,
+      "step": 9020
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.04686948284506798,
+      "learning_rate": 1.1400952380952382e-05,
+      "loss": 0.9012,
+      "step": 9030
+    },
+    {
+      "epoch": 1.2914285714285714,
+      "grad_norm": 0.26491132378578186,
+      "learning_rate": 1.1391428571428573e-05,
+      "loss": 0.0,
+      "step": 9040
+    },
+    {
+      "epoch": 1.292857142857143,
+      "grad_norm": 0.5649831295013428,
+      "learning_rate": 1.1381904761904763e-05,
+      "loss": 0.0,
+      "step": 9050
+    },
+    {
+      "epoch": 1.2942857142857143,
+      "grad_norm": 15282727.0,
+      "learning_rate": 1.1372380952380952e-05,
+      "loss": 1.7251,
+      "step": 9060
+    },
+    {
+      "epoch": 1.2957142857142858,
+      "grad_norm": 2754296.75,
+      "learning_rate": 1.1362857142857145e-05,
+      "loss": 1.841,
+      "step": 9070
+    },
+    {
+      "epoch": 1.2971428571428572,
+      "grad_norm": 0.1370944380760193,
+      "learning_rate": 1.1353333333333334e-05,
+      "loss": 0.9336,
+      "step": 9080
+    },
+    {
+      "epoch": 1.2985714285714285,
+      "grad_norm": 0.018099334090948105,
+      "learning_rate": 1.1343809523809526e-05,
+      "loss": 0.8593,
+      "step": 9090
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.12131799012422562,
+      "learning_rate": 1.1334285714285715e-05,
+      "loss": 0.0,
+      "step": 9100
+    },
+    {
+      "epoch": 1.3014285714285714,
+      "grad_norm": 0.0642264187335968,
+      "learning_rate": 1.1324761904761906e-05,
+      "loss": 0.0,
+      "step": 9110
+    },
+    {
+      "epoch": 1.302857142857143,
+      "grad_norm": 0.033341508358716965,
+      "learning_rate": 1.1315238095238096e-05,
+      "loss": 1.7358,
+      "step": 9120
+    },
+    {
+      "epoch": 1.3042857142857143,
+      "grad_norm": 0.04357098042964935,
+      "learning_rate": 1.1305714285714287e-05,
+      "loss": 0.0,
+      "step": 9130
+    },
+    {
+      "epoch": 1.3057142857142856,
+      "grad_norm": 0.03646727278828621,
+      "learning_rate": 1.1296190476190476e-05,
+      "loss": 0.0,
+      "step": 9140
+    },
+    {
+      "epoch": 1.3071428571428572,
+      "grad_norm": 0.06602869182825089,
+      "learning_rate": 1.1286666666666668e-05,
+      "loss": 0.7711,
+      "step": 9150
+    },
+    {
+      "epoch": 1.3085714285714285,
+      "grad_norm": 0.02178690768778324,
+      "learning_rate": 1.1277142857142857e-05,
+      "loss": 0.0,
+      "step": 9160
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.035533491522073746,
+      "learning_rate": 1.126761904761905e-05,
+      "loss": 0.0,
+      "step": 9170
+    },
+    {
+      "epoch": 1.3114285714285714,
+      "grad_norm": 0.10600371658802032,
+      "learning_rate": 1.1258095238095238e-05,
+      "loss": 1.6214,
+      "step": 9180
+    },
+    {
+      "epoch": 1.3128571428571427,
+      "grad_norm": 0.14072033762931824,
+      "learning_rate": 1.1248571428571429e-05,
+      "loss": 0.0,
+      "step": 9190
+    },
+    {
+      "epoch": 1.3142857142857143,
+      "grad_norm": 2.283571243286133,
+      "learning_rate": 1.123904761904762e-05,
+      "loss": 0.0,
+      "step": 9200
+    },
+    {
+      "epoch": 1.3157142857142858,
+      "grad_norm": 0.05963047221302986,
+      "learning_rate": 1.122952380952381e-05,
+      "loss": 0.7708,
+      "step": 9210
+    },
+    {
+      "epoch": 1.3171428571428572,
+      "grad_norm": 0.2830362021923065,
+      "learning_rate": 1.1220000000000003e-05,
+      "loss": 1.7425,
+      "step": 9220
+    },
+    {
+      "epoch": 1.3185714285714285,
+      "grad_norm": 0.010538119822740555,
+      "learning_rate": 1.1210476190476192e-05,
+      "loss": 0.9697,
+      "step": 9230
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3297023.75,
+      "learning_rate": 1.120095238095238e-05,
+      "loss": 2.5914,
+      "step": 9240
+    },
+    {
+      "epoch": 1.3214285714285714,
+      "grad_norm": 0.12161114811897278,
+      "learning_rate": 1.1191428571428573e-05,
+      "loss": 0.7982,
+      "step": 9250
+    },
+    {
+      "epoch": 1.322857142857143,
+      "grad_norm": 0.3002341091632843,
+      "learning_rate": 1.1181904761904762e-05,
+      "loss": 0.0,
+      "step": 9260
+    },
+    {
+      "epoch": 1.3242857142857143,
+      "grad_norm": 0.04993857815861702,
+      "learning_rate": 1.1172380952380953e-05,
+      "loss": 0.0,
+      "step": 9270
+    },
+    {
+      "epoch": 1.3257142857142856,
+      "grad_norm": 0.028125273063778877,
+      "learning_rate": 1.1162857142857145e-05,
+      "loss": 0.0,
+      "step": 9280
+    },
+    {
+      "epoch": 1.3271428571428572,
+      "grad_norm": 0.020571378991007805,
+      "learning_rate": 1.1153333333333334e-05,
+      "loss": 0.0,
+      "step": 9290
+    },
+    {
+      "epoch": 1.3285714285714285,
+      "grad_norm": 0.02745864912867546,
+      "learning_rate": 1.1143809523809526e-05,
+      "loss": 0.0,
+      "step": 9300
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.015707144513726234,
+      "learning_rate": 1.1134285714285715e-05,
+      "loss": 1.0189,
+      "step": 9310
+    },
+    {
+      "epoch": 1.3314285714285714,
+      "grad_norm": 0.024634873494505882,
+      "learning_rate": 1.1124761904761904e-05,
+      "loss": 0.9355,
+      "step": 9320
+    },
+    {
+      "epoch": 1.3328571428571427,
+      "grad_norm": 0.13222827017307281,
+      "learning_rate": 1.1115238095238096e-05,
+      "loss": 0.0,
+      "step": 9330
+    },
+    {
+      "epoch": 1.3342857142857143,
+      "grad_norm": 0.021714776754379272,
+      "learning_rate": 1.1105714285714287e-05,
+      "loss": 0.8535,
+      "step": 9340
+    },
+    {
+      "epoch": 1.3357142857142856,
+      "grad_norm": 0.016119644045829773,
+      "learning_rate": 1.1096190476190478e-05,
+      "loss": 0.0,
+      "step": 9350
+    },
+    {
+      "epoch": 1.3371428571428572,
+      "grad_norm": 0.007789562921971083,
+      "learning_rate": 1.1086666666666668e-05,
+      "loss": 0.0,
+      "step": 9360
+    },
+    {
+      "epoch": 1.3385714285714285,
+      "grad_norm": 3113352.75,
+      "learning_rate": 1.1077142857142857e-05,
+      "loss": 0.8593,
+      "step": 9370
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.04810258746147156,
+      "learning_rate": 1.106761904761905e-05,
+      "loss": 0.0,
+      "step": 9380
+    },
+    {
+      "epoch": 1.3414285714285714,
+      "grad_norm": 0.011734110303223133,
+      "learning_rate": 1.1058095238095239e-05,
+      "loss": 0.8568,
+      "step": 9390
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 0.012482531368732452,
+      "learning_rate": 1.104857142857143e-05,
+      "loss": 0.8351,
+      "step": 9400
+    },
+    {
+      "epoch": 1.3442857142857143,
+      "grad_norm": 6307.9296875,
+      "learning_rate": 1.103904761904762e-05,
+      "loss": 0.8565,
+      "step": 9410
+    },
+    {
+      "epoch": 1.3457142857142856,
+      "grad_norm": 0.022619159892201424,
+      "learning_rate": 1.102952380952381e-05,
+      "loss": 0.8384,
+      "step": 9420
+    },
+    {
+      "epoch": 1.3471428571428572,
+      "grad_norm": 0.043744947761297226,
+      "learning_rate": 1.1020000000000001e-05,
+      "loss": 0.9236,
+      "step": 9430
+    },
+    {
+      "epoch": 1.3485714285714285,
+      "grad_norm": 0.03145880252122879,
+      "learning_rate": 1.1010476190476192e-05,
+      "loss": 0.0,
+      "step": 9440
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.02551647275686264,
+      "learning_rate": 1.100095238095238e-05,
+      "loss": 0.0,
+      "step": 9450
+    },
+    {
+      "epoch": 1.3514285714285714,
+      "grad_norm": 0.055378690361976624,
+      "learning_rate": 1.0991428571428573e-05,
+      "loss": 0.0,
+      "step": 9460
+    },
+    {
+      "epoch": 1.3528571428571428,
+      "grad_norm": 0.02146647870540619,
+      "learning_rate": 1.0981904761904762e-05,
+      "loss": 0.0,
+      "step": 9470
+    },
+    {
+      "epoch": 1.3542857142857143,
+      "grad_norm": 0.006402506958693266,
+      "learning_rate": 1.0972380952380954e-05,
+      "loss": 0.844,
+      "step": 9480
+    },
+    {
+      "epoch": 1.3557142857142856,
+      "grad_norm": 0.019644366577267647,
+      "learning_rate": 1.0962857142857143e-05,
+      "loss": 1.8345,
+      "step": 9490
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.0326433889567852,
+      "learning_rate": 1.0953333333333334e-05,
+      "loss": 1.589,
+      "step": 9500
+    },
+    {
+      "epoch": 1.3585714285714285,
+      "grad_norm": 0.1283015012741089,
+      "learning_rate": 1.0943809523809525e-05,
+      "loss": 0.8055,
+      "step": 9510
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.03079349175095558,
+      "learning_rate": 1.0934285714285715e-05,
+      "loss": 0.0,
+      "step": 9520
+    },
+    {
+      "epoch": 1.3614285714285714,
+      "grad_norm": 0.0146969398483634,
+      "learning_rate": 1.0924761904761904e-05,
+      "loss": 1.6907,
+      "step": 9530
+    },
+    {
+      "epoch": 1.362857142857143,
+      "grad_norm": 0.0678119882941246,
+      "learning_rate": 1.0915238095238097e-05,
+      "loss": 0.0,
+      "step": 9540
+    },
+    {
+      "epoch": 1.3642857142857143,
+      "grad_norm": 0.017553478479385376,
+      "learning_rate": 1.0905714285714286e-05,
+      "loss": 0.0,
+      "step": 9550
+    },
+    {
+      "epoch": 1.3657142857142857,
+      "grad_norm": 0.018692970275878906,
+      "learning_rate": 1.0896190476190478e-05,
+      "loss": 0.9334,
+      "step": 9560
+    },
+    {
+      "epoch": 1.3671428571428572,
+      "grad_norm": 0.03647415712475777,
+      "learning_rate": 1.0886666666666667e-05,
+      "loss": 0.0,
+      "step": 9570
+    },
+    {
+      "epoch": 1.3685714285714285,
+      "grad_norm": 0.02425375208258629,
+      "learning_rate": 1.0877142857142858e-05,
+      "loss": 0.0,
+      "step": 9580
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2254489.25,
+      "learning_rate": 1.086761904761905e-05,
+      "loss": 2.716,
+      "step": 9590
+    },
+    {
+      "epoch": 1.3714285714285714,
+      "grad_norm": 0.031068919226527214,
+      "learning_rate": 1.0858095238095239e-05,
+      "loss": 0.0,
+      "step": 9600
+    },
+    {
+      "epoch": 1.3728571428571428,
+      "grad_norm": 0.03205320984125137,
+      "learning_rate": 1.0848571428571431e-05,
+      "loss": 0.8781,
+      "step": 9610
+    },
+    {
+      "epoch": 1.3742857142857143,
+      "grad_norm": 0.05405518040060997,
+      "learning_rate": 1.083904761904762e-05,
+      "loss": 0.0,
+      "step": 9620
+    },
+    {
+      "epoch": 1.3757142857142857,
+      "grad_norm": 0.2159292995929718,
+      "learning_rate": 1.0829523809523809e-05,
+      "loss": 0.8197,
+      "step": 9630
+    },
+    {
+      "epoch": 1.3771428571428572,
+      "grad_norm": 0.030176235362887383,
+      "learning_rate": 1.0820000000000001e-05,
+      "loss": 0.0,
+      "step": 9640
+    },
+    {
+      "epoch": 1.3785714285714286,
+      "grad_norm": 0.021075256168842316,
+      "learning_rate": 1.081047619047619e-05,
+      "loss": 1.8213,
+      "step": 9650
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2064819037914276,
+      "learning_rate": 1.0800952380952381e-05,
+      "loss": 0.6834,
+      "step": 9660
+    },
+    {
+      "epoch": 1.3814285714285715,
+      "grad_norm": 0.16339394450187683,
+      "learning_rate": 1.0791428571428573e-05,
+      "loss": 0.7987,
+      "step": 9670
+    },
+    {
+      "epoch": 1.3828571428571428,
+      "grad_norm": 0.0185325238853693,
+      "learning_rate": 1.0781904761904762e-05,
+      "loss": 0.7773,
+      "step": 9680
+    },
+    {
+      "epoch": 1.3842857142857143,
+      "grad_norm": 0.028442049399018288,
+      "learning_rate": 1.0772380952380955e-05,
+      "loss": 0.8279,
+      "step": 9690
+    },
+    {
+      "epoch": 1.3857142857142857,
+      "grad_norm": 374.2330017089844,
+      "learning_rate": 1.0762857142857144e-05,
+      "loss": 0.9022,
+      "step": 9700
+    },
+    {
+      "epoch": 1.387142857142857,
+      "grad_norm": 0.021085338667035103,
+      "learning_rate": 1.0753333333333333e-05,
+      "loss": 1.5727,
+      "step": 9710
+    },
+    {
+      "epoch": 1.3885714285714286,
+      "grad_norm": 0.061144813895225525,
+      "learning_rate": 1.0743809523809525e-05,
+      "loss": 0.0,
+      "step": 9720
+    },
+    {
+      "epoch": 1.3900000000000001,
+      "grad_norm": 0.22852379083633423,
+      "learning_rate": 1.0734285714285716e-05,
+      "loss": 1.6321,
+      "step": 9730
+    },
+    {
+      "epoch": 1.3914285714285715,
+      "grad_norm": 0.037348005920648575,
+      "learning_rate": 1.0724761904761906e-05,
+      "loss": 0.8798,
+      "step": 9740
+    },
+    {
+      "epoch": 1.3928571428571428,
+      "grad_norm": 0.07598963379859924,
+      "learning_rate": 1.0715238095238097e-05,
+      "loss": 0.0,
+      "step": 9750
+    },
+    {
+      "epoch": 1.3942857142857144,
+      "grad_norm": 0.01769055798649788,
+      "learning_rate": 1.0705714285714286e-05,
+      "loss": 0.0,
+      "step": 9760
+    },
+    {
+      "epoch": 1.3957142857142857,
+      "grad_norm": 0.012463989667594433,
+      "learning_rate": 1.0696190476190478e-05,
+      "loss": 0.0,
+      "step": 9770
+    },
+    {
+      "epoch": 1.3971428571428572,
+      "grad_norm": 0.014657414518296719,
+      "learning_rate": 1.0686666666666667e-05,
+      "loss": 0.879,
+      "step": 9780
+    },
+    {
+      "epoch": 1.3985714285714286,
+      "grad_norm": 0.09840326011180878,
+      "learning_rate": 1.0677142857142858e-05,
+      "loss": 2.7137,
+      "step": 9790
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.03041483834385872,
+      "learning_rate": 1.0667619047619048e-05,
+      "loss": 0.0,
+      "step": 9800
+    },
+    {
+      "epoch": 1.4014285714285715,
+      "grad_norm": 0.01784934476017952,
+      "learning_rate": 1.0658095238095239e-05,
+      "loss": 0.9505,
+      "step": 9810
+    },
+    {
+      "epoch": 1.4028571428571428,
+      "grad_norm": 0.013548986054956913,
+      "learning_rate": 1.064857142857143e-05,
+      "loss": 2.5361,
+      "step": 9820
+    },
+    {
+      "epoch": 1.4042857142857144,
+      "grad_norm": 0.06798744946718216,
+      "learning_rate": 1.063904761904762e-05,
+      "loss": 0.0,
+      "step": 9830
+    },
+    {
+      "epoch": 1.4057142857142857,
+      "grad_norm": 3214457.5,
+      "learning_rate": 1.062952380952381e-05,
+      "loss": 0.9116,
+      "step": 9840
+    },
+    {
+      "epoch": 1.407142857142857,
+      "grad_norm": 4608768.5,
+      "learning_rate": 1.0620000000000002e-05,
+      "loss": 2.5307,
+      "step": 9850
+    },
+    {
+      "epoch": 1.4085714285714286,
+      "grad_norm": 0.12910330295562744,
+      "learning_rate": 1.061047619047619e-05,
+      "loss": 1.5968,
+      "step": 9860
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9353005886077881,
+      "learning_rate": 1.0600952380952383e-05,
+      "loss": 0.0,
+      "step": 9870
+    },
+    {
+      "epoch": 1.4114285714285715,
+      "grad_norm": 0.09018875658512115,
+      "learning_rate": 1.0591428571428572e-05,
+      "loss": 0.0,
+      "step": 9880
+    },
+    {
+      "epoch": 1.4128571428571428,
+      "grad_norm": 0.011958391405642033,
+      "learning_rate": 1.0581904761904763e-05,
+      "loss": 0.0,
+      "step": 9890
+    },
+    {
+      "epoch": 1.4142857142857144,
+      "grad_norm": 0.024801990017294884,
+      "learning_rate": 1.0572380952380953e-05,
+      "loss": 0.0,
+      "step": 9900
+    },
+    {
+      "epoch": 1.4157142857142857,
+      "grad_norm": 0.016756532713770866,
+      "learning_rate": 1.0562857142857144e-05,
+      "loss": 0.0,
+      "step": 9910
+    },
+    {
+      "epoch": 1.4171428571428573,
+      "grad_norm": 0.03835080564022064,
+      "learning_rate": 1.0553333333333333e-05,
+      "loss": 0.0,
+      "step": 9920
+    },
+    {
+      "epoch": 1.4185714285714286,
+      "grad_norm": 0.020455079153180122,
+      "learning_rate": 1.0543809523809525e-05,
+      "loss": 0.9679,
+      "step": 9930
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.007498301565647125,
+      "learning_rate": 1.0534285714285714e-05,
+      "loss": 0.0,
+      "step": 9940
+    },
+    {
+      "epoch": 1.4214285714285715,
+      "grad_norm": 6159580.0,
+      "learning_rate": 1.0524761904761906e-05,
+      "loss": 0.8571,
+      "step": 9950
+    },
+    {
+      "epoch": 1.4228571428571428,
+      "grad_norm": 0.03572804108262062,
+      "learning_rate": 1.0515238095238095e-05,
+      "loss": 0.8435,
+      "step": 9960
+    },
+    {
+      "epoch": 1.4242857142857144,
+      "grad_norm": 4400301.5,
+      "learning_rate": 1.0505714285714286e-05,
+      "loss": 2.7984,
+      "step": 9970
+    },
+    {
+      "epoch": 1.4257142857142857,
+      "grad_norm": 0.1124780997633934,
+      "learning_rate": 1.0496190476190478e-05,
+      "loss": 0.9236,
+      "step": 9980
+    },
+    {
+      "epoch": 1.427142857142857,
+      "grad_norm": 0.29640457034111023,
+      "learning_rate": 1.0486666666666667e-05,
+      "loss": 1.8657,
+      "step": 9990
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.394471138715744,
+      "learning_rate": 1.047714285714286e-05,
+      "loss": 0.873,
+      "step": 10000
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.018534602597355843,
+      "learning_rate": 1.0467619047619049e-05,
+      "loss": 0.9428,
+      "step": 10010
+    },
+    {
+      "epoch": 1.4314285714285715,
+      "grad_norm": 0.05628011003136635,
+      "learning_rate": 1.0458095238095238e-05,
+      "loss": 2.6097,
+      "step": 10020
+    },
+    {
+      "epoch": 1.4328571428571428,
+      "grad_norm": 0.04259981960058212,
+      "learning_rate": 1.044857142857143e-05,
+      "loss": 0.8193,
+      "step": 10030
+    },
+    {
+      "epoch": 1.4342857142857142,
+      "grad_norm": 0.02374720573425293,
+      "learning_rate": 1.043904761904762e-05,
+      "loss": 0.9628,
+      "step": 10040
+    },
+    {
+      "epoch": 1.4357142857142857,
+      "grad_norm": 0.0744156688451767,
+      "learning_rate": 1.042952380952381e-05,
+      "loss": 0.8325,
+      "step": 10050
+    },
+    {
+      "epoch": 1.4371428571428573,
+      "grad_norm": 0.06819042563438416,
+      "learning_rate": 1.0420000000000002e-05,
+      "loss": 1.5474,
+      "step": 10060
+    },
+    {
+      "epoch": 1.4385714285714286,
+      "grad_norm": 151463440.0,
+      "learning_rate": 1.041047619047619e-05,
+      "loss": 0.7588,
+      "step": 10070
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.031236380338668823,
+      "learning_rate": 1.0400952380952383e-05,
+      "loss": 0.0,
+      "step": 10080
+    },
+    {
+      "epoch": 1.4414285714285715,
+      "grad_norm": 0.01712191104888916,
+      "learning_rate": 1.0391428571428572e-05,
+      "loss": 0.7758,
+      "step": 10090
+    },
+    {
+      "epoch": 1.4428571428571428,
+      "grad_norm": 0.013785792514681816,
+      "learning_rate": 1.0381904761904763e-05,
+      "loss": 0.0,
+      "step": 10100
+    },
+    {
+      "epoch": 1.4442857142857144,
+      "grad_norm": 0.02513999491930008,
+      "learning_rate": 1.0372380952380953e-05,
+      "loss": 1.6772,
+      "step": 10110
+    },
+    {
+      "epoch": 1.4457142857142857,
+      "grad_norm": 0.05135834962129593,
+      "learning_rate": 1.0362857142857144e-05,
+      "loss": 0.844,
+      "step": 10120
+    },
+    {
+      "epoch": 1.447142857142857,
+      "grad_norm": 4288965.5,
+      "learning_rate": 1.0353333333333335e-05,
+      "loss": 0.913,
+      "step": 10130
+    },
+    {
+      "epoch": 1.4485714285714286,
+      "grad_norm": 0.02619149349629879,
+      "learning_rate": 1.0343809523809525e-05,
+      "loss": 1.8649,
+      "step": 10140
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.11403828859329224,
+      "learning_rate": 1.0334285714285714e-05,
+      "loss": 0.0,
+      "step": 10150
+    },
+    {
+      "epoch": 1.4514285714285715,
+      "grad_norm": 0.10401636362075806,
+      "learning_rate": 1.0324761904761907e-05,
+      "loss": 0.8771,
+      "step": 10160
+    },
+    {
+      "epoch": 1.4528571428571428,
+      "grad_norm": 0.04993535205721855,
+      "learning_rate": 1.0315238095238096e-05,
+      "loss": 0.0,
+      "step": 10170
+    },
+    {
+      "epoch": 1.4542857142857142,
+      "grad_norm": 5836940.5,
+      "learning_rate": 1.0305714285714286e-05,
+      "loss": 1.7612,
+      "step": 10180
+    },
+    {
+      "epoch": 1.4557142857142857,
+      "grad_norm": 0.03801821172237396,
+      "learning_rate": 1.0296190476190477e-05,
+      "loss": 0.0,
+      "step": 10190
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 0.021374685689806938,
+      "learning_rate": 1.0286666666666668e-05,
+      "loss": 0.0,
+      "step": 10200
+    },
+    {
+      "epoch": 1.4585714285714286,
+      "grad_norm": 0.054252102971076965,
+      "learning_rate": 1.0277142857142858e-05,
+      "loss": 0.0,
+      "step": 10210
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.034760575741529465,
+      "learning_rate": 1.0267619047619049e-05,
+      "loss": 0.0,
+      "step": 10220
+    },
+    {
+      "epoch": 1.4614285714285713,
+      "grad_norm": 2333526.25,
+      "learning_rate": 1.0258095238095238e-05,
+      "loss": 1.7983,
+      "step": 10230
+    },
+    {
+      "epoch": 1.4628571428571429,
+      "grad_norm": 0.01770380698144436,
+      "learning_rate": 1.024857142857143e-05,
+      "loss": 0.9602,
+      "step": 10240
+    },
+    {
+      "epoch": 1.4642857142857144,
+      "grad_norm": 0.038541462272405624,
+      "learning_rate": 1.0239047619047619e-05,
+      "loss": 0.0,
+      "step": 10250
+    },
+    {
+      "epoch": 1.4657142857142857,
+      "grad_norm": 0.02205018885433674,
+      "learning_rate": 1.0229523809523811e-05,
+      "loss": 1.6874,
+      "step": 10260
+    },
+    {
+      "epoch": 1.467142857142857,
+      "grad_norm": 0.02114938572049141,
+      "learning_rate": 1.022e-05,
+      "loss": 0.8203,
+      "step": 10270
+    },
+    {
+      "epoch": 1.4685714285714286,
+      "grad_norm": 0.033991940319538116,
+      "learning_rate": 1.0210476190476191e-05,
+      "loss": 0.7779,
+      "step": 10280
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.048570018261671066,
+      "learning_rate": 1.0200952380952382e-05,
+      "loss": 0.0,
+      "step": 10290
+    },
+    {
+      "epoch": 1.4714285714285715,
+      "grad_norm": 0.027855537831783295,
+      "learning_rate": 1.0191428571428572e-05,
+      "loss": 0.9228,
+      "step": 10300
+    },
+    {
+      "epoch": 1.4728571428571429,
+      "grad_norm": 0.014924844726920128,
+      "learning_rate": 1.0181904761904761e-05,
+      "loss": 0.8669,
+      "step": 10310
+    },
+    {
+      "epoch": 1.4742857142857142,
+      "grad_norm": 2917559.0,
+      "learning_rate": 1.0172380952380954e-05,
+      "loss": 1.6435,
+      "step": 10320
+    },
+    {
+      "epoch": 1.4757142857142858,
+      "grad_norm": 0.055225372314453125,
+      "learning_rate": 1.0162857142857143e-05,
+      "loss": 1.7079,
+      "step": 10330
+    },
+    {
+      "epoch": 1.477142857142857,
+      "grad_norm": 0.02875506691634655,
+      "learning_rate": 1.0153333333333335e-05,
+      "loss": 0.0,
+      "step": 10340
+    },
+    {
+      "epoch": 1.4785714285714286,
+      "grad_norm": 9990818.0,
+      "learning_rate": 1.0143809523809524e-05,
+      "loss": 1.7626,
+      "step": 10350
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.018270818516612053,
+      "learning_rate": 1.0134285714285715e-05,
+      "loss": 0.0,
+      "step": 10360
+    },
+    {
+      "epoch": 1.4814285714285713,
+      "grad_norm": 0.03116474486887455,
+      "learning_rate": 1.0124761904761907e-05,
+      "loss": 2.766,
+      "step": 10370
+    },
+    {
+      "epoch": 1.4828571428571429,
+      "grad_norm": 0.06323779374361038,
+      "learning_rate": 1.0115238095238096e-05,
+      "loss": 0.8003,
+      "step": 10380
+    },
+    {
+      "epoch": 1.4842857142857142,
+      "grad_norm": 0.368496298789978,
+      "learning_rate": 1.0105714285714288e-05,
+      "loss": 1.5163,
+      "step": 10390
+    },
+    {
+      "epoch": 1.4857142857142858,
+      "grad_norm": 1.0995755195617676,
+      "learning_rate": 1.0096190476190477e-05,
+      "loss": 1.6704,
+      "step": 10400
+    },
+    {
+      "epoch": 1.487142857142857,
+      "grad_norm": 0.06878669559955597,
+      "learning_rate": 1.0086666666666666e-05,
+      "loss": 0.0,
+      "step": 10410
+    },
+    {
+      "epoch": 1.4885714285714284,
+      "grad_norm": 0.03460712358355522,
+      "learning_rate": 1.0077142857142858e-05,
+      "loss": 0.0,
+      "step": 10420
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.016890717670321465,
+      "learning_rate": 1.0067619047619049e-05,
+      "loss": 0.0,
+      "step": 10430
+    },
+    {
+      "epoch": 1.4914285714285715,
+      "grad_norm": 1.131646990776062,
+      "learning_rate": 1.0058095238095238e-05,
+      "loss": 1.7269,
+      "step": 10440
+    },
+    {
+      "epoch": 1.4928571428571429,
+      "grad_norm": 3955923.25,
+      "learning_rate": 1.004857142857143e-05,
+      "loss": 2.688,
+      "step": 10450
+    },
+    {
+      "epoch": 1.4942857142857142,
+      "grad_norm": 0.03988966345787048,
+      "learning_rate": 1.003904761904762e-05,
+      "loss": 0.0,
+      "step": 10460
+    },
+    {
+      "epoch": 1.4957142857142858,
+      "grad_norm": 0.07250599563121796,
+      "learning_rate": 1.0029523809523812e-05,
+      "loss": 0.911,
+      "step": 10470
+    },
+    {
+      "epoch": 1.497142857142857,
+      "grad_norm": 0.08377046883106232,
+      "learning_rate": 1.002e-05,
+      "loss": 0.86,
+      "step": 10480
+    },
+    {
+      "epoch": 1.4985714285714287,
+      "grad_norm": 0.09960997849702835,
+      "learning_rate": 1.0010476190476191e-05,
+      "loss": 0.8789,
+      "step": 10490
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2338420.75,
+      "learning_rate": 1.0000952380952382e-05,
+      "loss": 0.8194,
+      "step": 10500
+    },
+    {
+      "epoch": 1.5014285714285713,
+      "grad_norm": 0.09179136157035828,
+      "learning_rate": 9.991428571428573e-06,
+      "loss": 0.0,
+      "step": 10510
+    },
+    {
+      "epoch": 1.502857142857143,
+      "grad_norm": 0.02117820270359516,
+      "learning_rate": 9.981904761904763e-06,
+      "loss": 0.0,
+      "step": 10520
+    },
+    {
+      "epoch": 1.5042857142857144,
+      "grad_norm": 0.04899248480796814,
+      "learning_rate": 9.972380952380954e-06,
+      "loss": 0.0,
+      "step": 10530
+    },
+    {
+      "epoch": 1.5057142857142858,
+      "grad_norm": 0.03176189213991165,
+      "learning_rate": 9.962857142857145e-06,
+      "loss": 0.5263,
+      "step": 10540
+    },
+    {
+      "epoch": 1.5071428571428571,
+      "grad_norm": 0.05837094038724899,
+      "learning_rate": 9.953333333333333e-06,
+      "loss": 0.0,
+      "step": 10550
+    },
+    {
+      "epoch": 1.5085714285714285,
+      "grad_norm": 0.014203615486621857,
+      "learning_rate": 9.943809523809524e-06,
+      "loss": 2.8317,
+      "step": 10560
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4120672.5,
+      "learning_rate": 9.934285714285715e-06,
+      "loss": 0.9391,
+      "step": 10570
+    },
+    {
+      "epoch": 1.5114285714285716,
+      "grad_norm": 3843579.5,
+      "learning_rate": 9.924761904761905e-06,
+      "loss": 1.809,
+      "step": 10580
+    },
+    {
+      "epoch": 1.512857142857143,
+      "grad_norm": 0.04459831491112709,
+      "learning_rate": 9.915238095238096e-06,
+      "loss": 0.0,
+      "step": 10590
+    },
+    {
+      "epoch": 1.5142857142857142,
+      "grad_norm": 0.03221514821052551,
+      "learning_rate": 9.905714285714287e-06,
+      "loss": 0.8579,
+      "step": 10600
+    },
+    {
+      "epoch": 1.5157142857142856,
+      "grad_norm": 0.017137229442596436,
+      "learning_rate": 9.896190476190477e-06,
+      "loss": 0.0,
+      "step": 10610
+    },
+    {
+      "epoch": 1.5171428571428571,
+      "grad_norm": 0.01663975976407528,
+      "learning_rate": 9.886666666666668e-06,
+      "loss": 0.0,
+      "step": 10620
+    },
+    {
+      "epoch": 1.5185714285714287,
+      "grad_norm": 0.020876798778772354,
+      "learning_rate": 9.877142857142859e-06,
+      "loss": 0.8937,
+      "step": 10630
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5064546.0,
+      "learning_rate": 9.867619047619048e-06,
+      "loss": 3.472,
+      "step": 10640
+    },
+    {
+      "epoch": 1.5214285714285714,
+      "grad_norm": 0.03440290689468384,
+      "learning_rate": 9.858095238095238e-06,
+      "loss": 1.6396,
+      "step": 10650
+    },
+    {
+      "epoch": 1.522857142857143,
+      "grad_norm": 0.03516411408782005,
+      "learning_rate": 9.848571428571429e-06,
+      "loss": 0.0,
+      "step": 10660
+    },
+    {
+      "epoch": 1.5242857142857142,
+      "grad_norm": 0.15740182995796204,
+      "learning_rate": 9.83904761904762e-06,
+      "loss": 1.6889,
+      "step": 10670
+    },
+    {
+      "epoch": 1.5257142857142858,
+      "grad_norm": 0.014876801520586014,
+      "learning_rate": 9.82952380952381e-06,
+      "loss": 0.0,
+      "step": 10680
+    },
+    {
+      "epoch": 1.5271428571428571,
+      "grad_norm": 0.015245836228132248,
+      "learning_rate": 9.820000000000001e-06,
+      "loss": 1.9275,
+      "step": 10690
+    },
+    {
+      "epoch": 1.5285714285714285,
+      "grad_norm": 0.03401646763086319,
+      "learning_rate": 9.810476190476191e-06,
+      "loss": 0.798,
+      "step": 10700
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.020757881924510002,
+      "learning_rate": 9.800952380952382e-06,
+      "loss": 0.0,
+      "step": 10710
+    },
+    {
+      "epoch": 1.5314285714285716,
+      "grad_norm": 0.057735756039619446,
+      "learning_rate": 9.791428571428571e-06,
+      "loss": 0.796,
+      "step": 10720
+    },
+    {
+      "epoch": 1.532857142857143,
+      "grad_norm": 0.04900228977203369,
+      "learning_rate": 9.781904761904762e-06,
+      "loss": 0.0,
+      "step": 10730
+    },
+    {
+      "epoch": 1.5342857142857143,
+      "grad_norm": 0.024792909622192383,
+      "learning_rate": 9.772380952380952e-06,
+      "loss": 0.0,
+      "step": 10740
+    },
+    {
+      "epoch": 1.5357142857142856,
+      "grad_norm": 0.015745213255286217,
+      "learning_rate": 9.762857142857145e-06,
+      "loss": 0.0,
+      "step": 10750
+    },
+    {
+      "epoch": 1.5371428571428571,
+      "grad_norm": 0.024253172799944878,
+      "learning_rate": 9.753333333333335e-06,
+      "loss": 0.9589,
+      "step": 10760
+    },
+    {
+      "epoch": 1.5385714285714287,
+      "grad_norm": 0.009830779395997524,
+      "learning_rate": 9.743809523809524e-06,
+      "loss": 0.809,
+      "step": 10770
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.004012355115264654,
+      "learning_rate": 9.734285714285715e-06,
+      "loss": 0.0,
+      "step": 10780
+    },
+    {
+      "epoch": 1.5414285714285714,
+      "grad_norm": 0.29741397500038147,
+      "learning_rate": 9.724761904761906e-06,
+      "loss": 0.0,
+      "step": 10790
+    },
+    {
+      "epoch": 1.5428571428571427,
+      "grad_norm": 0.01610482484102249,
+      "learning_rate": 9.715238095238096e-06,
+      "loss": 1.5386,
+      "step": 10800
+    },
+    {
+      "epoch": 1.5442857142857143,
+      "grad_norm": 0.031117159873247147,
+      "learning_rate": 9.705714285714287e-06,
+      "loss": 0.9608,
+      "step": 10810
+    },
+    {
+      "epoch": 1.5457142857142858,
+      "grad_norm": 0.07643602043390274,
+      "learning_rate": 9.696190476190478e-06,
+      "loss": 0.9395,
+      "step": 10820
+    },
+    {
+      "epoch": 1.5471428571428572,
+      "grad_norm": 0.057319868355989456,
+      "learning_rate": 9.686666666666668e-06,
+      "loss": 0.0,
+      "step": 10830
+    },
+    {
+      "epoch": 1.5485714285714285,
+      "grad_norm": 0.06430930644273758,
+      "learning_rate": 9.677142857142859e-06,
+      "loss": 0.0,
+      "step": 10840
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.014286993071436882,
+      "learning_rate": 9.667619047619048e-06,
+      "loss": 0.8479,
+      "step": 10850
+    },
+    {
+      "epoch": 1.5514285714285714,
+      "grad_norm": 0.06223646178841591,
+      "learning_rate": 9.658095238095238e-06,
+      "loss": 0.8932,
+      "step": 10860
+    },
+    {
+      "epoch": 1.552857142857143,
+      "grad_norm": 0.02701249159872532,
+      "learning_rate": 9.648571428571429e-06,
+      "loss": 0.0,
+      "step": 10870
+    },
+    {
+      "epoch": 1.5542857142857143,
+      "grad_norm": 0.018355971202254295,
+      "learning_rate": 9.63904761904762e-06,
+      "loss": 0.0,
+      "step": 10880
+    },
+    {
+      "epoch": 1.5557142857142856,
+      "grad_norm": 0.03157276660203934,
+      "learning_rate": 9.62952380952381e-06,
+      "loss": 0.0,
+      "step": 10890
+    },
+    {
+      "epoch": 1.5571428571428572,
+      "grad_norm": 0.16080467402935028,
+      "learning_rate": 9.620000000000001e-06,
+      "loss": 0.1058,
+      "step": 10900
+    },
+    {
+      "epoch": 1.5585714285714287,
+      "grad_norm": 0.018533460795879364,
+      "learning_rate": 9.610476190476192e-06,
+      "loss": 0.9224,
+      "step": 10910
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.07870723307132721,
+      "learning_rate": 9.600952380952382e-06,
+      "loss": 1.6733,
+      "step": 10920
+    },
+    {
+      "epoch": 1.5614285714285714,
+      "grad_norm": 0.007128334604203701,
+      "learning_rate": 9.591428571428573e-06,
+      "loss": 0.0,
+      "step": 10930
+    },
+    {
+      "epoch": 1.5628571428571427,
+      "grad_norm": 0.012992989271879196,
+      "learning_rate": 9.581904761904762e-06,
+      "loss": 0.0,
+      "step": 10940
+    },
+    {
+      "epoch": 1.5642857142857143,
+      "grad_norm": 0.04158896207809448,
+      "learning_rate": 9.572380952380953e-06,
+      "loss": 3.8218,
+      "step": 10950
+    },
+    {
+      "epoch": 1.5657142857142858,
+      "grad_norm": 0.021905450150370598,
+      "learning_rate": 9.562857142857143e-06,
+      "loss": 0.8564,
+      "step": 10960
+    },
+    {
+      "epoch": 1.5671428571428572,
+      "grad_norm": 0.037827715277671814,
+      "learning_rate": 9.553333333333334e-06,
+      "loss": 0.7372,
+      "step": 10970
+    },
+    {
+      "epoch": 1.5685714285714285,
+      "grad_norm": 0.03636813908815384,
+      "learning_rate": 9.543809523809525e-06,
+      "loss": 0.9193,
+      "step": 10980
+    },
+    {
+      "epoch": 1.5699999999999998,
+      "grad_norm": 0.1919804960489273,
+      "learning_rate": 9.534285714285715e-06,
+      "loss": 1.757,
+      "step": 10990
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.0973452553153038,
+      "learning_rate": 9.524761904761906e-06,
+      "loss": 0.7944,
+      "step": 11000
+    },
+    {
+      "epoch": 1.572857142857143,
+      "grad_norm": 0.034837014973163605,
+      "learning_rate": 9.515238095238097e-06,
+      "loss": 0.0,
+      "step": 11010
+    },
+    {
+      "epoch": 1.5742857142857143,
+      "grad_norm": 0.020920246839523315,
+      "learning_rate": 9.505714285714287e-06,
+      "loss": 0.8301,
+      "step": 11020
+    },
+    {
+      "epoch": 1.5757142857142856,
+      "grad_norm": 0.07151384651660919,
+      "learning_rate": 9.496190476190476e-06,
+      "loss": 0.0,
+      "step": 11030
+    },
+    {
+      "epoch": 1.5771428571428572,
+      "grad_norm": 0.022104326635599136,
+      "learning_rate": 9.486666666666667e-06,
+      "loss": 0.8088,
+      "step": 11040
+    },
+    {
+      "epoch": 1.5785714285714287,
+      "grad_norm": 0.07792923599481583,
+      "learning_rate": 9.477142857142857e-06,
+      "loss": 0.8286,
+      "step": 11050
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.09004852175712585,
+      "learning_rate": 9.467619047619048e-06,
+      "loss": 0.0,
+      "step": 11060
+    },
+    {
+      "epoch": 1.5814285714285714,
+      "grad_norm": 2894938.0,
+      "learning_rate": 9.458095238095239e-06,
+      "loss": 3.5279,
+      "step": 11070
+    },
+    {
+      "epoch": 1.5828571428571427,
+      "grad_norm": 0.017983220517635345,
+      "learning_rate": 9.44857142857143e-06,
+      "loss": 1.8013,
+      "step": 11080
+    },
+    {
+      "epoch": 1.5842857142857143,
+      "grad_norm": 3101409.5,
+      "learning_rate": 9.43904761904762e-06,
+      "loss": 0.8419,
+      "step": 11090
+    },
+    {
+      "epoch": 1.5857142857142859,
+      "grad_norm": 0.018516412004828453,
+      "learning_rate": 9.42952380952381e-06,
+      "loss": 0.7651,
+      "step": 11100
+    },
+    {
+      "epoch": 1.5871428571428572,
+      "grad_norm": 1.3221491575241089,
+      "learning_rate": 9.42e-06,
+      "loss": 0.7717,
+      "step": 11110
+    },
+    {
+      "epoch": 1.5885714285714285,
+      "grad_norm": 0.06739199161529541,
+      "learning_rate": 9.41047619047619e-06,
+      "loss": 0.7517,
+      "step": 11120
+    },
+    {
+      "epoch": 1.5899999999999999,
+      "grad_norm": 0.23477089405059814,
+      "learning_rate": 9.400952380952381e-06,
+      "loss": 0.0,
+      "step": 11130
+    },
+    {
+      "epoch": 1.5914285714285714,
+      "grad_norm": 0.3631838262081146,
+      "learning_rate": 9.391428571428573e-06,
+      "loss": 1.7791,
+      "step": 11140
+    },
+    {
+      "epoch": 1.592857142857143,
+      "grad_norm": 0.03126092627644539,
+      "learning_rate": 9.381904761904764e-06,
+      "loss": 0.7602,
+      "step": 11150
+    },
+    {
+      "epoch": 1.5942857142857143,
+      "grad_norm": 0.11022724956274033,
+      "learning_rate": 9.372380952380953e-06,
+      "loss": 0.0,
+      "step": 11160
+    },
+    {
+      "epoch": 1.5957142857142856,
+      "grad_norm": 2556505.0,
+      "learning_rate": 9.362857142857143e-06,
+      "loss": 0.9204,
+      "step": 11170
+    },
+    {
+      "epoch": 1.5971428571428572,
+      "grad_norm": 4595861.0,
+      "learning_rate": 9.353333333333334e-06,
+      "loss": 1.8266,
+      "step": 11180
+    },
+    {
+      "epoch": 1.5985714285714285,
+      "grad_norm": 0.022031361237168312,
+      "learning_rate": 9.343809523809525e-06,
+      "loss": 0.8186,
+      "step": 11190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.03242994844913483,
+      "learning_rate": 9.334285714285715e-06,
+      "loss": 0.9774,
+      "step": 11200
+    },
+    {
+      "epoch": 1.6014285714285714,
+      "grad_norm": 0.13498488068580627,
+      "learning_rate": 9.324761904761906e-06,
+      "loss": 1.6948,
+      "step": 11210
+    },
+    {
+      "epoch": 1.6028571428571428,
+      "grad_norm": 0.014309341087937355,
+      "learning_rate": 9.315238095238097e-06,
+      "loss": 0.0,
+      "step": 11220
+    },
+    {
+      "epoch": 1.6042857142857143,
+      "grad_norm": 0.0475706122815609,
+      "learning_rate": 9.305714285714287e-06,
+      "loss": 0.7538,
+      "step": 11230
+    },
+    {
+      "epoch": 1.6057142857142859,
+      "grad_norm": 0.022850675508379936,
+      "learning_rate": 9.296190476190476e-06,
+      "loss": 0.9121,
+      "step": 11240
+    },
+    {
+      "epoch": 1.6071428571428572,
+      "grad_norm": 0.04371821507811546,
+      "learning_rate": 9.286666666666667e-06,
+      "loss": 0.8515,
+      "step": 11250
+    },
+    {
+      "epoch": 1.6085714285714285,
+      "grad_norm": 0.06175847724080086,
+      "learning_rate": 9.277142857142858e-06,
+      "loss": 0.0,
+      "step": 11260
+    },
+    {
+      "epoch": 1.6099999999999999,
+      "grad_norm": 0.043548502027988434,
+      "learning_rate": 9.267619047619048e-06,
+      "loss": 0.0,
+      "step": 11270
+    },
+    {
+      "epoch": 1.6114285714285714,
+      "grad_norm": 0.19634562730789185,
+      "learning_rate": 9.258095238095239e-06,
+      "loss": 1.6483,
+      "step": 11280
+    },
+    {
+      "epoch": 1.612857142857143,
+      "grad_norm": 0.0442969836294651,
+      "learning_rate": 9.24857142857143e-06,
+      "loss": 0.0,
+      "step": 11290
+    },
+    {
+      "epoch": 1.6142857142857143,
+      "grad_norm": 0.10007743537425995,
+      "learning_rate": 9.23904761904762e-06,
+      "loss": 1.71,
+      "step": 11300
+    },
+    {
+      "epoch": 1.6157142857142857,
+      "grad_norm": 0.011671596206724644,
+      "learning_rate": 9.229523809523811e-06,
+      "loss": 0.8455,
+      "step": 11310
+    },
+    {
+      "epoch": 1.617142857142857,
+      "grad_norm": 0.052139513194561005,
+      "learning_rate": 9.220000000000002e-06,
+      "loss": 2.3641,
+      "step": 11320
+    },
+    {
+      "epoch": 1.6185714285714285,
+      "grad_norm": 0.03550355136394501,
+      "learning_rate": 9.21047619047619e-06,
+      "loss": 0.0,
+      "step": 11330
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.030742991715669632,
+      "learning_rate": 9.200952380952381e-06,
+      "loss": 0.7314,
+      "step": 11340
+    },
+    {
+      "epoch": 1.6214285714285714,
+      "grad_norm": 0.024231290444731712,
+      "learning_rate": 9.191428571428572e-06,
+      "loss": 0.0,
+      "step": 11350
+    },
+    {
+      "epoch": 1.6228571428571428,
+      "grad_norm": 0.0076188212260603905,
+      "learning_rate": 9.181904761904762e-06,
+      "loss": 0.0,
+      "step": 11360
+    },
+    {
+      "epoch": 1.6242857142857143,
+      "grad_norm": 0.08596672117710114,
+      "learning_rate": 9.172380952380953e-06,
+      "loss": 0.0,
+      "step": 11370
+    },
+    {
+      "epoch": 1.6257142857142857,
+      "grad_norm": 0.020385991781949997,
+      "learning_rate": 9.162857142857144e-06,
+      "loss": 0.9792,
+      "step": 11380
+    },
+    {
+      "epoch": 1.6271428571428572,
+      "grad_norm": 0.03302644565701485,
+      "learning_rate": 9.153333333333334e-06,
+      "loss": 0.0,
+      "step": 11390
+    },
+    {
+      "epoch": 1.6285714285714286,
+      "grad_norm": 6145187.5,
+      "learning_rate": 9.143809523809525e-06,
+      "loss": 0.9114,
+      "step": 11400
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.008451025933027267,
+      "learning_rate": 9.134285714285716e-06,
+      "loss": 0.0008,
+      "step": 11410
+    },
+    {
+      "epoch": 1.6314285714285715,
+      "grad_norm": 0.07952427864074707,
+      "learning_rate": 9.124761904761905e-06,
+      "loss": 2.7717,
+      "step": 11420
+    },
+    {
+      "epoch": 1.632857142857143,
+      "grad_norm": 0.26180052757263184,
+      "learning_rate": 9.115238095238095e-06,
+      "loss": 1.6667,
+      "step": 11430
+    },
+    {
+      "epoch": 1.6342857142857143,
+      "grad_norm": 0.08318978548049927,
+      "learning_rate": 9.105714285714286e-06,
+      "loss": 0.8937,
+      "step": 11440
+    },
+    {
+      "epoch": 1.6357142857142857,
+      "grad_norm": 0.02828172966837883,
+      "learning_rate": 9.096190476190477e-06,
+      "loss": 0.0,
+      "step": 11450
+    },
+    {
+      "epoch": 1.637142857142857,
+      "grad_norm": 0.05926382169127464,
+      "learning_rate": 9.086666666666667e-06,
+      "loss": 0.0,
+      "step": 11460
+    },
+    {
+      "epoch": 1.6385714285714286,
+      "grad_norm": 0.03142034262418747,
+      "learning_rate": 9.077142857142858e-06,
+      "loss": 0.0,
+      "step": 11470
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.05101107805967331,
+      "learning_rate": 9.067619047619048e-06,
+      "loss": 1.8281,
+      "step": 11480
+    },
+    {
+      "epoch": 1.6414285714285715,
+      "grad_norm": 0.10932973027229309,
+      "learning_rate": 9.058095238095239e-06,
+      "loss": 0.8198,
+      "step": 11490
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.039454907178878784,
+      "learning_rate": 9.048571428571428e-06,
+      "loss": 0.8136,
+      "step": 11500
+    },
+    {
+      "epoch": 1.6442857142857141,
+      "grad_norm": 7504889.0,
+      "learning_rate": 9.039047619047619e-06,
+      "loss": 0.8696,
+      "step": 11510
+    },
+    {
+      "epoch": 1.6457142857142857,
+      "grad_norm": 0.06502719223499298,
+      "learning_rate": 9.029523809523811e-06,
+      "loss": 0.0,
+      "step": 11520
+    },
+    {
+      "epoch": 1.6471428571428572,
+      "grad_norm": 0.005630491767078638,
+      "learning_rate": 9.020000000000002e-06,
+      "loss": 0.0,
+      "step": 11530
+    },
+    {
+      "epoch": 1.6485714285714286,
+      "grad_norm": 0.04296800494194031,
+      "learning_rate": 9.010476190476192e-06,
+      "loss": 1.6754,
+      "step": 11540
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.03174138441681862,
+      "learning_rate": 9.000952380952381e-06,
+      "loss": 0.0,
+      "step": 11550
+    },
+    {
+      "epoch": 1.6514285714285715,
+      "grad_norm": 0.04477181285619736,
+      "learning_rate": 8.991428571428572e-06,
+      "loss": 0.9508,
+      "step": 11560
+    },
+    {
+      "epoch": 1.6528571428571428,
+      "grad_norm": 0.04433402046561241,
+      "learning_rate": 8.981904761904763e-06,
+      "loss": 0.0,
+      "step": 11570
+    },
+    {
+      "epoch": 1.6542857142857144,
+      "grad_norm": 0.02097504772245884,
+      "learning_rate": 8.972380952380953e-06,
+      "loss": 0.0,
+      "step": 11580
+    },
+    {
+      "epoch": 1.6557142857142857,
+      "grad_norm": 0.037330057471990585,
+      "learning_rate": 8.962857142857144e-06,
+      "loss": 0.9105,
+      "step": 11590
+    },
+    {
+      "epoch": 1.657142857142857,
+      "grad_norm": 0.02221691980957985,
+      "learning_rate": 8.953333333333335e-06,
+      "loss": 0.0,
+      "step": 11600
+    },
+    {
+      "epoch": 1.6585714285714286,
+      "grad_norm": 0.020200734958052635,
+      "learning_rate": 8.943809523809525e-06,
+      "loss": 0.0,
+      "step": 11610
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 0.0309230238199234,
+      "learning_rate": 8.934285714285716e-06,
+      "loss": 0.0,
+      "step": 11620
+    },
+    {
+      "epoch": 1.6614285714285715,
+      "grad_norm": 4660519.5,
+      "learning_rate": 8.924761904761905e-06,
+      "loss": 3.5208,
+      "step": 11630
+    },
+    {
+      "epoch": 1.6628571428571428,
+      "grad_norm": 0.04088440164923668,
+      "learning_rate": 8.915238095238095e-06,
+      "loss": 0.0,
+      "step": 11640
+    },
+    {
+      "epoch": 1.6642857142857141,
+      "grad_norm": 4753061.0,
+      "learning_rate": 8.905714285714286e-06,
+      "loss": 0.8386,
+      "step": 11650
+    },
+    {
+      "epoch": 1.6657142857142857,
+      "grad_norm": 0.029728995636105537,
+      "learning_rate": 8.896190476190477e-06,
+      "loss": 0.0,
+      "step": 11660
+    },
+    {
+      "epoch": 1.6671428571428573,
+      "grad_norm": 0.029220612719655037,
+      "learning_rate": 8.886666666666667e-06,
+      "loss": 0.8158,
+      "step": 11670
+    },
+    {
+      "epoch": 1.6685714285714286,
+      "grad_norm": 0.03511948511004448,
+      "learning_rate": 8.877142857142858e-06,
+      "loss": 1.3741,
+      "step": 11680
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3778995.5,
+      "learning_rate": 8.867619047619049e-06,
+      "loss": 1.8713,
+      "step": 11690
+    },
+    {
+      "epoch": 1.6714285714285713,
+      "grad_norm": 0.046670690178871155,
+      "learning_rate": 8.85809523809524e-06,
+      "loss": 1.6915,
+      "step": 11700
+    },
+    {
+      "epoch": 1.6728571428571428,
+      "grad_norm": 0.08629463613033295,
+      "learning_rate": 8.84857142857143e-06,
+      "loss": 1.6498,
+      "step": 11710
+    },
+    {
+      "epoch": 1.6742857142857144,
+      "grad_norm": 0.09224408119916916,
+      "learning_rate": 8.839047619047619e-06,
+      "loss": 0.7731,
+      "step": 11720
+    },
+    {
+      "epoch": 1.6757142857142857,
+      "grad_norm": 0.09708651900291443,
+      "learning_rate": 8.82952380952381e-06,
+      "loss": 0.257,
+      "step": 11730
+    },
+    {
+      "epoch": 1.677142857142857,
+      "grad_norm": 173.38900756835938,
+      "learning_rate": 8.82e-06,
+      "loss": 0.0003,
+      "step": 11740
+    },
+    {
+      "epoch": 1.6785714285714286,
+      "grad_norm": 0.05735960230231285,
+      "learning_rate": 8.810476190476191e-06,
+      "loss": 0.7941,
+      "step": 11750
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.02949138730764389,
+      "learning_rate": 8.800952380952382e-06,
+      "loss": 0.0,
+      "step": 11760
+    },
+    {
+      "epoch": 1.6814285714285715,
+      "grad_norm": 0.019309235736727715,
+      "learning_rate": 8.791428571428572e-06,
+      "loss": 0.9568,
+      "step": 11770
+    },
+    {
+      "epoch": 1.6828571428571428,
+      "grad_norm": 0.029960017651319504,
+      "learning_rate": 8.781904761904763e-06,
+      "loss": 0.0,
+      "step": 11780
+    },
+    {
+      "epoch": 1.6842857142857142,
+      "grad_norm": 0.12467370182275772,
+      "learning_rate": 8.772380952380953e-06,
+      "loss": 0.0,
+      "step": 11790
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.03524554520845413,
+      "learning_rate": 8.762857142857144e-06,
+      "loss": 0.901,
+      "step": 11800
+    },
+    {
+      "epoch": 1.6871428571428573,
+      "grad_norm": 0.02077874168753624,
+      "learning_rate": 8.753333333333333e-06,
+      "loss": 0.0,
+      "step": 11810
+    },
+    {
+      "epoch": 1.6885714285714286,
+      "grad_norm": 0.07136952131986618,
+      "learning_rate": 8.743809523809524e-06,
+      "loss": 0.9533,
+      "step": 11820
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.028592074289917946,
+      "learning_rate": 8.734285714285714e-06,
+      "loss": 0.0,
+      "step": 11830
+    },
+    {
+      "epoch": 1.6914285714285713,
+      "grad_norm": 0.04971682280302048,
+      "learning_rate": 8.724761904761905e-06,
+      "loss": 0.0,
+      "step": 11840
+    },
+    {
+      "epoch": 1.6928571428571428,
+      "grad_norm": 0.02126012183725834,
+      "learning_rate": 8.715238095238096e-06,
+      "loss": 0.8358,
+      "step": 11850
+    },
+    {
+      "epoch": 1.6942857142857144,
+      "grad_norm": 4750947.0,
+      "learning_rate": 8.705714285714286e-06,
+      "loss": 1.7417,
+      "step": 11860
+    },
+    {
+      "epoch": 1.6957142857142857,
+      "grad_norm": 0.04757643863558769,
+      "learning_rate": 8.696190476190477e-06,
+      "loss": 0.9111,
+      "step": 11870
+    },
+    {
+      "epoch": 1.697142857142857,
+      "grad_norm": 0.0325455516576767,
+      "learning_rate": 8.686666666666668e-06,
+      "loss": 1.7395,
+      "step": 11880
+    },
+    {
+      "epoch": 1.6985714285714286,
+      "grad_norm": 0.029710449278354645,
+      "learning_rate": 8.677142857142857e-06,
+      "loss": 0.8331,
+      "step": 11890
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.06584242731332779,
+      "learning_rate": 8.667619047619047e-06,
+      "loss": 0.0,
+      "step": 11900
+    },
+    {
+      "epoch": 1.7014285714285715,
+      "grad_norm": 0.041798193007707596,
+      "learning_rate": 8.65809523809524e-06,
+      "loss": 0.9795,
+      "step": 11910
+    },
+    {
+      "epoch": 1.7028571428571428,
+      "grad_norm": 0.024586506187915802,
+      "learning_rate": 8.64857142857143e-06,
+      "loss": 1.7472,
+      "step": 11920
+    },
+    {
+      "epoch": 1.7042857142857142,
+      "grad_norm": 0.4036490023136139,
+      "learning_rate": 8.639047619047621e-06,
+      "loss": 0.0001,
+      "step": 11930
+    },
+    {
+      "epoch": 1.7057142857142857,
+      "grad_norm": 0.052682604640722275,
+      "learning_rate": 8.62952380952381e-06,
+      "loss": 0.0,
+      "step": 11940
+    },
+    {
+      "epoch": 1.7071428571428573,
+      "grad_norm": 0.020886780694127083,
+      "learning_rate": 8.62e-06,
+      "loss": 0.8153,
+      "step": 11950
+    },
+    {
+      "epoch": 1.7085714285714286,
+      "grad_norm": 5013815.0,
+      "learning_rate": 8.610476190476191e-06,
+      "loss": 2.5194,
+      "step": 11960
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4492655.0,
+      "learning_rate": 8.600952380952382e-06,
+      "loss": 0.9916,
+      "step": 11970
+    },
+    {
+      "epoch": 1.7114285714285713,
+      "grad_norm": 0.05369066074490547,
+      "learning_rate": 8.591428571428572e-06,
+      "loss": 1.9118,
+      "step": 11980
+    },
+    {
+      "epoch": 1.7128571428571429,
+      "grad_norm": 4975093.0,
+      "learning_rate": 8.581904761904763e-06,
+      "loss": 1.7144,
+      "step": 11990
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.02870997227728367,
+      "learning_rate": 8.572380952380954e-06,
+      "loss": 1.8505,
+      "step": 12000
+    },
+    {
+      "epoch": 1.7157142857142857,
+      "grad_norm": 0.05842341110110283,
+      "learning_rate": 8.562857142857144e-06,
+      "loss": 0.8684,
+      "step": 12010
+    },
+    {
+      "epoch": 1.717142857142857,
+      "grad_norm": 0.08662697672843933,
+      "learning_rate": 8.553333333333333e-06,
+      "loss": 0.0,
+      "step": 12020
+    },
+    {
+      "epoch": 1.7185714285714284,
+      "grad_norm": 0.06977694481611252,
+      "learning_rate": 8.543809523809524e-06,
+      "loss": 0.0,
+      "step": 12030
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.05909581854939461,
+      "learning_rate": 8.534285714285715e-06,
+      "loss": 0.0,
+      "step": 12040
+    },
+    {
+      "epoch": 1.7214285714285715,
+      "grad_norm": 0.05366959050297737,
+      "learning_rate": 8.524761904761905e-06,
+      "loss": 1.7453,
+      "step": 12050
+    },
+    {
+      "epoch": 1.7228571428571429,
+      "grad_norm": 0.03931805118918419,
+      "learning_rate": 8.515238095238096e-06,
+      "loss": 0.7853,
+      "step": 12060
+    },
+    {
+      "epoch": 1.7242857142857142,
+      "grad_norm": 0.053184110671281815,
+      "learning_rate": 8.505714285714287e-06,
+      "loss": 0.0,
+      "step": 12070
+    },
+    {
+      "epoch": 1.7257142857142858,
+      "grad_norm": 0.057105839252471924,
+      "learning_rate": 8.496190476190477e-06,
+      "loss": 0.0,
+      "step": 12080
+    },
+    {
+      "epoch": 1.727142857142857,
+      "grad_norm": 0.04400743916630745,
+      "learning_rate": 8.486666666666668e-06,
+      "loss": 0.9426,
+      "step": 12090
+    },
+    {
+      "epoch": 1.7285714285714286,
+      "grad_norm": 0.02190248668193817,
+      "learning_rate": 8.477142857142859e-06,
+      "loss": 0.0,
+      "step": 12100
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.208927869796753,
+      "learning_rate": 8.467619047619047e-06,
+      "loss": 0.0,
+      "step": 12110
+    },
+    {
+      "epoch": 1.7314285714285713,
+      "grad_norm": 0.0405452698469162,
+      "learning_rate": 8.458095238095238e-06,
+      "loss": 2.5951,
+      "step": 12120
+    },
+    {
+      "epoch": 1.7328571428571429,
+      "grad_norm": 0.025026915594935417,
+      "learning_rate": 8.448571428571429e-06,
+      "loss": 0.0,
+      "step": 12130
+    },
+    {
+      "epoch": 1.7342857142857144,
+      "grad_norm": 0.05885399878025055,
+      "learning_rate": 8.43904761904762e-06,
+      "loss": 0.8928,
+      "step": 12140
+    },
+    {
+      "epoch": 1.7357142857142858,
+      "grad_norm": 0.05936788395047188,
+      "learning_rate": 8.42952380952381e-06,
+      "loss": 0.8258,
+      "step": 12150
+    },
+    {
+      "epoch": 1.737142857142857,
+      "grad_norm": 0.03433917462825775,
+      "learning_rate": 8.42e-06,
+      "loss": 0.0,
+      "step": 12160
+    },
+    {
+      "epoch": 1.7385714285714284,
+      "grad_norm": 0.07870560884475708,
+      "learning_rate": 8.410476190476191e-06,
+      "loss": 0.8094,
+      "step": 12170
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.046131033450365067,
+      "learning_rate": 8.400952380952382e-06,
+      "loss": 0.8421,
+      "step": 12180
+    },
+    {
+      "epoch": 1.7414285714285715,
+      "grad_norm": 0.034755297005176544,
+      "learning_rate": 8.391428571428573e-06,
+      "loss": 0.9294,
+      "step": 12190
+    },
+    {
+      "epoch": 1.7428571428571429,
+      "grad_norm": 1.1396067142486572,
+      "learning_rate": 8.381904761904762e-06,
+      "loss": 0.0,
+      "step": 12200
+    },
+    {
+      "epoch": 1.7442857142857142,
+      "grad_norm": 0.12503458559513092,
+      "learning_rate": 8.372380952380952e-06,
+      "loss": 0.9413,
+      "step": 12210
+    },
+    {
+      "epoch": 1.7457142857142856,
+      "grad_norm": 0.03785332292318344,
+      "learning_rate": 8.362857142857143e-06,
+      "loss": 0.654,
+      "step": 12220
+    },
+    {
+      "epoch": 1.747142857142857,
+      "grad_norm": 0.038935281336307526,
+      "learning_rate": 8.353333333333335e-06,
+      "loss": 0.0,
+      "step": 12230
+    },
+    {
+      "epoch": 1.7485714285714287,
+      "grad_norm": 0.016737395897507668,
+      "learning_rate": 8.343809523809524e-06,
+      "loss": 0.0,
+      "step": 12240
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.03153121471405029,
+      "learning_rate": 8.334285714285715e-06,
+      "loss": 0.0,
+      "step": 12250
+    },
+    {
+      "epoch": 1.7514285714285713,
+      "grad_norm": 0.08055660128593445,
+      "learning_rate": 8.324761904761905e-06,
+      "loss": 0.0,
+      "step": 12260
+    },
+    {
+      "epoch": 1.752857142857143,
+      "grad_norm": 0.010881300084292889,
+      "learning_rate": 8.315238095238096e-06,
+      "loss": 0.8748,
+      "step": 12270
+    },
+    {
+      "epoch": 1.7542857142857144,
+      "grad_norm": 0.014238235540688038,
+      "learning_rate": 8.305714285714285e-06,
+      "loss": 0.8942,
+      "step": 12280
+    },
+    {
+      "epoch": 1.7557142857142858,
+      "grad_norm": 0.01419445127248764,
+      "learning_rate": 8.296190476190476e-06,
+      "loss": 1.8275,
+      "step": 12290
+    },
+    {
+      "epoch": 1.7571428571428571,
+      "grad_norm": 0.016620777547359467,
+      "learning_rate": 8.286666666666668e-06,
+      "loss": 0.0,
+      "step": 12300
+    },
+    {
+      "epoch": 1.7585714285714285,
+      "grad_norm": 0.051889825612306595,
+      "learning_rate": 8.277142857142859e-06,
+      "loss": 0.951,
+      "step": 12310
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.04498285800218582,
+      "learning_rate": 8.26761904761905e-06,
+      "loss": 0.0,
+      "step": 12320
+    },
+    {
+      "epoch": 1.7614285714285716,
+      "grad_norm": 0.016710374504327774,
+      "learning_rate": 8.258095238095238e-06,
+      "loss": 0.8351,
+      "step": 12330
+    },
+    {
+      "epoch": 1.762857142857143,
+      "grad_norm": 0.0809824988245964,
+      "learning_rate": 8.248571428571429e-06,
+      "loss": 0.8234,
+      "step": 12340
+    },
+    {
+      "epoch": 1.7642857142857142,
+      "grad_norm": 0.013250447809696198,
+      "learning_rate": 8.23904761904762e-06,
+      "loss": 0.0,
+      "step": 12350
+    },
+    {
+      "epoch": 1.7657142857142856,
+      "grad_norm": 0.013528087176382542,
+      "learning_rate": 8.22952380952381e-06,
+      "loss": 0.8569,
+      "step": 12360
+    },
+    {
+      "epoch": 1.7671428571428571,
+      "grad_norm": 0.0186093058437109,
+      "learning_rate": 8.220000000000001e-06,
+      "loss": 0.0,
+      "step": 12370
+    },
+    {
+      "epoch": 1.7685714285714287,
+      "grad_norm": 0.021876858547329903,
+      "learning_rate": 8.210476190476192e-06,
+      "loss": 0.925,
+      "step": 12380
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.061624739319086075,
+      "learning_rate": 8.200952380952382e-06,
+      "loss": 0.8682,
+      "step": 12390
+    },
+    {
+      "epoch": 1.7714285714285714,
+      "grad_norm": 0.020552512258291245,
+      "learning_rate": 8.191428571428573e-06,
+      "loss": 0.0,
+      "step": 12400
+    },
+    {
+      "epoch": 1.772857142857143,
+      "grad_norm": 0.028076374903321266,
+      "learning_rate": 8.181904761904764e-06,
+      "loss": 0.0,
+      "step": 12410
+    },
+    {
+      "epoch": 1.7742857142857142,
+      "grad_norm": 0.1330912560224533,
+      "learning_rate": 8.172380952380952e-06,
+      "loss": 0.0,
+      "step": 12420
+    },
+    {
+      "epoch": 1.7757142857142858,
+      "grad_norm": 0.019502893090248108,
+      "learning_rate": 8.162857142857143e-06,
+      "loss": 2.8272,
+      "step": 12430
+    },
+    {
+      "epoch": 1.7771428571428571,
+      "grad_norm": 4954593.5,
+      "learning_rate": 8.153333333333334e-06,
+      "loss": 1.681,
+      "step": 12440
+    },
+    {
+      "epoch": 1.7785714285714285,
+      "grad_norm": 0.1823224574327469,
+      "learning_rate": 8.143809523809524e-06,
+      "loss": 0.0,
+      "step": 12450
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.031395167112350464,
+      "learning_rate": 8.134285714285715e-06,
+      "loss": 1.7186,
+      "step": 12460
+    },
+    {
+      "epoch": 1.7814285714285716,
+      "grad_norm": 0.04063650593161583,
+      "learning_rate": 8.124761904761906e-06,
+      "loss": 0.6777,
+      "step": 12470
+    },
+    {
+      "epoch": 1.782857142857143,
+      "grad_norm": 0.04235513508319855,
+      "learning_rate": 8.115238095238096e-06,
+      "loss": 1.7929,
+      "step": 12480
+    },
+    {
+      "epoch": 1.7842857142857143,
+      "grad_norm": 0.02930898778140545,
+      "learning_rate": 8.105714285714287e-06,
+      "loss": 0.0,
+      "step": 12490
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.08112237602472305,
+      "learning_rate": 8.096190476190476e-06,
+      "loss": 0.0,
+      "step": 12500
+    },
+    {
+      "epoch": 1.7871428571428571,
+      "grad_norm": 0.02050590142607689,
+      "learning_rate": 8.086666666666667e-06,
+      "loss": 0.0,
+      "step": 12510
+    },
+    {
+      "epoch": 1.7885714285714287,
+      "grad_norm": 0.022168094292283058,
+      "learning_rate": 8.077142857142857e-06,
+      "loss": 0.0,
+      "step": 12520
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.06000300496816635,
+      "learning_rate": 8.067619047619048e-06,
+      "loss": 0.0,
+      "step": 12530
+    },
+    {
+      "epoch": 1.7914285714285714,
+      "grad_norm": 0.01769920438528061,
+      "learning_rate": 8.058095238095239e-06,
+      "loss": 0.0,
+      "step": 12540
+    },
+    {
+      "epoch": 1.7928571428571427,
+      "grad_norm": 0.04867337644100189,
+      "learning_rate": 8.04857142857143e-06,
+      "loss": 2.6993,
+      "step": 12550
+    },
+    {
+      "epoch": 1.7942857142857143,
+      "grad_norm": 0.39821621775627136,
+      "learning_rate": 8.03904761904762e-06,
+      "loss": 0.0,
+      "step": 12560
+    },
+    {
+      "epoch": 1.7957142857142858,
+      "grad_norm": 0.03589576855301857,
+      "learning_rate": 8.02952380952381e-06,
+      "loss": 0.0,
+      "step": 12570
+    },
+    {
+      "epoch": 1.7971428571428572,
+      "grad_norm": 0.020575974136590958,
+      "learning_rate": 8.020000000000001e-06,
+      "loss": 0.9805,
+      "step": 12580
+    },
+    {
+      "epoch": 1.7985714285714285,
+      "grad_norm": 0.04015924409031868,
+      "learning_rate": 8.01047619047619e-06,
+      "loss": 2.7083,
+      "step": 12590
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.048210106790065765,
+      "learning_rate": 8.00095238095238e-06,
+      "loss": 1.5671,
+      "step": 12600
+    },
+    {
+      "epoch": 1.8014285714285714,
+      "grad_norm": 0.025007126852869987,
+      "learning_rate": 7.991428571428571e-06,
+      "loss": 0.0,
+      "step": 12610
+    },
+    {
+      "epoch": 1.802857142857143,
+      "grad_norm": 0.043515849858522415,
+      "learning_rate": 7.981904761904764e-06,
+      "loss": 0.8841,
+      "step": 12620
+    },
+    {
+      "epoch": 1.8042857142857143,
+      "grad_norm": 11400702.0,
+      "learning_rate": 7.972380952380953e-06,
+      "loss": 0.8572,
+      "step": 12630
+    },
+    {
+      "epoch": 1.8057142857142856,
+      "grad_norm": 0.1292562484741211,
+      "learning_rate": 7.962857142857143e-06,
+      "loss": 0.0,
+      "step": 12640
+    },
+    {
+      "epoch": 1.8071428571428572,
+      "grad_norm": 0.19864441454410553,
+      "learning_rate": 7.953333333333334e-06,
+      "loss": 0.9528,
+      "step": 12650
+    },
+    {
+      "epoch": 1.8085714285714287,
+      "grad_norm": 0.14438113570213318,
+      "learning_rate": 7.943809523809525e-06,
+      "loss": 0.8676,
+      "step": 12660
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.10513030737638474,
+      "learning_rate": 7.934285714285715e-06,
+      "loss": 0.0,
+      "step": 12670
+    },
+    {
+      "epoch": 1.8114285714285714,
+      "grad_norm": 3853691.25,
+      "learning_rate": 7.924761904761906e-06,
+      "loss": 0.8819,
+      "step": 12680
+    },
+    {
+      "epoch": 1.8128571428571427,
+      "grad_norm": 0.02937900274991989,
+      "learning_rate": 7.915238095238097e-06,
+      "loss": 0.0,
+      "step": 12690
+    },
+    {
+      "epoch": 1.8142857142857143,
+      "grad_norm": 0.007405418902635574,
+      "learning_rate": 7.905714285714287e-06,
+      "loss": 0.0,
+      "step": 12700
+    },
+    {
+      "epoch": 1.8157142857142858,
+      "grad_norm": 0.01287918258458376,
+      "learning_rate": 7.896190476190478e-06,
+      "loss": 0.9297,
+      "step": 12710
+    },
+    {
+      "epoch": 1.8171428571428572,
+      "grad_norm": 0.025693848729133606,
+      "learning_rate": 7.886666666666667e-06,
+      "loss": 1.7613,
+      "step": 12720
+    },
+    {
+      "epoch": 1.8185714285714285,
+      "grad_norm": 0.06837037205696106,
+      "learning_rate": 7.877142857142857e-06,
+      "loss": 0.0,
+      "step": 12730
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 0.05290854349732399,
+      "learning_rate": 7.867619047619048e-06,
+      "loss": 0.9853,
+      "step": 12740
+    },
+    {
+      "epoch": 1.8214285714285714,
+      "grad_norm": 0.017154382541775703,
+      "learning_rate": 7.858095238095239e-06,
+      "loss": 0.9079,
+      "step": 12750
+    },
+    {
+      "epoch": 1.822857142857143,
+      "grad_norm": 87.49150085449219,
+      "learning_rate": 7.84857142857143e-06,
+      "loss": 0.0,
+      "step": 12760
+    },
+    {
+      "epoch": 1.8242857142857143,
+      "grad_norm": 0.025130225345492363,
+      "learning_rate": 7.83904761904762e-06,
+      "loss": 0.886,
+      "step": 12770
+    },
+    {
+      "epoch": 1.8257142857142856,
+      "grad_norm": 0.03993659466505051,
+      "learning_rate": 7.82952380952381e-06,
+      "loss": 0.0,
+      "step": 12780
+    },
+    {
+      "epoch": 1.8271428571428572,
+      "grad_norm": 0.021225837990641594,
+      "learning_rate": 7.820000000000001e-06,
+      "loss": 0.0,
+      "step": 12790
+    },
+    {
+      "epoch": 1.8285714285714287,
+      "grad_norm": 4598252.5,
+      "learning_rate": 7.810476190476192e-06,
+      "loss": 1.6334,
+      "step": 12800
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.27047473192214966,
+      "learning_rate": 7.800952380952381e-06,
+      "loss": 0.9283,
+      "step": 12810
+    },
+    {
+      "epoch": 1.8314285714285714,
+      "grad_norm": 0.01126906555145979,
+      "learning_rate": 7.791428571428572e-06,
+      "loss": 1.7532,
+      "step": 12820
+    },
+    {
+      "epoch": 1.8328571428571427,
+      "grad_norm": 0.0347716398537159,
+      "learning_rate": 7.781904761904762e-06,
+      "loss": 0.0,
+      "step": 12830
+    },
+    {
+      "epoch": 1.8342857142857143,
+      "grad_norm": 1.7544127702713013,
+      "learning_rate": 7.772380952380953e-06,
+      "loss": 0.0,
+      "step": 12840
+    },
+    {
+      "epoch": 1.8357142857142859,
+      "grad_norm": 15578080.0,
+      "learning_rate": 7.762857142857144e-06,
+      "loss": 1.5278,
+      "step": 12850
+    },
+    {
+      "epoch": 1.8371428571428572,
+      "grad_norm": 0.023089904338121414,
+      "learning_rate": 7.753333333333334e-06,
+      "loss": 0.7302,
+      "step": 12860
+    },
+    {
+      "epoch": 1.8385714285714285,
+      "grad_norm": 0.03251311928033829,
+      "learning_rate": 7.743809523809525e-06,
+      "loss": 0.0,
+      "step": 12870
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.00599782494828105,
+      "learning_rate": 7.734285714285715e-06,
+      "loss": 0.0,
+      "step": 12880
+    },
+    {
+      "epoch": 1.8414285714285714,
+      "grad_norm": 0.07999641448259354,
+      "learning_rate": 7.724761904761904e-06,
+      "loss": 0.8788,
+      "step": 12890
+    },
+    {
+      "epoch": 1.842857142857143,
+      "grad_norm": 0.034568943083286285,
+      "learning_rate": 7.715238095238095e-06,
+      "loss": 0.0,
+      "step": 12900
+    },
+    {
+      "epoch": 1.8442857142857143,
+      "grad_norm": 0.03586239740252495,
+      "learning_rate": 7.705714285714286e-06,
+      "loss": 0.8989,
+      "step": 12910
+    },
+    {
+      "epoch": 1.8457142857142856,
+      "grad_norm": 532647744.0,
+      "learning_rate": 7.696190476190476e-06,
+      "loss": 1.3925,
+      "step": 12920
+    },
+    {
+      "epoch": 1.8471428571428572,
+      "grad_norm": 581642.5,
+      "learning_rate": 7.686666666666667e-06,
+      "loss": 0.0,
+      "step": 12930
+    },
+    {
+      "epoch": 1.8485714285714285,
+      "grad_norm": 0.017363866791129112,
+      "learning_rate": 7.677142857142858e-06,
+      "loss": 0.0,
+      "step": 12940
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.06800173968076706,
+      "learning_rate": 7.667619047619048e-06,
+      "loss": 0.8515,
+      "step": 12950
+    },
+    {
+      "epoch": 1.8514285714285714,
+      "grad_norm": 259845728.0,
+      "learning_rate": 7.658095238095239e-06,
+      "loss": 2.3724,
+      "step": 12960
+    },
+    {
+      "epoch": 1.8528571428571428,
+      "grad_norm": 0.022794008255004883,
+      "learning_rate": 7.64857142857143e-06,
+      "loss": 1.7457,
+      "step": 12970
+    },
+    {
+      "epoch": 1.8542857142857143,
+      "grad_norm": 0.16726438701152802,
+      "learning_rate": 7.639047619047619e-06,
+      "loss": 2.5499,
+      "step": 12980
+    },
+    {
+      "epoch": 1.8557142857142859,
+      "grad_norm": 0.06718578189611435,
+      "learning_rate": 7.62952380952381e-06,
+      "loss": 0.0,
+      "step": 12990
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.04962693154811859,
+      "learning_rate": 7.620000000000001e-06,
+      "loss": 0.0,
+      "step": 13000
+    },
+    {
+      "epoch": 1.8585714285714285,
+      "grad_norm": 0.0328340157866478,
+      "learning_rate": 7.610476190476191e-06,
+      "loss": 0.0,
+      "step": 13010
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 0.019031718373298645,
+      "learning_rate": 7.600952380952381e-06,
+      "loss": 0.9015,
+      "step": 13020
+    },
+    {
+      "epoch": 1.8614285714285714,
+      "grad_norm": 0.08938376605510712,
+      "learning_rate": 7.591428571428572e-06,
+      "loss": 0.8288,
+      "step": 13030
+    },
+    {
+      "epoch": 1.862857142857143,
+      "grad_norm": 0.041360948234796524,
+      "learning_rate": 7.5819047619047625e-06,
+      "loss": 0.0,
+      "step": 13040
+    },
+    {
+      "epoch": 1.8642857142857143,
+      "grad_norm": 0.008723510429263115,
+      "learning_rate": 7.572380952380953e-06,
+      "loss": 0.0,
+      "step": 13050
+    },
+    {
+      "epoch": 1.8657142857142857,
+      "grad_norm": 17061550.0,
+      "learning_rate": 7.562857142857144e-06,
+      "loss": 0.7525,
+      "step": 13060
+    },
+    {
+      "epoch": 1.867142857142857,
+      "grad_norm": 0.007334563415497541,
+      "learning_rate": 7.553333333333334e-06,
+      "loss": 0.0,
+      "step": 13070
+    },
+    {
+      "epoch": 1.8685714285714285,
+      "grad_norm": 0.02472829259932041,
+      "learning_rate": 7.543809523809524e-06,
+      "loss": 1.8733,
+      "step": 13080
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0108131542801857,
+      "learning_rate": 7.534285714285715e-06,
+      "loss": 0.0,
+      "step": 13090
+    },
+    {
+      "epoch": 1.8714285714285714,
+      "grad_norm": 0.02634693868458271,
+      "learning_rate": 7.5247619047619055e-06,
+      "loss": 0.0,
+      "step": 13100
+    },
+    {
+      "epoch": 1.8728571428571428,
+      "grad_norm": 0.011638377793133259,
+      "learning_rate": 7.515238095238095e-06,
+      "loss": 0.9291,
+      "step": 13110
+    },
+    {
+      "epoch": 1.8742857142857143,
+      "grad_norm": 4829373.0,
+      "learning_rate": 7.505714285714286e-06,
+      "loss": 0.8884,
+      "step": 13120
+    },
+    {
+      "epoch": 1.8757142857142857,
+      "grad_norm": 1.196772575378418,
+      "learning_rate": 7.496190476190477e-06,
+      "loss": 1.6337,
+      "step": 13130
+    },
+    {
+      "epoch": 1.8771428571428572,
+      "grad_norm": 0.030632255598902702,
+      "learning_rate": 7.486666666666667e-06,
+      "loss": 0.8071,
+      "step": 13140
+    },
+    {
+      "epoch": 1.8785714285714286,
+      "grad_norm": 0.04415116459131241,
+      "learning_rate": 7.477142857142857e-06,
+      "loss": 0.0,
+      "step": 13150
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.02758082002401352,
+      "learning_rate": 7.467619047619048e-06,
+      "loss": 1.6633,
+      "step": 13160
+    },
+    {
+      "epoch": 1.8814285714285715,
+      "grad_norm": 0.01741745136678219,
+      "learning_rate": 7.458095238095238e-06,
+      "loss": 1.6585,
+      "step": 13170
+    },
+    {
+      "epoch": 1.882857142857143,
+      "grad_norm": 0.317516565322876,
+      "learning_rate": 7.44857142857143e-06,
+      "loss": 0.0,
+      "step": 13180
+    },
+    {
+      "epoch": 1.8842857142857143,
+      "grad_norm": 0.024077218025922775,
+      "learning_rate": 7.4390476190476205e-06,
+      "loss": 1.8841,
+      "step": 13190
+    },
+    {
+      "epoch": 1.8857142857142857,
+      "grad_norm": 9307705.0,
+      "learning_rate": 7.4295238095238095e-06,
+      "loss": 0.8635,
+      "step": 13200
+    },
+    {
+      "epoch": 1.887142857142857,
+      "grad_norm": 0.026135461404919624,
+      "learning_rate": 7.420000000000001e-06,
+      "loss": 0.9786,
+      "step": 13210
+    },
+    {
+      "epoch": 1.8885714285714286,
+      "grad_norm": 0.033224478363990784,
+      "learning_rate": 7.410476190476192e-06,
+      "loss": 0.9618,
+      "step": 13220
+    },
+    {
+      "epoch": 1.8900000000000001,
+      "grad_norm": 0.06334014981985092,
+      "learning_rate": 7.400952380952382e-06,
+      "loss": 0.0,
+      "step": 13230
+    },
+    {
+      "epoch": 1.8914285714285715,
+      "grad_norm": 0.04852615296840668,
+      "learning_rate": 7.391428571428571e-06,
+      "loss": 0.0,
+      "step": 13240
+    },
+    {
+      "epoch": 1.8928571428571428,
+      "grad_norm": 4664189.0,
+      "learning_rate": 7.381904761904763e-06,
+      "loss": 0.9242,
+      "step": 13250
+    },
+    {
+      "epoch": 1.8942857142857141,
+      "grad_norm": 0.013565067201852798,
+      "learning_rate": 7.372380952380953e-06,
+      "loss": 0.0,
+      "step": 13260
+    },
+    {
+      "epoch": 1.8957142857142857,
+      "grad_norm": 0.026172904297709465,
+      "learning_rate": 7.362857142857144e-06,
+      "loss": 0.0,
+      "step": 13270
+    },
+    {
+      "epoch": 1.8971428571428572,
+      "grad_norm": 0.03383246064186096,
+      "learning_rate": 7.353333333333334e-06,
+      "loss": 0.0,
+      "step": 13280
+    },
+    {
+      "epoch": 1.8985714285714286,
+      "grad_norm": 0.03225620090961456,
+      "learning_rate": 7.3438095238095244e-06,
+      "loss": 0.0,
+      "step": 13290
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.01578551158308983,
+      "learning_rate": 7.334285714285715e-06,
+      "loss": 0.4979,
+      "step": 13300
+    },
+    {
+      "epoch": 1.9014285714285715,
+      "grad_norm": 0.037304289638996124,
+      "learning_rate": 7.324761904761906e-06,
+      "loss": 0.0,
+      "step": 13310
+    },
+    {
+      "epoch": 1.9028571428571428,
+      "grad_norm": 0.02372370846569538,
+      "learning_rate": 7.315238095238096e-06,
+      "loss": 0.8976,
+      "step": 13320
+    },
+    {
+      "epoch": 1.9042857142857144,
+      "grad_norm": 0.0276383887976408,
+      "learning_rate": 7.305714285714286e-06,
+      "loss": 0.954,
+      "step": 13330
+    },
+    {
+      "epoch": 1.9057142857142857,
+      "grad_norm": 0.02599337138235569,
+      "learning_rate": 7.296190476190477e-06,
+      "loss": 0.9404,
+      "step": 13340
+    },
+    {
+      "epoch": 1.907142857142857,
+      "grad_norm": 0.032843947410583496,
+      "learning_rate": 7.2866666666666675e-06,
+      "loss": 1.9172,
+      "step": 13350
+    },
+    {
+      "epoch": 1.9085714285714286,
+      "grad_norm": 0.05747692286968231,
+      "learning_rate": 7.277142857142858e-06,
+      "loss": 1.7163,
+      "step": 13360
+    },
+    {
+      "epoch": 1.9100000000000001,
+      "grad_norm": 0.08389802277088165,
+      "learning_rate": 7.267619047619048e-06,
+      "loss": 0.0,
+      "step": 13370
+    },
+    {
+      "epoch": 1.9114285714285715,
+      "grad_norm": 0.04973737895488739,
+      "learning_rate": 7.258095238095239e-06,
+      "loss": 1.8034,
+      "step": 13380
+    },
+    {
+      "epoch": 1.9128571428571428,
+      "grad_norm": 19002402.0,
+      "learning_rate": 7.248571428571429e-06,
+      "loss": 1.4182,
+      "step": 13390
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 0.03119681030511856,
+      "learning_rate": 7.23904761904762e-06,
+      "loss": 0.5672,
+      "step": 13400
+    },
+    {
+      "epoch": 1.9157142857142857,
+      "grad_norm": 0.07832422852516174,
+      "learning_rate": 7.22952380952381e-06,
+      "loss": 0.0,
+      "step": 13410
+    },
+    {
+      "epoch": 1.9171428571428573,
+      "grad_norm": 0.04946679621934891,
+      "learning_rate": 7.22e-06,
+      "loss": 2.5876,
+      "step": 13420
+    },
+    {
+      "epoch": 1.9185714285714286,
+      "grad_norm": 0.030742652714252472,
+      "learning_rate": 7.210476190476191e-06,
+      "loss": 0.0,
+      "step": 13430
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4490061.0,
+      "learning_rate": 7.200952380952382e-06,
+      "loss": 2.804,
+      "step": 13440
+    },
+    {
+      "epoch": 1.9214285714285713,
+      "grad_norm": 0.05265035480260849,
+      "learning_rate": 7.191428571428572e-06,
+      "loss": 0.881,
+      "step": 13450
+    },
+    {
+      "epoch": 1.9228571428571428,
+      "grad_norm": 0.10334628075361252,
+      "learning_rate": 7.181904761904762e-06,
+      "loss": 0.8537,
+      "step": 13460
+    },
+    {
+      "epoch": 1.9242857142857144,
+      "grad_norm": 0.03528710827231407,
+      "learning_rate": 7.172380952380953e-06,
+      "loss": 0.9015,
+      "step": 13470
+    },
+    {
+      "epoch": 1.9257142857142857,
+      "grad_norm": 0.10401737689971924,
+      "learning_rate": 7.162857142857143e-06,
+      "loss": 0.0,
+      "step": 13480
+    },
+    {
+      "epoch": 1.927142857142857,
+      "grad_norm": 0.23502996563911438,
+      "learning_rate": 7.153333333333334e-06,
+      "loss": 0.0,
+      "step": 13490
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.060353346168994904,
+      "learning_rate": 7.143809523809524e-06,
+      "loss": 0.0,
+      "step": 13500
+    },
+    {
+      "epoch": 1.9300000000000002,
+      "grad_norm": 0.06288421899080276,
+      "learning_rate": 7.1342857142857145e-06,
+      "loss": 0.0,
+      "step": 13510
+    },
+    {
+      "epoch": 1.9314285714285715,
+      "grad_norm": 3069222.5,
+      "learning_rate": 7.124761904761905e-06,
+      "loss": 0.9227,
+      "step": 13520
+    },
+    {
+      "epoch": 1.9328571428571428,
+      "grad_norm": 0.05926242843270302,
+      "learning_rate": 7.115238095238096e-06,
+      "loss": 0.8723,
+      "step": 13530
+    },
+    {
+      "epoch": 1.9342857142857142,
+      "grad_norm": 0.03476998209953308,
+      "learning_rate": 7.1057142857142856e-06,
+      "loss": 0.0,
+      "step": 13540
+    },
+    {
+      "epoch": 1.9357142857142857,
+      "grad_norm": 0.0370350256562233,
+      "learning_rate": 7.096190476190476e-06,
+      "loss": 0.0,
+      "step": 13550
+    },
+    {
+      "epoch": 1.9371428571428573,
+      "grad_norm": 0.025922702625393867,
+      "learning_rate": 7.086666666666667e-06,
+      "loss": 0.0,
+      "step": 13560
+    },
+    {
+      "epoch": 1.9385714285714286,
+      "grad_norm": 0.03708845376968384,
+      "learning_rate": 7.077142857142858e-06,
+      "loss": 0.8643,
+      "step": 13570
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.019589660689234734,
+      "learning_rate": 7.067619047619049e-06,
+      "loss": 0.0,
+      "step": 13580
+    },
+    {
+      "epoch": 1.9414285714285713,
+      "grad_norm": 0.011717982590198517,
+      "learning_rate": 7.058095238095238e-06,
+      "loss": 0.0,
+      "step": 13590
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "grad_norm": 0.02304036170244217,
+      "learning_rate": 7.0485714285714295e-06,
+      "loss": 1.8256,
+      "step": 13600
+    },
+    {
+      "epoch": 1.9442857142857144,
+      "grad_norm": 0.15395911037921906,
+      "learning_rate": 7.03904761904762e-06,
+      "loss": 0.0,
+      "step": 13610
+    },
+    {
+      "epoch": 1.9457142857142857,
+      "grad_norm": 0.015593756921589375,
+      "learning_rate": 7.029523809523811e-06,
+      "loss": 0.0,
+      "step": 13620
+    },
+    {
+      "epoch": 1.947142857142857,
+      "grad_norm": 0.052702855318784714,
+      "learning_rate": 7.0200000000000006e-06,
+      "loss": 0.9505,
+      "step": 13630
+    },
+    {
+      "epoch": 1.9485714285714286,
+      "grad_norm": 0.05286931246519089,
+      "learning_rate": 7.010476190476191e-06,
+      "loss": 0.8599,
+      "step": 13640
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.021530501544475555,
+      "learning_rate": 7.000952380952382e-06,
+      "loss": 1.7314,
+      "step": 13650
+    },
+    {
+      "epoch": 1.9514285714285715,
+      "grad_norm": 0.025508251041173935,
+      "learning_rate": 6.9914285714285725e-06,
+      "loss": 0.0,
+      "step": 13660
+    },
+    {
+      "epoch": 1.9528571428571428,
+      "grad_norm": 0.04209945723414421,
+      "learning_rate": 6.981904761904762e-06,
+      "loss": 0.0,
+      "step": 13670
+    },
+    {
+      "epoch": 1.9542857142857142,
+      "grad_norm": 0.1630469411611557,
+      "learning_rate": 6.972380952380953e-06,
+      "loss": 2.5708,
+      "step": 13680
+    },
+    {
+      "epoch": 1.9557142857142857,
+      "grad_norm": 0.07957037538290024,
+      "learning_rate": 6.962857142857144e-06,
+      "loss": 0.8845,
+      "step": 13690
+    },
+    {
+      "epoch": 1.9571428571428573,
+      "grad_norm": 0.07693076878786087,
+      "learning_rate": 6.953333333333334e-06,
+      "loss": 1.8996,
+      "step": 13700
+    },
+    {
+      "epoch": 1.9585714285714286,
+      "grad_norm": 0.02903757244348526,
+      "learning_rate": 6.943809523809525e-06,
+      "loss": 0.0,
+      "step": 13710
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.029366090893745422,
+      "learning_rate": 6.934285714285715e-06,
+      "loss": 0.0,
+      "step": 13720
+    },
+    {
+      "epoch": 1.9614285714285713,
+      "grad_norm": 0.03108065575361252,
+      "learning_rate": 6.924761904761905e-06,
+      "loss": 1.7357,
+      "step": 13730
+    },
+    {
+      "epoch": 1.9628571428571429,
+      "grad_norm": 0.01383767556399107,
+      "learning_rate": 6.915238095238096e-06,
+      "loss": 0.6418,
+      "step": 13740
+    },
+    {
+      "epoch": 1.9642857142857144,
+      "grad_norm": 3147418.0,
+      "learning_rate": 6.905714285714287e-06,
+      "loss": 2.7398,
+      "step": 13750
+    },
+    {
+      "epoch": 1.9657142857142857,
+      "grad_norm": 0.19594497978687286,
+      "learning_rate": 6.8961904761904764e-06,
+      "loss": 0.0,
+      "step": 13760
+    },
+    {
+      "epoch": 1.967142857142857,
+      "grad_norm": 7139092.0,
+      "learning_rate": 6.886666666666667e-06,
+      "loss": 2.6174,
+      "step": 13770
+    },
+    {
+      "epoch": 1.9685714285714284,
+      "grad_norm": 0.05144255980849266,
+      "learning_rate": 6.877142857142858e-06,
+      "loss": 0.8091,
+      "step": 13780
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.041355427354574203,
+      "learning_rate": 6.867619047619048e-06,
+      "loss": 0.0,
+      "step": 13790
+    },
+    {
+      "epoch": 1.9714285714285715,
+      "grad_norm": 0.05698074400424957,
+      "learning_rate": 6.858095238095238e-06,
+      "loss": 2.4876,
+      "step": 13800
+    },
+    {
+      "epoch": 1.9728571428571429,
+      "grad_norm": 0.024431632831692696,
+      "learning_rate": 6.848571428571429e-06,
+      "loss": 0.0,
+      "step": 13810
+    },
+    {
+      "epoch": 1.9742857142857142,
+      "grad_norm": 0.03249214217066765,
+      "learning_rate": 6.8390476190476195e-06,
+      "loss": 0.7012,
+      "step": 13820
+    },
+    {
+      "epoch": 1.9757142857142858,
+      "grad_norm": 0.025866640731692314,
+      "learning_rate": 6.82952380952381e-06,
+      "loss": 0.0,
+      "step": 13830
+    },
+    {
+      "epoch": 1.977142857142857,
+      "grad_norm": 0.0322066992521286,
+      "learning_rate": 6.820000000000001e-06,
+      "loss": 0.8694,
+      "step": 13840
+    },
+    {
+      "epoch": 1.9785714285714286,
+      "grad_norm": 0.03143410384654999,
+      "learning_rate": 6.8104761904761906e-06,
+      "loss": 0.0,
+      "step": 13850
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.03990085795521736,
+      "learning_rate": 6.800952380952381e-06,
+      "loss": 0.0,
+      "step": 13860
+    },
+    {
+      "epoch": 1.9814285714285713,
+      "grad_norm": 0.0176379531621933,
+      "learning_rate": 6.791428571428572e-06,
+      "loss": 0.0,
+      "step": 13870
+    },
+    {
+      "epoch": 1.9828571428571429,
+      "grad_norm": 6278807.5,
+      "learning_rate": 6.7819047619047625e-06,
+      "loss": 3.384,
+      "step": 13880
+    },
+    {
+      "epoch": 1.9842857142857144,
+      "grad_norm": 0.04752973094582558,
+      "learning_rate": 6.772380952380952e-06,
+      "loss": 0.8943,
+      "step": 13890
+    },
+    {
+      "epoch": 1.9857142857142858,
+      "grad_norm": 0.011762041598558426,
+      "learning_rate": 6.762857142857143e-06,
+      "loss": 0.811,
+      "step": 13900
+    },
+    {
+      "epoch": 1.987142857142857,
+      "grad_norm": 0.03045344538986683,
+      "learning_rate": 6.753333333333334e-06,
+      "loss": 0.0,
+      "step": 13910
+    },
+    {
+      "epoch": 1.9885714285714284,
+      "grad_norm": 0.15351426601409912,
+      "learning_rate": 6.743809523809525e-06,
+      "loss": 0.0,
+      "step": 13920
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4441822.0,
+      "learning_rate": 6.734285714285714e-06,
+      "loss": 0.8598,
+      "step": 13930
+    },
+    {
+      "epoch": 1.9914285714285715,
+      "grad_norm": 0.02928520366549492,
+      "learning_rate": 6.724761904761905e-06,
+      "loss": 0.0,
+      "step": 13940
+    },
+    {
+      "epoch": 1.9928571428571429,
+      "grad_norm": 0.016408968716859818,
+      "learning_rate": 6.715238095238095e-06,
+      "loss": 0.9091,
+      "step": 13950
+    },
+    {
+      "epoch": 1.9942857142857142,
+      "grad_norm": 0.011410136707127094,
+      "learning_rate": 6.705714285714287e-06,
+      "loss": 0.9257,
+      "step": 13960
+    },
+    {
+      "epoch": 1.9957142857142856,
+      "grad_norm": 0.02280694618821144,
+      "learning_rate": 6.6961904761904775e-06,
+      "loss": 0.0,
+      "step": 13970
+    },
+    {
+      "epoch": 1.997142857142857,
+      "grad_norm": 9090234.0,
+      "learning_rate": 6.6866666666666665e-06,
+      "loss": 1.7967,
+      "step": 13980
+    },
+    {
+      "epoch": 1.9985714285714287,
+      "grad_norm": 0.04230930656194687,
+      "learning_rate": 6.677142857142858e-06,
+      "loss": 1.5388,
+      "step": 13990
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.040579888969659805,
+      "learning_rate": 6.667619047619049e-06,
+      "loss": 1.5513,
+      "step": 14000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9411666666666667,
+      "eval_f1_macro": 0.9406174956166151,
+      "eval_f1_weighted": 0.9412113998371214,
+      "eval_loss": 1.0513261556625366,
+      "eval_runtime": 186.2241,
+      "eval_samples_per_second": 32.219,
+      "eval_steps_per_second": 16.11,
+      "step": 14000
+    },
+    {
+      "epoch": 2.0014285714285713,
+      "grad_norm": 0.040499672293663025,
+      "learning_rate": 6.658095238095239e-06,
+      "loss": 0.0,
+      "step": 14010
+    },
+    {
+      "epoch": 2.0028571428571427,
+      "grad_norm": 0.042962539941072464,
+      "learning_rate": 6.648571428571429e-06,
+      "loss": 0.9265,
+      "step": 14020
+    },
+    {
+      "epoch": 2.0042857142857144,
+      "grad_norm": 0.038135118782520294,
+      "learning_rate": 6.63904761904762e-06,
+      "loss": 0.9615,
+      "step": 14030
+    },
+    {
+      "epoch": 2.005714285714286,
+      "grad_norm": 0.035035379230976105,
+      "learning_rate": 6.62952380952381e-06,
+      "loss": 0.7945,
+      "step": 14040
+    },
+    {
+      "epoch": 2.007142857142857,
+      "grad_norm": 0.02678876370191574,
+      "learning_rate": 6.620000000000001e-06,
+      "loss": 0.0,
+      "step": 14050
+    },
+    {
+      "epoch": 2.0085714285714285,
+      "grad_norm": 0.05272085964679718,
+      "learning_rate": 6.610476190476191e-06,
+      "loss": 0.0,
+      "step": 14060
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.015692196786403656,
+      "learning_rate": 6.6009523809523814e-06,
+      "loss": 0.8969,
+      "step": 14070
+    },
+    {
+      "epoch": 2.0114285714285716,
+      "grad_norm": 0.02628430724143982,
+      "learning_rate": 6.591428571428572e-06,
+      "loss": 0.0,
+      "step": 14080
+    },
+    {
+      "epoch": 2.012857142857143,
+      "grad_norm": 0.052258994430303574,
+      "learning_rate": 6.581904761904763e-06,
+      "loss": 0.0,
+      "step": 14090
+    },
+    {
+      "epoch": 2.0142857142857142,
+      "grad_norm": 0.07236022502183914,
+      "learning_rate": 6.572380952380953e-06,
+      "loss": 0.0,
+      "step": 14100
+    },
+    {
+      "epoch": 2.0157142857142856,
+      "grad_norm": 0.04797660931944847,
+      "learning_rate": 6.562857142857143e-06,
+      "loss": 0.0058,
+      "step": 14110
+    },
+    {
+      "epoch": 2.0171428571428573,
+      "grad_norm": 0.006010725162923336,
+      "learning_rate": 6.553333333333334e-06,
+      "loss": 0.0,
+      "step": 14120
+    },
+    {
+      "epoch": 2.0185714285714287,
+      "grad_norm": 0.027171477675437927,
+      "learning_rate": 6.5438095238095245e-06,
+      "loss": 0.9373,
+      "step": 14130
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.02046150714159012,
+      "learning_rate": 6.534285714285715e-06,
+      "loss": 0.0,
+      "step": 14140
+    },
+    {
+      "epoch": 2.0214285714285714,
+      "grad_norm": 0.012259164825081825,
+      "learning_rate": 6.524761904761905e-06,
+      "loss": 0.9435,
+      "step": 14150
+    },
+    {
+      "epoch": 2.0228571428571427,
+      "grad_norm": 0.021372059360146523,
+      "learning_rate": 6.515238095238096e-06,
+      "loss": 0.8918,
+      "step": 14160
+    },
+    {
+      "epoch": 2.0242857142857145,
+      "grad_norm": 0.03328077495098114,
+      "learning_rate": 6.505714285714286e-06,
+      "loss": 0.0,
+      "step": 14170
+    },
+    {
+      "epoch": 2.025714285714286,
+      "grad_norm": 0.03952356427907944,
+      "learning_rate": 6.496190476190477e-06,
+      "loss": 0.9669,
+      "step": 14180
+    },
+    {
+      "epoch": 2.027142857142857,
+      "grad_norm": 0.03169762343168259,
+      "learning_rate": 6.486666666666667e-06,
+      "loss": 1.7534,
+      "step": 14190
+    },
+    {
+      "epoch": 2.0285714285714285,
+      "grad_norm": 0.036155182868242264,
+      "learning_rate": 6.477142857142857e-06,
+      "loss": 0.0,
+      "step": 14200
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.07257933914661407,
+      "learning_rate": 6.467619047619048e-06,
+      "loss": 0.0,
+      "step": 14210
+    },
+    {
+      "epoch": 2.0314285714285716,
+      "grad_norm": 0.11710529029369354,
+      "learning_rate": 6.458095238095239e-06,
+      "loss": 0.0,
+      "step": 14220
+    },
+    {
+      "epoch": 2.032857142857143,
+      "grad_norm": 0.0049024648033082485,
+      "learning_rate": 6.448571428571429e-06,
+      "loss": 0.0,
+      "step": 14230
+    },
+    {
+      "epoch": 2.0342857142857143,
+      "grad_norm": 0.0804293155670166,
+      "learning_rate": 6.439047619047619e-06,
+      "loss": 0.9177,
+      "step": 14240
+    },
+    {
+      "epoch": 2.0357142857142856,
+      "grad_norm": 0.024495337158441544,
+      "learning_rate": 6.42952380952381e-06,
+      "loss": 0.0,
+      "step": 14250
+    },
+    {
+      "epoch": 2.0371428571428574,
+      "grad_norm": 0.017586397007107735,
+      "learning_rate": 6.42e-06,
+      "loss": 0.7971,
+      "step": 14260
+    },
+    {
+      "epoch": 2.0385714285714287,
+      "grad_norm": 0.014519397169351578,
+      "learning_rate": 6.410476190476191e-06,
+      "loss": 0.7918,
+      "step": 14270
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.007483644876629114,
+      "learning_rate": 6.400952380952381e-06,
+      "loss": 0.0,
+      "step": 14280
+    },
+    {
+      "epoch": 2.0414285714285714,
+      "grad_norm": 0.020176388323307037,
+      "learning_rate": 6.3914285714285715e-06,
+      "loss": 0.0,
+      "step": 14290
+    },
+    {
+      "epoch": 2.0428571428571427,
+      "grad_norm": 0.0899249017238617,
+      "learning_rate": 6.381904761904762e-06,
+      "loss": 0.0,
+      "step": 14300
+    },
+    {
+      "epoch": 2.0442857142857145,
+      "grad_norm": 0.03339070826768875,
+      "learning_rate": 6.372380952380954e-06,
+      "loss": 0.3676,
+      "step": 14310
+    },
+    {
+      "epoch": 2.045714285714286,
+      "grad_norm": 0.01635129190981388,
+      "learning_rate": 6.3628571428571426e-06,
+      "loss": 0.0,
+      "step": 14320
+    },
+    {
+      "epoch": 2.047142857142857,
+      "grad_norm": 0.012603417970240116,
+      "learning_rate": 6.353333333333333e-06,
+      "loss": 0.0,
+      "step": 14330
+    },
+    {
+      "epoch": 2.0485714285714285,
+      "grad_norm": 5091128.0,
+      "learning_rate": 6.343809523809525e-06,
+      "loss": 0.9515,
+      "step": 14340
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.027420813217759132,
+      "learning_rate": 6.334285714285715e-06,
+      "loss": 0.9814,
+      "step": 14350
+    },
+    {
+      "epoch": 2.0514285714285716,
+      "grad_norm": 0.04156985878944397,
+      "learning_rate": 6.324761904761906e-06,
+      "loss": 0.0,
+      "step": 14360
+    },
+    {
+      "epoch": 2.052857142857143,
+      "grad_norm": 0.018319146707654,
+      "learning_rate": 6.315238095238096e-06,
+      "loss": 0.0,
+      "step": 14370
+    },
+    {
+      "epoch": 2.0542857142857143,
+      "grad_norm": 0.011170899495482445,
+      "learning_rate": 6.3057142857142864e-06,
+      "loss": 0.0,
+      "step": 14380
+    },
+    {
+      "epoch": 2.0557142857142856,
+      "grad_norm": 0.21745148301124573,
+      "learning_rate": 6.296190476190477e-06,
+      "loss": 0.0,
+      "step": 14390
+    },
+    {
+      "epoch": 2.057142857142857,
+      "grad_norm": 0.007084170822054148,
+      "learning_rate": 6.286666666666668e-06,
+      "loss": 0.8935,
+      "step": 14400
+    },
+    {
+      "epoch": 2.0585714285714287,
+      "grad_norm": 0.020290784537792206,
+      "learning_rate": 6.2771428571428575e-06,
+      "loss": 1.7453,
+      "step": 14410
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.015166763216257095,
+      "learning_rate": 6.267619047619048e-06,
+      "loss": 1.3303,
+      "step": 14420
+    },
+    {
+      "epoch": 2.0614285714285714,
+      "grad_norm": 8584750.0,
+      "learning_rate": 6.258095238095239e-06,
+      "loss": 0.846,
+      "step": 14430
+    },
+    {
+      "epoch": 2.0628571428571427,
+      "grad_norm": 0.02963012456893921,
+      "learning_rate": 6.2485714285714295e-06,
+      "loss": 1.8212,
+      "step": 14440
+    },
+    {
+      "epoch": 2.064285714285714,
+      "grad_norm": 0.2593902349472046,
+      "learning_rate": 6.239047619047619e-06,
+      "loss": 0.0,
+      "step": 14450
+    },
+    {
+      "epoch": 2.065714285714286,
+      "grad_norm": 0.044694364070892334,
+      "learning_rate": 6.22952380952381e-06,
+      "loss": 0.0,
+      "step": 14460
+    },
+    {
+      "epoch": 2.067142857142857,
+      "grad_norm": 0.029570017009973526,
+      "learning_rate": 6.220000000000001e-06,
+      "loss": 0.0,
+      "step": 14470
+    },
+    {
+      "epoch": 2.0685714285714285,
+      "grad_norm": 0.16961121559143066,
+      "learning_rate": 6.210476190476191e-06,
+      "loss": 0.8222,
+      "step": 14480
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.01637084409594536,
+      "learning_rate": 6.200952380952382e-06,
+      "loss": 0.0,
+      "step": 14490
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 0.011821471154689789,
+      "learning_rate": 6.191428571428572e-06,
+      "loss": 0.9404,
+      "step": 14500
+    },
+    {
+      "epoch": 2.072857142857143,
+      "grad_norm": 0.029488177970051765,
+      "learning_rate": 6.181904761904762e-06,
+      "loss": 0.9529,
+      "step": 14510
+    },
+    {
+      "epoch": 2.0742857142857143,
+      "grad_norm": 0.02430882677435875,
+      "learning_rate": 6.172380952380953e-06,
+      "loss": 0.868,
+      "step": 14520
+    },
+    {
+      "epoch": 2.0757142857142856,
+      "grad_norm": 0.00898817554116249,
+      "learning_rate": 6.162857142857144e-06,
+      "loss": 0.0,
+      "step": 14530
+    },
+    {
+      "epoch": 2.077142857142857,
+      "grad_norm": 0.00614017341285944,
+      "learning_rate": 6.1533333333333334e-06,
+      "loss": 0.0,
+      "step": 14540
+    },
+    {
+      "epoch": 2.0785714285714287,
+      "grad_norm": 0.021440107375383377,
+      "learning_rate": 6.143809523809524e-06,
+      "loss": 0.0,
+      "step": 14550
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.00791624654084444,
+      "learning_rate": 6.134285714285715e-06,
+      "loss": 1.86,
+      "step": 14560
+    },
+    {
+      "epoch": 2.0814285714285714,
+      "grad_norm": 0.019122779369354248,
+      "learning_rate": 6.124761904761905e-06,
+      "loss": 0.7588,
+      "step": 14570
+    },
+    {
+      "epoch": 2.0828571428571427,
+      "grad_norm": 0.028674675151705742,
+      "learning_rate": 6.115238095238095e-06,
+      "loss": 0.8987,
+      "step": 14580
+    },
+    {
+      "epoch": 2.084285714285714,
+      "grad_norm": 0.01149414200335741,
+      "learning_rate": 6.105714285714286e-06,
+      "loss": 0.0,
+      "step": 14590
+    },
+    {
+      "epoch": 2.085714285714286,
+      "grad_norm": 0.03224168345332146,
+      "learning_rate": 6.0961904761904765e-06,
+      "loss": 0.0,
+      "step": 14600
+    },
+    {
+      "epoch": 2.087142857142857,
+      "grad_norm": 0.06836597621440887,
+      "learning_rate": 6.086666666666667e-06,
+      "loss": 0.0,
+      "step": 14610
+    },
+    {
+      "epoch": 2.0885714285714285,
+      "grad_norm": 0.01457243412733078,
+      "learning_rate": 6.077142857142858e-06,
+      "loss": 0.8425,
+      "step": 14620
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.027177726849913597,
+      "learning_rate": 6.0676190476190476e-06,
+      "loss": 0.0,
+      "step": 14630
+    },
+    {
+      "epoch": 2.0914285714285716,
+      "grad_norm": 0.026359591633081436,
+      "learning_rate": 6.058095238095238e-06,
+      "loss": 0.8152,
+      "step": 14640
+    },
+    {
+      "epoch": 2.092857142857143,
+      "grad_norm": 0.01572665572166443,
+      "learning_rate": 6.048571428571429e-06,
+      "loss": 0.0,
+      "step": 14650
+    },
+    {
+      "epoch": 2.0942857142857143,
+      "grad_norm": 0.02452906407415867,
+      "learning_rate": 6.03904761904762e-06,
+      "loss": 0.0,
+      "step": 14660
+    },
+    {
+      "epoch": 2.0957142857142856,
+      "grad_norm": 0.015321512706577778,
+      "learning_rate": 6.029523809523809e-06,
+      "loss": 0.0,
+      "step": 14670
+    },
+    {
+      "epoch": 2.097142857142857,
+      "grad_norm": 0.030517295002937317,
+      "learning_rate": 6.02e-06,
+      "loss": 1.7372,
+      "step": 14680
+    },
+    {
+      "epoch": 2.0985714285714288,
+      "grad_norm": 0.014419672079384327,
+      "learning_rate": 6.010476190476191e-06,
+      "loss": 0.0,
+      "step": 14690
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.03204868733882904,
+      "learning_rate": 6.000952380952382e-06,
+      "loss": 0.0,
+      "step": 14700
+    },
+    {
+      "epoch": 2.1014285714285714,
+      "grad_norm": 0.016639873385429382,
+      "learning_rate": 5.991428571428571e-06,
+      "loss": 0.0,
+      "step": 14710
+    },
+    {
+      "epoch": 2.1028571428571428,
+      "grad_norm": 0.007913457229733467,
+      "learning_rate": 5.981904761904762e-06,
+      "loss": 0.0,
+      "step": 14720
+    },
+    {
+      "epoch": 2.104285714285714,
+      "grad_norm": 0.023547278717160225,
+      "learning_rate": 5.972380952380953e-06,
+      "loss": 0.0,
+      "step": 14730
+    },
+    {
+      "epoch": 2.105714285714286,
+      "grad_norm": 0.015420054085552692,
+      "learning_rate": 5.962857142857144e-06,
+      "loss": 0.0,
+      "step": 14740
+    },
+    {
+      "epoch": 2.107142857142857,
+      "grad_norm": 0.01140822097659111,
+      "learning_rate": 5.9533333333333345e-06,
+      "loss": 0.0,
+      "step": 14750
+    },
+    {
+      "epoch": 2.1085714285714285,
+      "grad_norm": 0.006182031240314245,
+      "learning_rate": 5.943809523809524e-06,
+      "loss": 0.0,
+      "step": 14760
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.10311761498451233,
+      "learning_rate": 5.934285714285715e-06,
+      "loss": 0.0,
+      "step": 14770
+    },
+    {
+      "epoch": 2.111428571428571,
+      "grad_norm": 0.07140810042619705,
+      "learning_rate": 5.924761904761906e-06,
+      "loss": 0.0,
+      "step": 14780
+    },
+    {
+      "epoch": 2.112857142857143,
+      "grad_norm": 0.02012266218662262,
+      "learning_rate": 5.915238095238096e-06,
+      "loss": 0.0,
+      "step": 14790
+    },
+    {
+      "epoch": 2.1142857142857143,
+      "grad_norm": 0.06497155129909515,
+      "learning_rate": 5.905714285714286e-06,
+      "loss": 0.0,
+      "step": 14800
+    },
+    {
+      "epoch": 2.1157142857142857,
+      "grad_norm": 0.017037099227309227,
+      "learning_rate": 5.896190476190477e-06,
+      "loss": 0.0,
+      "step": 14810
+    },
+    {
+      "epoch": 2.117142857142857,
+      "grad_norm": 0.025647088885307312,
+      "learning_rate": 5.886666666666667e-06,
+      "loss": 0.0,
+      "step": 14820
+    },
+    {
+      "epoch": 2.1185714285714288,
+      "grad_norm": 0.015479667112231255,
+      "learning_rate": 5.877142857142858e-06,
+      "loss": 0.0,
+      "step": 14830
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.006776072550565004,
+      "learning_rate": 5.867619047619048e-06,
+      "loss": 0.0,
+      "step": 14840
+    },
+    {
+      "epoch": 2.1214285714285714,
+      "grad_norm": 0.010106460191309452,
+      "learning_rate": 5.8580952380952384e-06,
+      "loss": 1.0385,
+      "step": 14850
+    },
+    {
+      "epoch": 2.1228571428571428,
+      "grad_norm": 0.03378785774111748,
+      "learning_rate": 5.848571428571429e-06,
+      "loss": 0.0,
+      "step": 14860
+    },
+    {
+      "epoch": 2.124285714285714,
+      "grad_norm": 0.007815182209014893,
+      "learning_rate": 5.83904761904762e-06,
+      "loss": 0.0,
+      "step": 14870
+    },
+    {
+      "epoch": 2.125714285714286,
+      "grad_norm": 0.013247417286038399,
+      "learning_rate": 5.82952380952381e-06,
+      "loss": 0.0,
+      "step": 14880
+    },
+    {
+      "epoch": 2.1271428571428572,
+      "grad_norm": 0.003985950257629156,
+      "learning_rate": 5.82e-06,
+      "loss": 0.0,
+      "step": 14890
+    },
+    {
+      "epoch": 2.1285714285714286,
+      "grad_norm": 0.08258320391178131,
+      "learning_rate": 5.810476190476191e-06,
+      "loss": 1.9723,
+      "step": 14900
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.00741097517311573,
+      "learning_rate": 5.8009523809523815e-06,
+      "loss": 1.8996,
+      "step": 14910
+    },
+    {
+      "epoch": 2.1314285714285712,
+      "grad_norm": 0.039227720350027084,
+      "learning_rate": 5.791428571428572e-06,
+      "loss": 0.0,
+      "step": 14920
+    },
+    {
+      "epoch": 2.132857142857143,
+      "grad_norm": 0.020858589559793472,
+      "learning_rate": 5.781904761904762e-06,
+      "loss": 0.0,
+      "step": 14930
+    },
+    {
+      "epoch": 2.1342857142857143,
+      "grad_norm": 0.020924564450979233,
+      "learning_rate": 5.7723809523809526e-06,
+      "loss": 0.0,
+      "step": 14940
+    },
+    {
+      "epoch": 2.1357142857142857,
+      "grad_norm": 0.01114263292402029,
+      "learning_rate": 5.762857142857143e-06,
+      "loss": 0.0,
+      "step": 14950
+    },
+    {
+      "epoch": 2.137142857142857,
+      "grad_norm": 0.02770884335041046,
+      "learning_rate": 5.753333333333334e-06,
+      "loss": 0.9501,
+      "step": 14960
+    },
+    {
+      "epoch": 2.138571428571429,
+      "grad_norm": 0.016802500933408737,
+      "learning_rate": 5.7438095238095245e-06,
+      "loss": 0.0,
+      "step": 14970
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.00748151820152998,
+      "learning_rate": 5.734285714285714e-06,
+      "loss": 0.0,
+      "step": 14980
+    },
+    {
+      "epoch": 2.1414285714285715,
+      "grad_norm": 0.3284965753555298,
+      "learning_rate": 5.724761904761905e-06,
+      "loss": 0.0,
+      "step": 14990
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.011496130377054214,
+      "learning_rate": 5.715238095238096e-06,
+      "loss": 0.9409,
+      "step": 15000
+    },
+    {
+      "epoch": 2.144285714285714,
+      "grad_norm": 0.010649595409631729,
+      "learning_rate": 5.705714285714286e-06,
+      "loss": 0.9179,
+      "step": 15010
+    },
+    {
+      "epoch": 2.145714285714286,
+      "grad_norm": 0.031590234488248825,
+      "learning_rate": 5.696190476190476e-06,
+      "loss": 0.0,
+      "step": 15020
+    },
+    {
+      "epoch": 2.1471428571428572,
+      "grad_norm": 33273278.0,
+      "learning_rate": 5.686666666666667e-06,
+      "loss": 1.6998,
+      "step": 15030
+    },
+    {
+      "epoch": 2.1485714285714286,
+      "grad_norm": 0.0236576609313488,
+      "learning_rate": 5.677142857142857e-06,
+      "loss": 0.0,
+      "step": 15040
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0441250316798687,
+      "learning_rate": 5.667619047619049e-06,
+      "loss": 0.0,
+      "step": 15050
+    },
+    {
+      "epoch": 2.1514285714285712,
+      "grad_norm": 0.01983768492937088,
+      "learning_rate": 5.658095238095238e-06,
+      "loss": 0.0,
+      "step": 15060
+    },
+    {
+      "epoch": 2.152857142857143,
+      "grad_norm": 0.023717420175671577,
+      "learning_rate": 5.6485714285714285e-06,
+      "loss": 0.0,
+      "step": 15070
+    },
+    {
+      "epoch": 2.1542857142857144,
+      "grad_norm": 0.009698578156530857,
+      "learning_rate": 5.63904761904762e-06,
+      "loss": 0.0,
+      "step": 15080
+    },
+    {
+      "epoch": 2.1557142857142857,
+      "grad_norm": 0.009750750847160816,
+      "learning_rate": 5.629523809523811e-06,
+      "loss": 0.8681,
+      "step": 15090
+    },
+    {
+      "epoch": 2.157142857142857,
+      "grad_norm": 0.016962634399533272,
+      "learning_rate": 5.620000000000001e-06,
+      "loss": 0.0,
+      "step": 15100
+    },
+    {
+      "epoch": 2.1585714285714284,
+      "grad_norm": 0.030302736908197403,
+      "learning_rate": 5.610476190476191e-06,
+      "loss": 0.0,
+      "step": 15110
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.013204126618802547,
+      "learning_rate": 5.600952380952382e-06,
+      "loss": 0.0,
+      "step": 15120
+    },
+    {
+      "epoch": 2.1614285714285715,
+      "grad_norm": 0.009623350575566292,
+      "learning_rate": 5.591428571428572e-06,
+      "loss": 0.8845,
+      "step": 15130
+    },
+    {
+      "epoch": 2.162857142857143,
+      "grad_norm": 0.008715730160474777,
+      "learning_rate": 5.581904761904763e-06,
+      "loss": 0.0,
+      "step": 15140
+    },
+    {
+      "epoch": 2.164285714285714,
+      "grad_norm": 0.010613787919282913,
+      "learning_rate": 5.572380952380953e-06,
+      "loss": 0.0,
+      "step": 15150
+    },
+    {
+      "epoch": 2.1657142857142855,
+      "grad_norm": 0.012956739403307438,
+      "learning_rate": 5.5628571428571434e-06,
+      "loss": 0.0,
+      "step": 15160
+    },
+    {
+      "epoch": 2.1671428571428573,
+      "grad_norm": 0.013021477498114109,
+      "learning_rate": 5.553333333333334e-06,
+      "loss": 0.0,
+      "step": 15170
+    },
+    {
+      "epoch": 2.1685714285714286,
+      "grad_norm": 3544762.5,
+      "learning_rate": 5.543809523809525e-06,
+      "loss": 0.9516,
+      "step": 15180
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.017944103106856346,
+      "learning_rate": 5.5342857142857145e-06,
+      "loss": 0.0,
+      "step": 15190
+    },
+    {
+      "epoch": 2.1714285714285713,
+      "grad_norm": 0.026982605457305908,
+      "learning_rate": 5.524761904761905e-06,
+      "loss": 0.0,
+      "step": 15200
+    },
+    {
+      "epoch": 2.172857142857143,
+      "grad_norm": 0.009403177537024021,
+      "learning_rate": 5.515238095238096e-06,
+      "loss": 0.0,
+      "step": 15210
+    },
+    {
+      "epoch": 2.1742857142857144,
+      "grad_norm": 0.04293425381183624,
+      "learning_rate": 5.5057142857142865e-06,
+      "loss": 0.8452,
+      "step": 15220
+    },
+    {
+      "epoch": 2.1757142857142857,
+      "grad_norm": 0.1375996172428131,
+      "learning_rate": 5.496190476190477e-06,
+      "loss": 1.3654,
+      "step": 15230
+    },
+    {
+      "epoch": 2.177142857142857,
+      "grad_norm": 0.013958347029983997,
+      "learning_rate": 5.486666666666667e-06,
+      "loss": 0.0,
+      "step": 15240
+    },
+    {
+      "epoch": 2.1785714285714284,
+      "grad_norm": 0.2042660266160965,
+      "learning_rate": 5.477142857142858e-06,
+      "loss": 0.0,
+      "step": 15250
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.084653340280056,
+      "learning_rate": 5.467619047619048e-06,
+      "loss": 0.0,
+      "step": 15260
+    },
+    {
+      "epoch": 2.1814285714285715,
+      "grad_norm": 0.015553503297269344,
+      "learning_rate": 5.458095238095239e-06,
+      "loss": 0.788,
+      "step": 15270
+    },
+    {
+      "epoch": 2.182857142857143,
+      "grad_norm": 0.02720208838582039,
+      "learning_rate": 5.448571428571429e-06,
+      "loss": 0.0,
+      "step": 15280
+    },
+    {
+      "epoch": 2.184285714285714,
+      "grad_norm": 0.020132893696427345,
+      "learning_rate": 5.439047619047619e-06,
+      "loss": 0.0,
+      "step": 15290
+    },
+    {
+      "epoch": 2.185714285714286,
+      "grad_norm": 0.006068873684853315,
+      "learning_rate": 5.42952380952381e-06,
+      "loss": 0.9559,
+      "step": 15300
+    },
+    {
+      "epoch": 2.1871428571428573,
+      "grad_norm": 7011777.0,
+      "learning_rate": 5.420000000000001e-06,
+      "loss": 1.8814,
+      "step": 15310
+    },
+    {
+      "epoch": 2.1885714285714286,
+      "grad_norm": 0.0083516426384449,
+      "learning_rate": 5.4104761904761904e-06,
+      "loss": 1.0396,
+      "step": 15320
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.759993076324463,
+      "learning_rate": 5.400952380952381e-06,
+      "loss": 1.0026,
+      "step": 15330
+    },
+    {
+      "epoch": 2.1914285714285713,
+      "grad_norm": 0.4637646973133087,
+      "learning_rate": 5.391428571428572e-06,
+      "loss": 1.018,
+      "step": 15340
+    },
+    {
+      "epoch": 2.192857142857143,
+      "grad_norm": 0.016094034537672997,
+      "learning_rate": 5.381904761904762e-06,
+      "loss": 0.9164,
+      "step": 15350
+    },
+    {
+      "epoch": 2.1942857142857144,
+      "grad_norm": 0.02648800052702427,
+      "learning_rate": 5.372380952380953e-06,
+      "loss": 0.0,
+      "step": 15360
+    },
+    {
+      "epoch": 2.1957142857142857,
+      "grad_norm": 0.004686668980866671,
+      "learning_rate": 5.362857142857143e-06,
+      "loss": 0.0,
+      "step": 15370
+    },
+    {
+      "epoch": 2.197142857142857,
+      "grad_norm": 0.03560120239853859,
+      "learning_rate": 5.3533333333333335e-06,
+      "loss": 0.0,
+      "step": 15380
+    },
+    {
+      "epoch": 2.1985714285714284,
+      "grad_norm": 6222867.5,
+      "learning_rate": 5.343809523809524e-06,
+      "loss": 0.8894,
+      "step": 15390
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.01654532365500927,
+      "learning_rate": 5.334285714285715e-06,
+      "loss": 0.0,
+      "step": 15400
+    },
+    {
+      "epoch": 2.2014285714285715,
+      "grad_norm": 0.030119221657514572,
+      "learning_rate": 5.3247619047619046e-06,
+      "loss": 0.0,
+      "step": 15410
+    },
+    {
+      "epoch": 2.202857142857143,
+      "grad_norm": 0.03396198898553848,
+      "learning_rate": 5.315238095238095e-06,
+      "loss": 0.0,
+      "step": 15420
+    },
+    {
+      "epoch": 2.204285714285714,
+      "grad_norm": 0.01072627492249012,
+      "learning_rate": 5.305714285714286e-06,
+      "loss": 0.0,
+      "step": 15430
+    },
+    {
+      "epoch": 2.2057142857142855,
+      "grad_norm": 0.011737393215298653,
+      "learning_rate": 5.296190476190477e-06,
+      "loss": 0.9739,
+      "step": 15440
+    },
+    {
+      "epoch": 2.2071428571428573,
+      "grad_norm": 8973664.0,
+      "learning_rate": 5.286666666666666e-06,
+      "loss": 0.7889,
+      "step": 15450
+    },
+    {
+      "epoch": 2.2085714285714286,
+      "grad_norm": 533.453857421875,
+      "learning_rate": 5.277142857142857e-06,
+      "loss": 0.9533,
+      "step": 15460
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.009765361435711384,
+      "learning_rate": 5.2676190476190485e-06,
+      "loss": 0.0,
+      "step": 15470
+    },
+    {
+      "epoch": 2.2114285714285713,
+      "grad_norm": 0.016572890803217888,
+      "learning_rate": 5.258095238095239e-06,
+      "loss": 1.9522,
+      "step": 15480
+    },
+    {
+      "epoch": 2.2128571428571426,
+      "grad_norm": 0.014318190515041351,
+      "learning_rate": 5.24857142857143e-06,
+      "loss": 0.8385,
+      "step": 15490
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 0.026814959943294525,
+      "learning_rate": 5.2390476190476195e-06,
+      "loss": 0.0,
+      "step": 15500
+    },
+    {
+      "epoch": 2.2157142857142857,
+      "grad_norm": 0.021901581436395645,
+      "learning_rate": 5.22952380952381e-06,
+      "loss": 0.0,
+      "step": 15510
+    },
+    {
+      "epoch": 2.217142857142857,
+      "grad_norm": 0.02863289602100849,
+      "learning_rate": 5.220000000000001e-06,
+      "loss": 0.0,
+      "step": 15520
+    },
+    {
+      "epoch": 2.2185714285714284,
+      "grad_norm": 0.013008411973714828,
+      "learning_rate": 5.2104761904761915e-06,
+      "loss": 0.0,
+      "step": 15530
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.018024584278464317,
+      "learning_rate": 5.200952380952381e-06,
+      "loss": 0.0,
+      "step": 15540
+    },
+    {
+      "epoch": 2.2214285714285715,
+      "grad_norm": 0.01259527262300253,
+      "learning_rate": 5.191428571428572e-06,
+      "loss": 0.0,
+      "step": 15550
+    },
+    {
+      "epoch": 2.222857142857143,
+      "grad_norm": 0.026787765324115753,
+      "learning_rate": 5.181904761904763e-06,
+      "loss": 0.0,
+      "step": 15560
+    },
+    {
+      "epoch": 2.224285714285714,
+      "grad_norm": 0.016977623105049133,
+      "learning_rate": 5.172380952380953e-06,
+      "loss": 0.0,
+      "step": 15570
+    },
+    {
+      "epoch": 2.2257142857142855,
+      "grad_norm": 0.015745533630251884,
+      "learning_rate": 5.162857142857143e-06,
+      "loss": 0.489,
+      "step": 15580
+    },
+    {
+      "epoch": 2.2271428571428573,
+      "grad_norm": 0.013185848481953144,
+      "learning_rate": 5.153333333333334e-06,
+      "loss": 0.0,
+      "step": 15590
+    },
+    {
+      "epoch": 2.2285714285714286,
+      "grad_norm": 0.016491331160068512,
+      "learning_rate": 5.143809523809524e-06,
+      "loss": 0.9684,
+      "step": 15600
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.015150822699069977,
+      "learning_rate": 5.134285714285715e-06,
+      "loss": 0.0,
+      "step": 15610
+    },
+    {
+      "epoch": 2.2314285714285713,
+      "grad_norm": 0.005874530877918005,
+      "learning_rate": 5.124761904761906e-06,
+      "loss": 0.0,
+      "step": 15620
+    },
+    {
+      "epoch": 2.2328571428571427,
+      "grad_norm": 0.004922645632177591,
+      "learning_rate": 5.1152380952380954e-06,
+      "loss": 0.0,
+      "step": 15630
+    },
+    {
+      "epoch": 2.2342857142857144,
+      "grad_norm": 0.008776701055467129,
+      "learning_rate": 5.105714285714286e-06,
+      "loss": 0.9014,
+      "step": 15640
+    },
+    {
+      "epoch": 2.2357142857142858,
+      "grad_norm": 0.008524585515260696,
+      "learning_rate": 5.096190476190477e-06,
+      "loss": 0.0,
+      "step": 15650
+    },
+    {
+      "epoch": 2.237142857142857,
+      "grad_norm": 0.030151639133691788,
+      "learning_rate": 5.086666666666667e-06,
+      "loss": 0.0,
+      "step": 15660
+    },
+    {
+      "epoch": 2.2385714285714284,
+      "grad_norm": 0.2200971394777298,
+      "learning_rate": 5.077142857142857e-06,
+      "loss": 0.0,
+      "step": 15670
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.014291862025856972,
+      "learning_rate": 5.067619047619048e-06,
+      "loss": 0.0,
+      "step": 15680
+    },
+    {
+      "epoch": 2.2414285714285715,
+      "grad_norm": 0.020342759788036346,
+      "learning_rate": 5.0580952380952385e-06,
+      "loss": 0.0,
+      "step": 15690
+    },
+    {
+      "epoch": 2.242857142857143,
+      "grad_norm": 0.015303666703402996,
+      "learning_rate": 5.048571428571429e-06,
+      "loss": 0.0,
+      "step": 15700
+    },
+    {
+      "epoch": 2.244285714285714,
+      "grad_norm": 0.007998144254088402,
+      "learning_rate": 5.039047619047619e-06,
+      "loss": 0.0,
+      "step": 15710
+    },
+    {
+      "epoch": 2.2457142857142856,
+      "grad_norm": 0.008900218643248081,
+      "learning_rate": 5.0295238095238096e-06,
+      "loss": 0.9481,
+      "step": 15720
+    },
+    {
+      "epoch": 2.2471428571428573,
+      "grad_norm": 0.01849919743835926,
+      "learning_rate": 5.02e-06,
+      "loss": 0.0,
+      "step": 15730
+    },
+    {
+      "epoch": 2.2485714285714287,
+      "grad_norm": 0.008505268022418022,
+      "learning_rate": 5.010476190476191e-06,
+      "loss": 1.0152,
+      "step": 15740
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3480314.0,
+      "learning_rate": 5.0009523809523815e-06,
+      "loss": 0.9936,
+      "step": 15750
+    },
+    {
+      "epoch": 2.2514285714285713,
+      "grad_norm": 0.014624119736254215,
+      "learning_rate": 4.991428571428572e-06,
+      "loss": 0.0,
+      "step": 15760
+    },
+    {
+      "epoch": 2.2528571428571427,
+      "grad_norm": 0.025296537205576897,
+      "learning_rate": 4.981904761904762e-06,
+      "loss": 0.0,
+      "step": 15770
+    },
+    {
+      "epoch": 2.2542857142857144,
+      "grad_norm": 0.009388687089085579,
+      "learning_rate": 4.972380952380953e-06,
+      "loss": 2.7277,
+      "step": 15780
+    },
+    {
+      "epoch": 2.255714285714286,
+      "grad_norm": 0.006550527177751064,
+      "learning_rate": 4.962857142857143e-06,
+      "loss": 0.0,
+      "step": 15790
+    },
+    {
+      "epoch": 2.257142857142857,
+      "grad_norm": 7358300.0,
+      "learning_rate": 4.953333333333334e-06,
+      "loss": 0.8764,
+      "step": 15800
+    },
+    {
+      "epoch": 2.2585714285714285,
+      "grad_norm": 0.023875588551163673,
+      "learning_rate": 4.943809523809524e-06,
+      "loss": 0.0,
+      "step": 15810
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.015785882249474525,
+      "learning_rate": 4.934285714285715e-06,
+      "loss": 0.8714,
+      "step": 15820
+    },
+    {
+      "epoch": 2.2614285714285716,
+      "grad_norm": 0.01754889264702797,
+      "learning_rate": 4.924761904761905e-06,
+      "loss": 0.0,
+      "step": 15830
+    },
+    {
+      "epoch": 2.262857142857143,
+      "grad_norm": 0.02094806171953678,
+      "learning_rate": 4.915238095238096e-06,
+      "loss": 0.0,
+      "step": 15840
+    },
+    {
+      "epoch": 2.2642857142857142,
+      "grad_norm": 0.0144230080768466,
+      "learning_rate": 4.9057142857142855e-06,
+      "loss": 0.0,
+      "step": 15850
+    },
+    {
+      "epoch": 2.2657142857142856,
+      "grad_norm": 0.014674977399408817,
+      "learning_rate": 4.896190476190477e-06,
+      "loss": 0.0,
+      "step": 15860
+    },
+    {
+      "epoch": 2.267142857142857,
+      "grad_norm": 0.013628990389406681,
+      "learning_rate": 4.886666666666668e-06,
+      "loss": 0.0,
+      "step": 15870
+    },
+    {
+      "epoch": 2.2685714285714287,
+      "grad_norm": 0.00806028489023447,
+      "learning_rate": 4.877142857142857e-06,
+      "loss": 0.0,
+      "step": 15880
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.012893387116491795,
+      "learning_rate": 4.867619047619048e-06,
+      "loss": 0.9631,
+      "step": 15890
+    },
+    {
+      "epoch": 2.2714285714285714,
+      "grad_norm": 0.03413955867290497,
+      "learning_rate": 4.858095238095239e-06,
+      "loss": 0.0,
+      "step": 15900
+    },
+    {
+      "epoch": 2.2728571428571427,
+      "grad_norm": 0.03138644993305206,
+      "learning_rate": 4.848571428571429e-06,
+      "loss": 0.0,
+      "step": 15910
+    },
+    {
+      "epoch": 2.2742857142857145,
+      "grad_norm": 0.03908616304397583,
+      "learning_rate": 4.839047619047619e-06,
+      "loss": 0.9806,
+      "step": 15920
+    },
+    {
+      "epoch": 2.275714285714286,
+      "grad_norm": 0.005133791361004114,
+      "learning_rate": 4.82952380952381e-06,
+      "loss": 0.7202,
+      "step": 15930
+    },
+    {
+      "epoch": 2.277142857142857,
+      "grad_norm": 0.0479804091155529,
+      "learning_rate": 4.8200000000000004e-06,
+      "loss": 0.9881,
+      "step": 15940
+    },
+    {
+      "epoch": 2.2785714285714285,
+      "grad_norm": 0.010117598809301853,
+      "learning_rate": 4.810476190476191e-06,
+      "loss": 0.0,
+      "step": 15950
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 0.009415602311491966,
+      "learning_rate": 4.800952380952381e-06,
+      "loss": 0.0,
+      "step": 15960
+    },
+    {
+      "epoch": 2.2814285714285716,
+      "grad_norm": 0.03315386548638344,
+      "learning_rate": 4.7914285714285715e-06,
+      "loss": 0.0,
+      "step": 15970
+    },
+    {
+      "epoch": 2.282857142857143,
+      "grad_norm": 0.006616033613681793,
+      "learning_rate": 4.781904761904762e-06,
+      "loss": 0.8437,
+      "step": 15980
+    },
+    {
+      "epoch": 2.2842857142857143,
+      "grad_norm": 0.017714357003569603,
+      "learning_rate": 4.772380952380953e-06,
+      "loss": 0.0,
+      "step": 15990
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.013708224520087242,
+      "learning_rate": 4.7628571428571435e-06,
+      "loss": 0.0,
+      "step": 16000
+    },
+    {
+      "epoch": 2.2871428571428574,
+      "grad_norm": 0.027172720059752464,
+      "learning_rate": 4.753333333333333e-06,
+      "loss": 0.0,
+      "step": 16010
+    },
+    {
+      "epoch": 2.2885714285714287,
+      "grad_norm": 0.01044511515647173,
+      "learning_rate": 4.743809523809525e-06,
+      "loss": 0.7768,
+      "step": 16020
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.022179150953888893,
+      "learning_rate": 4.734285714285715e-06,
+      "loss": 0.0,
+      "step": 16030
+    },
+    {
+      "epoch": 2.2914285714285714,
+      "grad_norm": 0.021245945245027542,
+      "learning_rate": 4.724761904761905e-06,
+      "loss": 0.0,
+      "step": 16040
+    },
+    {
+      "epoch": 2.2928571428571427,
+      "grad_norm": 0.08311065286397934,
+      "learning_rate": 4.715238095238096e-06,
+      "loss": 0.0,
+      "step": 16050
+    },
+    {
+      "epoch": 2.2942857142857145,
+      "grad_norm": 0.030629372224211693,
+      "learning_rate": 4.7057142857142865e-06,
+      "loss": 0.0,
+      "step": 16060
+    },
+    {
+      "epoch": 2.295714285714286,
+      "grad_norm": 0.03277376666665077,
+      "learning_rate": 4.696190476190476e-06,
+      "loss": 0.0,
+      "step": 16070
+    },
+    {
+      "epoch": 2.297142857142857,
+      "grad_norm": 0.008143771439790726,
+      "learning_rate": 4.686666666666667e-06,
+      "loss": 0.0,
+      "step": 16080
+    },
+    {
+      "epoch": 2.2985714285714285,
+      "grad_norm": 0.009722625836730003,
+      "learning_rate": 4.677142857142858e-06,
+      "loss": 0.0,
+      "step": 16090
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.971846878528595,
+      "learning_rate": 4.667619047619048e-06,
+      "loss": 1.8733,
+      "step": 16100
+    },
+    {
+      "epoch": 2.3014285714285716,
+      "grad_norm": 0.008375752717256546,
+      "learning_rate": 4.658095238095238e-06,
+      "loss": 1.7142,
+      "step": 16110
+    },
+    {
+      "epoch": 2.302857142857143,
+      "grad_norm": 0.017525505274534225,
+      "learning_rate": 4.648571428571429e-06,
+      "loss": 0.0,
+      "step": 16120
+    },
+    {
+      "epoch": 2.3042857142857143,
+      "grad_norm": 0.004345288034528494,
+      "learning_rate": 4.639047619047619e-06,
+      "loss": 0.0,
+      "step": 16130
+    },
+    {
+      "epoch": 2.3057142857142856,
+      "grad_norm": 0.016600633040070534,
+      "learning_rate": 4.62952380952381e-06,
+      "loss": 0.0,
+      "step": 16140
+    },
+    {
+      "epoch": 2.307142857142857,
+      "grad_norm": 0.01578422449529171,
+      "learning_rate": 4.620000000000001e-06,
+      "loss": 0.0001,
+      "step": 16150
+    },
+    {
+      "epoch": 2.3085714285714287,
+      "grad_norm": 0.007023902144283056,
+      "learning_rate": 4.6104761904761905e-06,
+      "loss": 0.0,
+      "step": 16160
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.012471306137740612,
+      "learning_rate": 4.600952380952381e-06,
+      "loss": 0.8609,
+      "step": 16170
+    },
+    {
+      "epoch": 2.3114285714285714,
+      "grad_norm": 0.015368607826530933,
+      "learning_rate": 4.591428571428572e-06,
+      "loss": 0.0,
+      "step": 16180
+    },
+    {
+      "epoch": 2.3128571428571427,
+      "grad_norm": 1.7062859535217285,
+      "learning_rate": 4.581904761904762e-06,
+      "loss": 0.0,
+      "step": 16190
+    },
+    {
+      "epoch": 2.314285714285714,
+      "grad_norm": 5304800.5,
+      "learning_rate": 4.572380952380952e-06,
+      "loss": 0.9685,
+      "step": 16200
+    },
+    {
+      "epoch": 2.315714285714286,
+      "grad_norm": 12489783.0,
+      "learning_rate": 4.562857142857144e-06,
+      "loss": 0.9111,
+      "step": 16210
+    },
+    {
+      "epoch": 2.317142857142857,
+      "grad_norm": 0.012120522558689117,
+      "learning_rate": 4.5533333333333335e-06,
+      "loss": 0.0,
+      "step": 16220
+    },
+    {
+      "epoch": 2.3185714285714285,
+      "grad_norm": 0.009377406910061836,
+      "learning_rate": 4.543809523809524e-06,
+      "loss": 0.0,
+      "step": 16230
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.011410492472350597,
+      "learning_rate": 4.534285714285715e-06,
+      "loss": 0.0,
+      "step": 16240
+    },
+    {
+      "epoch": 2.3214285714285716,
+      "grad_norm": 0.008047115989029408,
+      "learning_rate": 4.5247619047619054e-06,
+      "loss": 0.0,
+      "step": 16250
+    },
+    {
+      "epoch": 2.322857142857143,
+      "grad_norm": 0.004907062277197838,
+      "learning_rate": 4.515238095238096e-06,
+      "loss": 0.0,
+      "step": 16260
+    },
+    {
+      "epoch": 2.3242857142857143,
+      "grad_norm": 0.008754486218094826,
+      "learning_rate": 4.505714285714286e-06,
+      "loss": 0.8312,
+      "step": 16270
+    },
+    {
+      "epoch": 2.3257142857142856,
+      "grad_norm": 0.006838922388851643,
+      "learning_rate": 4.4961904761904765e-06,
+      "loss": 0.0,
+      "step": 16280
+    },
+    {
+      "epoch": 2.327142857142857,
+      "grad_norm": 0.005286881700158119,
+      "learning_rate": 4.486666666666667e-06,
+      "loss": 0.9961,
+      "step": 16290
+    },
+    {
+      "epoch": 2.3285714285714287,
+      "grad_norm": 0.01619318686425686,
+      "learning_rate": 4.477142857142858e-06,
+      "loss": 0.0,
+      "step": 16300
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.015769733116030693,
+      "learning_rate": 4.467619047619048e-06,
+      "loss": 0.0,
+      "step": 16310
+    },
+    {
+      "epoch": 2.3314285714285714,
+      "grad_norm": 0.005701390095055103,
+      "learning_rate": 4.458095238095238e-06,
+      "loss": 0.0,
+      "step": 16320
+    },
+    {
+      "epoch": 2.3328571428571427,
+      "grad_norm": 0.01904052309691906,
+      "learning_rate": 4.448571428571429e-06,
+      "loss": 0.0,
+      "step": 16330
+    },
+    {
+      "epoch": 2.3342857142857145,
+      "grad_norm": 0.01466944720596075,
+      "learning_rate": 4.43904761904762e-06,
+      "loss": 0.0,
+      "step": 16340
+    },
+    {
+      "epoch": 2.335714285714286,
+      "grad_norm": 0.0068887341767549515,
+      "learning_rate": 4.429523809523809e-06,
+      "loss": 0.938,
+      "step": 16350
+    },
+    {
+      "epoch": 2.337142857142857,
+      "grad_norm": 0.045321132987737656,
+      "learning_rate": 4.42e-06,
+      "loss": 0.0,
+      "step": 16360
+    },
+    {
+      "epoch": 2.3385714285714285,
+      "grad_norm": 0.008037515915930271,
+      "learning_rate": 4.410476190476191e-06,
+      "loss": 0.956,
+      "step": 16370
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.007663840893656015,
+      "learning_rate": 4.400952380952381e-06,
+      "loss": 0.0,
+      "step": 16380
+    },
+    {
+      "epoch": 2.3414285714285716,
+      "grad_norm": 0.005593115463852882,
+      "learning_rate": 4.391428571428572e-06,
+      "loss": 0.0,
+      "step": 16390
+    },
+    {
+      "epoch": 2.342857142857143,
+      "grad_norm": 0.018713518977165222,
+      "learning_rate": 4.381904761904763e-06,
+      "loss": 0.9499,
+      "step": 16400
+    },
+    {
+      "epoch": 2.3442857142857143,
+      "grad_norm": 0.013416334986686707,
+      "learning_rate": 4.372380952380953e-06,
+      "loss": 0.0,
+      "step": 16410
+    },
+    {
+      "epoch": 2.3457142857142856,
+      "grad_norm": 0.029729673638939857,
+      "learning_rate": 4.362857142857143e-06,
+      "loss": 0.0,
+      "step": 16420
+    },
+    {
+      "epoch": 2.347142857142857,
+      "grad_norm": 0.007621904369443655,
+      "learning_rate": 4.353333333333334e-06,
+      "loss": 0.0,
+      "step": 16430
+    },
+    {
+      "epoch": 2.3485714285714288,
+      "grad_norm": 0.023887479677796364,
+      "learning_rate": 4.343809523809524e-06,
+      "loss": 0.0,
+      "step": 16440
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.011513687670230865,
+      "learning_rate": 4.334285714285715e-06,
+      "loss": 0.8538,
+      "step": 16450
+    },
+    {
+      "epoch": 2.3514285714285714,
+      "grad_norm": 0.01611219346523285,
+      "learning_rate": 4.324761904761905e-06,
+      "loss": 0.0,
+      "step": 16460
+    },
+    {
+      "epoch": 2.3528571428571428,
+      "grad_norm": 0.015148641541600227,
+      "learning_rate": 4.3152380952380955e-06,
+      "loss": 0.0,
+      "step": 16470
+    },
+    {
+      "epoch": 2.354285714285714,
+      "grad_norm": 0.02332952246069908,
+      "learning_rate": 4.305714285714286e-06,
+      "loss": 0.0,
+      "step": 16480
+    },
+    {
+      "epoch": 2.355714285714286,
+      "grad_norm": 0.010906867682933807,
+      "learning_rate": 4.296190476190477e-06,
+      "loss": 0.0,
+      "step": 16490
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.008823190815746784,
+      "learning_rate": 4.2866666666666666e-06,
+      "loss": 0.0,
+      "step": 16500
+    },
+    {
+      "epoch": 2.3585714285714285,
+      "grad_norm": 0.005151278804987669,
+      "learning_rate": 4.277142857142857e-06,
+      "loss": 0.0,
+      "step": 16510
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.11744546890258789,
+      "learning_rate": 4.267619047619048e-06,
+      "loss": 0.0,
+      "step": 16520
+    },
+    {
+      "epoch": 2.361428571428571,
+      "grad_norm": 0.013963793404400349,
+      "learning_rate": 4.2580952380952385e-06,
+      "loss": 0.0,
+      "step": 16530
+    },
+    {
+      "epoch": 2.362857142857143,
+      "grad_norm": 0.007511726580560207,
+      "learning_rate": 4.248571428571429e-06,
+      "loss": 0.0,
+      "step": 16540
+    },
+    {
+      "epoch": 2.3642857142857143,
+      "grad_norm": 0.012398282997310162,
+      "learning_rate": 4.239047619047619e-06,
+      "loss": 0.0,
+      "step": 16550
+    },
+    {
+      "epoch": 2.3657142857142857,
+      "grad_norm": 0.03365852311253548,
+      "learning_rate": 4.2295238095238105e-06,
+      "loss": 1.0187,
+      "step": 16560
+    },
+    {
+      "epoch": 2.367142857142857,
+      "grad_norm": 0.00816162209957838,
+      "learning_rate": 4.22e-06,
+      "loss": 0.0,
+      "step": 16570
+    },
+    {
+      "epoch": 2.3685714285714283,
+      "grad_norm": 0.0244681965559721,
+      "learning_rate": 4.210476190476191e-06,
+      "loss": 0.0,
+      "step": 16580
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.003248831257224083,
+      "learning_rate": 4.200952380952381e-06,
+      "loss": 0.0,
+      "step": 16590
+    },
+    {
+      "epoch": 2.3714285714285714,
+      "grad_norm": 0.008639032021164894,
+      "learning_rate": 4.191428571428572e-06,
+      "loss": 0.8344,
+      "step": 16600
+    },
+    {
+      "epoch": 2.3728571428571428,
+      "grad_norm": 0.0164265688508749,
+      "learning_rate": 4.181904761904762e-06,
+      "loss": 0.0,
+      "step": 16610
+    },
+    {
+      "epoch": 2.374285714285714,
+      "grad_norm": 0.04277282580733299,
+      "learning_rate": 4.172380952380953e-06,
+      "loss": 1.6897,
+      "step": 16620
+    },
+    {
+      "epoch": 2.375714285714286,
+      "grad_norm": 0.017168715596199036,
+      "learning_rate": 4.162857142857143e-06,
+      "loss": 0.9594,
+      "step": 16630
+    },
+    {
+      "epoch": 2.3771428571428572,
+      "grad_norm": 0.008631149306893349,
+      "learning_rate": 4.153333333333334e-06,
+      "loss": 0.0,
+      "step": 16640
+    },
+    {
+      "epoch": 2.3785714285714286,
+      "grad_norm": 0.026826689019799232,
+      "learning_rate": 4.143809523809525e-06,
+      "loss": 0.0,
+      "step": 16650
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0017094006761908531,
+      "learning_rate": 4.134285714285714e-06,
+      "loss": 0.0,
+      "step": 16660
+    },
+    {
+      "epoch": 2.3814285714285717,
+      "grad_norm": 0.013639775104820728,
+      "learning_rate": 4.124761904761905e-06,
+      "loss": 0.0,
+      "step": 16670
+    },
+    {
+      "epoch": 2.382857142857143,
+      "grad_norm": 0.31601253151893616,
+      "learning_rate": 4.115238095238096e-06,
+      "loss": 0.0,
+      "step": 16680
+    },
+    {
+      "epoch": 2.3842857142857143,
+      "grad_norm": 0.011272274889051914,
+      "learning_rate": 4.105714285714286e-06,
+      "loss": 0.8166,
+      "step": 16690
+    },
+    {
+      "epoch": 2.3857142857142857,
+      "grad_norm": 0.041645489633083344,
+      "learning_rate": 4.096190476190476e-06,
+      "loss": 0.0,
+      "step": 16700
+    },
+    {
+      "epoch": 2.387142857142857,
+      "grad_norm": 0.008316690102219582,
+      "learning_rate": 4.086666666666667e-06,
+      "loss": 0.0,
+      "step": 16710
+    },
+    {
+      "epoch": 2.388571428571429,
+      "grad_norm": 3842169.5,
+      "learning_rate": 4.0771428571428574e-06,
+      "loss": 1.9159,
+      "step": 16720
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.011303669773042202,
+      "learning_rate": 4.067619047619048e-06,
+      "loss": 0.0,
+      "step": 16730
+    },
+    {
+      "epoch": 2.3914285714285715,
+      "grad_norm": 0.025521820411086082,
+      "learning_rate": 4.058095238095238e-06,
+      "loss": 0.0,
+      "step": 16740
+    },
+    {
+      "epoch": 2.392857142857143,
+      "grad_norm": 0.0064083244651556015,
+      "learning_rate": 4.0485714285714285e-06,
+      "loss": 0.9237,
+      "step": 16750
+    },
+    {
+      "epoch": 2.394285714285714,
+      "grad_norm": 0.607117772102356,
+      "learning_rate": 4.039047619047619e-06,
+      "loss": 0.0,
+      "step": 16760
+    },
+    {
+      "epoch": 2.395714285714286,
+      "grad_norm": 0.008776474744081497,
+      "learning_rate": 4.02952380952381e-06,
+      "loss": 0.0,
+      "step": 16770
+    },
+    {
+      "epoch": 2.3971428571428572,
+      "grad_norm": 0.010984569787979126,
+      "learning_rate": 4.0200000000000005e-06,
+      "loss": 0.8561,
+      "step": 16780
+    },
+    {
+      "epoch": 2.3985714285714286,
+      "grad_norm": 0.00283090234734118,
+      "learning_rate": 4.010476190476191e-06,
+      "loss": 0.0,
+      "step": 16790
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.026738466694951057,
+      "learning_rate": 4.000952380952382e-06,
+      "loss": 0.0,
+      "step": 16800
+    },
+    {
+      "epoch": 2.4014285714285712,
+      "grad_norm": 0.010282258503139019,
+      "learning_rate": 3.9914285714285716e-06,
+      "loss": 0.0,
+      "step": 16810
+    },
+    {
+      "epoch": 2.402857142857143,
+      "grad_norm": 0.003372363978996873,
+      "learning_rate": 3.981904761904762e-06,
+      "loss": 0.7963,
+      "step": 16820
+    },
+    {
+      "epoch": 2.4042857142857144,
+      "grad_norm": 0.00782694946974516,
+      "learning_rate": 3.972380952380953e-06,
+      "loss": 0.9732,
+      "step": 16830
+    },
+    {
+      "epoch": 2.4057142857142857,
+      "grad_norm": 0.02235283888876438,
+      "learning_rate": 3.9628571428571435e-06,
+      "loss": 0.0,
+      "step": 16840
+    },
+    {
+      "epoch": 2.407142857142857,
+      "grad_norm": 0.02207907848060131,
+      "learning_rate": 3.953333333333333e-06,
+      "loss": 0.0,
+      "step": 16850
+    },
+    {
+      "epoch": 2.4085714285714284,
+      "grad_norm": 0.018025755882263184,
+      "learning_rate": 3.943809523809524e-06,
+      "loss": 0.9364,
+      "step": 16860
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.002154488814994693,
+      "learning_rate": 3.934285714285715e-06,
+      "loss": 0.0,
+      "step": 16870
+    },
+    {
+      "epoch": 2.4114285714285715,
+      "grad_norm": 5516866.5,
+      "learning_rate": 3.924761904761905e-06,
+      "loss": 3.0092,
+      "step": 16880
+    },
+    {
+      "epoch": 2.412857142857143,
+      "grad_norm": 0.012341703288257122,
+      "learning_rate": 3.915238095238095e-06,
+      "loss": 0.0,
+      "step": 16890
+    },
+    {
+      "epoch": 2.414285714285714,
+      "grad_norm": 0.018630096688866615,
+      "learning_rate": 3.905714285714286e-06,
+      "loss": 0.0,
+      "step": 16900
+    },
+    {
+      "epoch": 2.4157142857142855,
+      "grad_norm": 0.010301096364855766,
+      "learning_rate": 3.896190476190476e-06,
+      "loss": 0.965,
+      "step": 16910
+    },
+    {
+      "epoch": 2.4171428571428573,
+      "grad_norm": 0.006753115449100733,
+      "learning_rate": 3.886666666666667e-06,
+      "loss": 0.0,
+      "step": 16920
+    },
+    {
+      "epoch": 2.4185714285714286,
+      "grad_norm": 0.024608319625258446,
+      "learning_rate": 3.877142857142858e-06,
+      "loss": 0.953,
+      "step": 16930
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.007527059409767389,
+      "learning_rate": 3.8676190476190475e-06,
+      "loss": 1.7447,
+      "step": 16940
+    },
+    {
+      "epoch": 2.4214285714285713,
+      "grad_norm": 0.017536232247948647,
+      "learning_rate": 3.858095238095239e-06,
+      "loss": 0.0004,
+      "step": 16950
+    },
+    {
+      "epoch": 2.422857142857143,
+      "grad_norm": 0.02406061254441738,
+      "learning_rate": 3.848571428571429e-06,
+      "loss": 0.0,
+      "step": 16960
+    },
+    {
+      "epoch": 2.4242857142857144,
+      "grad_norm": 0.042816922068595886,
+      "learning_rate": 3.839047619047619e-06,
+      "loss": 1.8194,
+      "step": 16970
+    },
+    {
+      "epoch": 2.4257142857142857,
+      "grad_norm": 0.025658391416072845,
+      "learning_rate": 3.82952380952381e-06,
+      "loss": 0.9241,
+      "step": 16980
+    },
+    {
+      "epoch": 2.427142857142857,
+      "grad_norm": 3.767086982727051,
+      "learning_rate": 3.820000000000001e-06,
+      "loss": 0.8909,
+      "step": 16990
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.03295642137527466,
+      "learning_rate": 3.8104761904761905e-06,
+      "loss": 0.0,
+      "step": 17000
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.010507110506296158,
+      "learning_rate": 3.800952380952381e-06,
+      "loss": 0.0,
+      "step": 17010
+    },
+    {
+      "epoch": 2.4314285714285715,
+      "grad_norm": 0.023496774956583977,
+      "learning_rate": 3.7914285714285722e-06,
+      "loss": 0.0,
+      "step": 17020
+    },
+    {
+      "epoch": 2.432857142857143,
+      "grad_norm": 0.014495810493826866,
+      "learning_rate": 3.781904761904762e-06,
+      "loss": 0.0,
+      "step": 17030
+    },
+    {
+      "epoch": 2.434285714285714,
+      "grad_norm": 0.020109763368964195,
+      "learning_rate": 3.772380952380953e-06,
+      "loss": 0.6797,
+      "step": 17040
+    },
+    {
+      "epoch": 2.435714285714286,
+      "grad_norm": 0.016561275348067284,
+      "learning_rate": 3.762857142857143e-06,
+      "loss": 0.0,
+      "step": 17050
+    },
+    {
+      "epoch": 2.4371428571428573,
+      "grad_norm": 0.007582352962344885,
+      "learning_rate": 3.753333333333334e-06,
+      "loss": 0.8447,
+      "step": 17060
+    },
+    {
+      "epoch": 2.4385714285714286,
+      "grad_norm": 0.04775860533118248,
+      "learning_rate": 3.743809523809524e-06,
+      "loss": 0.0,
+      "step": 17070
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.008977421559393406,
+      "learning_rate": 3.734285714285715e-06,
+      "loss": 0.0,
+      "step": 17080
+    },
+    {
+      "epoch": 2.4414285714285713,
+      "grad_norm": 0.008807310834527016,
+      "learning_rate": 3.724761904761905e-06,
+      "loss": 0.0,
+      "step": 17090
+    },
+    {
+      "epoch": 2.442857142857143,
+      "grad_norm": 0.028799952939152718,
+      "learning_rate": 3.7152380952380957e-06,
+      "loss": 0.0,
+      "step": 17100
+    },
+    {
+      "epoch": 2.4442857142857144,
+      "grad_norm": 0.003955888096243143,
+      "learning_rate": 3.705714285714286e-06,
+      "loss": 0.0,
+      "step": 17110
+    },
+    {
+      "epoch": 2.4457142857142857,
+      "grad_norm": 24559338.0,
+      "learning_rate": 3.6961904761904766e-06,
+      "loss": 0.909,
+      "step": 17120
+    },
+    {
+      "epoch": 2.447142857142857,
+      "grad_norm": 0.01052605640143156,
+      "learning_rate": 3.686666666666667e-06,
+      "loss": 0.0,
+      "step": 17130
+    },
+    {
+      "epoch": 2.4485714285714284,
+      "grad_norm": 0.005958448629826307,
+      "learning_rate": 3.6771428571428575e-06,
+      "loss": 0.0,
+      "step": 17140
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0114182373508811,
+      "learning_rate": 3.667619047619048e-06,
+      "loss": 1.0662,
+      "step": 17150
+    },
+    {
+      "epoch": 2.4514285714285715,
+      "grad_norm": 0.01767958514392376,
+      "learning_rate": 3.6580952380952383e-06,
+      "loss": 0.0,
+      "step": 17160
+    },
+    {
+      "epoch": 2.452857142857143,
+      "grad_norm": 0.033098526298999786,
+      "learning_rate": 3.648571428571429e-06,
+      "loss": 0.0,
+      "step": 17170
+    },
+    {
+      "epoch": 2.454285714285714,
+      "grad_norm": 0.0327240414917469,
+      "learning_rate": 3.639047619047619e-06,
+      "loss": 0.0,
+      "step": 17180
+    },
+    {
+      "epoch": 2.4557142857142855,
+      "grad_norm": 4357304.5,
+      "learning_rate": 3.62952380952381e-06,
+      "loss": 0.9249,
+      "step": 17190
+    },
+    {
+      "epoch": 2.4571428571428573,
+      "grad_norm": 0.025806942954659462,
+      "learning_rate": 3.62e-06,
+      "loss": 0.9452,
+      "step": 17200
+    },
+    {
+      "epoch": 2.4585714285714286,
+      "grad_norm": 0.009442337788641453,
+      "learning_rate": 3.6104761904761907e-06,
+      "loss": 0.9819,
+      "step": 17210
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.00507747707888484,
+      "learning_rate": 3.600952380952381e-06,
+      "loss": 0.9922,
+      "step": 17220
+    },
+    {
+      "epoch": 2.4614285714285713,
+      "grad_norm": 0.05158117413520813,
+      "learning_rate": 3.591428571428572e-06,
+      "loss": 0.8608,
+      "step": 17230
+    },
+    {
+      "epoch": 2.4628571428571426,
+      "grad_norm": 0.04027364030480385,
+      "learning_rate": 3.581904761904762e-06,
+      "loss": 0.7674,
+      "step": 17240
+    },
+    {
+      "epoch": 2.4642857142857144,
+      "grad_norm": 0.016667887568473816,
+      "learning_rate": 3.572380952380953e-06,
+      "loss": 0.0,
+      "step": 17250
+    },
+    {
+      "epoch": 2.4657142857142857,
+      "grad_norm": 0.008836090564727783,
+      "learning_rate": 3.562857142857143e-06,
+      "loss": 0.5208,
+      "step": 17260
+    },
+    {
+      "epoch": 2.467142857142857,
+      "grad_norm": 0.010174066759645939,
+      "learning_rate": 3.5533333333333338e-06,
+      "loss": 0.0,
+      "step": 17270
+    },
+    {
+      "epoch": 2.4685714285714284,
+      "grad_norm": 0.016187459230422974,
+      "learning_rate": 3.5438095238095244e-06,
+      "loss": 0.0,
+      "step": 17280
+    },
+    {
+      "epoch": 2.4699999999999998,
+      "grad_norm": 0.012634752318263054,
+      "learning_rate": 3.5342857142857146e-06,
+      "loss": 0.0,
+      "step": 17290
+    },
+    {
+      "epoch": 2.4714285714285715,
+      "grad_norm": 0.013091981410980225,
+      "learning_rate": 3.5247619047619053e-06,
+      "loss": 0.0,
+      "step": 17300
+    },
+    {
+      "epoch": 2.472857142857143,
+      "grad_norm": 0.0048146555200219154,
+      "learning_rate": 3.5152380952380955e-06,
+      "loss": 0.0,
+      "step": 17310
+    },
+    {
+      "epoch": 2.474285714285714,
+      "grad_norm": 10739664.0,
+      "learning_rate": 3.505714285714286e-06,
+      "loss": 0.8698,
+      "step": 17320
+    },
+    {
+      "epoch": 2.4757142857142855,
+      "grad_norm": 0.06704136729240417,
+      "learning_rate": 3.4961904761904764e-06,
+      "loss": 0.0,
+      "step": 17330
+    },
+    {
+      "epoch": 2.4771428571428573,
+      "grad_norm": 0.015293789096176624,
+      "learning_rate": 3.486666666666667e-06,
+      "loss": 0.0,
+      "step": 17340
+    },
+    {
+      "epoch": 2.4785714285714286,
+      "grad_norm": 0.005765389651060104,
+      "learning_rate": 3.4771428571428572e-06,
+      "loss": 0.9065,
+      "step": 17350
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.01050998643040657,
+      "learning_rate": 3.467619047619048e-06,
+      "loss": 0.0,
+      "step": 17360
+    },
+    {
+      "epoch": 2.4814285714285713,
+      "grad_norm": 0.011139916256070137,
+      "learning_rate": 3.458095238095238e-06,
+      "loss": 0.0,
+      "step": 17370
+    },
+    {
+      "epoch": 2.482857142857143,
+      "grad_norm": 0.03285232558846474,
+      "learning_rate": 3.4485714285714288e-06,
+      "loss": 0.862,
+      "step": 17380
+    },
+    {
+      "epoch": 2.4842857142857144,
+      "grad_norm": 0.0048422315157949924,
+      "learning_rate": 3.439047619047619e-06,
+      "loss": 0.0,
+      "step": 17390
+    },
+    {
+      "epoch": 2.4857142857142858,
+      "grad_norm": 0.032139815390110016,
+      "learning_rate": 3.4295238095238096e-06,
+      "loss": 0.0,
+      "step": 17400
+    },
+    {
+      "epoch": 2.487142857142857,
+      "grad_norm": 0.024096768349409103,
+      "learning_rate": 3.4200000000000007e-06,
+      "loss": 0.0,
+      "step": 17410
+    },
+    {
+      "epoch": 2.4885714285714284,
+      "grad_norm": 0.017785528674721718,
+      "learning_rate": 3.4104761904761905e-06,
+      "loss": 1.0059,
+      "step": 17420
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.028359048068523407,
+      "learning_rate": 3.4009523809523816e-06,
+      "loss": 1.5963,
+      "step": 17430
+    },
+    {
+      "epoch": 2.4914285714285715,
+      "grad_norm": 0.013235215097665787,
+      "learning_rate": 3.391428571428572e-06,
+      "loss": 0.0,
+      "step": 17440
+    },
+    {
+      "epoch": 2.492857142857143,
+      "grad_norm": 0.010767869651317596,
+      "learning_rate": 3.3819047619047625e-06,
+      "loss": 0.0,
+      "step": 17450
+    },
+    {
+      "epoch": 2.494285714285714,
+      "grad_norm": 0.04704238474369049,
+      "learning_rate": 3.3723809523809527e-06,
+      "loss": 0.0,
+      "step": 17460
+    },
+    {
+      "epoch": 2.4957142857142856,
+      "grad_norm": 0.014711592346429825,
+      "learning_rate": 3.3628571428571433e-06,
+      "loss": 0.0,
+      "step": 17470
+    },
+    {
+      "epoch": 2.4971428571428573,
+      "grad_norm": 0.002586189191788435,
+      "learning_rate": 3.3533333333333336e-06,
+      "loss": 0.0,
+      "step": 17480
+    },
+    {
+      "epoch": 2.4985714285714287,
+      "grad_norm": 0.2102186232805252,
+      "learning_rate": 3.343809523809524e-06,
+      "loss": 1.7684,
+      "step": 17490
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.043125588446855545,
+      "learning_rate": 3.3342857142857144e-06,
+      "loss": 0.0,
+      "step": 17500
+    },
+    {
+      "epoch": 2.5014285714285713,
+      "grad_norm": 0.02120296098291874,
+      "learning_rate": 3.324761904761905e-06,
+      "loss": 1.8074,
+      "step": 17510
+    },
+    {
+      "epoch": 2.5028571428571427,
+      "grad_norm": 0.010577156208455563,
+      "learning_rate": 3.3152380952380953e-06,
+      "loss": 0.977,
+      "step": 17520
+    },
+    {
+      "epoch": 2.5042857142857144,
+      "grad_norm": 0.026182152330875397,
+      "learning_rate": 3.305714285714286e-06,
+      "loss": 0.9179,
+      "step": 17530
+    },
+    {
+      "epoch": 2.505714285714286,
+      "grad_norm": 0.02287248708307743,
+      "learning_rate": 3.2961904761904766e-06,
+      "loss": 0.0,
+      "step": 17540
+    },
+    {
+      "epoch": 2.507142857142857,
+      "grad_norm": 0.018622037023305893,
+      "learning_rate": 3.286666666666667e-06,
+      "loss": 0.0,
+      "step": 17550
+    },
+    {
+      "epoch": 2.5085714285714285,
+      "grad_norm": 0.02060418203473091,
+      "learning_rate": 3.2771428571428575e-06,
+      "loss": 0.0,
+      "step": 17560
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.014186565764248371,
+      "learning_rate": 3.2676190476190477e-06,
+      "loss": 0.0,
+      "step": 17570
+    },
+    {
+      "epoch": 2.5114285714285716,
+      "grad_norm": 0.009235070087015629,
+      "learning_rate": 3.2580952380952383e-06,
+      "loss": 0.9135,
+      "step": 17580
+    },
+    {
+      "epoch": 2.512857142857143,
+      "grad_norm": 0.013410610146820545,
+      "learning_rate": 3.2485714285714286e-06,
+      "loss": 0.0,
+      "step": 17590
+    },
+    {
+      "epoch": 2.5142857142857142,
+      "grad_norm": 0.34238624572753906,
+      "learning_rate": 3.2390476190476196e-06,
+      "loss": 0.9065,
+      "step": 17600
+    },
+    {
+      "epoch": 2.5157142857142856,
+      "grad_norm": 0.04515593498945236,
+      "learning_rate": 3.2295238095238094e-06,
+      "loss": 0.0,
+      "step": 17610
+    },
+    {
+      "epoch": 2.517142857142857,
+      "grad_norm": 0.013686695136129856,
+      "learning_rate": 3.2200000000000005e-06,
+      "loss": 0.9177,
+      "step": 17620
+    },
+    {
+      "epoch": 2.5185714285714287,
+      "grad_norm": 0.02293088473379612,
+      "learning_rate": 3.2104761904761907e-06,
+      "loss": 0.0,
+      "step": 17630
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.008502194657921791,
+      "learning_rate": 3.2009523809523814e-06,
+      "loss": 0.0,
+      "step": 17640
+    },
+    {
+      "epoch": 2.5214285714285714,
+      "grad_norm": 0.008408510126173496,
+      "learning_rate": 3.1914285714285716e-06,
+      "loss": 0.0,
+      "step": 17650
+    },
+    {
+      "epoch": 2.522857142857143,
+      "grad_norm": 0.013963821344077587,
+      "learning_rate": 3.1819047619047623e-06,
+      "loss": 0.0,
+      "step": 17660
+    },
+    {
+      "epoch": 2.524285714285714,
+      "grad_norm": 0.03430933505296707,
+      "learning_rate": 3.172380952380953e-06,
+      "loss": 0.0,
+      "step": 17670
+    },
+    {
+      "epoch": 2.525714285714286,
+      "grad_norm": 0.009724419564008713,
+      "learning_rate": 3.162857142857143e-06,
+      "loss": 0.9189,
+      "step": 17680
+    },
+    {
+      "epoch": 2.527142857142857,
+      "grad_norm": 0.011253872886300087,
+      "learning_rate": 3.1533333333333338e-06,
+      "loss": 0.0,
+      "step": 17690
+    },
+    {
+      "epoch": 2.5285714285714285,
+      "grad_norm": 0.30557364225387573,
+      "learning_rate": 3.143809523809524e-06,
+      "loss": 0.0,
+      "step": 17700
+    },
+    {
+      "epoch": 2.5300000000000002,
+      "grad_norm": 0.005390917882323265,
+      "learning_rate": 3.1342857142857146e-06,
+      "loss": 0.0,
+      "step": 17710
+    },
+    {
+      "epoch": 2.5314285714285716,
+      "grad_norm": 0.012574554421007633,
+      "learning_rate": 3.124761904761905e-06,
+      "loss": 1.9137,
+      "step": 17720
+    },
+    {
+      "epoch": 2.532857142857143,
+      "grad_norm": 0.009668853133916855,
+      "learning_rate": 3.1152380952380955e-06,
+      "loss": 0.0,
+      "step": 17730
+    },
+    {
+      "epoch": 2.5342857142857143,
+      "grad_norm": 0.009268962778151035,
+      "learning_rate": 3.1057142857142857e-06,
+      "loss": 0.0,
+      "step": 17740
+    },
+    {
+      "epoch": 2.5357142857142856,
+      "grad_norm": 0.010846162214875221,
+      "learning_rate": 3.0961904761904764e-06,
+      "loss": 0.0,
+      "step": 17750
+    },
+    {
+      "epoch": 2.5371428571428574,
+      "grad_norm": 0.016054224222898483,
+      "learning_rate": 3.0866666666666666e-06,
+      "loss": 0.9977,
+      "step": 17760
+    },
+    {
+      "epoch": 2.5385714285714287,
+      "grad_norm": 0.0046564736403524876,
+      "learning_rate": 3.0771428571428573e-06,
+      "loss": 0.0,
+      "step": 17770
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3847250.75,
+      "learning_rate": 3.0676190476190475e-06,
+      "loss": 1.0141,
+      "step": 17780
+    },
+    {
+      "epoch": 2.5414285714285714,
+      "grad_norm": 0.008961065672338009,
+      "learning_rate": 3.058095238095238e-06,
+      "loss": 0.0,
+      "step": 17790
+    },
+    {
+      "epoch": 2.5428571428571427,
+      "grad_norm": 2477979904.0,
+      "learning_rate": 3.048571428571429e-06,
+      "loss": 1.6436,
+      "step": 17800
+    },
+    {
+      "epoch": 2.5442857142857145,
+      "grad_norm": 0.012956581078469753,
+      "learning_rate": 3.0390476190476194e-06,
+      "loss": 0.0,
+      "step": 17810
+    },
+    {
+      "epoch": 2.545714285714286,
+      "grad_norm": 0.03231925144791603,
+      "learning_rate": 3.02952380952381e-06,
+      "loss": 0.0,
+      "step": 17820
+    },
+    {
+      "epoch": 2.547142857142857,
+      "grad_norm": 0.010125134140253067,
+      "learning_rate": 3.0200000000000003e-06,
+      "loss": 0.0,
+      "step": 17830
+    },
+    {
+      "epoch": 2.5485714285714285,
+      "grad_norm": 0.007222610525786877,
+      "learning_rate": 3.010476190476191e-06,
+      "loss": 0.0,
+      "step": 17840
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.018849702551960945,
+      "learning_rate": 3.000952380952381e-06,
+      "loss": 0.0,
+      "step": 17850
+    },
+    {
+      "epoch": 2.5514285714285716,
+      "grad_norm": 0.010753105394542217,
+      "learning_rate": 2.991428571428572e-06,
+      "loss": 0.0,
+      "step": 17860
+    },
+    {
+      "epoch": 2.552857142857143,
+      "grad_norm": 0.023951411247253418,
+      "learning_rate": 2.981904761904762e-06,
+      "loss": 0.0,
+      "step": 17870
+    },
+    {
+      "epoch": 2.5542857142857143,
+      "grad_norm": 0.007318285759538412,
+      "learning_rate": 2.9723809523809527e-06,
+      "loss": 2.5627,
+      "step": 17880
+    },
+    {
+      "epoch": 2.5557142857142856,
+      "grad_norm": 0.01598944701254368,
+      "learning_rate": 2.962857142857143e-06,
+      "loss": 0.0,
+      "step": 17890
+    },
+    {
+      "epoch": 2.557142857142857,
+      "grad_norm": 0.011659009382128716,
+      "learning_rate": 2.9533333333333336e-06,
+      "loss": 0.0,
+      "step": 17900
+    },
+    {
+      "epoch": 2.5585714285714287,
+      "grad_norm": 0.018028831109404564,
+      "learning_rate": 2.943809523809524e-06,
+      "loss": 0.0,
+      "step": 17910
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.01006721705198288,
+      "learning_rate": 2.9342857142857144e-06,
+      "loss": 0.0,
+      "step": 17920
+    },
+    {
+      "epoch": 2.5614285714285714,
+      "grad_norm": 0.013284722343087196,
+      "learning_rate": 2.924761904761905e-06,
+      "loss": 0.0,
+      "step": 17930
+    },
+    {
+      "epoch": 2.5628571428571427,
+      "grad_norm": 0.015733370557427406,
+      "learning_rate": 2.9152380952380953e-06,
+      "loss": 0.0,
+      "step": 17940
+    },
+    {
+      "epoch": 2.564285714285714,
+      "grad_norm": 0.04628940671682358,
+      "learning_rate": 2.905714285714286e-06,
+      "loss": 0.0,
+      "step": 17950
+    },
+    {
+      "epoch": 2.565714285714286,
+      "grad_norm": 0.027789564803242683,
+      "learning_rate": 2.896190476190476e-06,
+      "loss": 0.0,
+      "step": 17960
+    },
+    {
+      "epoch": 2.567142857142857,
+      "grad_norm": 0.01429324597120285,
+      "learning_rate": 2.8866666666666673e-06,
+      "loss": 0.0,
+      "step": 17970
+    },
+    {
+      "epoch": 2.5685714285714285,
+      "grad_norm": 0.011450983583927155,
+      "learning_rate": 2.877142857142857e-06,
+      "loss": 0.0,
+      "step": 17980
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.02064792811870575,
+      "learning_rate": 2.867619047619048e-06,
+      "loss": 0.8546,
+      "step": 17990
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.00437431363388896,
+      "learning_rate": 2.858095238095238e-06,
+      "loss": 0.0,
+      "step": 18000
+    },
+    {
+      "epoch": 2.572857142857143,
+      "grad_norm": 33941464.0,
+      "learning_rate": 2.848571428571429e-06,
+      "loss": 0.7972,
+      "step": 18010
+    },
+    {
+      "epoch": 2.5742857142857143,
+      "grad_norm": 0.0035586010199040174,
+      "learning_rate": 2.8390476190476192e-06,
+      "loss": 0.652,
+      "step": 18020
+    },
+    {
+      "epoch": 2.5757142857142856,
+      "grad_norm": 0.015361207537353039,
+      "learning_rate": 2.82952380952381e-06,
+      "loss": 0.0,
+      "step": 18030
+    },
+    {
+      "epoch": 2.5771428571428574,
+      "grad_norm": 0.005621931049972773,
+      "learning_rate": 2.82e-06,
+      "loss": 0.0,
+      "step": 18040
+    },
+    {
+      "epoch": 2.5785714285714287,
+      "grad_norm": 0.005515384487807751,
+      "learning_rate": 2.8104761904761908e-06,
+      "loss": 0.8872,
+      "step": 18050
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.011529969982802868,
+      "learning_rate": 2.8009523809523814e-06,
+      "loss": 0.8428,
+      "step": 18060
+    },
+    {
+      "epoch": 2.5814285714285714,
+      "grad_norm": 0.00392519123852253,
+      "learning_rate": 2.7914285714285716e-06,
+      "loss": 0.8227,
+      "step": 18070
+    },
+    {
+      "epoch": 2.5828571428571427,
+      "grad_norm": 11.406058311462402,
+      "learning_rate": 2.7819047619047623e-06,
+      "loss": 0.0,
+      "step": 18080
+    },
+    {
+      "epoch": 2.5842857142857145,
+      "grad_norm": 0.033514171838760376,
+      "learning_rate": 2.7723809523809525e-06,
+      "loss": 2.0261,
+      "step": 18090
+    },
+    {
+      "epoch": 2.585714285714286,
+      "grad_norm": 0.005717436783015728,
+      "learning_rate": 2.762857142857143e-06,
+      "loss": 0.8778,
+      "step": 18100
+    },
+    {
+      "epoch": 2.587142857142857,
+      "grad_norm": 0.010702522471547127,
+      "learning_rate": 2.7533333333333334e-06,
+      "loss": 0.0,
+      "step": 18110
+    },
+    {
+      "epoch": 2.5885714285714285,
+      "grad_norm": 0.011322562582790852,
+      "learning_rate": 2.743809523809524e-06,
+      "loss": 0.0,
+      "step": 18120
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.0555827617645264,
+      "learning_rate": 2.7342857142857142e-06,
+      "loss": 0.0,
+      "step": 18130
+    },
+    {
+      "epoch": 2.5914285714285716,
+      "grad_norm": 0.008201966993510723,
+      "learning_rate": 2.724761904761905e-06,
+      "loss": 0.9853,
+      "step": 18140
+    },
+    {
+      "epoch": 2.592857142857143,
+      "grad_norm": 0.014410467818379402,
+      "learning_rate": 2.715238095238095e-06,
+      "loss": 0.0,
+      "step": 18150
+    },
+    {
+      "epoch": 2.5942857142857143,
+      "grad_norm": 0.025164291262626648,
+      "learning_rate": 2.7057142857142858e-06,
+      "loss": 0.0,
+      "step": 18160
+    },
+    {
+      "epoch": 2.5957142857142856,
+      "grad_norm": 0.016926314681768417,
+      "learning_rate": 2.696190476190476e-06,
+      "loss": 0.0,
+      "step": 18170
+    },
+    {
+      "epoch": 2.597142857142857,
+      "grad_norm": 0.10771900415420532,
+      "learning_rate": 2.686666666666667e-06,
+      "loss": 0.827,
+      "step": 18180
+    },
+    {
+      "epoch": 2.5985714285714288,
+      "grad_norm": 0.025340022519230843,
+      "learning_rate": 2.6771428571428577e-06,
+      "loss": 0.8228,
+      "step": 18190
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.005198781378567219,
+      "learning_rate": 2.667619047619048e-06,
+      "loss": 0.8813,
+      "step": 18200
+    },
+    {
+      "epoch": 2.6014285714285714,
+      "grad_norm": 0.013945966958999634,
+      "learning_rate": 2.6580952380952386e-06,
+      "loss": 0.8745,
+      "step": 18210
+    },
+    {
+      "epoch": 2.6028571428571428,
+      "grad_norm": 0.014230018481612206,
+      "learning_rate": 2.648571428571429e-06,
+      "loss": 0.9436,
+      "step": 18220
+    },
+    {
+      "epoch": 2.604285714285714,
+      "grad_norm": 0.03269869089126587,
+      "learning_rate": 2.6390476190476195e-06,
+      "loss": 0.8918,
+      "step": 18230
+    },
+    {
+      "epoch": 2.605714285714286,
+      "grad_norm": 0.024383826181292534,
+      "learning_rate": 2.6295238095238097e-06,
+      "loss": 0.0,
+      "step": 18240
+    },
+    {
+      "epoch": 2.607142857142857,
+      "grad_norm": 0.020895877853035927,
+      "learning_rate": 2.6200000000000003e-06,
+      "loss": 0.0,
+      "step": 18250
+    },
+    {
+      "epoch": 2.6085714285714285,
+      "grad_norm": 0.014836377464234829,
+      "learning_rate": 2.6104761904761906e-06,
+      "loss": 1.07,
+      "step": 18260
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.02389298751950264,
+      "learning_rate": 2.600952380952381e-06,
+      "loss": 0.0,
+      "step": 18270
+    },
+    {
+      "epoch": 2.611428571428571,
+      "grad_norm": 0.006141149438917637,
+      "learning_rate": 2.5914285714285714e-06,
+      "loss": 0.0,
+      "step": 18280
+    },
+    {
+      "epoch": 2.612857142857143,
+      "grad_norm": 0.009341922588646412,
+      "learning_rate": 2.581904761904762e-06,
+      "loss": 0.0,
+      "step": 18290
+    },
+    {
+      "epoch": 2.6142857142857143,
+      "grad_norm": 0.022810202091932297,
+      "learning_rate": 2.5723809523809527e-06,
+      "loss": 0.0,
+      "step": 18300
+    },
+    {
+      "epoch": 2.6157142857142857,
+      "grad_norm": 0.015304435975849628,
+      "learning_rate": 2.562857142857143e-06,
+      "loss": 0.0,
+      "step": 18310
+    },
+    {
+      "epoch": 2.617142857142857,
+      "grad_norm": 0.014176672324538231,
+      "learning_rate": 2.5533333333333336e-06,
+      "loss": 0.9668,
+      "step": 18320
+    },
+    {
+      "epoch": 2.6185714285714283,
+      "grad_norm": 0.0393512099981308,
+      "learning_rate": 2.543809523809524e-06,
+      "loss": 0.8901,
+      "step": 18330
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.05788866803050041,
+      "learning_rate": 2.534285714285715e-06,
+      "loss": 1.5625,
+      "step": 18340
+    },
+    {
+      "epoch": 2.6214285714285714,
+      "grad_norm": 0.010419961996376514,
+      "learning_rate": 2.5247619047619047e-06,
+      "loss": 0.0,
+      "step": 18350
+    },
+    {
+      "epoch": 2.6228571428571428,
+      "grad_norm": 0.013747953809797764,
+      "learning_rate": 2.5152380952380958e-06,
+      "loss": 0.9594,
+      "step": 18360
+    },
+    {
+      "epoch": 2.6242857142857146,
+      "grad_norm": 0.014468139037489891,
+      "learning_rate": 2.5057142857142856e-06,
+      "loss": 0.0,
+      "step": 18370
+    },
+    {
+      "epoch": 2.6257142857142854,
+      "grad_norm": 0.022457975894212723,
+      "learning_rate": 2.4961904761904766e-06,
+      "loss": 0.0,
+      "step": 18380
+    },
+    {
+      "epoch": 2.6271428571428572,
+      "grad_norm": 0.008654256351292133,
+      "learning_rate": 2.486666666666667e-06,
+      "loss": 0.0,
+      "step": 18390
+    },
+    {
+      "epoch": 2.6285714285714286,
+      "grad_norm": 5251791.5,
+      "learning_rate": 2.4771428571428575e-06,
+      "loss": 0.8251,
+      "step": 18400
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.03998098522424698,
+      "learning_rate": 2.4676190476190477e-06,
+      "loss": 0.0,
+      "step": 18410
+    },
+    {
+      "epoch": 2.6314285714285717,
+      "grad_norm": 0.010598020628094673,
+      "learning_rate": 2.4580952380952384e-06,
+      "loss": 0.0,
+      "step": 18420
+    },
+    {
+      "epoch": 2.632857142857143,
+      "grad_norm": 0.023571403697133064,
+      "learning_rate": 2.4485714285714286e-06,
+      "loss": 0.0,
+      "step": 18430
+    },
+    {
+      "epoch": 2.6342857142857143,
+      "grad_norm": 0.007361903320997953,
+      "learning_rate": 2.4390476190476193e-06,
+      "loss": 0.0,
+      "step": 18440
+    },
+    {
+      "epoch": 2.6357142857142857,
+      "grad_norm": 0.020980671048164368,
+      "learning_rate": 2.4295238095238095e-06,
+      "loss": 0.0,
+      "step": 18450
+    },
+    {
+      "epoch": 2.637142857142857,
+      "grad_norm": 0.011586438864469528,
+      "learning_rate": 2.42e-06,
+      "loss": 0.0,
+      "step": 18460
+    },
+    {
+      "epoch": 2.638571428571429,
+      "grad_norm": 0.020586147904396057,
+      "learning_rate": 2.4104761904761908e-06,
+      "loss": 0.0,
+      "step": 18470
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.017591768875718117,
+      "learning_rate": 2.400952380952381e-06,
+      "loss": 0.0,
+      "step": 18480
+    },
+    {
+      "epoch": 2.6414285714285715,
+      "grad_norm": 0.015565615147352219,
+      "learning_rate": 2.3914285714285716e-06,
+      "loss": 0.8999,
+      "step": 18490
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 0.028424086049199104,
+      "learning_rate": 2.3819047619047623e-06,
+      "loss": 0.0,
+      "step": 18500
+    },
+    {
+      "epoch": 2.644285714285714,
+      "grad_norm": 0.05599723756313324,
+      "learning_rate": 2.3723809523809525e-06,
+      "loss": 0.0,
+      "step": 18510
+    },
+    {
+      "epoch": 2.645714285714286,
+      "grad_norm": 0.010638006962835789,
+      "learning_rate": 2.362857142857143e-06,
+      "loss": 1.8079,
+      "step": 18520
+    },
+    {
+      "epoch": 2.6471428571428572,
+      "grad_norm": 0.009201262146234512,
+      "learning_rate": 2.3533333333333334e-06,
+      "loss": 0.0,
+      "step": 18530
+    },
+    {
+      "epoch": 2.6485714285714286,
+      "grad_norm": 0.039645228534936905,
+      "learning_rate": 2.343809523809524e-06,
+      "loss": 1.8253,
+      "step": 18540
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.011205514892935753,
+      "learning_rate": 2.3342857142857147e-06,
+      "loss": 0.0,
+      "step": 18550
+    },
+    {
+      "epoch": 2.6514285714285712,
+      "grad_norm": 0.01212307345122099,
+      "learning_rate": 2.324761904761905e-06,
+      "loss": 0.8559,
+      "step": 18560
+    },
+    {
+      "epoch": 2.652857142857143,
+      "grad_norm": 0.015934692695736885,
+      "learning_rate": 2.3152380952380956e-06,
+      "loss": 0.0,
+      "step": 18570
+    },
+    {
+      "epoch": 2.6542857142857144,
+      "grad_norm": 0.00802537053823471,
+      "learning_rate": 2.3057142857142858e-06,
+      "loss": 0.0,
+      "step": 18580
+    },
+    {
+      "epoch": 2.6557142857142857,
+      "grad_norm": 0.01648503914475441,
+      "learning_rate": 2.2961904761904764e-06,
+      "loss": 0.0,
+      "step": 18590
+    },
+    {
+      "epoch": 2.657142857142857,
+      "grad_norm": 0.03174847364425659,
+      "learning_rate": 2.2866666666666667e-06,
+      "loss": 0.0,
+      "step": 18600
+    },
+    {
+      "epoch": 2.6585714285714284,
+      "grad_norm": 0.016849342733621597,
+      "learning_rate": 2.2771428571428573e-06,
+      "loss": 0.0,
+      "step": 18610
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.006888090167194605,
+      "learning_rate": 2.267619047619048e-06,
+      "loss": 1.8244,
+      "step": 18620
+    },
+    {
+      "epoch": 2.6614285714285715,
+      "grad_norm": 0.017884982749819756,
+      "learning_rate": 2.2580952380952386e-06,
+      "loss": 0.883,
+      "step": 18630
+    },
+    {
+      "epoch": 2.662857142857143,
+      "grad_norm": 0.03172047436237335,
+      "learning_rate": 2.248571428571429e-06,
+      "loss": 0.0,
+      "step": 18640
+    },
+    {
+      "epoch": 2.664285714285714,
+      "grad_norm": 0.01442161574959755,
+      "learning_rate": 2.2390476190476195e-06,
+      "loss": 0.9971,
+      "step": 18650
+    },
+    {
+      "epoch": 2.6657142857142855,
+      "grad_norm": 0.029089923948049545,
+      "learning_rate": 2.2295238095238097e-06,
+      "loss": 0.0,
+      "step": 18660
+    },
+    {
+      "epoch": 2.6671428571428573,
+      "grad_norm": 0.016114570200443268,
+      "learning_rate": 2.2200000000000003e-06,
+      "loss": 0.0,
+      "step": 18670
+    },
+    {
+      "epoch": 2.6685714285714286,
+      "grad_norm": 0.032905373722314835,
+      "learning_rate": 2.2104761904761906e-06,
+      "loss": 0.0,
+      "step": 18680
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.008628591895103455,
+      "learning_rate": 2.2009523809523812e-06,
+      "loss": 0.9217,
+      "step": 18690
+    },
+    {
+      "epoch": 2.6714285714285713,
+      "grad_norm": 0.022052578628063202,
+      "learning_rate": 2.1914285714285714e-06,
+      "loss": 0.0,
+      "step": 18700
+    },
+    {
+      "epoch": 2.6728571428571426,
+      "grad_norm": 0.034719083458185196,
+      "learning_rate": 2.181904761904762e-06,
+      "loss": 0.0,
+      "step": 18710
+    },
+    {
+      "epoch": 2.6742857142857144,
+      "grad_norm": 0.039499010890722275,
+      "learning_rate": 2.1723809523809523e-06,
+      "loss": 0.0,
+      "step": 18720
+    },
+    {
+      "epoch": 2.6757142857142857,
+      "grad_norm": 0.024418560788035393,
+      "learning_rate": 2.162857142857143e-06,
+      "loss": 0.0,
+      "step": 18730
+    },
+    {
+      "epoch": 2.677142857142857,
+      "grad_norm": 0.05033542588353157,
+      "learning_rate": 2.153333333333333e-06,
+      "loss": 0.9078,
+      "step": 18740
+    },
+    {
+      "epoch": 2.678571428571429,
+      "grad_norm": 0.010121088474988937,
+      "learning_rate": 2.1438095238095243e-06,
+      "loss": 1.0094,
+      "step": 18750
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.022984053939580917,
+      "learning_rate": 2.1342857142857145e-06,
+      "loss": 1.0375,
+      "step": 18760
+    },
+    {
+      "epoch": 2.6814285714285715,
+      "grad_norm": 0.015825951471924782,
+      "learning_rate": 2.124761904761905e-06,
+      "loss": 0.0,
+      "step": 18770
+    },
+    {
+      "epoch": 2.682857142857143,
+      "grad_norm": 0.013215320184826851,
+      "learning_rate": 2.1152380952380954e-06,
+      "loss": 0.8843,
+      "step": 18780
+    },
+    {
+      "epoch": 2.684285714285714,
+      "grad_norm": 0.034561626613140106,
+      "learning_rate": 2.105714285714286e-06,
+      "loss": 0.9717,
+      "step": 18790
+    },
+    {
+      "epoch": 2.685714285714286,
+      "grad_norm": 0.012007675133645535,
+      "learning_rate": 2.0961904761904762e-06,
+      "loss": 0.8355,
+      "step": 18800
+    },
+    {
+      "epoch": 2.6871428571428573,
+      "grad_norm": 0.0071459063328802586,
+      "learning_rate": 2.086666666666667e-06,
+      "loss": 0.0,
+      "step": 18810
+    },
+    {
+      "epoch": 2.6885714285714286,
+      "grad_norm": 0.35165172815322876,
+      "learning_rate": 2.077142857142857e-06,
+      "loss": 0.0,
+      "step": 18820
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.044401660561561584,
+      "learning_rate": 2.0676190476190477e-06,
+      "loss": 0.0,
+      "step": 18830
+    },
+    {
+      "epoch": 2.6914285714285713,
+      "grad_norm": 0.3554089665412903,
+      "learning_rate": 2.0580952380952384e-06,
+      "loss": 0.0,
+      "step": 18840
+    },
+    {
+      "epoch": 2.692857142857143,
+      "grad_norm": 0.022895246744155884,
+      "learning_rate": 2.0485714285714286e-06,
+      "loss": 0.0,
+      "step": 18850
+    },
+    {
+      "epoch": 2.6942857142857144,
+      "grad_norm": 0.03616378456354141,
+      "learning_rate": 2.0390476190476193e-06,
+      "loss": 1.778,
+      "step": 18860
+    },
+    {
+      "epoch": 2.6957142857142857,
+      "grad_norm": 0.02744809351861477,
+      "learning_rate": 2.0295238095238095e-06,
+      "loss": 0.0,
+      "step": 18870
+    },
+    {
+      "epoch": 2.697142857142857,
+      "grad_norm": 0.06410019099712372,
+      "learning_rate": 2.02e-06,
+      "loss": 0.0,
+      "step": 18880
+    },
+    {
+      "epoch": 2.6985714285714284,
+      "grad_norm": 0.009056619368493557,
+      "learning_rate": 2.010476190476191e-06,
+      "loss": 0.0,
+      "step": 18890
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.025036513805389404,
+      "learning_rate": 2.000952380952381e-06,
+      "loss": 0.0,
+      "step": 18900
+    },
+    {
+      "epoch": 2.7014285714285715,
+      "grad_norm": 0.04669957980513573,
+      "learning_rate": 1.9914285714285717e-06,
+      "loss": 0.7697,
+      "step": 18910
+    },
+    {
+      "epoch": 2.702857142857143,
+      "grad_norm": 0.01690228469669819,
+      "learning_rate": 1.9819047619047623e-06,
+      "loss": 0.0,
+      "step": 18920
+    },
+    {
+      "epoch": 2.704285714285714,
+      "grad_norm": 0.013450556434690952,
+      "learning_rate": 1.9723809523809525e-06,
+      "loss": 0.0,
+      "step": 18930
+    },
+    {
+      "epoch": 2.7057142857142855,
+      "grad_norm": 0.009837672114372253,
+      "learning_rate": 1.962857142857143e-06,
+      "loss": 0.0,
+      "step": 18940
+    },
+    {
+      "epoch": 2.7071428571428573,
+      "grad_norm": 0.007727089338004589,
+      "learning_rate": 1.9533333333333334e-06,
+      "loss": 0.0,
+      "step": 18950
+    },
+    {
+      "epoch": 2.7085714285714286,
+      "grad_norm": 0.009401781484484673,
+      "learning_rate": 1.943809523809524e-06,
+      "loss": 0.9192,
+      "step": 18960
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.014666692353785038,
+      "learning_rate": 1.9342857142857143e-06,
+      "loss": 0.0,
+      "step": 18970
+    },
+    {
+      "epoch": 2.7114285714285713,
+      "grad_norm": 0.03761747479438782,
+      "learning_rate": 1.924761904761905e-06,
+      "loss": 0.0,
+      "step": 18980
+    },
+    {
+      "epoch": 2.7128571428571426,
+      "grad_norm": 0.017192600294947624,
+      "learning_rate": 1.915238095238095e-06,
+      "loss": 0.0,
+      "step": 18990
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.015906931832432747,
+      "learning_rate": 1.9057142857142858e-06,
+      "loss": 0.0,
+      "step": 19000
+    },
+    {
+      "epoch": 2.7157142857142857,
+      "grad_norm": 0.011816808022558689,
+      "learning_rate": 1.8961904761904764e-06,
+      "loss": 1.7602,
+      "step": 19010
+    },
+    {
+      "epoch": 2.717142857142857,
+      "grad_norm": 0.03680361807346344,
+      "learning_rate": 1.8866666666666669e-06,
+      "loss": 0.9778,
+      "step": 19020
+    },
+    {
+      "epoch": 2.7185714285714284,
+      "grad_norm": 0.014581674709916115,
+      "learning_rate": 1.8771428571428573e-06,
+      "loss": 0.0,
+      "step": 19030
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.009853411465883255,
+      "learning_rate": 1.8676190476190478e-06,
+      "loss": 0.9572,
+      "step": 19040
+    },
+    {
+      "epoch": 2.7214285714285715,
+      "grad_norm": 0.022520339116454124,
+      "learning_rate": 1.8580952380952382e-06,
+      "loss": 0.0,
+      "step": 19050
+    },
+    {
+      "epoch": 2.722857142857143,
+      "grad_norm": 0.037078455090522766,
+      "learning_rate": 1.8485714285714288e-06,
+      "loss": 0.0,
+      "step": 19060
+    },
+    {
+      "epoch": 2.724285714285714,
+      "grad_norm": 0.02080574631690979,
+      "learning_rate": 1.8390476190476193e-06,
+      "loss": 1.7618,
+      "step": 19070
+    },
+    {
+      "epoch": 2.725714285714286,
+      "grad_norm": 0.013858597725629807,
+      "learning_rate": 1.8295238095238097e-06,
+      "loss": 0.0,
+      "step": 19080
+    },
+    {
+      "epoch": 2.727142857142857,
+      "grad_norm": 0.01692185178399086,
+      "learning_rate": 1.8200000000000002e-06,
+      "loss": 0.0,
+      "step": 19090
+    },
+    {
+      "epoch": 2.7285714285714286,
+      "grad_norm": 0.014618800021708012,
+      "learning_rate": 1.8104761904761906e-06,
+      "loss": 0.0,
+      "step": 19100
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.013522250577807426,
+      "learning_rate": 1.800952380952381e-06,
+      "loss": 0.0,
+      "step": 19110
+    },
+    {
+      "epoch": 2.7314285714285713,
+      "grad_norm": 0.011585542932152748,
+      "learning_rate": 1.7914285714285715e-06,
+      "loss": 0.0,
+      "step": 19120
+    },
+    {
+      "epoch": 2.732857142857143,
+      "grad_norm": 0.004032991826534271,
+      "learning_rate": 1.781904761904762e-06,
+      "loss": 0.0,
+      "step": 19130
+    },
+    {
+      "epoch": 2.7342857142857144,
+      "grad_norm": 0.02088911458849907,
+      "learning_rate": 1.7723809523809525e-06,
+      "loss": 0.0,
+      "step": 19140
+    },
+    {
+      "epoch": 2.7357142857142858,
+      "grad_norm": 0.1235940158367157,
+      "learning_rate": 1.7628571428571432e-06,
+      "loss": 0.0,
+      "step": 19150
+    },
+    {
+      "epoch": 2.737142857142857,
+      "grad_norm": 0.012252864427864552,
+      "learning_rate": 1.7533333333333336e-06,
+      "loss": 0.0,
+      "step": 19160
+    },
+    {
+      "epoch": 2.7385714285714284,
+      "grad_norm": 0.007913474924862385,
+      "learning_rate": 1.743809523809524e-06,
+      "loss": 0.0,
+      "step": 19170
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.008999345824122429,
+      "learning_rate": 1.7342857142857145e-06,
+      "loss": 0.0,
+      "step": 19180
+    },
+    {
+      "epoch": 2.7414285714285715,
+      "grad_norm": 0.003593433415517211,
+      "learning_rate": 1.724761904761905e-06,
+      "loss": 0.0,
+      "step": 19190
+    },
+    {
+      "epoch": 2.742857142857143,
+      "grad_norm": 0.007274888921529055,
+      "learning_rate": 1.7152380952380954e-06,
+      "loss": 0.0,
+      "step": 19200
+    },
+    {
+      "epoch": 2.744285714285714,
+      "grad_norm": 0.02707606740295887,
+      "learning_rate": 1.7057142857142858e-06,
+      "loss": 1.0108,
+      "step": 19210
+    },
+    {
+      "epoch": 2.7457142857142856,
+      "grad_norm": 0.022561397403478622,
+      "learning_rate": 1.6961904761904762e-06,
+      "loss": 0.9517,
+      "step": 19220
+    },
+    {
+      "epoch": 2.7471428571428573,
+      "grad_norm": 0.010746422223746777,
+      "learning_rate": 1.6866666666666667e-06,
+      "loss": 0.0,
+      "step": 19230
+    },
+    {
+      "epoch": 2.7485714285714287,
+      "grad_norm": 0.0189865343272686,
+      "learning_rate": 1.6771428571428571e-06,
+      "loss": 0.0,
+      "step": 19240
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.01099209301173687,
+      "learning_rate": 1.6676190476190476e-06,
+      "loss": 0.9422,
+      "step": 19250
+    },
+    {
+      "epoch": 2.7514285714285713,
+      "grad_norm": 0.009334533475339413,
+      "learning_rate": 1.6580952380952384e-06,
+      "loss": 0.0,
+      "step": 19260
+    },
+    {
+      "epoch": 2.7528571428571427,
+      "grad_norm": 0.013022125698626041,
+      "learning_rate": 1.6485714285714289e-06,
+      "loss": 0.0,
+      "step": 19270
+    },
+    {
+      "epoch": 2.7542857142857144,
+      "grad_norm": 0.013493633829057217,
+      "learning_rate": 1.6390476190476193e-06,
+      "loss": 1.0402,
+      "step": 19280
+    },
+    {
+      "epoch": 2.755714285714286,
+      "grad_norm": 0.023864207789301872,
+      "learning_rate": 1.6295238095238097e-06,
+      "loss": 0.0,
+      "step": 19290
+    },
+    {
+      "epoch": 2.757142857142857,
+      "grad_norm": 0.013668490573763847,
+      "learning_rate": 1.6200000000000002e-06,
+      "loss": 0.0,
+      "step": 19300
+    },
+    {
+      "epoch": 2.7585714285714285,
+      "grad_norm": 0.004224321339279413,
+      "learning_rate": 1.6104761904761906e-06,
+      "loss": 0.0,
+      "step": 19310
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.005939226597547531,
+      "learning_rate": 1.600952380952381e-06,
+      "loss": 0.0,
+      "step": 19320
+    },
+    {
+      "epoch": 2.7614285714285716,
+      "grad_norm": 0.0042728241533041,
+      "learning_rate": 1.5914285714285715e-06,
+      "loss": 0.0,
+      "step": 19330
+    },
+    {
+      "epoch": 2.762857142857143,
+      "grad_norm": 0.006031028926372528,
+      "learning_rate": 1.581904761904762e-06,
+      "loss": 0.0,
+      "step": 19340
+    },
+    {
+      "epoch": 2.7642857142857142,
+      "grad_norm": 0.015320521779358387,
+      "learning_rate": 1.5723809523809526e-06,
+      "loss": 0.0,
+      "step": 19350
+    },
+    {
+      "epoch": 2.7657142857142856,
+      "grad_norm": 0.011915501207113266,
+      "learning_rate": 1.562857142857143e-06,
+      "loss": 0.0,
+      "step": 19360
+    },
+    {
+      "epoch": 2.767142857142857,
+      "grad_norm": 0.13586825132369995,
+      "learning_rate": 1.5533333333333334e-06,
+      "loss": 0.0,
+      "step": 19370
+    },
+    {
+      "epoch": 2.7685714285714287,
+      "grad_norm": 0.027230247855186462,
+      "learning_rate": 1.5438095238095239e-06,
+      "loss": 0.0,
+      "step": 19380
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.010185507126152515,
+      "learning_rate": 1.5342857142857145e-06,
+      "loss": 0.0,
+      "step": 19390
+    },
+    {
+      "epoch": 2.7714285714285714,
+      "grad_norm": 0.008883404545485973,
+      "learning_rate": 1.524761904761905e-06,
+      "loss": 0.0,
+      "step": 19400
+    },
+    {
+      "epoch": 2.772857142857143,
+      "grad_norm": 0.01675032451748848,
+      "learning_rate": 1.5152380952380954e-06,
+      "loss": 0.0,
+      "step": 19410
+    },
+    {
+      "epoch": 2.774285714285714,
+      "grad_norm": 0.01758684404194355,
+      "learning_rate": 1.5057142857142858e-06,
+      "loss": 0.9363,
+      "step": 19420
+    },
+    {
+      "epoch": 2.775714285714286,
+      "grad_norm": 0.03635222464799881,
+      "learning_rate": 1.4961904761904763e-06,
+      "loss": 1.003,
+      "step": 19430
+    },
+    {
+      "epoch": 2.777142857142857,
+      "grad_norm": 0.02365276776254177,
+      "learning_rate": 1.486666666666667e-06,
+      "loss": 0.0,
+      "step": 19440
+    },
+    {
+      "epoch": 2.7785714285714285,
+      "grad_norm": 0.009333426132798195,
+      "learning_rate": 1.4771428571428573e-06,
+      "loss": 0.0,
+      "step": 19450
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 0.009278497658669949,
+      "learning_rate": 1.4676190476190478e-06,
+      "loss": 0.9298,
+      "step": 19460
+    },
+    {
+      "epoch": 2.7814285714285716,
+      "grad_norm": 0.005123590584844351,
+      "learning_rate": 1.4580952380952382e-06,
+      "loss": 0.0,
+      "step": 19470
+    },
+    {
+      "epoch": 2.782857142857143,
+      "grad_norm": 0.017383050173521042,
+      "learning_rate": 1.4485714285714286e-06,
+      "loss": 0.0,
+      "step": 19480
+    },
+    {
+      "epoch": 2.7842857142857143,
+      "grad_norm": 0.05856950208544731,
+      "learning_rate": 1.439047619047619e-06,
+      "loss": 0.0,
+      "step": 19490
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.005129855126142502,
+      "learning_rate": 1.4295238095238095e-06,
+      "loss": 0.0,
+      "step": 19500
+    },
+    {
+      "epoch": 2.7871428571428574,
+      "grad_norm": 0.009891141206026077,
+      "learning_rate": 1.42e-06,
+      "loss": 0.0,
+      "step": 19510
+    },
+    {
+      "epoch": 2.7885714285714287,
+      "grad_norm": 0.00696169026196003,
+      "learning_rate": 1.4104761904761908e-06,
+      "loss": 0.8981,
+      "step": 19520
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.007330980151891708,
+      "learning_rate": 1.4009523809523813e-06,
+      "loss": 0.0,
+      "step": 19530
+    },
+    {
+      "epoch": 2.7914285714285714,
+      "grad_norm": 0.007694378960877657,
+      "learning_rate": 1.3914285714285717e-06,
+      "loss": 0.8513,
+      "step": 19540
+    },
+    {
+      "epoch": 2.7928571428571427,
+      "grad_norm": 0.027766233310103416,
+      "learning_rate": 1.3819047619047621e-06,
+      "loss": 0.9072,
+      "step": 19550
+    },
+    {
+      "epoch": 2.7942857142857145,
+      "grad_norm": 0.06495768576860428,
+      "learning_rate": 1.3723809523809526e-06,
+      "loss": 0.9283,
+      "step": 19560
+    },
+    {
+      "epoch": 2.795714285714286,
+      "grad_norm": 0.007060553878545761,
+      "learning_rate": 1.362857142857143e-06,
+      "loss": 0.886,
+      "step": 19570
+    },
+    {
+      "epoch": 2.797142857142857,
+      "grad_norm": 0.01171286404132843,
+      "learning_rate": 1.3533333333333334e-06,
+      "loss": 0.0,
+      "step": 19580
+    },
+    {
+      "epoch": 2.7985714285714285,
+      "grad_norm": 0.03173728287220001,
+      "learning_rate": 1.3438095238095239e-06,
+      "loss": 0.0,
+      "step": 19590
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.007293702103197575,
+      "learning_rate": 1.3342857142857143e-06,
+      "loss": 0.8402,
+      "step": 19600
+    },
+    {
+      "epoch": 2.8014285714285716,
+      "grad_norm": 0.019622651860117912,
+      "learning_rate": 1.3247619047619047e-06,
+      "loss": 0.0,
+      "step": 19610
+    },
+    {
+      "epoch": 2.802857142857143,
+      "grad_norm": 0.009242665953934193,
+      "learning_rate": 1.3152380952380952e-06,
+      "loss": 0.0,
+      "step": 19620
+    },
+    {
+      "epoch": 2.8042857142857143,
+      "grad_norm": 0.013806388713419437,
+      "learning_rate": 1.3057142857142856e-06,
+      "loss": 0.4822,
+      "step": 19630
+    },
+    {
+      "epoch": 2.8057142857142856,
+      "grad_norm": 0.042973000556230545,
+      "learning_rate": 1.2961904761904763e-06,
+      "loss": 0.0,
+      "step": 19640
+    },
+    {
+      "epoch": 2.807142857142857,
+      "grad_norm": 0.11540071666240692,
+      "learning_rate": 1.286666666666667e-06,
+      "loss": 1.7632,
+      "step": 19650
+    },
+    {
+      "epoch": 2.8085714285714287,
+      "grad_norm": 0.009889152832329273,
+      "learning_rate": 1.2771428571428573e-06,
+      "loss": 0.0,
+      "step": 19660
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.015030757524073124,
+      "learning_rate": 1.2676190476190478e-06,
+      "loss": 1.7649,
+      "step": 19670
+    },
+    {
+      "epoch": 2.8114285714285714,
+      "grad_norm": 0.008396998979151249,
+      "learning_rate": 1.2580952380952382e-06,
+      "loss": 1.0293,
+      "step": 19680
+    },
+    {
+      "epoch": 2.8128571428571427,
+      "grad_norm": 0.005490635521709919,
+      "learning_rate": 1.2485714285714287e-06,
+      "loss": 0.0,
+      "step": 19690
+    },
+    {
+      "epoch": 2.814285714285714,
+      "grad_norm": 0.020350586622953415,
+      "learning_rate": 1.239047619047619e-06,
+      "loss": 0.0,
+      "step": 19700
+    },
+    {
+      "epoch": 2.815714285714286,
+      "grad_norm": 0.01960679702460766,
+      "learning_rate": 1.2295238095238095e-06,
+      "loss": 1.06,
+      "step": 19710
+    },
+    {
+      "epoch": 2.817142857142857,
+      "grad_norm": 0.023930624127388,
+      "learning_rate": 1.2200000000000002e-06,
+      "loss": 0.0,
+      "step": 19720
+    },
+    {
+      "epoch": 2.8185714285714285,
+      "grad_norm": 0.022764993831515312,
+      "learning_rate": 1.2104761904761906e-06,
+      "loss": 0.0,
+      "step": 19730
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.008100210689008236,
+      "learning_rate": 1.200952380952381e-06,
+      "loss": 0.0,
+      "step": 19740
+    },
+    {
+      "epoch": 2.821428571428571,
+      "grad_norm": 0.052262354642152786,
+      "learning_rate": 1.1914285714285715e-06,
+      "loss": 0.884,
+      "step": 19750
+    },
+    {
+      "epoch": 2.822857142857143,
+      "grad_norm": 0.00824433658272028,
+      "learning_rate": 1.181904761904762e-06,
+      "loss": 0.0,
+      "step": 19760
+    },
+    {
+      "epoch": 2.8242857142857143,
+      "grad_norm": 0.011663568206131458,
+      "learning_rate": 1.1723809523809526e-06,
+      "loss": 0.0,
+      "step": 19770
+    },
+    {
+      "epoch": 2.8257142857142856,
+      "grad_norm": 0.010176041163504124,
+      "learning_rate": 1.162857142857143e-06,
+      "loss": 0.0,
+      "step": 19780
+    },
+    {
+      "epoch": 2.8271428571428574,
+      "grad_norm": 0.08770725131034851,
+      "learning_rate": 1.1533333333333334e-06,
+      "loss": 0.0,
+      "step": 19790
+    },
+    {
+      "epoch": 2.8285714285714287,
+      "grad_norm": 0.00560695119202137,
+      "learning_rate": 1.1438095238095239e-06,
+      "loss": 0.0,
+      "step": 19800
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.009446214884519577,
+      "learning_rate": 1.1342857142857145e-06,
+      "loss": 0.0,
+      "step": 19810
+    },
+    {
+      "epoch": 2.8314285714285714,
+      "grad_norm": 2.310727119445801,
+      "learning_rate": 1.124761904761905e-06,
+      "loss": 0.0,
+      "step": 19820
+    },
+    {
+      "epoch": 2.8328571428571427,
+      "grad_norm": 0.0208747461438179,
+      "learning_rate": 1.1152380952380954e-06,
+      "loss": 0.0,
+      "step": 19830
+    },
+    {
+      "epoch": 2.8342857142857145,
+      "grad_norm": 0.015969814732670784,
+      "learning_rate": 1.1057142857142858e-06,
+      "loss": 0.0,
+      "step": 19840
+    },
+    {
+      "epoch": 2.835714285714286,
+      "grad_norm": 0.013834877870976925,
+      "learning_rate": 1.0961904761904763e-06,
+      "loss": 0.8965,
+      "step": 19850
+    },
+    {
+      "epoch": 2.837142857142857,
+      "grad_norm": 0.008724192157387733,
+      "learning_rate": 1.0866666666666667e-06,
+      "loss": 0.0,
+      "step": 19860
+    },
+    {
+      "epoch": 2.8385714285714285,
+      "grad_norm": 0.007255997974425554,
+      "learning_rate": 1.0771428571428574e-06,
+      "loss": 0.0,
+      "step": 19870
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.014445670880377293,
+      "learning_rate": 1.0676190476190478e-06,
+      "loss": 0.0,
+      "step": 19880
+    },
+    {
+      "epoch": 2.8414285714285716,
+      "grad_norm": 0.02448618970811367,
+      "learning_rate": 1.0580952380952382e-06,
+      "loss": 0.0,
+      "step": 19890
+    },
+    {
+      "epoch": 2.842857142857143,
+      "grad_norm": 0.015276198275387287,
+      "learning_rate": 1.0485714285714287e-06,
+      "loss": 0.8793,
+      "step": 19900
+    },
+    {
+      "epoch": 2.8442857142857143,
+      "grad_norm": 0.00474558724090457,
+      "learning_rate": 1.039047619047619e-06,
+      "loss": 0.0,
+      "step": 19910
+    },
+    {
+      "epoch": 2.8457142857142856,
+      "grad_norm": 0.026547016575932503,
+      "learning_rate": 1.0295238095238095e-06,
+      "loss": 0.0,
+      "step": 19920
+    },
+    {
+      "epoch": 2.847142857142857,
+      "grad_norm": 0.01980089768767357,
+      "learning_rate": 1.02e-06,
+      "loss": 0.9575,
+      "step": 19930
+    },
+    {
+      "epoch": 2.8485714285714288,
+      "grad_norm": 0.01428477093577385,
+      "learning_rate": 1.0104761904761906e-06,
+      "loss": 0.9133,
+      "step": 19940
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.005189124029129744,
+      "learning_rate": 1.000952380952381e-06,
+      "loss": 0.0,
+      "step": 19950
+    },
+    {
+      "epoch": 2.8514285714285714,
+      "grad_norm": 0.014072846621274948,
+      "learning_rate": 9.914285714285715e-07,
+      "loss": 0.9134,
+      "step": 19960
+    },
+    {
+      "epoch": 2.8528571428571428,
+      "grad_norm": 0.03129696100950241,
+      "learning_rate": 9.81904761904762e-07,
+      "loss": 0.0,
+      "step": 19970
+    },
+    {
+      "epoch": 2.854285714285714,
+      "grad_norm": 0.022301360964775085,
+      "learning_rate": 9.723809523809524e-07,
+      "loss": 0.0,
+      "step": 19980
+    },
+    {
+      "epoch": 2.855714285714286,
+      "grad_norm": 0.0693112462759018,
+      "learning_rate": 9.628571428571428e-07,
+      "loss": 0.0,
+      "step": 19990
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.07488000392913818,
+      "learning_rate": 9.533333333333335e-07,
+      "loss": 1.0032,
+      "step": 20000
+    },
+    {
+      "epoch": 2.8585714285714285,
+      "grad_norm": 0.01510460302233696,
+      "learning_rate": 9.438095238095239e-07,
+      "loss": 0.0,
+      "step": 20010
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.01721733994781971,
+      "learning_rate": 9.342857142857144e-07,
+      "loss": 0.8428,
+      "step": 20020
+    },
+    {
+      "epoch": 2.861428571428571,
+      "grad_norm": 0.006324476562440395,
+      "learning_rate": 9.247619047619049e-07,
+      "loss": 0.0,
+      "step": 20030
+    },
+    {
+      "epoch": 2.862857142857143,
+      "grad_norm": 0.005335427820682526,
+      "learning_rate": 9.152380952380953e-07,
+      "loss": 0.0,
+      "step": 20040
+    },
+    {
+      "epoch": 2.8642857142857143,
+      "grad_norm": 0.07141967117786407,
+      "learning_rate": 9.057142857142857e-07,
+      "loss": 0.0,
+      "step": 20050
+    },
+    {
+      "epoch": 2.8657142857142857,
+      "grad_norm": 0.024915039539337158,
+      "learning_rate": 8.961904761904763e-07,
+      "loss": 0.9026,
+      "step": 20060
+    },
+    {
+      "epoch": 2.867142857142857,
+      "grad_norm": 0.030766287818551064,
+      "learning_rate": 8.866666666666668e-07,
+      "loss": 0.0,
+      "step": 20070
+    },
+    {
+      "epoch": 2.8685714285714283,
+      "grad_norm": 0.002713422989472747,
+      "learning_rate": 8.771428571428573e-07,
+      "loss": 0.0,
+      "step": 20080
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.012714025564491749,
+      "learning_rate": 8.676190476190477e-07,
+      "loss": 0.9809,
+      "step": 20090
+    },
+    {
+      "epoch": 2.8714285714285714,
+      "grad_norm": 0.01447638776153326,
+      "learning_rate": 8.580952380952381e-07,
+      "loss": 0.0,
+      "step": 20100
+    },
+    {
+      "epoch": 2.8728571428571428,
+      "grad_norm": 0.0069658150896430016,
+      "learning_rate": 8.485714285714286e-07,
+      "loss": 0.0,
+      "step": 20110
+    },
+    {
+      "epoch": 2.8742857142857146,
+      "grad_norm": 0.01258284505456686,
+      "learning_rate": 8.39047619047619e-07,
+      "loss": 0.907,
+      "step": 20120
+    },
+    {
+      "epoch": 2.8757142857142854,
+      "grad_norm": 0.007878626696765423,
+      "learning_rate": 8.295238095238097e-07,
+      "loss": 0.0,
+      "step": 20130
+    },
+    {
+      "epoch": 2.8771428571428572,
+      "grad_norm": 0.025965500622987747,
+      "learning_rate": 8.200000000000001e-07,
+      "loss": 0.0,
+      "step": 20140
+    },
+    {
+      "epoch": 2.8785714285714286,
+      "grad_norm": 0.02525782771408558,
+      "learning_rate": 8.104761904761905e-07,
+      "loss": 0.9372,
+      "step": 20150
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0058983089402318,
+      "learning_rate": 8.00952380952381e-07,
+      "loss": 0.9726,
+      "step": 20160
+    },
+    {
+      "epoch": 2.8814285714285717,
+      "grad_norm": 0.009704826399683952,
+      "learning_rate": 7.914285714285715e-07,
+      "loss": 0.9625,
+      "step": 20170
+    },
+    {
+      "epoch": 2.882857142857143,
+      "grad_norm": 0.024833304807543755,
+      "learning_rate": 7.819047619047619e-07,
+      "loss": 0.0,
+      "step": 20180
+    },
+    {
+      "epoch": 2.8842857142857143,
+      "grad_norm": 0.026112359017133713,
+      "learning_rate": 7.723809523809525e-07,
+      "loss": 0.0,
+      "step": 20190
+    },
+    {
+      "epoch": 2.8857142857142857,
+      "grad_norm": 0.018837688490748405,
+      "learning_rate": 7.628571428571429e-07,
+      "loss": 0.0,
+      "step": 20200
+    },
+    {
+      "epoch": 2.887142857142857,
+      "grad_norm": 0.03998300060629845,
+      "learning_rate": 7.533333333333335e-07,
+      "loss": 0.0,
+      "step": 20210
+    },
+    {
+      "epoch": 2.888571428571429,
+      "grad_norm": 0.21881887316703796,
+      "learning_rate": 7.438095238095239e-07,
+      "loss": 0.0,
+      "step": 20220
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4532130.0,
+      "learning_rate": 7.342857142857143e-07,
+      "loss": 0.9603,
+      "step": 20230
+    },
+    {
+      "epoch": 2.8914285714285715,
+      "grad_norm": 0.0052850255742669106,
+      "learning_rate": 7.247619047619048e-07,
+      "loss": 0.994,
+      "step": 20240
+    },
+    {
+      "epoch": 2.892857142857143,
+      "grad_norm": 0.04845016077160835,
+      "learning_rate": 7.152380952380952e-07,
+      "loss": 0.0,
+      "step": 20250
+    },
+    {
+      "epoch": 2.894285714285714,
+      "grad_norm": 0.012956401333212852,
+      "learning_rate": 7.057142857142859e-07,
+      "loss": 0.0,
+      "step": 20260
+    },
+    {
+      "epoch": 2.895714285714286,
+      "grad_norm": 0.059195440262556076,
+      "learning_rate": 6.961904761904763e-07,
+      "loss": 1.8616,
+      "step": 20270
+    },
+    {
+      "epoch": 2.8971428571428572,
+      "grad_norm": 0.023142732679843903,
+      "learning_rate": 6.866666666666667e-07,
+      "loss": 0.0,
+      "step": 20280
+    },
+    {
+      "epoch": 2.8985714285714286,
+      "grad_norm": 0.013157381676137447,
+      "learning_rate": 6.771428571428572e-07,
+      "loss": 0.0,
+      "step": 20290
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.02213357761502266,
+      "learning_rate": 6.676190476190476e-07,
+      "loss": 0.0,
+      "step": 20300
+    },
+    {
+      "epoch": 2.9014285714285712,
+      "grad_norm": 0.01996447518467903,
+      "learning_rate": 6.580952380952381e-07,
+      "loss": 0.0,
+      "step": 20310
+    },
+    {
+      "epoch": 2.902857142857143,
+      "grad_norm": 0.016634538769721985,
+      "learning_rate": 6.485714285714287e-07,
+      "loss": 0.0,
+      "step": 20320
+    },
+    {
+      "epoch": 2.9042857142857144,
+      "grad_norm": 0.020388424396514893,
+      "learning_rate": 6.390476190476191e-07,
+      "loss": 0.0,
+      "step": 20330
+    },
+    {
+      "epoch": 2.9057142857142857,
+      "grad_norm": 0.017914218828082085,
+      "learning_rate": 6.295238095238096e-07,
+      "loss": 0.0,
+      "step": 20340
+    },
+    {
+      "epoch": 2.907142857142857,
+      "grad_norm": 0.043481845408678055,
+      "learning_rate": 6.200000000000001e-07,
+      "loss": 0.0,
+      "step": 20350
+    },
+    {
+      "epoch": 2.9085714285714284,
+      "grad_norm": 0.012139342725276947,
+      "learning_rate": 6.104761904761905e-07,
+      "loss": 0.8219,
+      "step": 20360
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0166412815451622,
+      "learning_rate": 6.00952380952381e-07,
+      "loss": 0.0,
+      "step": 20370
+    },
+    {
+      "epoch": 2.9114285714285715,
+      "grad_norm": 0.01993827149271965,
+      "learning_rate": 5.914285714285715e-07,
+      "loss": 0.0,
+      "step": 20380
+    },
+    {
+      "epoch": 2.912857142857143,
+      "grad_norm": 0.019031496718525887,
+      "learning_rate": 5.81904761904762e-07,
+      "loss": 0.9489,
+      "step": 20390
+    },
+    {
+      "epoch": 2.914285714285714,
+      "grad_norm": 0.005138327833265066,
+      "learning_rate": 5.723809523809525e-07,
+      "loss": 0.0,
+      "step": 20400
+    },
+    {
+      "epoch": 2.9157142857142855,
+      "grad_norm": 0.015386278741061687,
+      "learning_rate": 5.628571428571429e-07,
+      "loss": 0.0107,
+      "step": 20410
+    },
+    {
+      "epoch": 2.9171428571428573,
+      "grad_norm": 0.010767322033643723,
+      "learning_rate": 5.533333333333334e-07,
+      "loss": 1.81,
+      "step": 20420
+    },
+    {
+      "epoch": 2.9185714285714286,
+      "grad_norm": 0.010464508086442947,
+      "learning_rate": 5.438095238095239e-07,
+      "loss": 0.0,
+      "step": 20430
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5853961.0,
+      "learning_rate": 5.342857142857143e-07,
+      "loss": 0.9433,
+      "step": 20440
+    },
+    {
+      "epoch": 2.9214285714285713,
+      "grad_norm": 0.004949202761054039,
+      "learning_rate": 5.247619047619048e-07,
+      "loss": 0.0,
+      "step": 20450
+    },
+    {
+      "epoch": 2.9228571428571426,
+      "grad_norm": 6101951.0,
+      "learning_rate": 5.152380952380953e-07,
+      "loss": 1.8357,
+      "step": 20460
+    },
+    {
+      "epoch": 2.9242857142857144,
+      "grad_norm": 0.018308518454432487,
+      "learning_rate": 5.057142857142858e-07,
+      "loss": 0.0,
+      "step": 20470
+    },
+    {
+      "epoch": 2.9257142857142857,
+      "grad_norm": 0.012956427410244942,
+      "learning_rate": 4.961904761904762e-07,
+      "loss": 0.0,
+      "step": 20480
+    },
+    {
+      "epoch": 2.927142857142857,
+      "grad_norm": 0.012743604369461536,
+      "learning_rate": 4.866666666666666e-07,
+      "loss": 0.0,
+      "step": 20490
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 0.06503531336784363,
+      "learning_rate": 4.771428571428572e-07,
+      "loss": 0.0,
+      "step": 20500
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.032919544726610184,
+      "learning_rate": 4.6761904761904766e-07,
+      "loss": 0.0,
+      "step": 20510
+    },
+    {
+      "epoch": 2.9314285714285715,
+      "grad_norm": 0.024278385564684868,
+      "learning_rate": 4.580952380952381e-07,
+      "loss": 0.8677,
+      "step": 20520
+    },
+    {
+      "epoch": 2.932857142857143,
+      "grad_norm": 0.015267795883119106,
+      "learning_rate": 4.4857142857142864e-07,
+      "loss": 0.0,
+      "step": 20530
+    },
+    {
+      "epoch": 2.934285714285714,
+      "grad_norm": 0.0032085177954286337,
+      "learning_rate": 4.390476190476191e-07,
+      "loss": 0.9858,
+      "step": 20540
+    },
+    {
+      "epoch": 2.935714285714286,
+      "grad_norm": 0.00807663332670927,
+      "learning_rate": 4.295238095238095e-07,
+      "loss": 0.0,
+      "step": 20550
+    },
+    {
+      "epoch": 2.9371428571428573,
+      "grad_norm": 0.02913852035999298,
+      "learning_rate": 4.2000000000000006e-07,
+      "loss": 1.0459,
+      "step": 20560
+    },
+    {
+      "epoch": 2.9385714285714286,
+      "grad_norm": 0.0063605885952711105,
+      "learning_rate": 4.104761904761905e-07,
+      "loss": 0.0,
+      "step": 20570
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.010251528583467007,
+      "learning_rate": 4.00952380952381e-07,
+      "loss": 0.0,
+      "step": 20580
+    },
+    {
+      "epoch": 2.9414285714285713,
+      "grad_norm": 0.010460608638823032,
+      "learning_rate": 3.9142857142857147e-07,
+      "loss": 0.0,
+      "step": 20590
+    },
+    {
+      "epoch": 2.942857142857143,
+      "grad_norm": 0.012397107668220997,
+      "learning_rate": 3.8190476190476196e-07,
+      "loss": 0.9394,
+      "step": 20600
+    },
+    {
+      "epoch": 2.9442857142857144,
+      "grad_norm": 0.016947537660598755,
+      "learning_rate": 3.723809523809524e-07,
+      "loss": 0.7463,
+      "step": 20610
+    },
+    {
+      "epoch": 2.9457142857142857,
+      "grad_norm": 0.032406315207481384,
+      "learning_rate": 3.6285714285714283e-07,
+      "loss": 0.0,
+      "step": 20620
+    },
+    {
+      "epoch": 2.947142857142857,
+      "grad_norm": 0.013167922385036945,
+      "learning_rate": 3.533333333333334e-07,
+      "loss": 0.0,
+      "step": 20630
+    },
+    {
+      "epoch": 2.9485714285714284,
+      "grad_norm": 0.016242461279034615,
+      "learning_rate": 3.438095238095238e-07,
+      "loss": 0.0,
+      "step": 20640
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.028596194460988045,
+      "learning_rate": 3.342857142857143e-07,
+      "loss": 0.0,
+      "step": 20650
+    },
+    {
+      "epoch": 2.9514285714285715,
+      "grad_norm": 0.013866441324353218,
+      "learning_rate": 3.247619047619048e-07,
+      "loss": 1.0181,
+      "step": 20660
+    },
+    {
+      "epoch": 2.952857142857143,
+      "grad_norm": 0.011294553056359291,
+      "learning_rate": 3.152380952380953e-07,
+      "loss": 0.0,
+      "step": 20670
+    },
+    {
+      "epoch": 2.954285714285714,
+      "grad_norm": 275854912.0,
+      "learning_rate": 3.0571428571428577e-07,
+      "loss": 0.687,
+      "step": 20680
+    },
+    {
+      "epoch": 2.9557142857142855,
+      "grad_norm": 0.032343145459890366,
+      "learning_rate": 2.961904761904762e-07,
+      "loss": 0.9455,
+      "step": 20690
+    },
+    {
+      "epoch": 2.9571428571428573,
+      "grad_norm": 0.0038426064420491457,
+      "learning_rate": 2.866666666666667e-07,
+      "loss": 0.9385,
+      "step": 20700
+    },
+    {
+      "epoch": 2.9585714285714286,
+      "grad_norm": 7.084019184112549,
+      "learning_rate": 2.771428571428572e-07,
+      "loss": 0.9835,
+      "step": 20710
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.00502736447378993,
+      "learning_rate": 2.676190476190476e-07,
+      "loss": 0.0,
+      "step": 20720
+    },
+    {
+      "epoch": 2.9614285714285713,
+      "grad_norm": 0.005130284931510687,
+      "learning_rate": 2.580952380952381e-07,
+      "loss": 0.0,
+      "step": 20730
+    },
+    {
+      "epoch": 2.9628571428571426,
+      "grad_norm": 0.019307216629385948,
+      "learning_rate": 2.485714285714286e-07,
+      "loss": 0.9563,
+      "step": 20740
+    },
+    {
+      "epoch": 2.9642857142857144,
+      "grad_norm": 0.2338123768568039,
+      "learning_rate": 2.390476190476191e-07,
+      "loss": 0.9178,
+      "step": 20750
+    },
+    {
+      "epoch": 2.9657142857142857,
+      "grad_norm": 0.06240418925881386,
+      "learning_rate": 2.2952380952380955e-07,
+      "loss": 0.0,
+      "step": 20760
+    },
+    {
+      "epoch": 2.967142857142857,
+      "grad_norm": 0.006811524275690317,
+      "learning_rate": 2.2e-07,
+      "loss": 0.0,
+      "step": 20770
+    },
+    {
+      "epoch": 2.9685714285714284,
+      "grad_norm": 0.024930506944656372,
+      "learning_rate": 2.104761904761905e-07,
+      "loss": 0.0,
+      "step": 20780
+    },
+    {
+      "epoch": 2.9699999999999998,
+      "grad_norm": 0.009307929314672947,
+      "learning_rate": 2.00952380952381e-07,
+      "loss": 0.0,
+      "step": 20790
+    },
+    {
+      "epoch": 2.9714285714285715,
+      "grad_norm": 0.018731560558080673,
+      "learning_rate": 1.9142857142857143e-07,
+      "loss": 0.7976,
+      "step": 20800
+    },
+    {
+      "epoch": 2.972857142857143,
+      "grad_norm": 0.023556828498840332,
+      "learning_rate": 1.8190476190476192e-07,
+      "loss": 0.0,
+      "step": 20810
+    },
+    {
+      "epoch": 2.974285714285714,
+      "grad_norm": 0.026655329391360283,
+      "learning_rate": 1.7238095238095238e-07,
+      "loss": 0.0,
+      "step": 20820
+    },
+    {
+      "epoch": 2.975714285714286,
+      "grad_norm": 0.6703856587409973,
+      "learning_rate": 1.6285714285714287e-07,
+      "loss": 0.0,
+      "step": 20830
+    },
+    {
+      "epoch": 2.977142857142857,
+      "grad_norm": 0.08300791680812836,
+      "learning_rate": 1.5333333333333333e-07,
+      "loss": 0.9486,
+      "step": 20840
+    },
+    {
+      "epoch": 2.9785714285714286,
+      "grad_norm": 0.009628263302147388,
+      "learning_rate": 1.4380952380952382e-07,
+      "loss": 0.0,
+      "step": 20850
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.04733583703637123,
+      "learning_rate": 1.342857142857143e-07,
+      "loss": 0.0,
+      "step": 20860
+    },
+    {
+      "epoch": 2.9814285714285713,
+      "grad_norm": 0.016383137553930283,
+      "learning_rate": 1.2476190476190477e-07,
+      "loss": 0.9942,
+      "step": 20870
+    },
+    {
+      "epoch": 2.982857142857143,
+      "grad_norm": 0.0073188296519219875,
+      "learning_rate": 1.1523809523809523e-07,
+      "loss": 0.0,
+      "step": 20880
+    },
+    {
+      "epoch": 2.9842857142857144,
+      "grad_norm": 0.01324001606553793,
+      "learning_rate": 1.0571428571428572e-07,
+      "loss": 0.8608,
+      "step": 20890
+    },
+    {
+      "epoch": 2.9857142857142858,
+      "grad_norm": 0.03589126467704773,
+      "learning_rate": 9.61904761904762e-08,
+      "loss": 0.0,
+      "step": 20900
+    },
+    {
+      "epoch": 2.987142857142857,
+      "grad_norm": 0.009757845662534237,
+      "learning_rate": 8.666666666666668e-08,
+      "loss": 1.0642,
+      "step": 20910
+    },
+    {
+      "epoch": 2.9885714285714284,
+      "grad_norm": 0.007351737003773451,
+      "learning_rate": 7.714285714285715e-08,
+      "loss": 0.9552,
+      "step": 20920
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.016089852899312973,
+      "learning_rate": 6.761904761904763e-08,
+      "loss": 0.0,
+      "step": 20930
+    },
+    {
+      "epoch": 2.9914285714285715,
+      "grad_norm": 0.006478155963122845,
+      "learning_rate": 5.80952380952381e-08,
+      "loss": 0.0,
+      "step": 20940
+    },
+    {
+      "epoch": 2.992857142857143,
+      "grad_norm": 0.008611363358795643,
+      "learning_rate": 4.857142857142857e-08,
+      "loss": 0.0,
+      "step": 20950
+    },
+    {
+      "epoch": 2.994285714285714,
+      "grad_norm": 0.014392401091754436,
+      "learning_rate": 3.904761904761905e-08,
+      "loss": 0.0,
+      "step": 20960
+    },
+    {
+      "epoch": 2.9957142857142856,
+      "grad_norm": 0.01103729847818613,
+      "learning_rate": 2.9523809523809528e-08,
+      "loss": 1.0153,
+      "step": 20970
+    },
+    {
+      "epoch": 2.9971428571428573,
+      "grad_norm": 0.018647512421011925,
+      "learning_rate": 2e-08,
+      "loss": 0.9694,
+      "step": 20980
+    },
+    {
+      "epoch": 2.9985714285714287,
+      "grad_norm": 0.01091854926198721,
+      "learning_rate": 1.0476190476190478e-08,
+      "loss": 0.5829,
+      "step": 20990
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.03248345106840134,
+      "learning_rate": 9.523809523809524e-10,
+      "loss": 0.0,
+      "step": 21000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9446666666666667,
+      "eval_f1_macro": 0.9441189767578706,
+      "eval_f1_weighted": 0.9446943277731716,
+      "eval_loss": 1.0492944717407227,
+      "eval_runtime": 185.8432,
+      "eval_samples_per_second": 32.285,
+      "eval_steps_per_second": 16.143,
+      "step": 21000
+    },
+    {
+      "epoch": 3.0,
+      "step": 21000,
+      "total_flos": 1372987486852800.0,
+      "train_loss": 0.8612068638761459,
+      "train_runtime": 3731.3178,
+      "train_samples_per_second": 11.256,
+      "train_steps_per_second": 5.628
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9446666666666667,
+      "eval_f1_macro": 0.9441189767578706,
+      "eval_f1_weighted": 0.9446943277731716,
+      "eval_loss": 1.0492944717407227,
+      "eval_runtime": 186.8464,
+      "eval_samples_per_second": 32.112,
+      "eval_steps_per_second": 16.056,
+      "step": 21000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 21000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1372987486852800.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}