diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24453 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3489,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4240590333938599,
+      "learning_rate": 4.7619047619047627e-08,
+      "loss": 0.3934,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.34474515914917,
+      "learning_rate": 9.523809523809525e-08,
+      "loss": 0.398,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3957951068878174,
+      "learning_rate": 1.4285714285714287e-07,
+      "loss": 0.394,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4374407529830933,
+      "learning_rate": 1.904761904761905e-07,
+      "loss": 0.4179,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4964826107025146,
+      "learning_rate": 2.3809523809523811e-07,
+      "loss": 0.4058,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.46127450466156,
+      "learning_rate": 2.8571428571428575e-07,
+      "loss": 0.4066,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4883235692977905,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 0.4137,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3953728675842285,
+      "learning_rate": 3.80952380952381e-07,
+      "loss": 0.4085,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3673226833343506,
+      "learning_rate": 4.285714285714286e-07,
+      "loss": 0.3985,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3896797895431519,
+      "learning_rate": 4.7619047619047623e-07,
+      "loss": 0.3972,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3578170537948608,
+      "learning_rate": 5.238095238095239e-07,
+      "loss": 0.3947,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4428050518035889,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 0.4101,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4399523735046387,
+      "learning_rate": 6.190476190476191e-07,
+      "loss": 0.4152,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.493770718574524,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 0.4092,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4112420082092285,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.3939,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4233614206314087,
+      "learning_rate": 7.61904761904762e-07,
+      "loss": 0.3973,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4595534801483154,
+      "learning_rate": 8.095238095238096e-07,
+      "loss": 0.402,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4608535766601562,
+      "learning_rate": 8.571428571428572e-07,
+      "loss": 0.3816,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4193949699401855,
+      "learning_rate": 9.047619047619048e-07,
+      "loss": 0.3909,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3311034440994263,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 0.3838,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.26755690574646,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.3648,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3362178802490234,
+      "learning_rate": 1.0476190476190478e-06,
+      "loss": 0.38,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3571105003356934,
+      "learning_rate": 1.0952380952380954e-06,
+      "loss": 0.3826,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3875077962875366,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 0.3928,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2882364988327026,
+      "learning_rate": 1.1904761904761906e-06,
+      "loss": 0.3794,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3290733098983765,
+      "learning_rate": 1.2380952380952382e-06,
+      "loss": 0.3787,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2560834884643555,
+      "learning_rate": 1.2857142857142856e-06,
+      "loss": 0.3583,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.171757221221924,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 0.3462,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5028345584869385,
+      "learning_rate": 1.3809523809523812e-06,
+      "loss": 0.3423,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6555448770523071,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.3314,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2543309926986694,
+      "learning_rate": 1.4761904761904762e-06,
+      "loss": 0.3164,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9398919343948364,
+      "learning_rate": 1.523809523809524e-06,
+      "loss": 0.3221,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8991830945014954,
+      "learning_rate": 1.5714285714285714e-06,
+      "loss": 0.3312,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.060887336730957,
+      "learning_rate": 1.6190476190476193e-06,
+      "loss": 0.3266,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0378755331039429,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.3179,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0888737440109253,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 0.3095,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2707414627075195,
+      "learning_rate": 1.761904761904762e-06,
+      "loss": 0.3047,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7934312224388123,
+      "learning_rate": 1.8095238095238097e-06,
+      "loss": 0.2916,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.72513747215271,
+      "learning_rate": 1.8571428571428573e-06,
+      "loss": 0.2754,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7445487976074219,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 0.2763,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7709694504737854,
+      "learning_rate": 1.9523809523809527e-06,
+      "loss": 0.2788,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.73027104139328,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.2789,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.979809582233429,
+      "learning_rate": 2.047619047619048e-06,
+      "loss": 0.2789,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9427708983421326,
+      "learning_rate": 2.0952380952380955e-06,
+      "loss": 0.2711,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7210543751716614,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 0.2671,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4684184789657593,
+      "learning_rate": 2.1904761904761908e-06,
+      "loss": 0.2631,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38682615756988525,
+      "learning_rate": 2.2380952380952384e-06,
+      "loss": 0.2568,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38365742564201355,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 0.2583,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3978930711746216,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 0.2682,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4151419401168823,
+      "learning_rate": 2.380952380952381e-06,
+      "loss": 0.2581,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40405192971229553,
+      "learning_rate": 2.428571428571429e-06,
+      "loss": 0.2584,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3843757212162018,
+      "learning_rate": 2.4761904761904764e-06,
+      "loss": 0.2475,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43965137004852295,
+      "learning_rate": 2.523809523809524e-06,
+      "loss": 0.2577,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38827186822891235,
+      "learning_rate": 2.571428571428571e-06,
+      "loss": 0.2561,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45033612847328186,
+      "learning_rate": 2.6190476190476192e-06,
+      "loss": 0.2556,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4053068161010742,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 0.2482,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39862483739852905,
+      "learning_rate": 2.7142857142857144e-06,
+      "loss": 0.235,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3886047899723053,
+      "learning_rate": 2.7619047619047625e-06,
+      "loss": 0.2383,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3530617356300354,
+      "learning_rate": 2.8095238095238096e-06,
+      "loss": 0.2462,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3844414949417114,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.241,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38413164019584656,
+      "learning_rate": 2.9047619047619053e-06,
+      "loss": 0.2374,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44040414690971375,
+      "learning_rate": 2.9523809523809525e-06,
+      "loss": 0.2332,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3900178074836731,
+      "learning_rate": 3e-06,
+      "loss": 0.2278,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31994953751564026,
+      "learning_rate": 3.047619047619048e-06,
+      "loss": 0.2327,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3429587781429291,
+      "learning_rate": 3.0952380952380957e-06,
+      "loss": 0.231,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3350449502468109,
+      "learning_rate": 3.142857142857143e-06,
+      "loss": 0.2277,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3268193304538727,
+      "learning_rate": 3.1904761904761905e-06,
+      "loss": 0.2323,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3028046488761902,
+      "learning_rate": 3.2380952380952385e-06,
+      "loss": 0.2311,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.311612069606781,
+      "learning_rate": 3.285714285714286e-06,
+      "loss": 0.2256,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32060110569000244,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2291,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30482083559036255,
+      "learning_rate": 3.3809523809523814e-06,
+      "loss": 0.2144,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.317109078168869,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 0.2292,
+      "step": 72
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30233991146087646,
+      "learning_rate": 3.476190476190476e-06,
+      "loss": 0.2187,
+      "step": 73
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3236132264137268,
+      "learning_rate": 3.523809523809524e-06,
+      "loss": 0.2116,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3278987407684326,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 0.2293,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3173205256462097,
+      "learning_rate": 3.6190476190476194e-06,
+      "loss": 0.2136,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29916858673095703,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 0.2156,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2961813509464264,
+      "learning_rate": 3.7142857142857146e-06,
+      "loss": 0.212,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2890182435512543,
+      "learning_rate": 3.761904761904762e-06,
+      "loss": 0.2155,
+      "step": 79
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3079325556755066,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.2097,
+      "step": 80
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30932968854904175,
+      "learning_rate": 3.857142857142858e-06,
+      "loss": 0.2114,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2830033302307129,
+      "learning_rate": 3.9047619047619055e-06,
+      "loss": 0.2164,
+      "step": 82
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3003292381763458,
+      "learning_rate": 3.952380952380952e-06,
+      "loss": 0.2098,
+      "step": 83
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.303975373506546,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.2042,
+      "step": 84
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2991327941417694,
+      "learning_rate": 4.047619047619048e-06,
+      "loss": 0.2036,
+      "step": 85
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31630221009254456,
+      "learning_rate": 4.095238095238096e-06,
+      "loss": 0.2159,
+      "step": 86
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28796082735061646,
+      "learning_rate": 4.1428571428571435e-06,
+      "loss": 0.1949,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31407997012138367,
+      "learning_rate": 4.190476190476191e-06,
+      "loss": 0.2101,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3446134030818939,
+      "learning_rate": 4.238095238095239e-06,
+      "loss": 0.1997,
+      "step": 89
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3045324981212616,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.2032,
+      "step": 90
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3583052158355713,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 0.21,
+      "step": 91
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2765870988368988,
+      "learning_rate": 4.3809523809523815e-06,
+      "loss": 0.2032,
+      "step": 92
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32289519906044006,
+      "learning_rate": 4.428571428571429e-06,
+      "loss": 0.2062,
+      "step": 93
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2939968705177307,
+      "learning_rate": 4.476190476190477e-06,
+      "loss": 0.2021,
+      "step": 94
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34030574560165405,
+      "learning_rate": 4.523809523809524e-06,
+      "loss": 0.2062,
+      "step": 95
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3662828207015991,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 0.2149,
+      "step": 96
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29985854029655457,
+      "learning_rate": 4.6190476190476196e-06,
+      "loss": 0.2047,
+      "step": 97
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2857242524623871,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.2032,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3669258654117584,
+      "learning_rate": 4.714285714285715e-06,
+      "loss": 0.199,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.373848021030426,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.1991,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.293122261762619,
+      "learning_rate": 4.80952380952381e-06,
+      "loss": 0.1913,
+      "step": 101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3194025158882141,
+      "learning_rate": 4.857142857142858e-06,
+      "loss": 0.2,
+      "step": 102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39969396591186523,
+      "learning_rate": 4.904761904761905e-06,
+      "loss": 0.1934,
+      "step": 103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2863703966140747,
+      "learning_rate": 4.952380952380953e-06,
+      "loss": 0.1927,
+      "step": 104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3207889497280121,
+      "learning_rate": 5e-06,
+      "loss": 0.2045,
+      "step": 105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3438229560852051,
+      "learning_rate": 4.998522458628842e-06,
+      "loss": 0.1903,
+      "step": 106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.35166507959365845,
+      "learning_rate": 4.9970449172576835e-06,
+      "loss": 0.1904,
+      "step": 107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2950548231601715,
+      "learning_rate": 4.995567375886525e-06,
+      "loss": 0.1935,
+      "step": 108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.285917192697525,
+      "learning_rate": 4.994089834515367e-06,
+      "loss": 0.1902,
+      "step": 109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3664076626300812,
+      "learning_rate": 4.992612293144208e-06,
+      "loss": 0.1854,
+      "step": 110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2948754131793976,
+      "learning_rate": 4.9911347517730505e-06,
+      "loss": 0.186,
+      "step": 111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3147653639316559,
+      "learning_rate": 4.989657210401891e-06,
+      "loss": 0.2003,
+      "step": 112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33585259318351746,
+      "learning_rate": 4.9881796690307335e-06,
+      "loss": 0.1889,
+      "step": 113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2930387854576111,
+      "learning_rate": 4.986702127659575e-06,
+      "loss": 0.1854,
+      "step": 114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3020271360874176,
+      "learning_rate": 4.985224586288417e-06,
+      "loss": 0.1927,
+      "step": 115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3940936028957367,
+      "learning_rate": 4.983747044917258e-06,
+      "loss": 0.1898,
+      "step": 116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3081377148628235,
+      "learning_rate": 4.9822695035461e-06,
+      "loss": 0.1836,
+      "step": 117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3002438545227051,
+      "learning_rate": 4.980791962174941e-06,
+      "loss": 0.1911,
+      "step": 118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3444017767906189,
+      "learning_rate": 4.979314420803783e-06,
+      "loss": 0.1986,
+      "step": 119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3336763083934784,
+      "learning_rate": 4.977836879432624e-06,
+      "loss": 0.1889,
+      "step": 120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2972318232059479,
+      "learning_rate": 4.976359338061466e-06,
+      "loss": 0.1859,
+      "step": 121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30811452865600586,
+      "learning_rate": 4.974881796690308e-06,
+      "loss": 0.1863,
+      "step": 122
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3365159332752228,
+      "learning_rate": 4.973404255319149e-06,
+      "loss": 0.1819,
+      "step": 123
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.29487839341163635,
+      "learning_rate": 4.971926713947991e-06,
+      "loss": 0.1864,
+      "step": 124
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3374158442020416,
+      "learning_rate": 4.970449172576833e-06,
+      "loss": 0.194,
+      "step": 125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35937610268592834,
+      "learning_rate": 4.968971631205674e-06,
+      "loss": 0.1873,
+      "step": 126
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35758790373802185,
+      "learning_rate": 4.967494089834516e-06,
+      "loss": 0.1945,
+      "step": 127
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3236318826675415,
+      "learning_rate": 4.966016548463357e-06,
+      "loss": 0.1888,
+      "step": 128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32535436749458313,
+      "learning_rate": 4.964539007092199e-06,
+      "loss": 0.1893,
+      "step": 129
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3190808594226837,
+      "learning_rate": 4.9630614657210404e-06,
+      "loss": 0.1954,
+      "step": 130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27906644344329834,
+      "learning_rate": 4.961583924349882e-06,
+      "loss": 0.1854,
+      "step": 131
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32667556405067444,
+      "learning_rate": 4.9601063829787235e-06,
+      "loss": 0.1906,
+      "step": 132
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3335009515285492,
+      "learning_rate": 4.958628841607566e-06,
+      "loss": 0.1865,
+      "step": 133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3046165108680725,
+      "learning_rate": 4.957151300236407e-06,
+      "loss": 0.1863,
+      "step": 134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3031184673309326,
+      "learning_rate": 4.955673758865249e-06,
+      "loss": 0.1907,
+      "step": 135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3069320619106293,
+      "learning_rate": 4.95419621749409e-06,
+      "loss": 0.1899,
+      "step": 136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32536405324935913,
+      "learning_rate": 4.952718676122932e-06,
+      "loss": 0.1925,
+      "step": 137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29151833057403564,
+      "learning_rate": 4.9512411347517736e-06,
+      "loss": 0.1861,
+      "step": 138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31043440103530884,
+      "learning_rate": 4.949763593380615e-06,
+      "loss": 0.1796,
+      "step": 139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3155137002468109,
+      "learning_rate": 4.948286052009457e-06,
+      "loss": 0.1818,
+      "step": 140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3326542377471924,
+      "learning_rate": 4.946808510638298e-06,
+      "loss": 0.1852,
+      "step": 141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3079502284526825,
+      "learning_rate": 4.94533096926714e-06,
+      "loss": 0.1888,
+      "step": 142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29642048478126526,
+      "learning_rate": 4.943853427895981e-06,
+      "loss": 0.1861,
+      "step": 143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32214003801345825,
+      "learning_rate": 4.942375886524824e-06,
+      "loss": 0.188,
+      "step": 144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2959584891796112,
+      "learning_rate": 4.940898345153664e-06,
+      "loss": 0.1831,
+      "step": 145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3107849359512329,
+      "learning_rate": 4.939420803782507e-06,
+      "loss": 0.1885,
+      "step": 146
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28456783294677734,
+      "learning_rate": 4.937943262411347e-06,
+      "loss": 0.1865,
+      "step": 147
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.29512566328048706,
+      "learning_rate": 4.93646572104019e-06,
+      "loss": 0.1834,
+      "step": 148
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35392776131629944,
+      "learning_rate": 4.934988179669031e-06,
+      "loss": 0.1812,
+      "step": 149
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3354708254337311,
+      "learning_rate": 4.933510638297873e-06,
+      "loss": 0.1846,
+      "step": 150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3243681788444519,
+      "learning_rate": 4.932033096926714e-06,
+      "loss": 0.1752,
+      "step": 151
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35909929871559143,
+      "learning_rate": 4.930555555555556e-06,
+      "loss": 0.1771,
+      "step": 152
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32351794838905334,
+      "learning_rate": 4.929078014184397e-06,
+      "loss": 0.1814,
+      "step": 153
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2922566831111908,
+      "learning_rate": 4.927600472813239e-06,
+      "loss": 0.1808,
+      "step": 154
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.29026326537132263,
+      "learning_rate": 4.9261229314420805e-06,
+      "loss": 0.1783,
+      "step": 155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3708924353122711,
+      "learning_rate": 4.924645390070922e-06,
+      "loss": 0.1795,
+      "step": 156
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3833075165748596,
+      "learning_rate": 4.923167848699764e-06,
+      "loss": 0.1806,
+      "step": 157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31332194805145264,
+      "learning_rate": 4.921690307328605e-06,
+      "loss": 0.1777,
+      "step": 158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32022660970687866,
+      "learning_rate": 4.9202127659574475e-06,
+      "loss": 0.1741,
+      "step": 159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29785865545272827,
+      "learning_rate": 4.918735224586289e-06,
+      "loss": 0.1781,
+      "step": 160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31883567571640015,
+      "learning_rate": 4.9172576832151305e-06,
+      "loss": 0.1828,
+      "step": 161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.320041686296463,
+      "learning_rate": 4.915780141843972e-06,
+      "loss": 0.1886,
+      "step": 162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32226115465164185,
+      "learning_rate": 4.914302600472814e-06,
+      "loss": 0.1821,
+      "step": 163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3467138707637787,
+      "learning_rate": 4.912825059101655e-06,
+      "loss": 0.1818,
+      "step": 164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29609978199005127,
+      "learning_rate": 4.911347517730497e-06,
+      "loss": 0.1735,
+      "step": 165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.30035585165023804,
+      "learning_rate": 4.909869976359338e-06,
+      "loss": 0.173,
+      "step": 166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3052808940410614,
+      "learning_rate": 4.90839243498818e-06,
+      "loss": 0.1806,
+      "step": 167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3003518581390381,
+      "learning_rate": 4.906914893617022e-06,
+      "loss": 0.1726,
+      "step": 168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2941701412200928,
+      "learning_rate": 4.905437352245863e-06,
+      "loss": 0.1784,
+      "step": 169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29058170318603516,
+      "learning_rate": 4.903959810874705e-06,
+      "loss": 0.1832,
+      "step": 170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2866921126842499,
+      "learning_rate": 4.902482269503547e-06,
+      "loss": 0.1782,
+      "step": 171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.286066472530365,
+      "learning_rate": 4.901004728132388e-06,
+      "loss": 0.1744,
+      "step": 172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2761238217353821,
+      "learning_rate": 4.89952718676123e-06,
+      "loss": 0.1716,
+      "step": 173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3069080114364624,
+      "learning_rate": 4.898049645390071e-06,
+      "loss": 0.1778,
+      "step": 174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2990931570529938,
+      "learning_rate": 4.896572104018913e-06,
+      "loss": 0.1737,
+      "step": 175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30002954602241516,
+      "learning_rate": 4.895094562647754e-06,
+      "loss": 0.1771,
+      "step": 176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29616260528564453,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 0.1795,
+      "step": 177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2952573895454407,
+      "learning_rate": 4.8921394799054374e-06,
+      "loss": 0.1741,
+      "step": 178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29204702377319336,
+      "learning_rate": 4.89066193853428e-06,
+      "loss": 0.1723,
+      "step": 179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2875840663909912,
+      "learning_rate": 4.8891843971631205e-06,
+      "loss": 0.1733,
+      "step": 180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2945385277271271,
+      "learning_rate": 4.887706855791963e-06,
+      "loss": 0.1701,
+      "step": 181
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3164166212081909,
+      "learning_rate": 4.886229314420804e-06,
+      "loss": 0.1748,
+      "step": 182
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2994084060192108,
+      "learning_rate": 4.884751773049646e-06,
+      "loss": 0.1778,
+      "step": 183
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31515756249427795,
+      "learning_rate": 4.8832742316784875e-06,
+      "loss": 0.1742,
+      "step": 184
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30128535628318787,
+      "learning_rate": 4.881796690307329e-06,
+      "loss": 0.1732,
+      "step": 185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3000538945198059,
+      "learning_rate": 4.8803191489361706e-06,
+      "loss": 0.1793,
+      "step": 186
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31238311529159546,
+      "learning_rate": 4.878841607565012e-06,
+      "loss": 0.1806,
+      "step": 187
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2775610685348511,
+      "learning_rate": 4.877364066193854e-06,
+      "loss": 0.1738,
+      "step": 188
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2933353781700134,
+      "learning_rate": 4.875886524822695e-06,
+      "loss": 0.1697,
+      "step": 189
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31295475363731384,
+      "learning_rate": 4.874408983451537e-06,
+      "loss": 0.178,
+      "step": 190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2728680372238159,
+      "learning_rate": 4.872931442080378e-06,
+      "loss": 0.1694,
+      "step": 191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2815025746822357,
+      "learning_rate": 4.871453900709221e-06,
+      "loss": 0.1658,
+      "step": 192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3167736232280731,
+      "learning_rate": 4.869976359338061e-06,
+      "loss": 0.1678,
+      "step": 193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3157294690608978,
+      "learning_rate": 4.868498817966904e-06,
+      "loss": 0.17,
+      "step": 194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2905581593513489,
+      "learning_rate": 4.867021276595745e-06,
+      "loss": 0.1793,
+      "step": 195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27497538924217224,
+      "learning_rate": 4.865543735224587e-06,
+      "loss": 0.1759,
+      "step": 196
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32698947191238403,
+      "learning_rate": 4.864066193853428e-06,
+      "loss": 0.1682,
+      "step": 197
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3027385473251343,
+      "learning_rate": 4.86258865248227e-06,
+      "loss": 0.174,
+      "step": 198
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30821678042411804,
+      "learning_rate": 4.861111111111111e-06,
+      "loss": 0.1739,
+      "step": 199
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3796403408050537,
+      "learning_rate": 4.859633569739953e-06,
+      "loss": 0.164,
+      "step": 200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27821823954582214,
+      "learning_rate": 4.858156028368794e-06,
+      "loss": 0.1724,
+      "step": 201
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32828235626220703,
+      "learning_rate": 4.856678486997636e-06,
+      "loss": 0.1839,
+      "step": 202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3202100694179535,
+      "learning_rate": 4.855200945626478e-06,
+      "loss": 0.1791,
+      "step": 203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31503674387931824,
+      "learning_rate": 4.853723404255319e-06,
+      "loss": 0.1626,
+      "step": 204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3114008903503418,
+      "learning_rate": 4.852245862884161e-06,
+      "loss": 0.1747,
+      "step": 205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2976803779602051,
+      "learning_rate": 4.850768321513003e-06,
+      "loss": 0.1739,
+      "step": 206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29751133918762207,
+      "learning_rate": 4.8492907801418445e-06,
+      "loss": 0.1717,
+      "step": 207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3073213994503021,
+      "learning_rate": 4.847813238770686e-06,
+      "loss": 0.1714,
+      "step": 208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31057098507881165,
+      "learning_rate": 4.8463356973995275e-06,
+      "loss": 0.1766,
+      "step": 209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.301988422870636,
+      "learning_rate": 4.844858156028369e-06,
+      "loss": 0.1747,
+      "step": 210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2876450717449188,
+      "learning_rate": 4.843380614657211e-06,
+      "loss": 0.1663,
+      "step": 211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2984127104282379,
+      "learning_rate": 4.841903073286052e-06,
+      "loss": 0.1623,
+      "step": 212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2993881106376648,
+      "learning_rate": 4.840425531914894e-06,
+      "loss": 0.1653,
+      "step": 213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2957998812198639,
+      "learning_rate": 4.838947990543736e-06,
+      "loss": 0.1719,
+      "step": 214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28871169686317444,
+      "learning_rate": 4.837470449172577e-06,
+      "loss": 0.1741,
+      "step": 215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28450289368629456,
+      "learning_rate": 4.835992907801419e-06,
+      "loss": 0.1675,
+      "step": 216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3095954358577728,
+      "learning_rate": 4.83451536643026e-06,
+      "loss": 0.1743,
+      "step": 217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28771740198135376,
+      "learning_rate": 4.833037825059102e-06,
+      "loss": 0.1699,
+      "step": 218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28795990347862244,
+      "learning_rate": 4.831560283687944e-06,
+      "loss": 0.1719,
+      "step": 219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2883548140525818,
+      "learning_rate": 4.830082742316785e-06,
+      "loss": 0.1715,
+      "step": 220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2784866988658905,
+      "learning_rate": 4.828605200945627e-06,
+      "loss": 0.1633,
+      "step": 221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2864377498626709,
+      "learning_rate": 4.827127659574468e-06,
+      "loss": 0.1723,
+      "step": 222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31789928674697876,
+      "learning_rate": 4.82565011820331e-06,
+      "loss": 0.1703,
+      "step": 223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28810226917266846,
+      "learning_rate": 4.824172576832151e-06,
+      "loss": 0.1701,
+      "step": 224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28274592757225037,
+      "learning_rate": 4.822695035460993e-06,
+      "loss": 0.1692,
+      "step": 225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.294897198677063,
+      "learning_rate": 4.8212174940898344e-06,
+      "loss": 0.1643,
+      "step": 226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29387393593788147,
+      "learning_rate": 4.819739952718677e-06,
+      "loss": 0.1741,
+      "step": 227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3175782561302185,
+      "learning_rate": 4.8182624113475175e-06,
+      "loss": 0.1684,
+      "step": 228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3241492211818695,
+      "learning_rate": 4.81678486997636e-06,
+      "loss": 0.1712,
+      "step": 229
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28357645869255066,
+      "learning_rate": 4.8153073286052014e-06,
+      "loss": 0.1657,
+      "step": 230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3193873167037964,
+      "learning_rate": 4.813829787234043e-06,
+      "loss": 0.1693,
+      "step": 231
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2756381034851074,
+      "learning_rate": 4.8123522458628845e-06,
+      "loss": 0.1583,
+      "step": 232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26709064841270447,
+      "learning_rate": 4.810874704491726e-06,
+      "loss": 0.1619,
+      "step": 233
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2902539372444153,
+      "learning_rate": 4.8093971631205676e-06,
+      "loss": 0.1664,
+      "step": 234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30566853284835815,
+      "learning_rate": 4.807919621749409e-06,
+      "loss": 0.1678,
+      "step": 235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30061808228492737,
+      "learning_rate": 4.806442080378251e-06,
+      "loss": 0.1666,
+      "step": 236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27073654532432556,
+      "learning_rate": 4.804964539007093e-06,
+      "loss": 0.1621,
+      "step": 237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27716368436813354,
+      "learning_rate": 4.8034869976359345e-06,
+      "loss": 0.1662,
+      "step": 238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30569934844970703,
+      "learning_rate": 4.802009456264776e-06,
+      "loss": 0.1781,
+      "step": 239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2923128306865692,
+      "learning_rate": 4.800531914893618e-06,
+      "loss": 0.1598,
+      "step": 240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33011940121650696,
+      "learning_rate": 4.799054373522459e-06,
+      "loss": 0.1631,
+      "step": 241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.32699519395828247,
+      "learning_rate": 4.797576832151301e-06,
+      "loss": 0.1625,
+      "step": 242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3068077266216278,
+      "learning_rate": 4.796099290780142e-06,
+      "loss": 0.1588,
+      "step": 243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2872140407562256,
+      "learning_rate": 4.794621749408984e-06,
+      "loss": 0.1724,
+      "step": 244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28918853402137756,
+      "learning_rate": 4.793144208037825e-06,
+      "loss": 0.1682,
+      "step": 245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29438456892967224,
+      "learning_rate": 4.791666666666668e-06,
+      "loss": 0.168,
+      "step": 246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31224462389945984,
+      "learning_rate": 4.790189125295508e-06,
+      "loss": 0.1703,
+      "step": 247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2956463396549225,
+      "learning_rate": 4.788711583924351e-06,
+      "loss": 0.1631,
+      "step": 248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2723066508769989,
+      "learning_rate": 4.787234042553192e-06,
+      "loss": 0.1559,
+      "step": 249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2921085059642792,
+      "learning_rate": 4.785756501182034e-06,
+      "loss": 0.1677,
+      "step": 250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27501416206359863,
+      "learning_rate": 4.784278959810875e-06,
+      "loss": 0.1585,
+      "step": 251
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28951090574264526,
+      "learning_rate": 4.782801418439717e-06,
+      "loss": 0.1668,
+      "step": 252
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3022778630256653,
+      "learning_rate": 4.781323877068558e-06,
+      "loss": 0.1695,
+      "step": 253
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3108287751674652,
+      "learning_rate": 4.7798463356974e-06,
+      "loss": 0.1674,
+      "step": 254
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30567386746406555,
+      "learning_rate": 4.7783687943262415e-06,
+      "loss": 0.172,
+      "step": 255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3010265529155731,
+      "learning_rate": 4.776891252955083e-06,
+      "loss": 0.162,
+      "step": 256
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3006778657436371,
+      "learning_rate": 4.775413711583925e-06,
+      "loss": 0.1631,
+      "step": 257
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2913971245288849,
+      "learning_rate": 4.773936170212766e-06,
+      "loss": 0.1596,
+      "step": 258
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26215067505836487,
+      "learning_rate": 4.7724586288416084e-06,
+      "loss": 0.1575,
+      "step": 259
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3001871705055237,
+      "learning_rate": 4.770981087470449e-06,
+      "loss": 0.1695,
+      "step": 260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29229936003685,
+      "learning_rate": 4.7695035460992915e-06,
+      "loss": 0.1587,
+      "step": 261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.281704306602478,
+      "learning_rate": 4.768026004728133e-06,
+      "loss": 0.161,
+      "step": 262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2904156744480133,
+      "learning_rate": 4.766548463356975e-06,
+      "loss": 0.1656,
+      "step": 263
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29582810401916504,
+      "learning_rate": 4.765070921985816e-06,
+      "loss": 0.1614,
+      "step": 264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2894721031188965,
+      "learning_rate": 4.763593380614658e-06,
+      "loss": 0.1635,
+      "step": 265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29529014229774475,
+      "learning_rate": 4.762115839243499e-06,
+      "loss": 0.1623,
+      "step": 266
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3229948878288269,
+      "learning_rate": 4.760638297872341e-06,
+      "loss": 0.1683,
+      "step": 267
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2796555161476135,
+      "learning_rate": 4.759160756501183e-06,
+      "loss": 0.1549,
+      "step": 268
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28925955295562744,
+      "learning_rate": 4.757683215130024e-06,
+      "loss": 0.1633,
+      "step": 269
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3275362551212311,
+      "learning_rate": 4.756205673758866e-06,
+      "loss": 0.1609,
+      "step": 270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30417779088020325,
+      "learning_rate": 4.754728132387707e-06,
+      "loss": 0.1638,
+      "step": 271
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28757476806640625,
+      "learning_rate": 4.753250591016549e-06,
+      "loss": 0.1547,
+      "step": 272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30827224254608154,
+      "learning_rate": 4.751773049645391e-06,
+      "loss": 0.1584,
+      "step": 273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2874588668346405,
+      "learning_rate": 4.750295508274232e-06,
+      "loss": 0.1583,
+      "step": 274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3285459578037262,
+      "learning_rate": 4.748817966903074e-06,
+      "loss": 0.1616,
+      "step": 275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31066420674324036,
+      "learning_rate": 4.747340425531915e-06,
+      "loss": 0.1694,
+      "step": 276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29990261793136597,
+      "learning_rate": 4.745862884160757e-06,
+      "loss": 0.163,
+      "step": 277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3014674484729767,
+      "learning_rate": 4.7443853427895984e-06,
+      "loss": 0.1692,
+      "step": 278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2811316251754761,
+      "learning_rate": 4.74290780141844e-06,
+      "loss": 0.1549,
+      "step": 279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30627551674842834,
+      "learning_rate": 4.7414302600472815e-06,
+      "loss": 0.1615,
+      "step": 280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3162028193473816,
+      "learning_rate": 4.739952718676124e-06,
+      "loss": 0.157,
+      "step": 281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29642122983932495,
+      "learning_rate": 4.7384751773049646e-06,
+      "loss": 0.1602,
+      "step": 282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3453240692615509,
+      "learning_rate": 4.736997635933807e-06,
+      "loss": 0.1685,
+      "step": 283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31275737285614014,
+      "learning_rate": 4.7355200945626485e-06,
+      "loss": 0.1628,
+      "step": 284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30649885535240173,
+      "learning_rate": 4.73404255319149e-06,
+      "loss": 0.1568,
+      "step": 285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2804078161716461,
+      "learning_rate": 4.7325650118203315e-06,
+      "loss": 0.1569,
+      "step": 286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3204512894153595,
+      "learning_rate": 4.731087470449173e-06,
+      "loss": 0.1567,
+      "step": 287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2934107780456543,
+      "learning_rate": 4.729609929078015e-06,
+      "loss": 0.1639,
+      "step": 288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30983462929725647,
+      "learning_rate": 4.728132387706856e-06,
+      "loss": 0.1667,
+      "step": 289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3167484700679779,
+      "learning_rate": 4.726654846335698e-06,
+      "loss": 0.1564,
+      "step": 290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28580886125564575,
+      "learning_rate": 4.725177304964539e-06,
+      "loss": 0.1601,
+      "step": 291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2920008897781372,
+      "learning_rate": 4.723699763593382e-06,
+      "loss": 0.1604,
+      "step": 292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3266066610813141,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 0.1539,
+      "step": 293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3400053083896637,
+      "learning_rate": 4.720744680851065e-06,
+      "loss": 0.1599,
+      "step": 294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29672902822494507,
+      "learning_rate": 4.719267139479905e-06,
+      "loss": 0.1677,
+      "step": 295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2885327637195587,
+      "learning_rate": 4.717789598108748e-06,
+      "loss": 0.1611,
+      "step": 296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31177258491516113,
+      "learning_rate": 4.716312056737589e-06,
+      "loss": 0.1601,
+      "step": 297
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31969717144966125,
+      "learning_rate": 4.714834515366431e-06,
+      "loss": 0.163,
+      "step": 298
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3055081069469452,
+      "learning_rate": 4.713356973995272e-06,
+      "loss": 0.1547,
+      "step": 299
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28702032566070557,
+      "learning_rate": 4.711879432624114e-06,
+      "loss": 0.1637,
+      "step": 300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3091968595981598,
+      "learning_rate": 4.710401891252955e-06,
+      "loss": 0.1667,
+      "step": 301
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30867868661880493,
+      "learning_rate": 4.708924349881797e-06,
+      "loss": 0.1623,
+      "step": 302
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29452067613601685,
+      "learning_rate": 4.707446808510639e-06,
+      "loss": 0.162,
+      "step": 303
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30876031517982483,
+      "learning_rate": 4.70596926713948e-06,
+      "loss": 0.1626,
+      "step": 304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29180067777633667,
+      "learning_rate": 4.704491725768322e-06,
+      "loss": 0.1564,
+      "step": 305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3025527894496918,
+      "learning_rate": 4.703014184397163e-06,
+      "loss": 0.1598,
+      "step": 306
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3138538599014282,
+      "learning_rate": 4.7015366430260054e-06,
+      "loss": 0.1612,
+      "step": 307
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3038712739944458,
+      "learning_rate": 4.700059101654847e-06,
+      "loss": 0.1538,
+      "step": 308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3020678758621216,
+      "learning_rate": 4.6985815602836885e-06,
+      "loss": 0.1605,
+      "step": 309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32480180263519287,
+      "learning_rate": 4.69710401891253e-06,
+      "loss": 0.1576,
+      "step": 310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.326655775308609,
+      "learning_rate": 4.695626477541372e-06,
+      "loss": 0.1613,
+      "step": 311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3322456181049347,
+      "learning_rate": 4.694148936170213e-06,
+      "loss": 0.1583,
+      "step": 312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33443212509155273,
+      "learning_rate": 4.692671394799055e-06,
+      "loss": 0.1705,
+      "step": 313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30078256130218506,
+      "learning_rate": 4.691193853427896e-06,
+      "loss": 0.1599,
+      "step": 314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3066742420196533,
+      "learning_rate": 4.689716312056738e-06,
+      "loss": 0.156,
+      "step": 315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3422672152519226,
+      "learning_rate": 4.68823877068558e-06,
+      "loss": 0.1581,
+      "step": 316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3084907829761505,
+      "learning_rate": 4.686761229314421e-06,
+      "loss": 0.1558,
+      "step": 317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2913210391998291,
+      "learning_rate": 4.685283687943263e-06,
+      "loss": 0.1639,
+      "step": 318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31148403882980347,
+      "learning_rate": 4.683806146572105e-06,
+      "loss": 0.1718,
+      "step": 319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.33954447507858276,
+      "learning_rate": 4.682328605200946e-06,
+      "loss": 0.1663,
+      "step": 320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2902482748031616,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 0.1569,
+      "step": 321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.287404865026474,
+      "learning_rate": 4.679373522458629e-06,
+      "loss": 0.1571,
+      "step": 322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27859029173851013,
+      "learning_rate": 4.677895981087471e-06,
+      "loss": 0.1559,
+      "step": 323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30411607027053833,
+      "learning_rate": 4.676418439716312e-06,
+      "loss": 0.1553,
+      "step": 324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3301770091056824,
+      "learning_rate": 4.674940898345154e-06,
+      "loss": 0.1608,
+      "step": 325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30002135038375854,
+      "learning_rate": 4.6734633569739954e-06,
+      "loss": 0.1576,
+      "step": 326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29414889216423035,
+      "learning_rate": 4.671985815602838e-06,
+      "loss": 0.1627,
+      "step": 327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28249481320381165,
+      "learning_rate": 4.6705082742316785e-06,
+      "loss": 0.1514,
+      "step": 328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28770700097084045,
+      "learning_rate": 4.669030732860521e-06,
+      "loss": 0.152,
+      "step": 329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30502554774284363,
+      "learning_rate": 4.6675531914893616e-06,
+      "loss": 0.1577,
+      "step": 330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28528234362602234,
+      "learning_rate": 4.666075650118204e-06,
+      "loss": 0.1591,
+      "step": 331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.293910413980484,
+      "learning_rate": 4.6645981087470455e-06,
+      "loss": 0.1589,
+      "step": 332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.294941246509552,
+      "learning_rate": 4.663120567375887e-06,
+      "loss": 0.1611,
+      "step": 333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31550052762031555,
+      "learning_rate": 4.6616430260047285e-06,
+      "loss": 0.1612,
+      "step": 334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2951798439025879,
+      "learning_rate": 4.66016548463357e-06,
+      "loss": 0.1513,
+      "step": 335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29904624819755554,
+      "learning_rate": 4.658687943262412e-06,
+      "loss": 0.1611,
+      "step": 336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29253438115119934,
+      "learning_rate": 4.657210401891253e-06,
+      "loss": 0.1589,
+      "step": 337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28482910990715027,
+      "learning_rate": 4.6557328605200955e-06,
+      "loss": 0.1577,
+      "step": 338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28554195165634155,
+      "learning_rate": 4.654255319148936e-06,
+      "loss": 0.1592,
+      "step": 339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2944633960723877,
+      "learning_rate": 4.652777777777779e-06,
+      "loss": 0.1575,
+      "step": 340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31336843967437744,
+      "learning_rate": 4.651300236406619e-06,
+      "loss": 0.1579,
+      "step": 341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2994990050792694,
+      "learning_rate": 4.649822695035462e-06,
+      "loss": 0.1548,
+      "step": 342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2991165816783905,
+      "learning_rate": 4.648345153664303e-06,
+      "loss": 0.1579,
+      "step": 343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2960761785507202,
+      "learning_rate": 4.646867612293145e-06,
+      "loss": 0.1566,
+      "step": 344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2933996021747589,
+      "learning_rate": 4.645390070921986e-06,
+      "loss": 0.1517,
+      "step": 345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3018741011619568,
+      "learning_rate": 4.643912529550828e-06,
+      "loss": 0.1499,
+      "step": 346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2859538793563843,
+      "learning_rate": 4.642434988179669e-06,
+      "loss": 0.1542,
+      "step": 347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2974526286125183,
+      "learning_rate": 4.640957446808511e-06,
+      "loss": 0.1559,
+      "step": 348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3103669583797455,
+      "learning_rate": 4.639479905437352e-06,
+      "loss": 0.1619,
+      "step": 349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3094070553779602,
+      "learning_rate": 4.638002364066194e-06,
+      "loss": 0.1599,
+      "step": 350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27510109543800354,
+      "learning_rate": 4.636524822695036e-06,
+      "loss": 0.1545,
+      "step": 351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3178057074546814,
+      "learning_rate": 4.635047281323877e-06,
+      "loss": 0.1575,
+      "step": 352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28596967458724976,
+      "learning_rate": 4.633569739952719e-06,
+      "loss": 0.155,
+      "step": 353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2940841615200043,
+      "learning_rate": 4.632092198581561e-06,
+      "loss": 0.1548,
+      "step": 354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28446516394615173,
+      "learning_rate": 4.6306146572104024e-06,
+      "loss": 0.1528,
+      "step": 355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30511361360549927,
+      "learning_rate": 4.629137115839244e-06,
+      "loss": 0.1555,
+      "step": 356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3373614549636841,
+      "learning_rate": 4.6276595744680855e-06,
+      "loss": 0.1558,
+      "step": 357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2949771583080292,
+      "learning_rate": 4.626182033096927e-06,
+      "loss": 0.1591,
+      "step": 358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3137626349925995,
+      "learning_rate": 4.624704491725769e-06,
+      "loss": 0.1624,
+      "step": 359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35294073820114136,
+      "learning_rate": 4.62322695035461e-06,
+      "loss": 0.1584,
+      "step": 360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3194246292114258,
+      "learning_rate": 4.621749408983452e-06,
+      "loss": 0.1532,
+      "step": 361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3209877908229828,
+      "learning_rate": 4.620271867612294e-06,
+      "loss": 0.1481,
+      "step": 362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29048117995262146,
+      "learning_rate": 4.618794326241135e-06,
+      "loss": 0.157,
+      "step": 363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3318732976913452,
+      "learning_rate": 4.617316784869977e-06,
+      "loss": 0.1627,
+      "step": 364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32037222385406494,
+      "learning_rate": 4.615839243498818e-06,
+      "loss": 0.1596,
+      "step": 365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3264883756637573,
+      "learning_rate": 4.61436170212766e-06,
+      "loss": 0.1584,
+      "step": 366
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2828350067138672,
+      "learning_rate": 4.612884160756502e-06,
+      "loss": 0.1598,
+      "step": 367
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3008638620376587,
+      "learning_rate": 4.611406619385343e-06,
+      "loss": 0.1514,
+      "step": 368
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2930569648742676,
+      "learning_rate": 4.609929078014185e-06,
+      "loss": 0.1516,
+      "step": 369
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27801671624183655,
+      "learning_rate": 4.608451536643026e-06,
+      "loss": 0.1514,
+      "step": 370
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2964909076690674,
+      "learning_rate": 4.606973995271868e-06,
+      "loss": 0.1511,
+      "step": 371
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29584363102912903,
+      "learning_rate": 4.605496453900709e-06,
+      "loss": 0.1593,
+      "step": 372
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29735302925109863,
+      "learning_rate": 4.604018912529552e-06,
+      "loss": 0.1518,
+      "step": 373
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31003421545028687,
+      "learning_rate": 4.6025413711583924e-06,
+      "loss": 0.1521,
+      "step": 374
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29468318819999695,
+      "learning_rate": 4.601063829787235e-06,
+      "loss": 0.1525,
+      "step": 375
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3082709014415741,
+      "learning_rate": 4.5995862884160755e-06,
+      "loss": 0.1535,
+      "step": 376
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30345532298088074,
+      "learning_rate": 4.598108747044918e-06,
+      "loss": 0.1492,
+      "step": 377
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2913258969783783,
+      "learning_rate": 4.596631205673759e-06,
+      "loss": 0.1507,
+      "step": 378
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31036365032196045,
+      "learning_rate": 4.595153664302601e-06,
+      "loss": 0.1549,
+      "step": 379
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3078053295612335,
+      "learning_rate": 4.5936761229314425e-06,
+      "loss": 0.1529,
+      "step": 380
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27835437655448914,
+      "learning_rate": 4.592198581560284e-06,
+      "loss": 0.1534,
+      "step": 381
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3037719130516052,
+      "learning_rate": 4.5907210401891255e-06,
+      "loss": 0.1584,
+      "step": 382
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2875194251537323,
+      "learning_rate": 4.589243498817967e-06,
+      "loss": 0.1496,
+      "step": 383
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3025689125061035,
+      "learning_rate": 4.587765957446809e-06,
+      "loss": 0.1523,
+      "step": 384
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30861130356788635,
+      "learning_rate": 4.58628841607565e-06,
+      "loss": 0.1549,
+      "step": 385
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2984231412410736,
+      "learning_rate": 4.5848108747044925e-06,
+      "loss": 0.1538,
+      "step": 386
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2919422686100006,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.1612,
+      "step": 387
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2814089357852936,
+      "learning_rate": 4.581855791962176e-06,
+      "loss": 0.1507,
+      "step": 388
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32090383768081665,
+      "learning_rate": 4.580378250591017e-06,
+      "loss": 0.155,
+      "step": 389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27118828892707825,
+      "learning_rate": 4.578900709219859e-06,
+      "loss": 0.1564,
+      "step": 390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28908324241638184,
+      "learning_rate": 4.5774231678487e-06,
+      "loss": 0.1517,
+      "step": 391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31309959292411804,
+      "learning_rate": 4.575945626477542e-06,
+      "loss": 0.15,
+      "step": 392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29398292303085327,
+      "learning_rate": 4.574468085106383e-06,
+      "loss": 0.1489,
+      "step": 393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30536139011383057,
+      "learning_rate": 4.572990543735225e-06,
+      "loss": 0.1611,
+      "step": 394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3120265305042267,
+      "learning_rate": 4.571513002364066e-06,
+      "loss": 0.1629,
+      "step": 395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29656338691711426,
+      "learning_rate": 4.570035460992908e-06,
+      "loss": 0.1534,
+      "step": 396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3127426505088806,
+      "learning_rate": 4.56855791962175e-06,
+      "loss": 0.153,
+      "step": 397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28530940413475037,
+      "learning_rate": 4.567080378250591e-06,
+      "loss": 0.1517,
+      "step": 398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31527185440063477,
+      "learning_rate": 4.565602836879433e-06,
+      "loss": 0.156,
+      "step": 399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30909857153892517,
+      "learning_rate": 4.564125295508275e-06,
+      "loss": 0.1518,
+      "step": 400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28175467252731323,
+      "learning_rate": 4.562647754137116e-06,
+      "loss": 0.1478,
+      "step": 401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.318112313747406,
+      "learning_rate": 4.561170212765958e-06,
+      "loss": 0.146,
+      "step": 402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2834266126155853,
+      "learning_rate": 4.5596926713947994e-06,
+      "loss": 0.1506,
+      "step": 403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2920738160610199,
+      "learning_rate": 4.558215130023641e-06,
+      "loss": 0.1531,
+      "step": 404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3048083782196045,
+      "learning_rate": 4.5567375886524825e-06,
+      "loss": 0.1521,
+      "step": 405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3016625642776489,
+      "learning_rate": 4.555260047281324e-06,
+      "loss": 0.157,
+      "step": 406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31861746311187744,
+      "learning_rate": 4.553782505910166e-06,
+      "loss": 0.1574,
+      "step": 407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2838152050971985,
+      "learning_rate": 4.552304964539008e-06,
+      "loss": 0.1603,
+      "step": 408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3126676082611084,
+      "learning_rate": 4.550827423167849e-06,
+      "loss": 0.1518,
+      "step": 409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31386828422546387,
+      "learning_rate": 4.549349881796691e-06,
+      "loss": 0.1526,
+      "step": 410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2845028042793274,
+      "learning_rate": 4.547872340425532e-06,
+      "loss": 0.1451,
+      "step": 411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29181551933288574,
+      "learning_rate": 4.546394799054374e-06,
+      "loss": 0.157,
+      "step": 412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29376184940338135,
+      "learning_rate": 4.544917257683216e-06,
+      "loss": 0.1442,
+      "step": 413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28759056329727173,
+      "learning_rate": 4.543439716312057e-06,
+      "loss": 0.1547,
+      "step": 414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.294232577085495,
+      "learning_rate": 4.541962174940899e-06,
+      "loss": 0.1544,
+      "step": 415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2921401858329773,
+      "learning_rate": 4.54048463356974e-06,
+      "loss": 0.1417,
+      "step": 416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2905738949775696,
+      "learning_rate": 4.539007092198582e-06,
+      "loss": 0.155,
+      "step": 417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29432037472724915,
+      "learning_rate": 4.537529550827423e-06,
+      "loss": 0.148,
+      "step": 418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28889697790145874,
+      "learning_rate": 4.536052009456265e-06,
+      "loss": 0.1495,
+      "step": 419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2967394292354584,
+      "learning_rate": 4.534574468085106e-06,
+      "loss": 0.1438,
+      "step": 420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2960224151611328,
+      "learning_rate": 4.533096926713949e-06,
+      "loss": 0.1466,
+      "step": 421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28826141357421875,
+      "learning_rate": 4.5316193853427894e-06,
+      "loss": 0.1485,
+      "step": 422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.290597528219223,
+      "learning_rate": 4.530141843971632e-06,
+      "loss": 0.151,
+      "step": 423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2847577631473541,
+      "learning_rate": 4.528664302600473e-06,
+      "loss": 0.1501,
+      "step": 424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2888893485069275,
+      "learning_rate": 4.527186761229315e-06,
+      "loss": 0.1481,
+      "step": 425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29155224561691284,
+      "learning_rate": 4.525709219858156e-06,
+      "loss": 0.1461,
+      "step": 426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2908017039299011,
+      "learning_rate": 4.524231678486998e-06,
+      "loss": 0.1526,
+      "step": 427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2779986560344696,
+      "learning_rate": 4.5227541371158395e-06,
+      "loss": 0.1486,
+      "step": 428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3022577166557312,
+      "learning_rate": 4.521276595744681e-06,
+      "loss": 0.1591,
+      "step": 429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28442084789276123,
+      "learning_rate": 4.5197990543735225e-06,
+      "loss": 0.1541,
+      "step": 430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30304408073425293,
+      "learning_rate": 4.518321513002364e-06,
+      "loss": 0.1594,
+      "step": 431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3168414831161499,
+      "learning_rate": 4.5168439716312065e-06,
+      "loss": 0.1547,
+      "step": 432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29197704792022705,
+      "learning_rate": 4.515366430260047e-06,
+      "loss": 0.1538,
+      "step": 433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.31468144059181213,
+      "learning_rate": 4.5138888888888895e-06,
+      "loss": 0.1512,
+      "step": 434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2896561920642853,
+      "learning_rate": 4.512411347517731e-06,
+      "loss": 0.1463,
+      "step": 435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30411049723625183,
+      "learning_rate": 4.510933806146573e-06,
+      "loss": 0.1513,
+      "step": 436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2907703220844269,
+      "learning_rate": 4.509456264775414e-06,
+      "loss": 0.1461,
+      "step": 437
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2915017604827881,
+      "learning_rate": 4.507978723404256e-06,
+      "loss": 0.1458,
+      "step": 438
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28700292110443115,
+      "learning_rate": 4.506501182033097e-06,
+      "loss": 0.1464,
+      "step": 439
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3054389953613281,
+      "learning_rate": 4.505023640661939e-06,
+      "loss": 0.1493,
+      "step": 440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31200307607650757,
+      "learning_rate": 4.50354609929078e-06,
+      "loss": 0.1512,
+      "step": 441
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3107149004936218,
+      "learning_rate": 4.502068557919622e-06,
+      "loss": 0.1508,
+      "step": 442
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30153799057006836,
+      "learning_rate": 4.500591016548464e-06,
+      "loss": 0.1532,
+      "step": 443
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3003308176994324,
+      "learning_rate": 4.499113475177305e-06,
+      "loss": 0.1562,
+      "step": 444
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3179265260696411,
+      "learning_rate": 4.497635933806147e-06,
+      "loss": 0.156,
+      "step": 445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28566238284111023,
+      "learning_rate": 4.496158392434988e-06,
+      "loss": 0.1488,
+      "step": 446
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28496724367141724,
+      "learning_rate": 4.49468085106383e-06,
+      "loss": 0.1531,
+      "step": 447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31209829449653625,
+      "learning_rate": 4.493203309692672e-06,
+      "loss": 0.1433,
+      "step": 448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32855620980262756,
+      "learning_rate": 4.491725768321513e-06,
+      "loss": 0.1474,
+      "step": 449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3074558973312378,
+      "learning_rate": 4.490248226950355e-06,
+      "loss": 0.1443,
+      "step": 450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2998528480529785,
+      "learning_rate": 4.4887706855791964e-06,
+      "loss": 0.1522,
+      "step": 451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28594866394996643,
+      "learning_rate": 4.487293144208038e-06,
+      "loss": 0.1457,
+      "step": 452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3104887902736664,
+      "learning_rate": 4.4858156028368795e-06,
+      "loss": 0.1469,
+      "step": 453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30967381596565247,
+      "learning_rate": 4.484338061465721e-06,
+      "loss": 0.1488,
+      "step": 454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2705971300601959,
+      "learning_rate": 4.482860520094563e-06,
+      "loss": 0.1434,
+      "step": 455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3046035170555115,
+      "learning_rate": 4.481382978723405e-06,
+      "loss": 0.1481,
+      "step": 456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29295024275779724,
+      "learning_rate": 4.479905437352246e-06,
+      "loss": 0.1537,
+      "step": 457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2762710452079773,
+      "learning_rate": 4.478427895981088e-06,
+      "loss": 0.147,
+      "step": 458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.289006769657135,
+      "learning_rate": 4.4769503546099296e-06,
+      "loss": 0.1482,
+      "step": 459
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3073202967643738,
+      "learning_rate": 4.475472813238771e-06,
+      "loss": 0.1516,
+      "step": 460
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3140169084072113,
+      "learning_rate": 4.473995271867613e-06,
+      "loss": 0.1512,
+      "step": 461
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28994452953338623,
+      "learning_rate": 4.472517730496454e-06,
+      "loss": 0.1471,
+      "step": 462
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29187649488449097,
+      "learning_rate": 4.471040189125296e-06,
+      "loss": 0.1478,
+      "step": 463
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29420313239097595,
+      "learning_rate": 4.469562647754137e-06,
+      "loss": 0.1477,
+      "step": 464
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28859058022499084,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 0.1499,
+      "step": 465
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2996075451374054,
+      "learning_rate": 4.46660756501182e-06,
+      "loss": 0.1507,
+      "step": 466
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2754644453525543,
+      "learning_rate": 4.465130023640663e-06,
+      "loss": 0.1455,
+      "step": 467
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29327189922332764,
+      "learning_rate": 4.463652482269503e-06,
+      "loss": 0.1483,
+      "step": 468
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2946658730506897,
+      "learning_rate": 4.462174940898346e-06,
+      "loss": 0.1537,
+      "step": 469
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2911798357963562,
+      "learning_rate": 4.460697399527187e-06,
+      "loss": 0.1503,
+      "step": 470
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2931766211986542,
+      "learning_rate": 4.459219858156029e-06,
+      "loss": 0.1472,
+      "step": 471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30041730403900146,
+      "learning_rate": 4.45774231678487e-06,
+      "loss": 0.1514,
+      "step": 472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28402015566825867,
+      "learning_rate": 4.456264775413712e-06,
+      "loss": 0.146,
+      "step": 473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29181718826293945,
+      "learning_rate": 4.454787234042553e-06,
+      "loss": 0.1484,
+      "step": 474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3052833676338196,
+      "learning_rate": 4.453309692671395e-06,
+      "loss": 0.1509,
+      "step": 475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28091058135032654,
+      "learning_rate": 4.4518321513002365e-06,
+      "loss": 0.1529,
+      "step": 476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30159497261047363,
+      "learning_rate": 4.450354609929078e-06,
+      "loss": 0.1508,
+      "step": 477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.274497389793396,
+      "learning_rate": 4.44887706855792e-06,
+      "loss": 0.1468,
+      "step": 478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30651533603668213,
+      "learning_rate": 4.447399527186761e-06,
+      "loss": 0.1502,
+      "step": 479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3123912811279297,
+      "learning_rate": 4.4459219858156035e-06,
+      "loss": 0.1474,
+      "step": 480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28743690252304077,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.1516,
+      "step": 481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2844238579273224,
+      "learning_rate": 4.4429669030732865e-06,
+      "loss": 0.1405,
+      "step": 482
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29450908303260803,
+      "learning_rate": 4.441489361702128e-06,
+      "loss": 0.15,
+      "step": 483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.308852881193161,
+      "learning_rate": 4.44001182033097e-06,
+      "loss": 0.1523,
+      "step": 484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2966323494911194,
+      "learning_rate": 4.438534278959811e-06,
+      "loss": 0.1426,
+      "step": 485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31234011054039,
+      "learning_rate": 4.437056737588653e-06,
+      "loss": 0.1511,
+      "step": 486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.295285701751709,
+      "learning_rate": 4.435579196217494e-06,
+      "loss": 0.1471,
+      "step": 487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29665645956993103,
+      "learning_rate": 4.434101654846336e-06,
+      "loss": 0.1508,
+      "step": 488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2952776253223419,
+      "learning_rate": 4.432624113475177e-06,
+      "loss": 0.146,
+      "step": 489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2941940128803253,
+      "learning_rate": 4.431146572104019e-06,
+      "loss": 0.1491,
+      "step": 490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2857465147972107,
+      "learning_rate": 4.429669030732861e-06,
+      "loss": 0.1519,
+      "step": 491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3003840744495392,
+      "learning_rate": 4.428191489361702e-06,
+      "loss": 0.1484,
+      "step": 492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27938753366470337,
+      "learning_rate": 4.426713947990544e-06,
+      "loss": 0.1471,
+      "step": 493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30460497736930847,
+      "learning_rate": 4.425236406619386e-06,
+      "loss": 0.1435,
+      "step": 494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2832207679748535,
+      "learning_rate": 4.423758865248227e-06,
+      "loss": 0.1442,
+      "step": 495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30536526441574097,
+      "learning_rate": 4.422281323877069e-06,
+      "loss": 0.1517,
+      "step": 496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29736849665641785,
+      "learning_rate": 4.42080378250591e-06,
+      "loss": 0.15,
+      "step": 497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2862701416015625,
+      "learning_rate": 4.419326241134752e-06,
+      "loss": 0.1451,
+      "step": 498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2746068239212036,
+      "learning_rate": 4.417848699763594e-06,
+      "loss": 0.1403,
+      "step": 499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3144620954990387,
+      "learning_rate": 4.416371158392435e-06,
+      "loss": 0.1522,
+      "step": 500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2791362404823303,
+      "learning_rate": 4.414893617021277e-06,
+      "loss": 0.1415,
+      "step": 501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2950166165828705,
+      "learning_rate": 4.413416075650119e-06,
+      "loss": 0.1514,
+      "step": 502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30711832642555237,
+      "learning_rate": 4.4119385342789604e-06,
+      "loss": 0.1468,
+      "step": 503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.290014386177063,
+      "learning_rate": 4.410460992907802e-06,
+      "loss": 0.1519,
+      "step": 504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27445051074028015,
+      "learning_rate": 4.4089834515366435e-06,
+      "loss": 0.1399,
+      "step": 505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28429850935935974,
+      "learning_rate": 4.407505910165485e-06,
+      "loss": 0.1474,
+      "step": 506
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29753240942955017,
+      "learning_rate": 4.4060283687943266e-06,
+      "loss": 0.1485,
+      "step": 507
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2955227792263031,
+      "learning_rate": 4.404550827423168e-06,
+      "loss": 0.1526,
+      "step": 508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29859885573387146,
+      "learning_rate": 4.40307328605201e-06,
+      "loss": 0.151,
+      "step": 509
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27583009004592896,
+      "learning_rate": 4.401595744680852e-06,
+      "loss": 0.147,
+      "step": 510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28273990750312805,
+      "learning_rate": 4.400118203309693e-06,
+      "loss": 0.15,
+      "step": 511
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31554752588272095,
+      "learning_rate": 4.398640661938535e-06,
+      "loss": 0.1521,
+      "step": 512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3001977801322937,
+      "learning_rate": 4.397163120567377e-06,
+      "loss": 0.1483,
+      "step": 513
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30147096514701843,
+      "learning_rate": 4.395685579196218e-06,
+      "loss": 0.1487,
+      "step": 514
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28131401538848877,
+      "learning_rate": 4.39420803782506e-06,
+      "loss": 0.1475,
+      "step": 515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2952650487422943,
+      "learning_rate": 4.392730496453901e-06,
+      "loss": 0.1489,
+      "step": 516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2945556938648224,
+      "learning_rate": 4.391252955082743e-06,
+      "loss": 0.1395,
+      "step": 517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29689210653305054,
+      "learning_rate": 4.389775413711584e-06,
+      "loss": 0.1467,
+      "step": 518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3008095622062683,
+      "learning_rate": 4.388297872340426e-06,
+      "loss": 0.149,
+      "step": 519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28883975744247437,
+      "learning_rate": 4.386820330969267e-06,
+      "loss": 0.1478,
+      "step": 520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31527507305145264,
+      "learning_rate": 4.38534278959811e-06,
+      "loss": 0.1438,
+      "step": 521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3065716624259949,
+      "learning_rate": 4.38386524822695e-06,
+      "loss": 0.1502,
+      "step": 522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3094262182712555,
+      "learning_rate": 4.382387706855793e-06,
+      "loss": 0.1463,
+      "step": 523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28747671842575073,
+      "learning_rate": 4.3809101654846335e-06,
+      "loss": 0.1433,
+      "step": 524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3082694113254547,
+      "learning_rate": 4.379432624113476e-06,
+      "loss": 0.1403,
+      "step": 525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.283964604139328,
+      "learning_rate": 4.377955082742317e-06,
+      "loss": 0.1415,
+      "step": 526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29084843397140503,
+      "learning_rate": 4.376477541371159e-06,
+      "loss": 0.153,
+      "step": 527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29000577330589294,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.145,
+      "step": 528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2643091678619385,
+      "learning_rate": 4.373522458628842e-06,
+      "loss": 0.1436,
+      "step": 529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2977960407733917,
+      "learning_rate": 4.3720449172576835e-06,
+      "loss": 0.1421,
+      "step": 530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2904430031776428,
+      "learning_rate": 4.370567375886525e-06,
+      "loss": 0.1471,
+      "step": 531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28084251284599304,
+      "learning_rate": 4.3690898345153674e-06,
+      "loss": 0.1469,
+      "step": 532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2758401930332184,
+      "learning_rate": 4.367612293144208e-06,
+      "loss": 0.1444,
+      "step": 533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3013249635696411,
+      "learning_rate": 4.3661347517730505e-06,
+      "loss": 0.1447,
+      "step": 534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27041032910346985,
+      "learning_rate": 4.364657210401891e-06,
+      "loss": 0.1402,
+      "step": 535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2782222032546997,
+      "learning_rate": 4.363179669030734e-06,
+      "loss": 0.1493,
+      "step": 536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2675043046474457,
+      "learning_rate": 4.361702127659575e-06,
+      "loss": 0.1428,
+      "step": 537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29422909021377563,
+      "learning_rate": 4.360224586288417e-06,
+      "loss": 0.1468,
+      "step": 538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2878548204898834,
+      "learning_rate": 4.358747044917258e-06,
+      "loss": 0.1488,
+      "step": 539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2842232584953308,
+      "learning_rate": 4.3572695035461e-06,
+      "loss": 0.1472,
+      "step": 540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2957862615585327,
+      "learning_rate": 4.355791962174941e-06,
+      "loss": 0.1515,
+      "step": 541
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29424312710762024,
+      "learning_rate": 4.354314420803783e-06,
+      "loss": 0.147,
+      "step": 542
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2823403775691986,
+      "learning_rate": 4.352836879432624e-06,
+      "loss": 0.1491,
+      "step": 543
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28637003898620605,
+      "learning_rate": 4.351359338061466e-06,
+      "loss": 0.1501,
+      "step": 544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27705931663513184,
+      "learning_rate": 4.349881796690308e-06,
+      "loss": 0.1439,
+      "step": 545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2889389097690582,
+      "learning_rate": 4.348404255319149e-06,
+      "loss": 0.1428,
+      "step": 546
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2772579789161682,
+      "learning_rate": 4.346926713947991e-06,
+      "loss": 0.1448,
+      "step": 547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28431400656700134,
+      "learning_rate": 4.345449172576833e-06,
+      "loss": 0.1512,
+      "step": 548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.287362664937973,
+      "learning_rate": 4.343971631205674e-06,
+      "loss": 0.1473,
+      "step": 549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3061977028846741,
+      "learning_rate": 4.342494089834516e-06,
+      "loss": 0.145,
+      "step": 550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2795885503292084,
+      "learning_rate": 4.3410165484633574e-06,
+      "loss": 0.1417,
+      "step": 551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2760257124900818,
+      "learning_rate": 4.339539007092199e-06,
+      "loss": 0.1418,
+      "step": 552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29454538226127625,
+      "learning_rate": 4.3380614657210405e-06,
+      "loss": 0.1414,
+      "step": 553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28224900364875793,
+      "learning_rate": 4.336583924349882e-06,
+      "loss": 0.1462,
+      "step": 554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29166895151138306,
+      "learning_rate": 4.3351063829787236e-06,
+      "loss": 0.1456,
+      "step": 555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28543752431869507,
+      "learning_rate": 4.333628841607566e-06,
+      "loss": 0.1427,
+      "step": 556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2974565625190735,
+      "learning_rate": 4.332151300236407e-06,
+      "loss": 0.1421,
+      "step": 557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27903100848197937,
+      "learning_rate": 4.330673758865249e-06,
+      "loss": 0.1402,
+      "step": 558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31764036417007446,
+      "learning_rate": 4.32919621749409e-06,
+      "loss": 0.1484,
+      "step": 559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2867543399333954,
+      "learning_rate": 4.327718676122932e-06,
+      "loss": 0.1444,
+      "step": 560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28913015127182007,
+      "learning_rate": 4.326241134751774e-06,
+      "loss": 0.1488,
+      "step": 561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2910228371620178,
+      "learning_rate": 4.324763593380615e-06,
+      "loss": 0.1425,
+      "step": 562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28646206855773926,
+      "learning_rate": 4.323286052009457e-06,
+      "loss": 0.1469,
+      "step": 563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28175657987594604,
+      "learning_rate": 4.321808510638298e-06,
+      "loss": 0.1472,
+      "step": 564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2771183252334595,
+      "learning_rate": 4.32033096926714e-06,
+      "loss": 0.1427,
+      "step": 565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26769471168518066,
+      "learning_rate": 4.318853427895981e-06,
+      "loss": 0.1387,
+      "step": 566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28029361367225647,
+      "learning_rate": 4.317375886524824e-06,
+      "loss": 0.1373,
+      "step": 567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2851644456386566,
+      "learning_rate": 4.315898345153664e-06,
+      "loss": 0.1451,
+      "step": 568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28463372588157654,
+      "learning_rate": 4.314420803782507e-06,
+      "loss": 0.1431,
+      "step": 569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2878042757511139,
+      "learning_rate": 4.312943262411347e-06,
+      "loss": 0.1462,
+      "step": 570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2978696823120117,
+      "learning_rate": 4.31146572104019e-06,
+      "loss": 0.1472,
+      "step": 571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2990373373031616,
+      "learning_rate": 4.309988179669031e-06,
+      "loss": 0.1454,
+      "step": 572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30174702405929565,
+      "learning_rate": 4.308510638297873e-06,
+      "loss": 0.1524,
+      "step": 573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2798447012901306,
+      "learning_rate": 4.307033096926714e-06,
+      "loss": 0.144,
+      "step": 574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.31803098320961,
+      "learning_rate": 4.305555555555556e-06,
+      "loss": 0.1412,
+      "step": 575
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29011499881744385,
+      "learning_rate": 4.3040780141843975e-06,
+      "loss": 0.1457,
+      "step": 576
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2867281436920166,
+      "learning_rate": 4.302600472813239e-06,
+      "loss": 0.1417,
+      "step": 577
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30410465598106384,
+      "learning_rate": 4.3011229314420805e-06,
+      "loss": 0.1473,
+      "step": 578
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28360387682914734,
+      "learning_rate": 4.299645390070922e-06,
+      "loss": 0.1403,
+      "step": 579
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2861372232437134,
+      "learning_rate": 4.2981678486997644e-06,
+      "loss": 0.1441,
+      "step": 580
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27881407737731934,
+      "learning_rate": 4.296690307328605e-06,
+      "loss": 0.1476,
+      "step": 581
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2853590250015259,
+      "learning_rate": 4.2952127659574475e-06,
+      "loss": 0.1444,
+      "step": 582
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33486029505729675,
+      "learning_rate": 4.293735224586289e-06,
+      "loss": 0.1533,
+      "step": 583
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31548529863357544,
+      "learning_rate": 4.292257683215131e-06,
+      "loss": 0.1468,
+      "step": 584
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29202157258987427,
+      "learning_rate": 4.290780141843972e-06,
+      "loss": 0.1424,
+      "step": 585
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29105034470558167,
+      "learning_rate": 4.289302600472814e-06,
+      "loss": 0.149,
+      "step": 586
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27002692222595215,
+      "learning_rate": 4.287825059101655e-06,
+      "loss": 0.1398,
+      "step": 587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3047403395175934,
+      "learning_rate": 4.286347517730497e-06,
+      "loss": 0.1454,
+      "step": 588
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3209473788738251,
+      "learning_rate": 4.284869976359338e-06,
+      "loss": 0.144,
+      "step": 589
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2747456133365631,
+      "learning_rate": 4.28339243498818e-06,
+      "loss": 0.1402,
+      "step": 590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29369720816612244,
+      "learning_rate": 4.281914893617022e-06,
+      "loss": 0.1424,
+      "step": 591
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29347583651542664,
+      "learning_rate": 4.280437352245863e-06,
+      "loss": 0.1383,
+      "step": 592
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2955438792705536,
+      "learning_rate": 4.278959810874705e-06,
+      "loss": 0.149,
+      "step": 593
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2783682346343994,
+      "learning_rate": 4.277482269503547e-06,
+      "loss": 0.1385,
+      "step": 594
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28423672914505005,
+      "learning_rate": 4.276004728132388e-06,
+      "loss": 0.1458,
+      "step": 595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27834346890449524,
+      "learning_rate": 4.27452718676123e-06,
+      "loss": 0.1461,
+      "step": 596
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3076930642127991,
+      "learning_rate": 4.273049645390071e-06,
+      "loss": 0.1511,
+      "step": 597
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2920694351196289,
+      "learning_rate": 4.271572104018913e-06,
+      "loss": 0.1414,
+      "step": 598
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2842082381248474,
+      "learning_rate": 4.2700945626477544e-06,
+      "loss": 0.1419,
+      "step": 599
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2883601784706116,
+      "learning_rate": 4.268617021276596e-06,
+      "loss": 0.1384,
+      "step": 600
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2891918420791626,
+      "learning_rate": 4.2671394799054375e-06,
+      "loss": 0.1475,
+      "step": 601
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2962299585342407,
+      "learning_rate": 4.26566193853428e-06,
+      "loss": 0.1466,
+      "step": 602
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2975197434425354,
+      "learning_rate": 4.2641843971631206e-06,
+      "loss": 0.1498,
+      "step": 603
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2721555233001709,
+      "learning_rate": 4.262706855791963e-06,
+      "loss": 0.1359,
+      "step": 604
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3061368763446808,
+      "learning_rate": 4.261229314420804e-06,
+      "loss": 0.1407,
+      "step": 605
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27989351749420166,
+      "learning_rate": 4.259751773049646e-06,
+      "loss": 0.138,
+      "step": 606
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29154035449028015,
+      "learning_rate": 4.2582742316784875e-06,
+      "loss": 0.139,
+      "step": 607
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30996108055114746,
+      "learning_rate": 4.256796690307329e-06,
+      "loss": 0.1514,
+      "step": 608
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31094786524772644,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 0.1407,
+      "step": 609
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2861872613430023,
+      "learning_rate": 4.253841607565012e-06,
+      "loss": 0.1436,
+      "step": 610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2934266924858093,
+      "learning_rate": 4.252364066193854e-06,
+      "loss": 0.1443,
+      "step": 611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27589017152786255,
+      "learning_rate": 4.250886524822695e-06,
+      "loss": 0.1394,
+      "step": 612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27942389249801636,
+      "learning_rate": 4.249408983451537e-06,
+      "loss": 0.1435,
+      "step": 613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31265994906425476,
+      "learning_rate": 4.247931442080378e-06,
+      "loss": 0.1435,
+      "step": 614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2708578109741211,
+      "learning_rate": 4.246453900709221e-06,
+      "loss": 0.1399,
+      "step": 615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29065263271331787,
+      "learning_rate": 4.244976359338061e-06,
+      "loss": 0.1529,
+      "step": 616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27975520491600037,
+      "learning_rate": 4.243498817966904e-06,
+      "loss": 0.1449,
+      "step": 617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2896137535572052,
+      "learning_rate": 4.242021276595745e-06,
+      "loss": 0.1426,
+      "step": 618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2777178883552551,
+      "learning_rate": 4.240543735224587e-06,
+      "loss": 0.1409,
+      "step": 619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2812087833881378,
+      "learning_rate": 4.239066193853428e-06,
+      "loss": 0.1468,
+      "step": 620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30926844477653503,
+      "learning_rate": 4.23758865248227e-06,
+      "loss": 0.1415,
+      "step": 621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2796373665332794,
+      "learning_rate": 4.236111111111111e-06,
+      "loss": 0.1439,
+      "step": 622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2899523377418518,
+      "learning_rate": 4.234633569739953e-06,
+      "loss": 0.1443,
+      "step": 623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2777215242385864,
+      "learning_rate": 4.2331560283687945e-06,
+      "loss": 0.1409,
+      "step": 624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2829614579677582,
+      "learning_rate": 4.231678486997636e-06,
+      "loss": 0.1417,
+      "step": 625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2784309685230255,
+      "learning_rate": 4.230200945626478e-06,
+      "loss": 0.1359,
+      "step": 626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2891806662082672,
+      "learning_rate": 4.228723404255319e-06,
+      "loss": 0.1445,
+      "step": 627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.291933536529541,
+      "learning_rate": 4.2272458628841614e-06,
+      "loss": 0.1416,
+      "step": 628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2994377315044403,
+      "learning_rate": 4.225768321513003e-06,
+      "loss": 0.1451,
+      "step": 629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32365095615386963,
+      "learning_rate": 4.2242907801418445e-06,
+      "loss": 0.1462,
+      "step": 630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30153635144233704,
+      "learning_rate": 4.222813238770686e-06,
+      "loss": 0.1487,
+      "step": 631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3056314289569855,
+      "learning_rate": 4.221335697399528e-06,
+      "loss": 0.1445,
+      "step": 632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30135276913642883,
+      "learning_rate": 4.219858156028369e-06,
+      "loss": 0.1423,
+      "step": 633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.293809711933136,
+      "learning_rate": 4.218380614657211e-06,
+      "loss": 0.1402,
+      "step": 634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28270187973976135,
+      "learning_rate": 4.216903073286052e-06,
+      "loss": 0.14,
+      "step": 635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3005997836589813,
+      "learning_rate": 4.215425531914894e-06,
+      "loss": 0.1402,
+      "step": 636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2902088463306427,
+      "learning_rate": 4.213947990543736e-06,
+      "loss": 0.1412,
+      "step": 637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3023683726787567,
+      "learning_rate": 4.212470449172577e-06,
+      "loss": 0.1441,
+      "step": 638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27745363116264343,
+      "learning_rate": 4.210992907801419e-06,
+      "loss": 0.1443,
+      "step": 639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2926003038883209,
+      "learning_rate": 4.20951536643026e-06,
+      "loss": 0.1374,
+      "step": 640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29988622665405273,
+      "learning_rate": 4.208037825059102e-06,
+      "loss": 0.1476,
+      "step": 641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29424041509628296,
+      "learning_rate": 4.206560283687944e-06,
+      "loss": 0.1408,
+      "step": 642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29958370327949524,
+      "learning_rate": 4.205082742316785e-06,
+      "loss": 0.1371,
+      "step": 643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3100656270980835,
+      "learning_rate": 4.203605200945627e-06,
+      "loss": 0.1482,
+      "step": 644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2951706647872925,
+      "learning_rate": 4.202127659574468e-06,
+      "loss": 0.1441,
+      "step": 645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29400336742401123,
+      "learning_rate": 4.20065011820331e-06,
+      "loss": 0.1471,
+      "step": 646
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30137714743614197,
+      "learning_rate": 4.1991725768321514e-06,
+      "loss": 0.1381,
+      "step": 647
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3026235103607178,
+      "learning_rate": 4.197695035460993e-06,
+      "loss": 0.1484,
+      "step": 648
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30929815769195557,
+      "learning_rate": 4.1962174940898345e-06,
+      "loss": 0.1494,
+      "step": 649
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28666308522224426,
+      "learning_rate": 4.194739952718677e-06,
+      "loss": 0.1414,
+      "step": 650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2805979251861572,
+      "learning_rate": 4.1932624113475176e-06,
+      "loss": 0.1375,
+      "step": 651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30657702684402466,
+      "learning_rate": 4.19178486997636e-06,
+      "loss": 0.1394,
+      "step": 652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30025872588157654,
+      "learning_rate": 4.1903073286052015e-06,
+      "loss": 0.1432,
+      "step": 653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30459463596343994,
+      "learning_rate": 4.188829787234043e-06,
+      "loss": 0.1481,
+      "step": 654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3065933585166931,
+      "learning_rate": 4.1873522458628845e-06,
+      "loss": 0.1415,
+      "step": 655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2998373508453369,
+      "learning_rate": 4.185874704491726e-06,
+      "loss": 0.1492,
+      "step": 656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2844208776950836,
+      "learning_rate": 4.184397163120568e-06,
+      "loss": 0.1409,
+      "step": 657
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27822598814964294,
+      "learning_rate": 4.182919621749409e-06,
+      "loss": 0.14,
+      "step": 658
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27559739351272583,
+      "learning_rate": 4.181442080378251e-06,
+      "loss": 0.1415,
+      "step": 659
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28201180696487427,
+      "learning_rate": 4.179964539007092e-06,
+      "loss": 0.1428,
+      "step": 660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2911911904811859,
+      "learning_rate": 4.178486997635935e-06,
+      "loss": 0.1464,
+      "step": 661
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3098863661289215,
+      "learning_rate": 4.177009456264775e-06,
+      "loss": 0.1415,
+      "step": 662
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2919551730155945,
+      "learning_rate": 4.175531914893618e-06,
+      "loss": 0.1401,
+      "step": 663
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29123958945274353,
+      "learning_rate": 4.174054373522459e-06,
+      "loss": 0.142,
+      "step": 664
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29178133606910706,
+      "learning_rate": 4.172576832151301e-06,
+      "loss": 0.143,
+      "step": 665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3129192888736725,
+      "learning_rate": 4.171099290780142e-06,
+      "loss": 0.1396,
+      "step": 666
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2876558005809784,
+      "learning_rate": 4.169621749408984e-06,
+      "loss": 0.1431,
+      "step": 667
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2821044623851776,
+      "learning_rate": 4.168144208037825e-06,
+      "loss": 0.1433,
+      "step": 668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2949013113975525,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.1416,
+      "step": 669
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3373323678970337,
+      "learning_rate": 4.165189125295508e-06,
+      "loss": 0.1422,
+      "step": 670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3034970462322235,
+      "learning_rate": 4.16371158392435e-06,
+      "loss": 0.1449,
+      "step": 671
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2797060012817383,
+      "learning_rate": 4.162234042553192e-06,
+      "loss": 0.1441,
+      "step": 672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2875618636608124,
+      "learning_rate": 4.160756501182033e-06,
+      "loss": 0.1416,
+      "step": 673
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3061351776123047,
+      "learning_rate": 4.159278959810875e-06,
+      "loss": 0.1377,
+      "step": 674
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2849779725074768,
+      "learning_rate": 4.157801418439716e-06,
+      "loss": 0.1426,
+      "step": 675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2907719314098358,
+      "learning_rate": 4.1563238770685584e-06,
+      "loss": 0.1401,
+      "step": 676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29381370544433594,
+      "learning_rate": 4.1548463356974e-06,
+      "loss": 0.1434,
+      "step": 677
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2804498076438904,
+      "learning_rate": 4.1533687943262415e-06,
+      "loss": 0.1408,
+      "step": 678
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.275444895029068,
+      "learning_rate": 4.151891252955083e-06,
+      "loss": 0.1417,
+      "step": 679
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27341586351394653,
+      "learning_rate": 4.150413711583925e-06,
+      "loss": 0.1378,
+      "step": 680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2919824421405792,
+      "learning_rate": 4.148936170212766e-06,
+      "loss": 0.1446,
+      "step": 681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28316348791122437,
+      "learning_rate": 4.147458628841608e-06,
+      "loss": 0.1362,
+      "step": 682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2960835099220276,
+      "learning_rate": 4.145981087470449e-06,
+      "loss": 0.1372,
+      "step": 683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2949550747871399,
+      "learning_rate": 4.144503546099291e-06,
+      "loss": 0.1402,
+      "step": 684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28910842537879944,
+      "learning_rate": 4.143026004728133e-06,
+      "loss": 0.1406,
+      "step": 685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2909477651119232,
+      "learning_rate": 4.141548463356974e-06,
+      "loss": 0.1414,
+      "step": 686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30360540747642517,
+      "learning_rate": 4.140070921985816e-06,
+      "loss": 0.1457,
+      "step": 687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2855706512928009,
+      "learning_rate": 4.138593380614658e-06,
+      "loss": 0.1356,
+      "step": 688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2964598536491394,
+      "learning_rate": 4.137115839243499e-06,
+      "loss": 0.1358,
+      "step": 689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28832998871803284,
+      "learning_rate": 4.135638297872341e-06,
+      "loss": 0.1329,
+      "step": 690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2996040880680084,
+      "learning_rate": 4.134160756501182e-06,
+      "loss": 0.1426,
+      "step": 691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28137755393981934,
+      "learning_rate": 4.132683215130024e-06,
+      "loss": 0.1422,
+      "step": 692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27830860018730164,
+      "learning_rate": 4.131205673758865e-06,
+      "loss": 0.1397,
+      "step": 693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30324214696884155,
+      "learning_rate": 4.129728132387707e-06,
+      "loss": 0.1447,
+      "step": 694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3021329939365387,
+      "learning_rate": 4.1282505910165484e-06,
+      "loss": 0.1438,
+      "step": 695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30520015954971313,
+      "learning_rate": 4.126773049645391e-06,
+      "loss": 0.1421,
+      "step": 696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31569889187812805,
+      "learning_rate": 4.1252955082742315e-06,
+      "loss": 0.1451,
+      "step": 697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27519088983535767,
+      "learning_rate": 4.123817966903074e-06,
+      "loss": 0.1396,
+      "step": 698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3046962320804596,
+      "learning_rate": 4.122340425531915e-06,
+      "loss": 0.1451,
+      "step": 699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2749660611152649,
+      "learning_rate": 4.120862884160757e-06,
+      "loss": 0.1393,
+      "step": 700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3148858845233917,
+      "learning_rate": 4.1193853427895985e-06,
+      "loss": 0.1449,
+      "step": 701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2860223054885864,
+      "learning_rate": 4.11790780141844e-06,
+      "loss": 0.1466,
+      "step": 702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2958418130874634,
+      "learning_rate": 4.1164302600472815e-06,
+      "loss": 0.1407,
+      "step": 703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3074707090854645,
+      "learning_rate": 4.114952718676123e-06,
+      "loss": 0.1463,
+      "step": 704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2819007635116577,
+      "learning_rate": 4.113475177304965e-06,
+      "loss": 0.1429,
+      "step": 705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28355252742767334,
+      "learning_rate": 4.111997635933806e-06,
+      "loss": 0.141,
+      "step": 706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3014814853668213,
+      "learning_rate": 4.1105200945626485e-06,
+      "loss": 0.1441,
+      "step": 707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2808004319667816,
+      "learning_rate": 4.109042553191489e-06,
+      "loss": 0.1378,
+      "step": 708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30655500292778015,
+      "learning_rate": 4.107565011820332e-06,
+      "loss": 0.147,
+      "step": 709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3061468303203583,
+      "learning_rate": 4.106087470449172e-06,
+      "loss": 0.1403,
+      "step": 710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2948397696018219,
+      "learning_rate": 4.104609929078015e-06,
+      "loss": 0.1418,
+      "step": 711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28176161646842957,
+      "learning_rate": 4.103132387706856e-06,
+      "loss": 0.138,
+      "step": 712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2762830853462219,
+      "learning_rate": 4.101654846335698e-06,
+      "loss": 0.1434,
+      "step": 713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2915934920310974,
+      "learning_rate": 4.100177304964539e-06,
+      "loss": 0.1379,
+      "step": 714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2960168421268463,
+      "learning_rate": 4.098699763593381e-06,
+      "loss": 0.1462,
+      "step": 715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27794840931892395,
+      "learning_rate": 4.097222222222222e-06,
+      "loss": 0.1477,
+      "step": 716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29872578382492065,
+      "learning_rate": 4.095744680851064e-06,
+      "loss": 0.1443,
+      "step": 717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2777767777442932,
+      "learning_rate": 4.094267139479905e-06,
+      "loss": 0.1335,
+      "step": 718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2738760709762573,
+      "learning_rate": 4.092789598108747e-06,
+      "loss": 0.1389,
+      "step": 719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2833902835845947,
+      "learning_rate": 4.091312056737589e-06,
+      "loss": 0.1371,
+      "step": 720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3024248480796814,
+      "learning_rate": 4.08983451536643e-06,
+      "loss": 0.1365,
+      "step": 721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2744184732437134,
+      "learning_rate": 4.088356973995272e-06,
+      "loss": 0.1402,
+      "step": 722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2921009063720703,
+      "learning_rate": 4.086879432624114e-06,
+      "loss": 0.1425,
+      "step": 723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28722310066223145,
+      "learning_rate": 4.0854018912529554e-06,
+      "loss": 0.136,
+      "step": 724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2708045244216919,
+      "learning_rate": 4.083924349881797e-06,
+      "loss": 0.1395,
+      "step": 725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28971266746520996,
+      "learning_rate": 4.0824468085106385e-06,
+      "loss": 0.1435,
+      "step": 726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29892614483833313,
+      "learning_rate": 4.08096926713948e-06,
+      "loss": 0.1405,
+      "step": 727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.293684184551239,
+      "learning_rate": 4.079491725768322e-06,
+      "loss": 0.1408,
+      "step": 728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29649442434310913,
+      "learning_rate": 4.078014184397163e-06,
+      "loss": 0.1403,
+      "step": 729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2769696116447449,
+      "learning_rate": 4.076536643026005e-06,
+      "loss": 0.1423,
+      "step": 730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29418668150901794,
+      "learning_rate": 4.075059101654847e-06,
+      "loss": 0.1492,
+      "step": 731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2974129319190979,
+      "learning_rate": 4.073581560283688e-06,
+      "loss": 0.1452,
+      "step": 732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2840820252895355,
+      "learning_rate": 4.07210401891253e-06,
+      "loss": 0.1421,
+      "step": 733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30748966336250305,
+      "learning_rate": 4.070626477541372e-06,
+      "loss": 0.1378,
+      "step": 734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2942531406879425,
+      "learning_rate": 4.069148936170213e-06,
+      "loss": 0.143,
+      "step": 735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2716488838195801,
+      "learning_rate": 4.067671394799055e-06,
+      "loss": 0.1339,
+      "step": 736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28804972767829895,
+      "learning_rate": 4.066193853427896e-06,
+      "loss": 0.141,
+      "step": 737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3043764531612396,
+      "learning_rate": 4.064716312056738e-06,
+      "loss": 0.1448,
+      "step": 738
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29227134585380554,
+      "learning_rate": 4.063238770685579e-06,
+      "loss": 0.1363,
+      "step": 739
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30961889028549194,
+      "learning_rate": 4.061761229314421e-06,
+      "loss": 0.1402,
+      "step": 740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2954045534133911,
+      "learning_rate": 4.060283687943262e-06,
+      "loss": 0.1393,
+      "step": 741
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2852894961833954,
+      "learning_rate": 4.058806146572105e-06,
+      "loss": 0.1399,
+      "step": 742
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3018682301044464,
+      "learning_rate": 4.0573286052009454e-06,
+      "loss": 0.1406,
+      "step": 743
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32532432675361633,
+      "learning_rate": 4.055851063829788e-06,
+      "loss": 0.1446,
+      "step": 744
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3091467022895813,
+      "learning_rate": 4.0543735224586285e-06,
+      "loss": 0.1441,
+      "step": 745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2878701984882355,
+      "learning_rate": 4.052895981087471e-06,
+      "loss": 0.139,
+      "step": 746
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31025537848472595,
+      "learning_rate": 4.051418439716312e-06,
+      "loss": 0.1393,
+      "step": 747
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29234030842781067,
+      "learning_rate": 4.049940898345154e-06,
+      "loss": 0.138,
+      "step": 748
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28872326016426086,
+      "learning_rate": 4.0484633569739955e-06,
+      "loss": 0.1373,
+      "step": 749
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3074430823326111,
+      "learning_rate": 4.046985815602837e-06,
+      "loss": 0.138,
+      "step": 750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2947796583175659,
+      "learning_rate": 4.0455082742316786e-06,
+      "loss": 0.1428,
+      "step": 751
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3205578327178955,
+      "learning_rate": 4.04403073286052e-06,
+      "loss": 0.1423,
+      "step": 752
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2928149104118347,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 0.1403,
+      "step": 753
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2662462890148163,
+      "learning_rate": 4.041075650118203e-06,
+      "loss": 0.1388,
+      "step": 754
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26826053857803345,
+      "learning_rate": 4.0395981087470455e-06,
+      "loss": 0.1417,
+      "step": 755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28748512268066406,
+      "learning_rate": 4.038120567375887e-06,
+      "loss": 0.131,
+      "step": 756
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29525142908096313,
+      "learning_rate": 4.036643026004729e-06,
+      "loss": 0.1401,
+      "step": 757
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28854089975357056,
+      "learning_rate": 4.03516548463357e-06,
+      "loss": 0.1387,
+      "step": 758
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27222731709480286,
+      "learning_rate": 4.033687943262412e-06,
+      "loss": 0.1408,
+      "step": 759
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2828696668148041,
+      "learning_rate": 4.032210401891253e-06,
+      "loss": 0.1405,
+      "step": 760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3017677366733551,
+      "learning_rate": 4.030732860520096e-06,
+      "loss": 0.1407,
+      "step": 761
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27210113406181335,
+      "learning_rate": 4.029255319148936e-06,
+      "loss": 0.1376,
+      "step": 762
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29224178194999695,
+      "learning_rate": 4.027777777777779e-06,
+      "loss": 0.1438,
+      "step": 763
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29064929485321045,
+      "learning_rate": 4.026300236406619e-06,
+      "loss": 0.1382,
+      "step": 764
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27730539441108704,
+      "learning_rate": 4.024822695035462e-06,
+      "loss": 0.1325,
+      "step": 765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28551870584487915,
+      "learning_rate": 4.023345153664303e-06,
+      "loss": 0.1347,
+      "step": 766
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29955190420150757,
+      "learning_rate": 4.021867612293145e-06,
+      "loss": 0.1377,
+      "step": 767
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3010675609111786,
+      "learning_rate": 4.020390070921986e-06,
+      "loss": 0.1445,
+      "step": 768
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2839129567146301,
+      "learning_rate": 4.018912529550828e-06,
+      "loss": 0.1357,
+      "step": 769
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28525620698928833,
+      "learning_rate": 4.017434988179669e-06,
+      "loss": 0.1392,
+      "step": 770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27335551381111145,
+      "learning_rate": 4.015957446808511e-06,
+      "loss": 0.139,
+      "step": 771
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.269033282995224,
+      "learning_rate": 4.0144799054373525e-06,
+      "loss": 0.1379,
+      "step": 772
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2774978578090668,
+      "learning_rate": 4.013002364066194e-06,
+      "loss": 0.1351,
+      "step": 773
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30730363726615906,
+      "learning_rate": 4.011524822695036e-06,
+      "loss": 0.1371,
+      "step": 774
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28533050417900085,
+      "learning_rate": 4.010047281323877e-06,
+      "loss": 0.1398,
+      "step": 775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28852036595344543,
+      "learning_rate": 4.0085697399527194e-06,
+      "loss": 0.1373,
+      "step": 776
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2799692749977112,
+      "learning_rate": 4.007092198581561e-06,
+      "loss": 0.1403,
+      "step": 777
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2960667014122009,
+      "learning_rate": 4.0056146572104025e-06,
+      "loss": 0.1404,
+      "step": 778
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2785898745059967,
+      "learning_rate": 4.004137115839244e-06,
+      "loss": 0.137,
+      "step": 779
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29664477705955505,
+      "learning_rate": 4.0026595744680856e-06,
+      "loss": 0.141,
+      "step": 780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29324209690093994,
+      "learning_rate": 4.001182033096927e-06,
+      "loss": 0.1477,
+      "step": 781
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29571622610092163,
+      "learning_rate": 3.999704491725769e-06,
+      "loss": 0.1353,
+      "step": 782
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2941003143787384,
+      "learning_rate": 3.99822695035461e-06,
+      "loss": 0.1403,
+      "step": 783
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3000856935977936,
+      "learning_rate": 3.996749408983452e-06,
+      "loss": 0.141,
+      "step": 784
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2789742350578308,
+      "learning_rate": 3.995271867612294e-06,
+      "loss": 0.1386,
+      "step": 785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2732674479484558,
+      "learning_rate": 3.993794326241135e-06,
+      "loss": 0.1378,
+      "step": 786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28325265645980835,
+      "learning_rate": 3.992316784869977e-06,
+      "loss": 0.1398,
+      "step": 787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28878849744796753,
+      "learning_rate": 3.990839243498818e-06,
+      "loss": 0.1373,
+      "step": 788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27129220962524414,
+      "learning_rate": 3.98936170212766e-06,
+      "loss": 0.1416,
+      "step": 789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28575021028518677,
+      "learning_rate": 3.987884160756502e-06,
+      "loss": 0.1422,
+      "step": 790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2822369635105133,
+      "learning_rate": 3.986406619385343e-06,
+      "loss": 0.1341,
+      "step": 791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2711465060710907,
+      "learning_rate": 3.984929078014185e-06,
+      "loss": 0.1386,
+      "step": 792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2980666160583496,
+      "learning_rate": 3.983451536643026e-06,
+      "loss": 0.1385,
+      "step": 793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2950401306152344,
+      "learning_rate": 3.981973995271868e-06,
+      "loss": 0.1338,
+      "step": 794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29184702038764954,
+      "learning_rate": 3.980496453900709e-06,
+      "loss": 0.1434,
+      "step": 795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27543723583221436,
+      "learning_rate": 3.979018912529552e-06,
+      "loss": 0.1379,
+      "step": 796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2907606065273285,
+      "learning_rate": 3.9775413711583925e-06,
+      "loss": 0.1341,
+      "step": 797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.291451096534729,
+      "learning_rate": 3.976063829787235e-06,
+      "loss": 0.1362,
+      "step": 798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31102070212364197,
+      "learning_rate": 3.9745862884160756e-06,
+      "loss": 0.1344,
+      "step": 799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2901325821876526,
+      "learning_rate": 3.973108747044918e-06,
+      "loss": 0.141,
+      "step": 800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29426053166389465,
+      "learning_rate": 3.9716312056737595e-06,
+      "loss": 0.1337,
+      "step": 801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28762757778167725,
+      "learning_rate": 3.970153664302601e-06,
+      "loss": 0.1404,
+      "step": 802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2861717641353607,
+      "learning_rate": 3.9686761229314425e-06,
+      "loss": 0.1421,
+      "step": 803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29405269026756287,
+      "learning_rate": 3.967198581560284e-06,
+      "loss": 0.1368,
+      "step": 804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.284153014421463,
+      "learning_rate": 3.965721040189126e-06,
+      "loss": 0.1371,
+      "step": 805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29278764128685,
+      "learning_rate": 3.964243498817967e-06,
+      "loss": 0.146,
+      "step": 806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28584909439086914,
+      "learning_rate": 3.962765957446809e-06,
+      "loss": 0.1386,
+      "step": 807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26947590708732605,
+      "learning_rate": 3.96128841607565e-06,
+      "loss": 0.134,
+      "step": 808
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28455930948257446,
+      "learning_rate": 3.959810874704493e-06,
+      "loss": 0.1286,
+      "step": 809
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29374200105667114,
+      "learning_rate": 3.958333333333333e-06,
+      "loss": 0.1413,
+      "step": 810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2807880640029907,
+      "learning_rate": 3.956855791962176e-06,
+      "loss": 0.1353,
+      "step": 811
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2899993658065796,
+      "learning_rate": 3.955378250591017e-06,
+      "loss": 0.141,
+      "step": 812
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28099220991134644,
+      "learning_rate": 3.953900709219859e-06,
+      "loss": 0.1401,
+      "step": 813
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28571051359176636,
+      "learning_rate": 3.9524231678487e-06,
+      "loss": 0.1369,
+      "step": 814
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27926862239837646,
+      "learning_rate": 3.950945626477542e-06,
+      "loss": 0.1406,
+      "step": 815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29341039061546326,
+      "learning_rate": 3.949468085106383e-06,
+      "loss": 0.1422,
+      "step": 816
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3273724317550659,
+      "learning_rate": 3.947990543735225e-06,
+      "loss": 0.1366,
+      "step": 817
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2864057123661041,
+      "learning_rate": 3.946513002364066e-06,
+      "loss": 0.1384,
+      "step": 818
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2901333272457123,
+      "learning_rate": 3.945035460992908e-06,
+      "loss": 0.1384,
+      "step": 819
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2691146731376648,
+      "learning_rate": 3.94355791962175e-06,
+      "loss": 0.1356,
+      "step": 820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.307944118976593,
+      "learning_rate": 3.942080378250591e-06,
+      "loss": 0.1388,
+      "step": 821
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31508874893188477,
+      "learning_rate": 3.940602836879433e-06,
+      "loss": 0.1364,
+      "step": 822
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29270583391189575,
+      "learning_rate": 3.939125295508275e-06,
+      "loss": 0.1388,
+      "step": 823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3098102807998657,
+      "learning_rate": 3.9376477541371164e-06,
+      "loss": 0.1389,
+      "step": 824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2766059637069702,
+      "learning_rate": 3.936170212765958e-06,
+      "loss": 0.1322,
+      "step": 825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2818599343299866,
+      "learning_rate": 3.9346926713947995e-06,
+      "loss": 0.1395,
+      "step": 826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2783481776714325,
+      "learning_rate": 3.933215130023641e-06,
+      "loss": 0.1353,
+      "step": 827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2850911617279053,
+      "learning_rate": 3.9317375886524826e-06,
+      "loss": 0.1398,
+      "step": 828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29799264669418335,
+      "learning_rate": 3.930260047281324e-06,
+      "loss": 0.1342,
+      "step": 829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2793444097042084,
+      "learning_rate": 3.928782505910166e-06,
+      "loss": 0.1364,
+      "step": 830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3010134994983673,
+      "learning_rate": 3.927304964539008e-06,
+      "loss": 0.1417,
+      "step": 831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2874111831188202,
+      "learning_rate": 3.925827423167849e-06,
+      "loss": 0.1383,
+      "step": 832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29187455773353577,
+      "learning_rate": 3.924349881796691e-06,
+      "loss": 0.1394,
+      "step": 833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28898483514785767,
+      "learning_rate": 3.922872340425532e-06,
+      "loss": 0.1399,
+      "step": 834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2808525264263153,
+      "learning_rate": 3.921394799054374e-06,
+      "loss": 0.1339,
+      "step": 835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32117101550102234,
+      "learning_rate": 3.919917257683216e-06,
+      "loss": 0.141,
+      "step": 836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2918335199356079,
+      "learning_rate": 3.918439716312057e-06,
+      "loss": 0.1392,
+      "step": 837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2987302541732788,
+      "learning_rate": 3.916962174940899e-06,
+      "loss": 0.1376,
+      "step": 838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28396928310394287,
+      "learning_rate": 3.91548463356974e-06,
+      "loss": 0.1366,
+      "step": 839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2748284339904785,
+      "learning_rate": 3.914007092198582e-06,
+      "loss": 0.1386,
+      "step": 840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2981834411621094,
+      "learning_rate": 3.912529550827423e-06,
+      "loss": 0.1385,
+      "step": 841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2808974087238312,
+      "learning_rate": 3.911052009456265e-06,
+      "loss": 0.1413,
+      "step": 842
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26401397585868835,
+      "learning_rate": 3.909574468085106e-06,
+      "loss": 0.1297,
+      "step": 843
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30584636330604553,
+      "learning_rate": 3.908096926713949e-06,
+      "loss": 0.1423,
+      "step": 844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3290936052799225,
+      "learning_rate": 3.9066193853427895e-06,
+      "loss": 0.1322,
+      "step": 845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28212642669677734,
+      "learning_rate": 3.905141843971632e-06,
+      "loss": 0.1349,
+      "step": 846
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29007601737976074,
+      "learning_rate": 3.903664302600473e-06,
+      "loss": 0.1392,
+      "step": 847
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29995259642601013,
+      "learning_rate": 3.902186761229315e-06,
+      "loss": 0.141,
+      "step": 848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.32899153232574463,
+      "learning_rate": 3.9007092198581565e-06,
+      "loss": 0.1387,
+      "step": 849
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30242908000946045,
+      "learning_rate": 3.899231678486998e-06,
+      "loss": 0.1421,
+      "step": 850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2836970388889313,
+      "learning_rate": 3.8977541371158395e-06,
+      "loss": 0.1363,
+      "step": 851
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29928794503211975,
+      "learning_rate": 3.896276595744681e-06,
+      "loss": 0.1404,
+      "step": 852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30467918515205383,
+      "learning_rate": 3.894799054373523e-06,
+      "loss": 0.135,
+      "step": 853
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3045257329940796,
+      "learning_rate": 3.893321513002364e-06,
+      "loss": 0.1398,
+      "step": 854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28601783514022827,
+      "learning_rate": 3.8918439716312065e-06,
+      "loss": 0.1436,
+      "step": 855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27304503321647644,
+      "learning_rate": 3.890366430260047e-06,
+      "loss": 0.1379,
+      "step": 856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2887423634529114,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.1369,
+      "step": 857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2947304844856262,
+      "learning_rate": 3.887411347517731e-06,
+      "loss": 0.1353,
+      "step": 858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3028261661529541,
+      "learning_rate": 3.885933806146573e-06,
+      "loss": 0.1386,
+      "step": 859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2955579459667206,
+      "learning_rate": 3.884456264775414e-06,
+      "loss": 0.1397,
+      "step": 860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2759718596935272,
+      "learning_rate": 3.882978723404256e-06,
+      "loss": 0.138,
+      "step": 861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29022523760795593,
+      "learning_rate": 3.881501182033097e-06,
+      "loss": 0.1364,
+      "step": 862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30592575669288635,
+      "learning_rate": 3.880023640661939e-06,
+      "loss": 0.1402,
+      "step": 863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3098193109035492,
+      "learning_rate": 3.87854609929078e-06,
+      "loss": 0.1388,
+      "step": 864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28497958183288574,
+      "learning_rate": 3.877068557919622e-06,
+      "loss": 0.135,
+      "step": 865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27340012788772583,
+      "learning_rate": 3.875591016548464e-06,
+      "loss": 0.1349,
+      "step": 866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.296935111284256,
+      "learning_rate": 3.874113475177305e-06,
+      "loss": 0.1389,
+      "step": 867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30188870429992676,
+      "learning_rate": 3.872635933806147e-06,
+      "loss": 0.132,
+      "step": 868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28601494431495667,
+      "learning_rate": 3.871158392434988e-06,
+      "loss": 0.1415,
+      "step": 869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27707338333129883,
+      "learning_rate": 3.86968085106383e-06,
+      "loss": 0.1351,
+      "step": 870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2665709853172302,
+      "learning_rate": 3.868203309692672e-06,
+      "loss": 0.1361,
+      "step": 871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2707883417606354,
+      "learning_rate": 3.8667257683215134e-06,
+      "loss": 0.1416,
+      "step": 872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2736751139163971,
+      "learning_rate": 3.865248226950355e-06,
+      "loss": 0.1335,
+      "step": 873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29390883445739746,
+      "learning_rate": 3.8637706855791965e-06,
+      "loss": 0.1351,
+      "step": 874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2881859540939331,
+      "learning_rate": 3.862293144208038e-06,
+      "loss": 0.1351,
+      "step": 875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2820168733596802,
+      "learning_rate": 3.8608156028368796e-06,
+      "loss": 0.1337,
+      "step": 876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2927733361721039,
+      "learning_rate": 3.859338061465721e-06,
+      "loss": 0.1446,
+      "step": 877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28589174151420593,
+      "learning_rate": 3.857860520094563e-06,
+      "loss": 0.1378,
+      "step": 878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30567967891693115,
+      "learning_rate": 3.856382978723405e-06,
+      "loss": 0.1341,
+      "step": 879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2780076563358307,
+      "learning_rate": 3.854905437352246e-06,
+      "loss": 0.1313,
+      "step": 880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.286905974149704,
+      "learning_rate": 3.853427895981088e-06,
+      "loss": 0.1341,
+      "step": 881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2779265344142914,
+      "learning_rate": 3.85195035460993e-06,
+      "loss": 0.133,
+      "step": 882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29259464144706726,
+      "learning_rate": 3.850472813238771e-06,
+      "loss": 0.1371,
+      "step": 883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2758921682834625,
+      "learning_rate": 3.848995271867613e-06,
+      "loss": 0.1381,
+      "step": 884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2837747633457184,
+      "learning_rate": 3.847517730496454e-06,
+      "loss": 0.1366,
+      "step": 885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28353196382522583,
+      "learning_rate": 3.846040189125296e-06,
+      "loss": 0.1369,
+      "step": 886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2868976891040802,
+      "learning_rate": 3.844562647754137e-06,
+      "loss": 0.1416,
+      "step": 887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28620126843452454,
+      "learning_rate": 3.843085106382979e-06,
+      "loss": 0.1358,
+      "step": 888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2767955958843231,
+      "learning_rate": 3.84160756501182e-06,
+      "loss": 0.1308,
+      "step": 889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27798473834991455,
+      "learning_rate": 3.840130023640663e-06,
+      "loss": 0.1388,
+      "step": 890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27987080812454224,
+      "learning_rate": 3.838652482269503e-06,
+      "loss": 0.1412,
+      "step": 891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.287905216217041,
+      "learning_rate": 3.837174940898346e-06,
+      "loss": 0.1307,
+      "step": 892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3017888367176056,
+      "learning_rate": 3.835697399527187e-06,
+      "loss": 0.1363,
+      "step": 893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2866717278957367,
+      "learning_rate": 3.834219858156029e-06,
+      "loss": 0.1369,
+      "step": 894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3042897880077362,
+      "learning_rate": 3.83274231678487e-06,
+      "loss": 0.1352,
+      "step": 895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2696511745452881,
+      "learning_rate": 3.831264775413712e-06,
+      "loss": 0.1343,
+      "step": 896
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29576727747917175,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 0.134,
+      "step": 897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3012337386608124,
+      "learning_rate": 3.828309692671395e-06,
+      "loss": 0.1406,
+      "step": 898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2837299108505249,
+      "learning_rate": 3.8268321513002365e-06,
+      "loss": 0.1343,
+      "step": 899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2881999909877777,
+      "learning_rate": 3.825354609929078e-06,
+      "loss": 0.1378,
+      "step": 900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29455244541168213,
+      "learning_rate": 3.8238770685579205e-06,
+      "loss": 0.1403,
+      "step": 901
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3030615746974945,
+      "learning_rate": 3.822399527186761e-06,
+      "loss": 0.137,
+      "step": 902
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2646782696247101,
+      "learning_rate": 3.8209219858156035e-06,
+      "loss": 0.1326,
+      "step": 903
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29197821021080017,
+      "learning_rate": 3.819444444444444e-06,
+      "loss": 0.1372,
+      "step": 904
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2664817273616791,
+      "learning_rate": 3.817966903073287e-06,
+      "loss": 0.1263,
+      "step": 905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2930912971496582,
+      "learning_rate": 3.816489361702128e-06,
+      "loss": 0.1367,
+      "step": 906
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2724604904651642,
+      "learning_rate": 3.81501182033097e-06,
+      "loss": 0.1405,
+      "step": 907
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29101988673210144,
+      "learning_rate": 3.813534278959811e-06,
+      "loss": 0.138,
+      "step": 908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28704380989074707,
+      "learning_rate": 3.812056737588653e-06,
+      "loss": 0.135,
+      "step": 909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2912169098854065,
+      "learning_rate": 3.8105791962174943e-06,
+      "loss": 0.141,
+      "step": 910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27939173579216003,
+      "learning_rate": 3.809101654846336e-06,
+      "loss": 0.1307,
+      "step": 911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3109738528728485,
+      "learning_rate": 3.8076241134751773e-06,
+      "loss": 0.1391,
+      "step": 912
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2832997143268585,
+      "learning_rate": 3.8061465721040193e-06,
+      "loss": 0.1344,
+      "step": 913
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2858584523200989,
+      "learning_rate": 3.804669030732861e-06,
+      "loss": 0.1353,
+      "step": 914
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2957686483860016,
+      "learning_rate": 3.8031914893617023e-06,
+      "loss": 0.1363,
+      "step": 915
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2822445333003998,
+      "learning_rate": 3.801713947990544e-06,
+      "loss": 0.1332,
+      "step": 916
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28698527812957764,
+      "learning_rate": 3.800236406619386e-06,
+      "loss": 0.137,
+      "step": 917
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29175078868865967,
+      "learning_rate": 3.798758865248227e-06,
+      "loss": 0.1346,
+      "step": 918
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.285359650850296,
+      "learning_rate": 3.797281323877069e-06,
+      "loss": 0.135,
+      "step": 919
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2824167311191559,
+      "learning_rate": 3.795803782505911e-06,
+      "loss": 0.1346,
+      "step": 920
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29884740710258484,
+      "learning_rate": 3.794326241134752e-06,
+      "loss": 0.1363,
+      "step": 921
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29446691274642944,
+      "learning_rate": 3.792848699763594e-06,
+      "loss": 0.1357,
+      "step": 922
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3002645969390869,
+      "learning_rate": 3.791371158392435e-06,
+      "loss": 0.1374,
+      "step": 923
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3043748438358307,
+      "learning_rate": 3.789893617021277e-06,
+      "loss": 0.1394,
+      "step": 924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2786545753479004,
+      "learning_rate": 3.7884160756501185e-06,
+      "loss": 0.1298,
+      "step": 925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2876208424568176,
+      "learning_rate": 3.78693853427896e-06,
+      "loss": 0.1381,
+      "step": 926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2887865900993347,
+      "learning_rate": 3.7854609929078016e-06,
+      "loss": 0.1328,
+      "step": 927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2904776930809021,
+      "learning_rate": 3.7839834515366436e-06,
+      "loss": 0.1354,
+      "step": 928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27404069900512695,
+      "learning_rate": 3.7825059101654847e-06,
+      "loss": 0.1334,
+      "step": 929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29318732023239136,
+      "learning_rate": 3.7810283687943266e-06,
+      "loss": 0.1332,
+      "step": 930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2716560661792755,
+      "learning_rate": 3.7795508274231677e-06,
+      "loss": 0.1319,
+      "step": 931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28456130623817444,
+      "learning_rate": 3.7780732860520097e-06,
+      "loss": 0.1369,
+      "step": 932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2826967239379883,
+      "learning_rate": 3.7765957446808516e-06,
+      "loss": 0.1338,
+      "step": 933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3071872889995575,
+      "learning_rate": 3.7751182033096928e-06,
+      "loss": 0.1413,
+      "step": 934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.286178857088089,
+      "learning_rate": 3.7736406619385347e-06,
+      "loss": 0.1321,
+      "step": 935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3212622106075287,
+      "learning_rate": 3.7721631205673762e-06,
+      "loss": 0.1403,
+      "step": 936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29869797825813293,
+      "learning_rate": 3.7706855791962178e-06,
+      "loss": 0.1366,
+      "step": 937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28863587975502014,
+      "learning_rate": 3.7692080378250593e-06,
+      "loss": 0.1361,
+      "step": 938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28116104006767273,
+      "learning_rate": 3.767730496453901e-06,
+      "loss": 0.1372,
+      "step": 939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2870403528213501,
+      "learning_rate": 3.7662529550827424e-06,
+      "loss": 0.1322,
+      "step": 940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3276669979095459,
+      "learning_rate": 3.7647754137115843e-06,
+      "loss": 0.1402,
+      "step": 941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29946812987327576,
+      "learning_rate": 3.7632978723404254e-06,
+      "loss": 0.1374,
+      "step": 942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28571459650993347,
+      "learning_rate": 3.7618203309692674e-06,
+      "loss": 0.1291,
+      "step": 943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29421335458755493,
+      "learning_rate": 3.7603427895981094e-06,
+      "loss": 0.1324,
+      "step": 944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2994726002216339,
+      "learning_rate": 3.7588652482269505e-06,
+      "loss": 0.1412,
+      "step": 945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2997495234012604,
+      "learning_rate": 3.7573877068557924e-06,
+      "loss": 0.1399,
+      "step": 946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2945842742919922,
+      "learning_rate": 3.7559101654846335e-06,
+      "loss": 0.1325,
+      "step": 947
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27185142040252686,
+      "learning_rate": 3.7544326241134755e-06,
+      "loss": 0.1319,
+      "step": 948
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2994024455547333,
+      "learning_rate": 3.752955082742317e-06,
+      "loss": 0.1405,
+      "step": 949
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3049403131008148,
+      "learning_rate": 3.7514775413711586e-06,
+      "loss": 0.1361,
+      "step": 950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30622977018356323,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.1363,
+      "step": 951
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29032963514328003,
+      "learning_rate": 3.748522458628842e-06,
+      "loss": 0.1372,
+      "step": 952
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29330381751060486,
+      "learning_rate": 3.7470449172576836e-06,
+      "loss": 0.1389,
+      "step": 953
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2871014475822449,
+      "learning_rate": 3.745567375886525e-06,
+      "loss": 0.1345,
+      "step": 954
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29108837246894836,
+      "learning_rate": 3.744089834515367e-06,
+      "loss": 0.13,
+      "step": 955
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29835349321365356,
+      "learning_rate": 3.742612293144208e-06,
+      "loss": 0.1358,
+      "step": 956
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30320024490356445,
+      "learning_rate": 3.74113475177305e-06,
+      "loss": 0.1341,
+      "step": 957
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29408392310142517,
+      "learning_rate": 3.7396572104018913e-06,
+      "loss": 0.137,
+      "step": 958
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2863946259021759,
+      "learning_rate": 3.738179669030733e-06,
+      "loss": 0.1305,
+      "step": 959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29572248458862305,
+      "learning_rate": 3.736702127659575e-06,
+      "loss": 0.1387,
+      "step": 960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27945277094841003,
+      "learning_rate": 3.7352245862884163e-06,
+      "loss": 0.1327,
+      "step": 961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28291401267051697,
+      "learning_rate": 3.7337470449172582e-06,
+      "loss": 0.135,
+      "step": 962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2841552793979645,
+      "learning_rate": 3.7322695035460998e-06,
+      "loss": 0.1311,
+      "step": 963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2827630937099457,
+      "learning_rate": 3.7307919621749413e-06,
+      "loss": 0.1385,
+      "step": 964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29116177558898926,
+      "learning_rate": 3.729314420803783e-06,
+      "loss": 0.132,
+      "step": 965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30451032519340515,
+      "learning_rate": 3.7278368794326244e-06,
+      "loss": 0.136,
+      "step": 966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2894097864627838,
+      "learning_rate": 3.726359338061466e-06,
+      "loss": 0.1375,
+      "step": 967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27166372537612915,
+      "learning_rate": 3.724881796690308e-06,
+      "loss": 0.1358,
+      "step": 968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30084818601608276,
+      "learning_rate": 3.723404255319149e-06,
+      "loss": 0.1374,
+      "step": 969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2727263569831848,
+      "learning_rate": 3.721926713947991e-06,
+      "loss": 0.1328,
+      "step": 970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.284184992313385,
+      "learning_rate": 3.720449172576833e-06,
+      "loss": 0.1355,
+      "step": 971
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28824329376220703,
+      "learning_rate": 3.718971631205674e-06,
+      "loss": 0.1352,
+      "step": 972
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27840736508369446,
+      "learning_rate": 3.717494089834516e-06,
+      "loss": 0.1305,
+      "step": 973
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2759610116481781,
+      "learning_rate": 3.716016548463357e-06,
+      "loss": 0.1347,
+      "step": 974
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2921728193759918,
+      "learning_rate": 3.714539007092199e-06,
+      "loss": 0.1378,
+      "step": 975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2897787094116211,
+      "learning_rate": 3.7130614657210406e-06,
+      "loss": 0.1355,
+      "step": 976
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2836354374885559,
+      "learning_rate": 3.711583924349882e-06,
+      "loss": 0.1365,
+      "step": 977
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29157429933547974,
+      "learning_rate": 3.7101063829787236e-06,
+      "loss": 0.136,
+      "step": 978
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2900297939777374,
+      "learning_rate": 3.7086288416075656e-06,
+      "loss": 0.1323,
+      "step": 979
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3050970137119293,
+      "learning_rate": 3.7071513002364067e-06,
+      "loss": 0.1396,
+      "step": 980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2850404977798462,
+      "learning_rate": 3.7056737588652486e-06,
+      "loss": 0.1361,
+      "step": 981
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27573466300964355,
+      "learning_rate": 3.7041962174940898e-06,
+      "loss": 0.1306,
+      "step": 982
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2913948595523834,
+      "learning_rate": 3.7027186761229317e-06,
+      "loss": 0.13,
+      "step": 983
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29537466168403625,
+      "learning_rate": 3.7012411347517737e-06,
+      "loss": 0.1373,
+      "step": 984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2757754325866699,
+      "learning_rate": 3.6997635933806148e-06,
+      "loss": 0.1307,
+      "step": 985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29556775093078613,
+      "learning_rate": 3.6982860520094567e-06,
+      "loss": 0.1398,
+      "step": 986
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27424076199531555,
+      "learning_rate": 3.6968085106382983e-06,
+      "loss": 0.133,
+      "step": 987
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2808970510959625,
+      "learning_rate": 3.69533096926714e-06,
+      "loss": 0.132,
+      "step": 988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.301440566778183,
+      "learning_rate": 3.6938534278959813e-06,
+      "loss": 0.1417,
+      "step": 989
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2951166331768036,
+      "learning_rate": 3.6923758865248233e-06,
+      "loss": 0.1385,
+      "step": 990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2841529846191406,
+      "learning_rate": 3.6908983451536644e-06,
+      "loss": 0.1344,
+      "step": 991
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28439974784851074,
+      "learning_rate": 3.6894208037825064e-06,
+      "loss": 0.1371,
+      "step": 992
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2776329219341278,
+      "learning_rate": 3.6879432624113475e-06,
+      "loss": 0.1393,
+      "step": 993
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28933215141296387,
+      "learning_rate": 3.6864657210401894e-06,
+      "loss": 0.1376,
+      "step": 994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3078655004501343,
+      "learning_rate": 3.6849881796690314e-06,
+      "loss": 0.1308,
+      "step": 995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2879551351070404,
+      "learning_rate": 3.6835106382978725e-06,
+      "loss": 0.1385,
+      "step": 996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27674123644828796,
+      "learning_rate": 3.6820330969267145e-06,
+      "loss": 0.1365,
+      "step": 997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2867019474506378,
+      "learning_rate": 3.680555555555556e-06,
+      "loss": 0.1325,
+      "step": 998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2820514142513275,
+      "learning_rate": 3.6790780141843975e-06,
+      "loss": 0.1374,
+      "step": 999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30006104707717896,
+      "learning_rate": 3.677600472813239e-06,
+      "loss": 0.1373,
+      "step": 1000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.292003870010376,
+      "learning_rate": 3.6761229314420806e-06,
+      "loss": 0.1405,
+      "step": 1001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3007003962993622,
+      "learning_rate": 3.674645390070922e-06,
+      "loss": 0.1358,
+      "step": 1002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2698974609375,
+      "learning_rate": 3.673167848699764e-06,
+      "loss": 0.1362,
+      "step": 1003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3111836612224579,
+      "learning_rate": 3.671690307328605e-06,
+      "loss": 0.1331,
+      "step": 1004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30400344729423523,
+      "learning_rate": 3.670212765957447e-06,
+      "loss": 0.1386,
+      "step": 1005
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2933863699436188,
+      "learning_rate": 3.668735224586289e-06,
+      "loss": 0.1351,
+      "step": 1006
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30335384607315063,
+      "learning_rate": 3.6672576832151302e-06,
+      "loss": 0.1405,
+      "step": 1007
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29099488258361816,
+      "learning_rate": 3.665780141843972e-06,
+      "loss": 0.1348,
+      "step": 1008
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2905612587928772,
+      "learning_rate": 3.6643026004728133e-06,
+      "loss": 0.1368,
+      "step": 1009
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2859121859073639,
+      "learning_rate": 3.6628250591016552e-06,
+      "loss": 0.1321,
+      "step": 1010
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2831714451313019,
+      "learning_rate": 3.6613475177304968e-06,
+      "loss": 0.13,
+      "step": 1011
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28861790895462036,
+      "learning_rate": 3.6598699763593383e-06,
+      "loss": 0.134,
+      "step": 1012
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.293247789144516,
+      "learning_rate": 3.65839243498818e-06,
+      "loss": 0.1311,
+      "step": 1013
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29096201062202454,
+      "learning_rate": 3.656914893617022e-06,
+      "loss": 0.1347,
+      "step": 1014
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2839386761188507,
+      "learning_rate": 3.655437352245863e-06,
+      "loss": 0.1348,
+      "step": 1015
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29057833552360535,
+      "learning_rate": 3.653959810874705e-06,
+      "loss": 0.1378,
+      "step": 1016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30836057662963867,
+      "learning_rate": 3.652482269503547e-06,
+      "loss": 0.1372,
+      "step": 1017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28458189964294434,
+      "learning_rate": 3.651004728132388e-06,
+      "loss": 0.1337,
+      "step": 1018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29900088906288147,
+      "learning_rate": 3.64952718676123e-06,
+      "loss": 0.1352,
+      "step": 1019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30995261669158936,
+      "learning_rate": 3.648049645390071e-06,
+      "loss": 0.1461,
+      "step": 1020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2833074927330017,
+      "learning_rate": 3.646572104018913e-06,
+      "loss": 0.1332,
+      "step": 1021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2832055985927582,
+      "learning_rate": 3.6450945626477545e-06,
+      "loss": 0.1296,
+      "step": 1022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27648308873176575,
+      "learning_rate": 3.643617021276596e-06,
+      "loss": 0.135,
+      "step": 1023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2958298921585083,
+      "learning_rate": 3.6421394799054376e-06,
+      "loss": 0.1366,
+      "step": 1024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2732112407684326,
+      "learning_rate": 3.6406619385342795e-06,
+      "loss": 0.1353,
+      "step": 1025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30360180139541626,
+      "learning_rate": 3.6391843971631206e-06,
+      "loss": 0.1355,
+      "step": 1026
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3016113042831421,
+      "learning_rate": 3.6377068557919626e-06,
+      "loss": 0.1291,
+      "step": 1027
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2888234853744507,
+      "learning_rate": 3.6362293144208037e-06,
+      "loss": 0.1373,
+      "step": 1028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2844870388507843,
+      "learning_rate": 3.6347517730496456e-06,
+      "loss": 0.1423,
+      "step": 1029
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2887290120124817,
+      "learning_rate": 3.6332742316784876e-06,
+      "loss": 0.14,
+      "step": 1030
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28420794010162354,
+      "learning_rate": 3.6317966903073287e-06,
+      "loss": 0.1332,
+      "step": 1031
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25794556736946106,
+      "learning_rate": 3.6303191489361707e-06,
+      "loss": 0.1303,
+      "step": 1032
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26992279291152954,
+      "learning_rate": 3.628841607565012e-06,
+      "loss": 0.1331,
+      "step": 1033
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2968750596046448,
+      "learning_rate": 3.6273640661938537e-06,
+      "loss": 0.1346,
+      "step": 1034
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2863878905773163,
+      "learning_rate": 3.6258865248226953e-06,
+      "loss": 0.1352,
+      "step": 1035
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2711179852485657,
+      "learning_rate": 3.624408983451537e-06,
+      "loss": 0.1317,
+      "step": 1036
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28469786047935486,
+      "learning_rate": 3.6229314420803783e-06,
+      "loss": 0.1397,
+      "step": 1037
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26579034328460693,
+      "learning_rate": 3.6214539007092203e-06,
+      "loss": 0.1297,
+      "step": 1038
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2866062521934509,
+      "learning_rate": 3.6199763593380614e-06,
+      "loss": 0.1375,
+      "step": 1039
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27711060643196106,
+      "learning_rate": 3.6184988179669034e-06,
+      "loss": 0.1367,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2737373113632202,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 0.1313,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28793638944625854,
+      "learning_rate": 3.6155437352245864e-06,
+      "loss": 0.1348,
+      "step": 1042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30166417360305786,
+      "learning_rate": 3.6140661938534284e-06,
+      "loss": 0.1391,
+      "step": 1043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2914607524871826,
+      "learning_rate": 3.6125886524822695e-06,
+      "loss": 0.1367,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28190135955810547,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 0.1321,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.279227614402771,
+      "learning_rate": 3.609633569739953e-06,
+      "loss": 0.1364,
+      "step": 1046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3139227628707886,
+      "learning_rate": 3.6081560283687945e-06,
+      "loss": 0.1381,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28012219071388245,
+      "learning_rate": 3.606678486997636e-06,
+      "loss": 0.1324,
+      "step": 1048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26524314284324646,
+      "learning_rate": 3.605200945626478e-06,
+      "loss": 0.1256,
+      "step": 1049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2778818607330322,
+      "learning_rate": 3.603723404255319e-06,
+      "loss": 0.131,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28027287125587463,
+      "learning_rate": 3.602245862884161e-06,
+      "loss": 0.1352,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2705986201763153,
+      "learning_rate": 3.600768321513003e-06,
+      "loss": 0.1359,
+      "step": 1052
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28018131852149963,
+      "learning_rate": 3.599290780141844e-06,
+      "loss": 0.1356,
+      "step": 1053
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3001031279563904,
+      "learning_rate": 3.597813238770686e-06,
+      "loss": 0.1328,
+      "step": 1054
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28043779730796814,
+      "learning_rate": 3.5963356973995272e-06,
+      "loss": 0.1312,
+      "step": 1055
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2602504789829254,
+      "learning_rate": 3.594858156028369e-06,
+      "loss": 0.1246,
+      "step": 1056
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2868635654449463,
+      "learning_rate": 3.5933806146572107e-06,
+      "loss": 0.1331,
+      "step": 1057
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2979179322719574,
+      "learning_rate": 3.5919030732860522e-06,
+      "loss": 0.1349,
+      "step": 1058
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27787864208221436,
+      "learning_rate": 3.5904255319148938e-06,
+      "loss": 0.1347,
+      "step": 1059
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3022499680519104,
+      "learning_rate": 3.5889479905437357e-06,
+      "loss": 0.138,
+      "step": 1060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28299573063850403,
+      "learning_rate": 3.587470449172577e-06,
+      "loss": 0.1325,
+      "step": 1061
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2782522737979889,
+      "learning_rate": 3.585992907801419e-06,
+      "loss": 0.129,
+      "step": 1062
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28664371371269226,
+      "learning_rate": 3.58451536643026e-06,
+      "loss": 0.1382,
+      "step": 1063
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2890227138996124,
+      "learning_rate": 3.583037825059102e-06,
+      "loss": 0.1312,
+      "step": 1064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29686394333839417,
+      "learning_rate": 3.581560283687944e-06,
+      "loss": 0.1329,
+      "step": 1065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27337437868118286,
+      "learning_rate": 3.580082742316785e-06,
+      "loss": 0.14,
+      "step": 1066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28173407912254333,
+      "learning_rate": 3.578605200945627e-06,
+      "loss": 0.1335,
+      "step": 1067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28702786564826965,
+      "learning_rate": 3.5771276595744684e-06,
+      "loss": 0.1337,
+      "step": 1068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28728947043418884,
+      "learning_rate": 3.57565011820331e-06,
+      "loss": 0.1387,
+      "step": 1069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2808537781238556,
+      "learning_rate": 3.5741725768321515e-06,
+      "loss": 0.1334,
+      "step": 1070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2801794111728668,
+      "learning_rate": 3.572695035460993e-06,
+      "loss": 0.1371,
+      "step": 1071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2781788110733032,
+      "learning_rate": 3.5712174940898346e-06,
+      "loss": 0.135,
+      "step": 1072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27386489510536194,
+      "learning_rate": 3.5697399527186765e-06,
+      "loss": 0.1318,
+      "step": 1073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2744510769844055,
+      "learning_rate": 3.5682624113475176e-06,
+      "loss": 0.128,
+      "step": 1074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2918887436389923,
+      "learning_rate": 3.5667848699763596e-06,
+      "loss": 0.1339,
+      "step": 1075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28798773884773254,
+      "learning_rate": 3.5653073286052015e-06,
+      "loss": 0.1335,
+      "step": 1076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2954133152961731,
+      "learning_rate": 3.5638297872340426e-06,
+      "loss": 0.1336,
+      "step": 1077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27675577998161316,
+      "learning_rate": 3.5623522458628846e-06,
+      "loss": 0.1372,
+      "step": 1078
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.264598548412323,
+      "learning_rate": 3.5608747044917257e-06,
+      "loss": 0.1271,
+      "step": 1079
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27909743785858154,
+      "learning_rate": 3.5593971631205677e-06,
+      "loss": 0.1288,
+      "step": 1080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27577412128448486,
+      "learning_rate": 3.557919621749409e-06,
+      "loss": 0.1387,
+      "step": 1081
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2805110216140747,
+      "learning_rate": 3.5564420803782507e-06,
+      "loss": 0.1335,
+      "step": 1082
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39830732345581055,
+      "learning_rate": 3.5549645390070927e-06,
+      "loss": 0.137,
+      "step": 1083
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28425392508506775,
+      "learning_rate": 3.5534869976359342e-06,
+      "loss": 0.1357,
+      "step": 1084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28320419788360596,
+      "learning_rate": 3.5520094562647758e-06,
+      "loss": 0.1374,
+      "step": 1085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2737938463687897,
+      "learning_rate": 3.5505319148936173e-06,
+      "loss": 0.1317,
+      "step": 1086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29593971371650696,
+      "learning_rate": 3.5490543735224593e-06,
+      "loss": 0.1392,
+      "step": 1087
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2826647460460663,
+      "learning_rate": 3.5475768321513004e-06,
+      "loss": 0.1279,
+      "step": 1088
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28226131200790405,
+      "learning_rate": 3.5460992907801423e-06,
+      "loss": 0.1359,
+      "step": 1089
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2747083306312561,
+      "learning_rate": 3.5446217494089834e-06,
+      "loss": 0.135,
+      "step": 1090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2708107829093933,
+      "learning_rate": 3.5431442080378254e-06,
+      "loss": 0.1328,
+      "step": 1091
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28234735131263733,
+      "learning_rate": 3.5416666666666673e-06,
+      "loss": 0.1307,
+      "step": 1092
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2894611060619354,
+      "learning_rate": 3.5401891252955085e-06,
+      "loss": 0.1323,
+      "step": 1093
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.276606947183609,
+      "learning_rate": 3.5387115839243504e-06,
+      "loss": 0.1316,
+      "step": 1094
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27292677760124207,
+      "learning_rate": 3.537234042553192e-06,
+      "loss": 0.1303,
+      "step": 1095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30232924222946167,
+      "learning_rate": 3.5357565011820335e-06,
+      "loss": 0.1388,
+      "step": 1096
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2754041254520416,
+      "learning_rate": 3.534278959810875e-06,
+      "loss": 0.129,
+      "step": 1097
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2740183174610138,
+      "learning_rate": 3.5328014184397165e-06,
+      "loss": 0.1349,
+      "step": 1098
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30713197588920593,
+      "learning_rate": 3.531323877068558e-06,
+      "loss": 0.137,
+      "step": 1099
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29255861043930054,
+      "learning_rate": 3.5298463356974e-06,
+      "loss": 0.1307,
+      "step": 1100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2927381694316864,
+      "learning_rate": 3.528368794326241e-06,
+      "loss": 0.1372,
+      "step": 1101
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2747761011123657,
+      "learning_rate": 3.526891252955083e-06,
+      "loss": 0.1347,
+      "step": 1102
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29051080346107483,
+      "learning_rate": 3.525413711583925e-06,
+      "loss": 0.1348,
+      "step": 1103
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2967738211154938,
+      "learning_rate": 3.523936170212766e-06,
+      "loss": 0.1405,
+      "step": 1104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29093116521835327,
+      "learning_rate": 3.522458628841608e-06,
+      "loss": 0.1375,
+      "step": 1105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26452237367630005,
+      "learning_rate": 3.5209810874704492e-06,
+      "loss": 0.1293,
+      "step": 1106
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2963341176509857,
+      "learning_rate": 3.519503546099291e-06,
+      "loss": 0.1353,
+      "step": 1107
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2842029333114624,
+      "learning_rate": 3.5180260047281327e-06,
+      "loss": 0.1369,
+      "step": 1108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2843793034553528,
+      "learning_rate": 3.5165484633569743e-06,
+      "loss": 0.1293,
+      "step": 1109
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2648527920246124,
+      "learning_rate": 3.515070921985816e-06,
+      "loss": 0.1294,
+      "step": 1110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2785526216030121,
+      "learning_rate": 3.5135933806146578e-06,
+      "loss": 0.1316,
+      "step": 1111
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2812475562095642,
+      "learning_rate": 3.512115839243499e-06,
+      "loss": 0.1302,
+      "step": 1112
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2896616458892822,
+      "learning_rate": 3.510638297872341e-06,
+      "loss": 0.1373,
+      "step": 1113
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29601964354515076,
+      "learning_rate": 3.5091607565011828e-06,
+      "loss": 0.136,
+      "step": 1114
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29470157623291016,
+      "learning_rate": 3.507683215130024e-06,
+      "loss": 0.1379,
+      "step": 1115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29352888464927673,
+      "learning_rate": 3.506205673758866e-06,
+      "loss": 0.137,
+      "step": 1116
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2891894280910492,
+      "learning_rate": 3.504728132387707e-06,
+      "loss": 0.1332,
+      "step": 1117
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31228727102279663,
+      "learning_rate": 3.503250591016549e-06,
+      "loss": 0.1354,
+      "step": 1118
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2830002009868622,
+      "learning_rate": 3.5017730496453904e-06,
+      "loss": 0.1375,
+      "step": 1119
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30769455432891846,
+      "learning_rate": 3.500295508274232e-06,
+      "loss": 0.1353,
+      "step": 1120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.268492728471756,
+      "learning_rate": 3.4988179669030735e-06,
+      "loss": 0.1329,
+      "step": 1121
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29373034834861755,
+      "learning_rate": 3.4973404255319155e-06,
+      "loss": 0.1322,
+      "step": 1122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27865368127822876,
+      "learning_rate": 3.4958628841607566e-06,
+      "loss": 0.1282,
+      "step": 1123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2929169833660126,
+      "learning_rate": 3.4943853427895985e-06,
+      "loss": 0.1372,
+      "step": 1124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31391963362693787,
+      "learning_rate": 3.4929078014184396e-06,
+      "loss": 0.1386,
+      "step": 1125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2919042706489563,
+      "learning_rate": 3.4914302600472816e-06,
+      "loss": 0.1372,
+      "step": 1126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.276607871055603,
+      "learning_rate": 3.4899527186761236e-06,
+      "loss": 0.1343,
+      "step": 1127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2895742356777191,
+      "learning_rate": 3.4884751773049647e-06,
+      "loss": 0.1324,
+      "step": 1128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28830260038375854,
+      "learning_rate": 3.4869976359338066e-06,
+      "loss": 0.1294,
+      "step": 1129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2916291654109955,
+      "learning_rate": 3.485520094562648e-06,
+      "loss": 0.1301,
+      "step": 1130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28904253244400024,
+      "learning_rate": 3.4840425531914897e-06,
+      "loss": 0.1378,
+      "step": 1131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27836576104164124,
+      "learning_rate": 3.4825650118203312e-06,
+      "loss": 0.1324,
+      "step": 1132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2714270055294037,
+      "learning_rate": 3.4810874704491728e-06,
+      "loss": 0.1249,
+      "step": 1133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.290276437997818,
+      "learning_rate": 3.4796099290780143e-06,
+      "loss": 0.1382,
+      "step": 1134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2902819812297821,
+      "learning_rate": 3.4781323877068563e-06,
+      "loss": 0.1362,
+      "step": 1135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2710173726081848,
+      "learning_rate": 3.4766548463356974e-06,
+      "loss": 0.1338,
+      "step": 1136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28448423743247986,
+      "learning_rate": 3.4751773049645393e-06,
+      "loss": 0.1351,
+      "step": 1137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28090378642082214,
+      "learning_rate": 3.4736997635933813e-06,
+      "loss": 0.1344,
+      "step": 1138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28663182258605957,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 0.1292,
+      "step": 1139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2905632555484772,
+      "learning_rate": 3.4707446808510643e-06,
+      "loss": 0.1271,
+      "step": 1140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29615235328674316,
+      "learning_rate": 3.4692671394799055e-06,
+      "loss": 0.1314,
+      "step": 1141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2951762080192566,
+      "learning_rate": 3.4677895981087474e-06,
+      "loss": 0.1355,
+      "step": 1142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3019202649593353,
+      "learning_rate": 3.466312056737589e-06,
+      "loss": 0.1356,
+      "step": 1143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29882511496543884,
+      "learning_rate": 3.4648345153664305e-06,
+      "loss": 0.1323,
+      "step": 1144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26828858256340027,
+      "learning_rate": 3.463356973995272e-06,
+      "loss": 0.1294,
+      "step": 1145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2803989052772522,
+      "learning_rate": 3.461879432624114e-06,
+      "loss": 0.1352,
+      "step": 1146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27997419238090515,
+      "learning_rate": 3.460401891252955e-06,
+      "loss": 0.1303,
+      "step": 1147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28511661291122437,
+      "learning_rate": 3.458924349881797e-06,
+      "loss": 0.1323,
+      "step": 1148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3100005090236664,
+      "learning_rate": 3.457446808510639e-06,
+      "loss": 0.1345,
+      "step": 1149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28533250093460083,
+      "learning_rate": 3.45596926713948e-06,
+      "loss": 0.1338,
+      "step": 1150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2826535999774933,
+      "learning_rate": 3.454491725768322e-06,
+      "loss": 0.125,
+      "step": 1151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29771560430526733,
+      "learning_rate": 3.453014184397163e-06,
+      "loss": 0.1305,
+      "step": 1152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2910444438457489,
+      "learning_rate": 3.451536643026005e-06,
+      "loss": 0.1357,
+      "step": 1153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2832120954990387,
+      "learning_rate": 3.4500591016548467e-06,
+      "loss": 0.1299,
+      "step": 1154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.316199392080307,
+      "learning_rate": 3.448581560283688e-06,
+      "loss": 0.1351,
+      "step": 1155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28114989399909973,
+      "learning_rate": 3.4471040189125297e-06,
+      "loss": 0.1275,
+      "step": 1156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30223196744918823,
+      "learning_rate": 3.4456264775413717e-06,
+      "loss": 0.1359,
+      "step": 1157
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2844015657901764,
+      "learning_rate": 3.444148936170213e-06,
+      "loss": 0.1307,
+      "step": 1158
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2777929902076721,
+      "learning_rate": 3.4426713947990548e-06,
+      "loss": 0.1312,
+      "step": 1159
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2758617699146271,
+      "learning_rate": 3.441193853427896e-06,
+      "loss": 0.1302,
+      "step": 1160
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2958165109157562,
+      "learning_rate": 3.439716312056738e-06,
+      "loss": 0.1318,
+      "step": 1161
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2866310477256775,
+      "learning_rate": 3.4382387706855798e-06,
+      "loss": 0.1272,
+      "step": 1162
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.280868798494339,
+      "learning_rate": 3.436761229314421e-06,
+      "loss": 0.1318,
+      "step": 1163
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2805544137954712,
+      "learning_rate": 3.435283687943263e-06,
+      "loss": 0.1216,
+      "step": 1164
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2837851941585541,
+      "learning_rate": 3.4338061465721044e-06,
+      "loss": 0.1197,
+      "step": 1165
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26824888586997986,
+      "learning_rate": 3.432328605200946e-06,
+      "loss": 0.1234,
+      "step": 1166
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26774662733078003,
+      "learning_rate": 3.4308510638297874e-06,
+      "loss": 0.116,
+      "step": 1167
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27060431241989136,
+      "learning_rate": 3.429373522458629e-06,
+      "loss": 0.1203,
+      "step": 1168
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.26172682642936707,
+      "learning_rate": 3.4278959810874705e-06,
+      "loss": 0.1152,
+      "step": 1169
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2787811756134033,
+      "learning_rate": 3.4264184397163125e-06,
+      "loss": 0.1169,
+      "step": 1170
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2766242027282715,
+      "learning_rate": 3.4249408983451536e-06,
+      "loss": 0.1118,
+      "step": 1171
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2759280204772949,
+      "learning_rate": 3.4234633569739955e-06,
+      "loss": 0.1135,
+      "step": 1172
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.29733550548553467,
+      "learning_rate": 3.4219858156028375e-06,
+      "loss": 0.1161,
+      "step": 1173
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2840992510318756,
+      "learning_rate": 3.4205082742316786e-06,
+      "loss": 0.1162,
+      "step": 1174
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3058291971683502,
+      "learning_rate": 3.4190307328605206e-06,
+      "loss": 0.1187,
+      "step": 1175
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2796337306499481,
+      "learning_rate": 3.4175531914893617e-06,
+      "loss": 0.1193,
+      "step": 1176
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3049129247665405,
+      "learning_rate": 3.4160756501182036e-06,
+      "loss": 0.1173,
+      "step": 1177
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2753768265247345,
+      "learning_rate": 3.414598108747045e-06,
+      "loss": 0.1168,
+      "step": 1178
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2855167090892792,
+      "learning_rate": 3.4131205673758867e-06,
+      "loss": 0.1177,
+      "step": 1179
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.28942784667015076,
+      "learning_rate": 3.4116430260047282e-06,
+      "loss": 0.115,
+      "step": 1180
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.27855753898620605,
+      "learning_rate": 3.41016548463357e-06,
+      "loss": 0.12,
+      "step": 1181
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2691517174243927,
+      "learning_rate": 3.4086879432624113e-06,
+      "loss": 0.1158,
+      "step": 1182
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2867049276828766,
+      "learning_rate": 3.4072104018912533e-06,
+      "loss": 0.1142,
+      "step": 1183
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28901442885398865,
+      "learning_rate": 3.4057328605200952e-06,
+      "loss": 0.1244,
+      "step": 1184
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2759941518306732,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 0.1177,
+      "step": 1185
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2885506749153137,
+      "learning_rate": 3.4027777777777783e-06,
+      "loss": 0.1223,
+      "step": 1186
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.28299227356910706,
+      "learning_rate": 3.4013002364066194e-06,
+      "loss": 0.1185,
+      "step": 1187
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2685181498527527,
+      "learning_rate": 3.3998226950354613e-06,
+      "loss": 0.1151,
+      "step": 1188
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2795426845550537,
+      "learning_rate": 3.398345153664303e-06,
+      "loss": 0.113,
+      "step": 1189
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3159796893596649,
+      "learning_rate": 3.3968676122931444e-06,
+      "loss": 0.1183,
+      "step": 1190
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2900717258453369,
+      "learning_rate": 3.395390070921986e-06,
+      "loss": 0.1207,
+      "step": 1191
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3061547577381134,
+      "learning_rate": 3.393912529550828e-06,
+      "loss": 0.1209,
+      "step": 1192
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2868369221687317,
+      "learning_rate": 3.392434988179669e-06,
+      "loss": 0.1152,
+      "step": 1193
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.28293681144714355,
+      "learning_rate": 3.390957446808511e-06,
+      "loss": 0.1152,
+      "step": 1194
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2693897783756256,
+      "learning_rate": 3.389479905437352e-06,
+      "loss": 0.1162,
+      "step": 1195
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.28829285502433777,
+      "learning_rate": 3.388002364066194e-06,
+      "loss": 0.1174,
+      "step": 1196
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2776600122451782,
+      "learning_rate": 3.386524822695036e-06,
+      "loss": 0.1155,
+      "step": 1197
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2848743796348572,
+      "learning_rate": 3.385047281323877e-06,
+      "loss": 0.114,
+      "step": 1198
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2962537705898285,
+      "learning_rate": 3.383569739952719e-06,
+      "loss": 0.119,
+      "step": 1199
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2776596248149872,
+      "learning_rate": 3.3820921985815606e-06,
+      "loss": 0.1171,
+      "step": 1200
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.27688196301460266,
+      "learning_rate": 3.380614657210402e-06,
+      "loss": 0.1142,
+      "step": 1201
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2855364978313446,
+      "learning_rate": 3.3791371158392437e-06,
+      "loss": 0.1195,
+      "step": 1202
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.29750436544418335,
+      "learning_rate": 3.377659574468085e-06,
+      "loss": 0.1229,
+      "step": 1203
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.30565622448921204,
+      "learning_rate": 3.3761820330969267e-06,
+      "loss": 0.1212,
+      "step": 1204
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.28108736872673035,
+      "learning_rate": 3.3747044917257687e-06,
+      "loss": 0.1151,
+      "step": 1205
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2884421944618225,
+      "learning_rate": 3.37322695035461e-06,
+      "loss": 0.1153,
+      "step": 1206
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2928875684738159,
+      "learning_rate": 3.3717494089834518e-06,
+      "loss": 0.1215,
+      "step": 1207
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2739309072494507,
+      "learning_rate": 3.3702718676122937e-06,
+      "loss": 0.113,
+      "step": 1208
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3110334277153015,
+      "learning_rate": 3.368794326241135e-06,
+      "loss": 0.125,
+      "step": 1209
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2945943772792816,
+      "learning_rate": 3.3673167848699768e-06,
+      "loss": 0.1179,
+      "step": 1210
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2857518494129181,
+      "learning_rate": 3.365839243498818e-06,
+      "loss": 0.118,
+      "step": 1211
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3058679401874542,
+      "learning_rate": 3.36436170212766e-06,
+      "loss": 0.1258,
+      "step": 1212
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.30979791283607483,
+      "learning_rate": 3.3628841607565014e-06,
+      "loss": 0.1171,
+      "step": 1213
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2900080680847168,
+      "learning_rate": 3.361406619385343e-06,
+      "loss": 0.1212,
+      "step": 1214
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3327723741531372,
+      "learning_rate": 3.359929078014185e-06,
+      "loss": 0.1158,
+      "step": 1215
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2705560624599457,
+      "learning_rate": 3.3584515366430264e-06,
+      "loss": 0.1173,
+      "step": 1216
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.27654606103897095,
+      "learning_rate": 3.356973995271868e-06,
+      "loss": 0.1148,
+      "step": 1217
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.27573952078819275,
+      "learning_rate": 3.3554964539007095e-06,
+      "loss": 0.115,
+      "step": 1218
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2880498766899109,
+      "learning_rate": 3.3540189125295514e-06,
+      "loss": 0.1185,
+      "step": 1219
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.26650840044021606,
+      "learning_rate": 3.3525413711583925e-06,
+      "loss": 0.1136,
+      "step": 1220
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.28205397725105286,
+      "learning_rate": 3.3510638297872345e-06,
+      "loss": 0.1173,
+      "step": 1221
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2864522337913513,
+      "learning_rate": 3.3495862884160756e-06,
+      "loss": 0.1191,
+      "step": 1222
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2774849236011505,
+      "learning_rate": 3.3481087470449176e-06,
+      "loss": 0.1159,
+      "step": 1223
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2812179625034332,
+      "learning_rate": 3.3466312056737595e-06,
+      "loss": 0.1184,
+      "step": 1224
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2862336337566376,
+      "learning_rate": 3.3451536643026006e-06,
+      "loss": 0.1212,
+      "step": 1225
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2784669101238251,
+      "learning_rate": 3.3436761229314426e-06,
+      "loss": 0.1163,
+      "step": 1226
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.27852392196655273,
+      "learning_rate": 3.342198581560284e-06,
+      "loss": 0.1178,
+      "step": 1227
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.31831488013267517,
+      "learning_rate": 3.3407210401891257e-06,
+      "loss": 0.1169,
+      "step": 1228
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.28438419103622437,
+      "learning_rate": 3.339243498817967e-06,
+      "loss": 0.1165,
+      "step": 1229
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2789434790611267,
+      "learning_rate": 3.3377659574468087e-06,
+      "loss": 0.116,
+      "step": 1230
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2784711718559265,
+      "learning_rate": 3.3362884160756503e-06,
+      "loss": 0.114,
+      "step": 1231
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.30339953303337097,
+      "learning_rate": 3.3348108747044922e-06,
+      "loss": 0.1178,
+      "step": 1232
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2773449420928955,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.1134,
+      "step": 1233
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3101428151130676,
+      "learning_rate": 3.3318557919621753e-06,
+      "loss": 0.1166,
+      "step": 1234
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3029414415359497,
+      "learning_rate": 3.3303782505910172e-06,
+      "loss": 0.1169,
+      "step": 1235
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3047308623790741,
+      "learning_rate": 3.3289007092198583e-06,
+      "loss": 0.1175,
+      "step": 1236
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.288804829120636,
+      "learning_rate": 3.3274231678487003e-06,
+      "loss": 0.1211,
+      "step": 1237
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2894182801246643,
+      "learning_rate": 3.3259456264775414e-06,
+      "loss": 0.1176,
+      "step": 1238
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.28906863927841187,
+      "learning_rate": 3.3244680851063834e-06,
+      "loss": 0.1198,
+      "step": 1239
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.28565263748168945,
+      "learning_rate": 3.322990543735225e-06,
+      "loss": 0.1196,
+      "step": 1240
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2917105555534363,
+      "learning_rate": 3.3215130023640664e-06,
+      "loss": 0.1208,
+      "step": 1241
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2863866090774536,
+      "learning_rate": 3.320035460992908e-06,
+      "loss": 0.1168,
+      "step": 1242
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.28636300563812256,
+      "learning_rate": 3.31855791962175e-06,
+      "loss": 0.1229,
+      "step": 1243
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.27634963393211365,
+      "learning_rate": 3.317080378250591e-06,
+      "loss": 0.1164,
+      "step": 1244
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.27865663170814514,
+      "learning_rate": 3.315602836879433e-06,
+      "loss": 0.1139,
+      "step": 1245
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2749914824962616,
+      "learning_rate": 3.314125295508275e-06,
+      "loss": 0.1178,
+      "step": 1246
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2907851040363312,
+      "learning_rate": 3.312647754137116e-06,
+      "loss": 0.1234,
+      "step": 1247
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2856595814228058,
+      "learning_rate": 3.311170212765958e-06,
+      "loss": 0.1168,
+      "step": 1248
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2901923656463623,
+      "learning_rate": 3.309692671394799e-06,
+      "loss": 0.1186,
+      "step": 1249
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2835504412651062,
+      "learning_rate": 3.308215130023641e-06,
+      "loss": 0.1112,
+      "step": 1250
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3015967011451721,
+      "learning_rate": 3.3067375886524826e-06,
+      "loss": 0.1201,
+      "step": 1251
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2766036093235016,
+      "learning_rate": 3.305260047281324e-06,
+      "loss": 0.1158,
+      "step": 1252
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2668551504611969,
+      "learning_rate": 3.3037825059101657e-06,
+      "loss": 0.1109,
+      "step": 1253
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.30145105719566345,
+      "learning_rate": 3.3023049645390076e-06,
+      "loss": 0.1193,
+      "step": 1254
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.29527950286865234,
+      "learning_rate": 3.3008274231678488e-06,
+      "loss": 0.1143,
+      "step": 1255
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2928523123264313,
+      "learning_rate": 3.2993498817966907e-06,
+      "loss": 0.1214,
+      "step": 1256
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.295089989900589,
+      "learning_rate": 3.297872340425532e-06,
+      "loss": 0.1149,
+      "step": 1257
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2784702479839325,
+      "learning_rate": 3.2963947990543738e-06,
+      "loss": 0.1156,
+      "step": 1258
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.29712343215942383,
+      "learning_rate": 3.2949172576832157e-06,
+      "loss": 0.1181,
+      "step": 1259
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2917318642139435,
+      "learning_rate": 3.293439716312057e-06,
+      "loss": 0.1246,
+      "step": 1260
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2938932180404663,
+      "learning_rate": 3.291962174940899e-06,
+      "loss": 0.1241,
+      "step": 1261
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.27165210247039795,
+      "learning_rate": 3.2904846335697403e-06,
+      "loss": 0.1141,
+      "step": 1262
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2818037271499634,
+      "learning_rate": 3.289007092198582e-06,
+      "loss": 0.1151,
+      "step": 1263
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2902469038963318,
+      "learning_rate": 3.2875295508274234e-06,
+      "loss": 0.115,
+      "step": 1264
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.27915653586387634,
+      "learning_rate": 3.286052009456265e-06,
+      "loss": 0.1118,
+      "step": 1265
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29197487235069275,
+      "learning_rate": 3.2845744680851065e-06,
+      "loss": 0.118,
+      "step": 1266
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2951302230358124,
+      "learning_rate": 3.2830969267139484e-06,
+      "loss": 0.1162,
+      "step": 1267
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29289188981056213,
+      "learning_rate": 3.2816193853427895e-06,
+      "loss": 0.1146,
+      "step": 1268
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3035213053226471,
+      "learning_rate": 3.2801418439716315e-06,
+      "loss": 0.1171,
+      "step": 1269
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3015804588794708,
+      "learning_rate": 3.2786643026004735e-06,
+      "loss": 0.1164,
+      "step": 1270
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2791287899017334,
+      "learning_rate": 3.2771867612293146e-06,
+      "loss": 0.1151,
+      "step": 1271
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29158976674079895,
+      "learning_rate": 3.2757092198581565e-06,
+      "loss": 0.1131,
+      "step": 1272
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.29008054733276367,
+      "learning_rate": 3.2742316784869976e-06,
+      "loss": 0.1166,
+      "step": 1273
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.28635627031326294,
+      "learning_rate": 3.2727541371158396e-06,
+      "loss": 0.1181,
+      "step": 1274
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.29918143153190613,
+      "learning_rate": 3.271276595744681e-06,
+      "loss": 0.1176,
+      "step": 1275
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2840932607650757,
+      "learning_rate": 3.2697990543735227e-06,
+      "loss": 0.1138,
+      "step": 1276
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3288319408893585,
+      "learning_rate": 3.268321513002364e-06,
+      "loss": 0.1172,
+      "step": 1277
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2990798354148865,
+      "learning_rate": 3.266843971631206e-06,
+      "loss": 0.1255,
+      "step": 1278
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.29733723402023315,
+      "learning_rate": 3.2653664302600473e-06,
+      "loss": 0.1167,
+      "step": 1279
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2832564115524292,
+      "learning_rate": 3.2638888888888892e-06,
+      "loss": 0.1163,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2958621382713318,
+      "learning_rate": 3.262411347517731e-06,
+      "loss": 0.1146,
+      "step": 1281
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.30595654249191284,
+      "learning_rate": 3.2609338061465723e-06,
+      "loss": 0.116,
+      "step": 1282
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.27691686153411865,
+      "learning_rate": 3.2594562647754142e-06,
+      "loss": 0.1165,
+      "step": 1283
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.30024150013923645,
+      "learning_rate": 3.2579787234042553e-06,
+      "loss": 0.1156,
+      "step": 1284
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.29381880164146423,
+      "learning_rate": 3.2565011820330973e-06,
+      "loss": 0.1225,
+      "step": 1285
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.27220723032951355,
+      "learning_rate": 3.255023640661939e-06,
+      "loss": 0.1115,
+      "step": 1286
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.28586480021476746,
+      "learning_rate": 3.2535460992907804e-06,
+      "loss": 0.1134,
+      "step": 1287
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2946845591068268,
+      "learning_rate": 3.252068557919622e-06,
+      "loss": 0.1144,
+      "step": 1288
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2757661044597626,
+      "learning_rate": 3.250591016548464e-06,
+      "loss": 0.1198,
+      "step": 1289
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.28749310970306396,
+      "learning_rate": 3.249113475177305e-06,
+      "loss": 0.1163,
+      "step": 1290
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.28712525963783264,
+      "learning_rate": 3.247635933806147e-06,
+      "loss": 0.1172,
+      "step": 1291
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.28629475831985474,
+      "learning_rate": 3.246158392434988e-06,
+      "loss": 0.1159,
+      "step": 1292
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.28771135210990906,
+      "learning_rate": 3.24468085106383e-06,
+      "loss": 0.1176,
+      "step": 1293
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.31358328461647034,
+      "learning_rate": 3.243203309692672e-06,
+      "loss": 0.1178,
+      "step": 1294
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2862052619457245,
+      "learning_rate": 3.241725768321513e-06,
+      "loss": 0.1137,
+      "step": 1295
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2851705551147461,
+      "learning_rate": 3.240248226950355e-06,
+      "loss": 0.1156,
+      "step": 1296
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.31489095091819763,
+      "learning_rate": 3.2387706855791966e-06,
+      "loss": 0.1193,
+      "step": 1297
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.27680546045303345,
+      "learning_rate": 3.237293144208038e-06,
+      "loss": 0.1164,
+      "step": 1298
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2815192937850952,
+      "learning_rate": 3.2358156028368796e-06,
+      "loss": 0.1156,
+      "step": 1299
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2700101435184479,
+      "learning_rate": 3.234338061465721e-06,
+      "loss": 0.1134,
+      "step": 1300
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3014972507953644,
+      "learning_rate": 3.2328605200945627e-06,
+      "loss": 0.1185,
+      "step": 1301
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.293879896402359,
+      "learning_rate": 3.2313829787234046e-06,
+      "loss": 0.1219,
+      "step": 1302
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2994425594806671,
+      "learning_rate": 3.2299054373522458e-06,
+      "loss": 0.122,
+      "step": 1303
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2938888669013977,
+      "learning_rate": 3.2284278959810877e-06,
+      "loss": 0.1181,
+      "step": 1304
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3070870339870453,
+      "learning_rate": 3.2269503546099297e-06,
+      "loss": 0.1158,
+      "step": 1305
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.29079729318618774,
+      "learning_rate": 3.2254728132387708e-06,
+      "loss": 0.1094,
+      "step": 1306
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2898445129394531,
+      "learning_rate": 3.2239952718676127e-06,
+      "loss": 0.1146,
+      "step": 1307
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2930821180343628,
+      "learning_rate": 3.222517730496454e-06,
+      "loss": 0.1169,
+      "step": 1308
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3003076910972595,
+      "learning_rate": 3.221040189125296e-06,
+      "loss": 0.1193,
+      "step": 1309
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.285940945148468,
+      "learning_rate": 3.2195626477541373e-06,
+      "loss": 0.1182,
+      "step": 1310
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.28318527340888977,
+      "learning_rate": 3.218085106382979e-06,
+      "loss": 0.115,
+      "step": 1311
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.28807568550109863,
+      "learning_rate": 3.2166075650118204e-06,
+      "loss": 0.113,
+      "step": 1312
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2884466350078583,
+      "learning_rate": 3.2151300236406624e-06,
+      "loss": 0.1144,
+      "step": 1313
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.31572261452674866,
+      "learning_rate": 3.2136524822695035e-06,
+      "loss": 0.118,
+      "step": 1314
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2980378270149231,
+      "learning_rate": 3.2121749408983454e-06,
+      "loss": 0.1181,
+      "step": 1315
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2927550673484802,
+      "learning_rate": 3.2106973995271874e-06,
+      "loss": 0.1172,
+      "step": 1316
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.29872190952301025,
+      "learning_rate": 3.2092198581560285e-06,
+      "loss": 0.1171,
+      "step": 1317
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.297059029340744,
+      "learning_rate": 3.2077423167848705e-06,
+      "loss": 0.1194,
+      "step": 1318
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.27931931614875793,
+      "learning_rate": 3.2062647754137116e-06,
+      "loss": 0.1222,
+      "step": 1319
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.29945141077041626,
+      "learning_rate": 3.2047872340425535e-06,
+      "loss": 0.1171,
+      "step": 1320
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.28636157512664795,
+      "learning_rate": 3.203309692671395e-06,
+      "loss": 0.1162,
+      "step": 1321
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.27227190136909485,
+      "learning_rate": 3.2018321513002366e-06,
+      "loss": 0.1143,
+      "step": 1322
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2920862138271332,
+      "learning_rate": 3.200354609929078e-06,
+      "loss": 0.118,
+      "step": 1323
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2938075065612793,
+      "learning_rate": 3.19887706855792e-06,
+      "loss": 0.118,
+      "step": 1324
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.28550827503204346,
+      "learning_rate": 3.197399527186761e-06,
+      "loss": 0.1141,
+      "step": 1325
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2871941328048706,
+      "learning_rate": 3.195921985815603e-06,
+      "loss": 0.114,
+      "step": 1326
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2798915505409241,
+      "learning_rate": 3.1944444444444443e-06,
+      "loss": 0.1144,
+      "step": 1327
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.28505071997642517,
+      "learning_rate": 3.1929669030732862e-06,
+      "loss": 0.1125,
+      "step": 1328
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2890617251396179,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 0.116,
+      "step": 1329
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.29411768913269043,
+      "learning_rate": 3.1900118203309693e-06,
+      "loss": 0.117,
+      "step": 1330
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2877364158630371,
+      "learning_rate": 3.1885342789598112e-06,
+      "loss": 0.1152,
+      "step": 1331
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2889752686023712,
+      "learning_rate": 3.1870567375886528e-06,
+      "loss": 0.1138,
+      "step": 1332
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.296420156955719,
+      "learning_rate": 3.1855791962174943e-06,
+      "loss": 0.1173,
+      "step": 1333
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.30378302931785583,
+      "learning_rate": 3.184101654846336e-06,
+      "loss": 0.1141,
+      "step": 1334
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2937588393688202,
+      "learning_rate": 3.1826241134751774e-06,
+      "loss": 0.1207,
+      "step": 1335
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.28481218218803406,
+      "learning_rate": 3.181146572104019e-06,
+      "loss": 0.1166,
+      "step": 1336
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.29722270369529724,
+      "learning_rate": 3.179669030732861e-06,
+      "loss": 0.1153,
+      "step": 1337
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.28433772921562195,
+      "learning_rate": 3.178191489361702e-06,
+      "loss": 0.1138,
+      "step": 1338
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2734004259109497,
+      "learning_rate": 3.176713947990544e-06,
+      "loss": 0.1119,
+      "step": 1339
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2874317169189453,
+      "learning_rate": 3.175236406619386e-06,
+      "loss": 0.1187,
+      "step": 1340
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.29149115085601807,
+      "learning_rate": 3.173758865248227e-06,
+      "loss": 0.1171,
+      "step": 1341
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2909756600856781,
+      "learning_rate": 3.172281323877069e-06,
+      "loss": 0.115,
+      "step": 1342
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.28182148933410645,
+      "learning_rate": 3.1708037825059105e-06,
+      "loss": 0.1113,
+      "step": 1343
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.28708261251449585,
+      "learning_rate": 3.169326241134752e-06,
+      "loss": 0.1113,
+      "step": 1344
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.29698339104652405,
+      "learning_rate": 3.167848699763594e-06,
+      "loss": 0.1158,
+      "step": 1345
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2869333028793335,
+      "learning_rate": 3.166371158392435e-06,
+      "loss": 0.1107,
+      "step": 1346
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3225131928920746,
+      "learning_rate": 3.164893617021277e-06,
+      "loss": 0.1166,
+      "step": 1347
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.28339093923568726,
+      "learning_rate": 3.1634160756501186e-06,
+      "loss": 0.1142,
+      "step": 1348
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2819472551345825,
+      "learning_rate": 3.16193853427896e-06,
+      "loss": 0.1169,
+      "step": 1349
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.29281067848205566,
+      "learning_rate": 3.1604609929078016e-06,
+      "loss": 0.1157,
+      "step": 1350
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.29216694831848145,
+      "learning_rate": 3.1589834515366436e-06,
+      "loss": 0.1184,
+      "step": 1351
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.29027190804481506,
+      "learning_rate": 3.1575059101654847e-06,
+      "loss": 0.1163,
+      "step": 1352
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.29170477390289307,
+      "learning_rate": 3.1560283687943267e-06,
+      "loss": 0.1222,
+      "step": 1353
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.274657279253006,
+      "learning_rate": 3.1545508274231678e-06,
+      "loss": 0.1133,
+      "step": 1354
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.29631003737449646,
+      "learning_rate": 3.1530732860520097e-06,
+      "loss": 0.1224,
+      "step": 1355
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.293163537979126,
+      "learning_rate": 3.1515957446808517e-06,
+      "loss": 0.1163,
+      "step": 1356
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2975763976573944,
+      "learning_rate": 3.150118203309693e-06,
+      "loss": 0.1158,
+      "step": 1357
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2844805121421814,
+      "learning_rate": 3.1486406619385348e-06,
+      "loss": 0.1176,
+      "step": 1358
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.29071927070617676,
+      "learning_rate": 3.1471631205673763e-06,
+      "loss": 0.1154,
+      "step": 1359
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2897449731826782,
+      "learning_rate": 3.145685579196218e-06,
+      "loss": 0.1159,
+      "step": 1360
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2943699061870575,
+      "learning_rate": 3.1442080378250594e-06,
+      "loss": 0.1227,
+      "step": 1361
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2778528928756714,
+      "learning_rate": 3.142730496453901e-06,
+      "loss": 0.116,
+      "step": 1362
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.26813358068466187,
+      "learning_rate": 3.1412529550827424e-06,
+      "loss": 0.1141,
+      "step": 1363
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2888466417789459,
+      "learning_rate": 3.1397754137115844e-06,
+      "loss": 0.1153,
+      "step": 1364
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3141460716724396,
+      "learning_rate": 3.1382978723404255e-06,
+      "loss": 0.1193,
+      "step": 1365
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2987734079360962,
+      "learning_rate": 3.1368203309692675e-06,
+      "loss": 0.1191,
+      "step": 1366
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2839658260345459,
+      "learning_rate": 3.1353427895981094e-06,
+      "loss": 0.1168,
+      "step": 1367
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.27612942457199097,
+      "learning_rate": 3.1338652482269505e-06,
+      "loss": 0.1149,
+      "step": 1368
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.27499064803123474,
+      "learning_rate": 3.1323877068557925e-06,
+      "loss": 0.1169,
+      "step": 1369
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2873942255973816,
+      "learning_rate": 3.1309101654846336e-06,
+      "loss": 0.1163,
+      "step": 1370
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2744005620479584,
+      "learning_rate": 3.1294326241134755e-06,
+      "loss": 0.1115,
+      "step": 1371
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2780010402202606,
+      "learning_rate": 3.127955082742317e-06,
+      "loss": 0.1197,
+      "step": 1372
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2969529628753662,
+      "learning_rate": 3.1264775413711586e-06,
+      "loss": 0.119,
+      "step": 1373
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2948795258998871,
+      "learning_rate": 3.125e-06,
+      "loss": 0.1157,
+      "step": 1374
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.28745150566101074,
+      "learning_rate": 3.123522458628842e-06,
+      "loss": 0.1164,
+      "step": 1375
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2909741699695587,
+      "learning_rate": 3.1220449172576832e-06,
+      "loss": 0.1177,
+      "step": 1376
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.29358652234077454,
+      "learning_rate": 3.120567375886525e-06,
+      "loss": 0.1187,
+      "step": 1377
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.28857558965682983,
+      "learning_rate": 3.119089834515367e-06,
+      "loss": 0.1153,
+      "step": 1378
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.29617875814437866,
+      "learning_rate": 3.1176122931442082e-06,
+      "loss": 0.1125,
+      "step": 1379
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.288015216588974,
+      "learning_rate": 3.11613475177305e-06,
+      "loss": 0.1164,
+      "step": 1380
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.28551438450813293,
+      "learning_rate": 3.1146572104018913e-06,
+      "loss": 0.1167,
+      "step": 1381
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.31125447154045105,
+      "learning_rate": 3.1131796690307333e-06,
+      "loss": 0.1206,
+      "step": 1382
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.27913016080856323,
+      "learning_rate": 3.111702127659575e-06,
+      "loss": 0.1148,
+      "step": 1383
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.29549098014831543,
+      "learning_rate": 3.1102245862884163e-06,
+      "loss": 0.1155,
+      "step": 1384
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.285230427980423,
+      "learning_rate": 3.108747044917258e-06,
+      "loss": 0.1169,
+      "step": 1385
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.29498451948165894,
+      "learning_rate": 3.1072695035461e-06,
+      "loss": 0.1164,
+      "step": 1386
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.312173068523407,
+      "learning_rate": 3.105791962174941e-06,
+      "loss": 0.1183,
+      "step": 1387
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.29572004079818726,
+      "learning_rate": 3.104314420803783e-06,
+      "loss": 0.1151,
+      "step": 1388
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.28423842787742615,
+      "learning_rate": 3.102836879432624e-06,
+      "loss": 0.1155,
+      "step": 1389
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2809999883174896,
+      "learning_rate": 3.101359338061466e-06,
+      "loss": 0.1192,
+      "step": 1390
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2858547568321228,
+      "learning_rate": 3.099881796690308e-06,
+      "loss": 0.1179,
+      "step": 1391
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2808387875556946,
+      "learning_rate": 3.098404255319149e-06,
+      "loss": 0.1144,
+      "step": 1392
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.27905139327049255,
+      "learning_rate": 3.096926713947991e-06,
+      "loss": 0.1203,
+      "step": 1393
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.28271642327308655,
+      "learning_rate": 3.0954491725768325e-06,
+      "loss": 0.1176,
+      "step": 1394
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.26291361451148987,
+      "learning_rate": 3.093971631205674e-06,
+      "loss": 0.1109,
+      "step": 1395
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.28382012248039246,
+      "learning_rate": 3.0924940898345156e-06,
+      "loss": 0.1174,
+      "step": 1396
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2760292887687683,
+      "learning_rate": 3.091016548463357e-06,
+      "loss": 0.1163,
+      "step": 1397
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2830650806427002,
+      "learning_rate": 3.0895390070921987e-06,
+      "loss": 0.1181,
+      "step": 1398
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.28073850274086,
+      "learning_rate": 3.0880614657210406e-06,
+      "loss": 0.117,
+      "step": 1399
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2902558743953705,
+      "learning_rate": 3.0865839243498817e-06,
+      "loss": 0.1201,
+      "step": 1400
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3109429180622101,
+      "learning_rate": 3.0851063829787237e-06,
+      "loss": 0.1175,
+      "step": 1401
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2945936620235443,
+      "learning_rate": 3.0836288416075656e-06,
+      "loss": 0.1243,
+      "step": 1402
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2797575294971466,
+      "learning_rate": 3.0821513002364067e-06,
+      "loss": 0.116,
+      "step": 1403
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.27586570382118225,
+      "learning_rate": 3.0806737588652487e-06,
+      "loss": 0.1144,
+      "step": 1404
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2825601100921631,
+      "learning_rate": 3.07919621749409e-06,
+      "loss": 0.1117,
+      "step": 1405
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2946073114871979,
+      "learning_rate": 3.0777186761229318e-06,
+      "loss": 0.1168,
+      "step": 1406
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2827392816543579,
+      "learning_rate": 3.0762411347517733e-06,
+      "loss": 0.1147,
+      "step": 1407
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.28599753975868225,
+      "learning_rate": 3.074763593380615e-06,
+      "loss": 0.1168,
+      "step": 1408
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2710088789463043,
+      "learning_rate": 3.0732860520094564e-06,
+      "loss": 0.1139,
+      "step": 1409
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.30704912543296814,
+      "learning_rate": 3.0718085106382983e-06,
+      "loss": 0.1184,
+      "step": 1410
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2830369472503662,
+      "learning_rate": 3.0703309692671394e-06,
+      "loss": 0.113,
+      "step": 1411
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2859196364879608,
+      "learning_rate": 3.0688534278959814e-06,
+      "loss": 0.1146,
+      "step": 1412
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.30158355832099915,
+      "learning_rate": 3.0673758865248233e-06,
+      "loss": 0.1214,
+      "step": 1413
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2923552989959717,
+      "learning_rate": 3.0658983451536645e-06,
+      "loss": 0.1156,
+      "step": 1414
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.28293243050575256,
+      "learning_rate": 3.0644208037825064e-06,
+      "loss": 0.1157,
+      "step": 1415
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.26670971512794495,
+      "learning_rate": 3.0629432624113475e-06,
+      "loss": 0.112,
+      "step": 1416
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.29871100187301636,
+      "learning_rate": 3.0614657210401895e-06,
+      "loss": 0.12,
+      "step": 1417
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.27904000878334045,
+      "learning_rate": 3.059988179669031e-06,
+      "loss": 0.1164,
+      "step": 1418
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.314921110868454,
+      "learning_rate": 3.0585106382978726e-06,
+      "loss": 0.1169,
+      "step": 1419
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.28920137882232666,
+      "learning_rate": 3.057033096926714e-06,
+      "loss": 0.1163,
+      "step": 1420
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2904154658317566,
+      "learning_rate": 3.055555555555556e-06,
+      "loss": 0.1169,
+      "step": 1421
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2875415086746216,
+      "learning_rate": 3.054078014184397e-06,
+      "loss": 0.1163,
+      "step": 1422
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.291875422000885,
+      "learning_rate": 3.052600472813239e-06,
+      "loss": 0.1202,
+      "step": 1423
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.29350513219833374,
+      "learning_rate": 3.0511229314420802e-06,
+      "loss": 0.1161,
+      "step": 1424
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3095230758190155,
+      "learning_rate": 3.049645390070922e-06,
+      "loss": 0.1169,
+      "step": 1425
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3327563405036926,
+      "learning_rate": 3.048167848699764e-06,
+      "loss": 0.1174,
+      "step": 1426
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.28892141580581665,
+      "learning_rate": 3.0466903073286052e-06,
+      "loss": 0.1169,
+      "step": 1427
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.28614214062690735,
+      "learning_rate": 3.045212765957447e-06,
+      "loss": 0.1164,
+      "step": 1428
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.28711581230163574,
+      "learning_rate": 3.0437352245862887e-06,
+      "loss": 0.1108,
+      "step": 1429
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.28866907954216003,
+      "learning_rate": 3.0422576832151303e-06,
+      "loss": 0.1183,
+      "step": 1430
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.28616321086883545,
+      "learning_rate": 3.040780141843972e-06,
+      "loss": 0.1196,
+      "step": 1431
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2817334830760956,
+      "learning_rate": 3.0393026004728133e-06,
+      "loss": 0.1157,
+      "step": 1432
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.28420883417129517,
+      "learning_rate": 3.037825059101655e-06,
+      "loss": 0.1184,
+      "step": 1433
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2866156995296478,
+      "learning_rate": 3.036347517730497e-06,
+      "loss": 0.1153,
+      "step": 1434
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2792140543460846,
+      "learning_rate": 3.034869976359338e-06,
+      "loss": 0.1154,
+      "step": 1435
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.28171539306640625,
+      "learning_rate": 3.03339243498818e-06,
+      "loss": 0.1164,
+      "step": 1436
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3061675429344177,
+      "learning_rate": 3.031914893617022e-06,
+      "loss": 0.12,
+      "step": 1437
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2839822769165039,
+      "learning_rate": 3.030437352245863e-06,
+      "loss": 0.1143,
+      "step": 1438
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2823212146759033,
+      "learning_rate": 3.028959810874705e-06,
+      "loss": 0.1157,
+      "step": 1439
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.29776132106781006,
+      "learning_rate": 3.0274822695035464e-06,
+      "loss": 0.1166,
+      "step": 1440
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2864833176136017,
+      "learning_rate": 3.026004728132388e-06,
+      "loss": 0.111,
+      "step": 1441
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.30720943212509155,
+      "learning_rate": 3.0245271867612295e-06,
+      "loss": 0.1152,
+      "step": 1442
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3302256166934967,
+      "learning_rate": 3.023049645390071e-06,
+      "loss": 0.1168,
+      "step": 1443
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.28479889035224915,
+      "learning_rate": 3.0215721040189126e-06,
+      "loss": 0.1133,
+      "step": 1444
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2852557301521301,
+      "learning_rate": 3.0200945626477545e-06,
+      "loss": 0.1158,
+      "step": 1445
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.30822867155075073,
+      "learning_rate": 3.0186170212765957e-06,
+      "loss": 0.1208,
+      "step": 1446
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.289862722158432,
+      "learning_rate": 3.0171394799054376e-06,
+      "loss": 0.115,
+      "step": 1447
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2934655249118805,
+      "learning_rate": 3.0156619385342796e-06,
+      "loss": 0.116,
+      "step": 1448
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.298281192779541,
+      "learning_rate": 3.0141843971631207e-06,
+      "loss": 0.1192,
+      "step": 1449
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.311158150434494,
+      "learning_rate": 3.0127068557919626e-06,
+      "loss": 0.1169,
+      "step": 1450
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.30020779371261597,
+      "learning_rate": 3.0112293144208037e-06,
+      "loss": 0.1148,
+      "step": 1451
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3068092167377472,
+      "learning_rate": 3.0097517730496457e-06,
+      "loss": 0.115,
+      "step": 1452
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2619610130786896,
+      "learning_rate": 3.0082742316784872e-06,
+      "loss": 0.1167,
+      "step": 1453
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.27616551518440247,
+      "learning_rate": 3.0067966903073288e-06,
+      "loss": 0.1152,
+      "step": 1454
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.28502732515335083,
+      "learning_rate": 3.0053191489361703e-06,
+      "loss": 0.1168,
+      "step": 1455
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3234563171863556,
+      "learning_rate": 3.0038416075650123e-06,
+      "loss": 0.1209,
+      "step": 1456
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3120412826538086,
+      "learning_rate": 3.0023640661938534e-06,
+      "loss": 0.1175,
+      "step": 1457
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.28492650389671326,
+      "learning_rate": 3.0008865248226953e-06,
+      "loss": 0.1137,
+      "step": 1458
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2917848825454712,
+      "learning_rate": 2.9994089834515364e-06,
+      "loss": 0.1174,
+      "step": 1459
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2962537109851837,
+      "learning_rate": 2.9979314420803784e-06,
+      "loss": 0.1126,
+      "step": 1460
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2825567126274109,
+      "learning_rate": 2.9964539007092203e-06,
+      "loss": 0.1136,
+      "step": 1461
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.31072762608528137,
+      "learning_rate": 2.9949763593380615e-06,
+      "loss": 0.1124,
+      "step": 1462
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.29849717020988464,
+      "learning_rate": 2.9934988179669034e-06,
+      "loss": 0.1226,
+      "step": 1463
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3104895055294037,
+      "learning_rate": 2.992021276595745e-06,
+      "loss": 0.1183,
+      "step": 1464
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2977125346660614,
+      "learning_rate": 2.9905437352245865e-06,
+      "loss": 0.1127,
+      "step": 1465
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3035031259059906,
+      "learning_rate": 2.989066193853428e-06,
+      "loss": 0.1183,
+      "step": 1466
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.29253464937210083,
+      "learning_rate": 2.9875886524822696e-06,
+      "loss": 0.1208,
+      "step": 1467
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.29931047558784485,
+      "learning_rate": 2.986111111111111e-06,
+      "loss": 0.1162,
+      "step": 1468
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3146195411682129,
+      "learning_rate": 2.984633569739953e-06,
+      "loss": 0.118,
+      "step": 1469
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3003896474838257,
+      "learning_rate": 2.9831560283687946e-06,
+      "loss": 0.118,
+      "step": 1470
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2987372577190399,
+      "learning_rate": 2.981678486997636e-06,
+      "loss": 0.1175,
+      "step": 1471
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.32163006067276,
+      "learning_rate": 2.980200945626478e-06,
+      "loss": 0.1132,
+      "step": 1472
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.29624757170677185,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 0.1133,
+      "step": 1473
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.322757750749588,
+      "learning_rate": 2.977245862884161e-06,
+      "loss": 0.1244,
+      "step": 1474
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.31568971276283264,
+      "learning_rate": 2.9757683215130027e-06,
+      "loss": 0.1227,
+      "step": 1475
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.283010333776474,
+      "learning_rate": 2.974290780141844e-06,
+      "loss": 0.1142,
+      "step": 1476
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.29099997878074646,
+      "learning_rate": 2.972813238770686e-06,
+      "loss": 0.1183,
+      "step": 1477
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.31473010778427124,
+      "learning_rate": 2.9713356973995273e-06,
+      "loss": 0.1189,
+      "step": 1478
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2891026735305786,
+      "learning_rate": 2.9698581560283692e-06,
+      "loss": 0.1167,
+      "step": 1479
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2985316216945648,
+      "learning_rate": 2.9683806146572108e-06,
+      "loss": 0.1196,
+      "step": 1480
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.27723076939582825,
+      "learning_rate": 2.9669030732860523e-06,
+      "loss": 0.1146,
+      "step": 1481
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30301421880722046,
+      "learning_rate": 2.965425531914894e-06,
+      "loss": 0.1194,
+      "step": 1482
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2973959147930145,
+      "learning_rate": 2.9639479905437358e-06,
+      "loss": 0.1143,
+      "step": 1483
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.33804571628570557,
+      "learning_rate": 2.962470449172577e-06,
+      "loss": 0.1219,
+      "step": 1484
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.30722787976264954,
+      "learning_rate": 2.960992907801419e-06,
+      "loss": 0.1183,
+      "step": 1485
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2922459542751312,
+      "learning_rate": 2.95951536643026e-06,
+      "loss": 0.1139,
+      "step": 1486
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.28375548124313354,
+      "learning_rate": 2.958037825059102e-06,
+      "loss": 0.1132,
+      "step": 1487
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3067808449268341,
+      "learning_rate": 2.956560283687944e-06,
+      "loss": 0.1142,
+      "step": 1488
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3010546565055847,
+      "learning_rate": 2.955082742316785e-06,
+      "loss": 0.1167,
+      "step": 1489
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2854355275630951,
+      "learning_rate": 2.953605200945627e-06,
+      "loss": 0.1122,
+      "step": 1490
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2954955995082855,
+      "learning_rate": 2.9521276595744685e-06,
+      "loss": 0.1207,
+      "step": 1491
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2881651222705841,
+      "learning_rate": 2.95065011820331e-06,
+      "loss": 0.1188,
+      "step": 1492
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2822742462158203,
+      "learning_rate": 2.9491725768321515e-06,
+      "loss": 0.1118,
+      "step": 1493
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2936824560165405,
+      "learning_rate": 2.947695035460993e-06,
+      "loss": 0.1193,
+      "step": 1494
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3326472043991089,
+      "learning_rate": 2.9462174940898346e-06,
+      "loss": 0.1187,
+      "step": 1495
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2815712094306946,
+      "learning_rate": 2.9447399527186766e-06,
+      "loss": 0.1089,
+      "step": 1496
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2846650183200836,
+      "learning_rate": 2.9432624113475177e-06,
+      "loss": 0.1166,
+      "step": 1497
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.291079044342041,
+      "learning_rate": 2.9417848699763596e-06,
+      "loss": 0.1114,
+      "step": 1498
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.28491029143333435,
+      "learning_rate": 2.9403073286052016e-06,
+      "loss": 0.1194,
+      "step": 1499
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.29006442427635193,
+      "learning_rate": 2.9388297872340427e-06,
+      "loss": 0.1199,
+      "step": 1500
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2869240939617157,
+      "learning_rate": 2.9373522458628847e-06,
+      "loss": 0.1154,
+      "step": 1501
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2912914454936981,
+      "learning_rate": 2.9358747044917258e-06,
+      "loss": 0.1208,
+      "step": 1502
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2944636046886444,
+      "learning_rate": 2.9343971631205677e-06,
+      "loss": 0.1169,
+      "step": 1503
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.26580101251602173,
+      "learning_rate": 2.9329196217494093e-06,
+      "loss": 0.111,
+      "step": 1504
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3002028465270996,
+      "learning_rate": 2.931442080378251e-06,
+      "loss": 0.1128,
+      "step": 1505
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2896462678909302,
+      "learning_rate": 2.9299645390070923e-06,
+      "loss": 0.115,
+      "step": 1506
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.28620320558547974,
+      "learning_rate": 2.9284869976359343e-06,
+      "loss": 0.1152,
+      "step": 1507
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.29929423332214355,
+      "learning_rate": 2.9270094562647754e-06,
+      "loss": 0.1178,
+      "step": 1508
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.27214059233665466,
+      "learning_rate": 2.9255319148936174e-06,
+      "loss": 0.1159,
+      "step": 1509
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2829926013946533,
+      "learning_rate": 2.9240543735224593e-06,
+      "loss": 0.1144,
+      "step": 1510
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2765200436115265,
+      "learning_rate": 2.9225768321513004e-06,
+      "loss": 0.1175,
+      "step": 1511
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2721102237701416,
+      "learning_rate": 2.9210992907801424e-06,
+      "loss": 0.1122,
+      "step": 1512
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2918475270271301,
+      "learning_rate": 2.9196217494089835e-06,
+      "loss": 0.1193,
+      "step": 1513
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.28817111253738403,
+      "learning_rate": 2.9181442080378254e-06,
+      "loss": 0.117,
+      "step": 1514
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.28446123003959656,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.1151,
+      "step": 1515
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2811199128627777,
+      "learning_rate": 2.9151891252955085e-06,
+      "loss": 0.1124,
+      "step": 1516
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.27345141768455505,
+      "learning_rate": 2.91371158392435e-06,
+      "loss": 0.1143,
+      "step": 1517
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2884834110736847,
+      "learning_rate": 2.912234042553192e-06,
+      "loss": 0.1125,
+      "step": 1518
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.28277963399887085,
+      "learning_rate": 2.910756501182033e-06,
+      "loss": 0.1142,
+      "step": 1519
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3021559715270996,
+      "learning_rate": 2.909278959810875e-06,
+      "loss": 0.1207,
+      "step": 1520
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3026861548423767,
+      "learning_rate": 2.907801418439716e-06,
+      "loss": 0.1192,
+      "step": 1521
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.28176048398017883,
+      "learning_rate": 2.906323877068558e-06,
+      "loss": 0.1177,
+      "step": 1522
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.28341928124427795,
+      "learning_rate": 2.9048463356974e-06,
+      "loss": 0.1156,
+      "step": 1523
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2923755943775177,
+      "learning_rate": 2.903368794326241e-06,
+      "loss": 0.1137,
+      "step": 1524
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2984902560710907,
+      "learning_rate": 2.901891252955083e-06,
+      "loss": 0.1184,
+      "step": 1525
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.27732157707214355,
+      "learning_rate": 2.9004137115839247e-06,
+      "loss": 0.1163,
+      "step": 1526
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2703724503517151,
+      "learning_rate": 2.8989361702127662e-06,
+      "loss": 0.111,
+      "step": 1527
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.27413684129714966,
+      "learning_rate": 2.8974586288416078e-06,
+      "loss": 0.1129,
+      "step": 1528
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.29157984256744385,
+      "learning_rate": 2.8959810874704493e-06,
+      "loss": 0.118,
+      "step": 1529
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2907381057739258,
+      "learning_rate": 2.894503546099291e-06,
+      "loss": 0.117,
+      "step": 1530
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.292966365814209,
+      "learning_rate": 2.8930260047281328e-06,
+      "loss": 0.1202,
+      "step": 1531
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.28155919909477234,
+      "learning_rate": 2.891548463356974e-06,
+      "loss": 0.1188,
+      "step": 1532
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3022976219654083,
+      "learning_rate": 2.890070921985816e-06,
+      "loss": 0.114,
+      "step": 1533
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.28788459300994873,
+      "learning_rate": 2.888593380614658e-06,
+      "loss": 0.1145,
+      "step": 1534
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.27871569991111755,
+      "learning_rate": 2.887115839243499e-06,
+      "loss": 0.1138,
+      "step": 1535
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.28395092487335205,
+      "learning_rate": 2.885638297872341e-06,
+      "loss": 0.1178,
+      "step": 1536
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.27866610884666443,
+      "learning_rate": 2.8841607565011824e-06,
+      "loss": 0.1153,
+      "step": 1537
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.30172938108444214,
+      "learning_rate": 2.882683215130024e-06,
+      "loss": 0.1217,
+      "step": 1538
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2854112684726715,
+      "learning_rate": 2.8812056737588655e-06,
+      "loss": 0.1156,
+      "step": 1539
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.27580493688583374,
+      "learning_rate": 2.879728132387707e-06,
+      "loss": 0.1082,
+      "step": 1540
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.28709912300109863,
+      "learning_rate": 2.8782505910165485e-06,
+      "loss": 0.1162,
+      "step": 1541
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2951715290546417,
+      "learning_rate": 2.8767730496453905e-06,
+      "loss": 0.1146,
+      "step": 1542
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3011291027069092,
+      "learning_rate": 2.8752955082742316e-06,
+      "loss": 0.1156,
+      "step": 1543
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.31616729497909546,
+      "learning_rate": 2.8738179669030736e-06,
+      "loss": 0.1185,
+      "step": 1544
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.301300048828125,
+      "learning_rate": 2.8723404255319155e-06,
+      "loss": 0.1162,
+      "step": 1545
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2878454327583313,
+      "learning_rate": 2.8708628841607566e-06,
+      "loss": 0.1137,
+      "step": 1546
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.29604196548461914,
+      "learning_rate": 2.8693853427895986e-06,
+      "loss": 0.1183,
+      "step": 1547
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.30757060647010803,
+      "learning_rate": 2.8679078014184397e-06,
+      "loss": 0.1165,
+      "step": 1548
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.27922627329826355,
+      "learning_rate": 2.8664302600472817e-06,
+      "loss": 0.1099,
+      "step": 1549
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3090381920337677,
+      "learning_rate": 2.864952718676123e-06,
+      "loss": 0.1154,
+      "step": 1550
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2988344430923462,
+      "learning_rate": 2.8634751773049647e-06,
+      "loss": 0.1085,
+      "step": 1551
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.31521934270858765,
+      "learning_rate": 2.8619976359338063e-06,
+      "loss": 0.1196,
+      "step": 1552
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.296637624502182,
+      "learning_rate": 2.8605200945626482e-06,
+      "loss": 0.1175,
+      "step": 1553
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.31049177050590515,
+      "learning_rate": 2.8590425531914893e-06,
+      "loss": 0.1193,
+      "step": 1554
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2768094837665558,
+      "learning_rate": 2.8575650118203313e-06,
+      "loss": 0.1119,
+      "step": 1555
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3143351674079895,
+      "learning_rate": 2.8560874704491724e-06,
+      "loss": 0.1183,
+      "step": 1556
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3294629454612732,
+      "learning_rate": 2.8546099290780144e-06,
+      "loss": 0.1163,
+      "step": 1557
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.30514198541641235,
+      "learning_rate": 2.8531323877068563e-06,
+      "loss": 0.1162,
+      "step": 1558
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2866133451461792,
+      "learning_rate": 2.8516548463356974e-06,
+      "loss": 0.1168,
+      "step": 1559
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2778390944004059,
+      "learning_rate": 2.8501773049645394e-06,
+      "loss": 0.1146,
+      "step": 1560
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.30379387736320496,
+      "learning_rate": 2.848699763593381e-06,
+      "loss": 0.1167,
+      "step": 1561
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.31135454773902893,
+      "learning_rate": 2.8472222222222224e-06,
+      "loss": 0.1174,
+      "step": 1562
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.30998438596725464,
+      "learning_rate": 2.845744680851064e-06,
+      "loss": 0.1179,
+      "step": 1563
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.29471930861473083,
+      "learning_rate": 2.8442671394799055e-06,
+      "loss": 0.1187,
+      "step": 1564
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2846895754337311,
+      "learning_rate": 2.842789598108747e-06,
+      "loss": 0.1111,
+      "step": 1565
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2967732846736908,
+      "learning_rate": 2.841312056737589e-06,
+      "loss": 0.1134,
+      "step": 1566
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.31540733575820923,
+      "learning_rate": 2.83983451536643e-06,
+      "loss": 0.1141,
+      "step": 1567
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3168407678604126,
+      "learning_rate": 2.838356973995272e-06,
+      "loss": 0.1193,
+      "step": 1568
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.32081612944602966,
+      "learning_rate": 2.836879432624114e-06,
+      "loss": 0.1201,
+      "step": 1569
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2976883351802826,
+      "learning_rate": 2.835401891252955e-06,
+      "loss": 0.1185,
+      "step": 1570
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3068239390850067,
+      "learning_rate": 2.833924349881797e-06,
+      "loss": 0.1178,
+      "step": 1571
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.27740710973739624,
+      "learning_rate": 2.8324468085106386e-06,
+      "loss": 0.1107,
+      "step": 1572
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.30583858489990234,
+      "learning_rate": 2.83096926713948e-06,
+      "loss": 0.1176,
+      "step": 1573
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.28787466883659363,
+      "learning_rate": 2.8294917257683217e-06,
+      "loss": 0.1114,
+      "step": 1574
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3005199730396271,
+      "learning_rate": 2.8280141843971632e-06,
+      "loss": 0.1203,
+      "step": 1575
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.30560943484306335,
+      "learning_rate": 2.8265366430260048e-06,
+      "loss": 0.1238,
+      "step": 1576
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.29109084606170654,
+      "learning_rate": 2.8250591016548467e-06,
+      "loss": 0.1172,
+      "step": 1577
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.28717246651649475,
+      "learning_rate": 2.823581560283688e-06,
+      "loss": 0.1168,
+      "step": 1578
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.29949572682380676,
+      "learning_rate": 2.8221040189125298e-06,
+      "loss": 0.119,
+      "step": 1579
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2828640043735504,
+      "learning_rate": 2.8206264775413717e-06,
+      "loss": 0.1152,
+      "step": 1580
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3093225955963135,
+      "learning_rate": 2.819148936170213e-06,
+      "loss": 0.1153,
+      "step": 1581
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3024139106273651,
+      "learning_rate": 2.817671394799055e-06,
+      "loss": 0.1193,
+      "step": 1582
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2919802963733673,
+      "learning_rate": 2.816193853427896e-06,
+      "loss": 0.113,
+      "step": 1583
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.28762662410736084,
+      "learning_rate": 2.814716312056738e-06,
+      "loss": 0.1161,
+      "step": 1584
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2896806299686432,
+      "learning_rate": 2.8132387706855794e-06,
+      "loss": 0.1172,
+      "step": 1585
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2917637526988983,
+      "learning_rate": 2.811761229314421e-06,
+      "loss": 0.1161,
+      "step": 1586
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.286636084318161,
+      "learning_rate": 2.8102836879432625e-06,
+      "loss": 0.1161,
+      "step": 1587
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2838037610054016,
+      "learning_rate": 2.8088061465721044e-06,
+      "loss": 0.1169,
+      "step": 1588
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.286459356546402,
+      "learning_rate": 2.8073286052009455e-06,
+      "loss": 0.1146,
+      "step": 1589
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3094099760055542,
+      "learning_rate": 2.8058510638297875e-06,
+      "loss": 0.1199,
+      "step": 1590
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2887307107448578,
+      "learning_rate": 2.8043735224586286e-06,
+      "loss": 0.1157,
+      "step": 1591
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3044376075267792,
+      "learning_rate": 2.8028959810874706e-06,
+      "loss": 0.1198,
+      "step": 1592
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2991600036621094,
+      "learning_rate": 2.8014184397163125e-06,
+      "loss": 0.1184,
+      "step": 1593
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.292858362197876,
+      "learning_rate": 2.7999408983451536e-06,
+      "loss": 0.1176,
+      "step": 1594
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.29214680194854736,
+      "learning_rate": 2.7984633569739956e-06,
+      "loss": 0.1194,
+      "step": 1595
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2950979173183441,
+      "learning_rate": 2.796985815602837e-06,
+      "loss": 0.1163,
+      "step": 1596
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.280679851770401,
+      "learning_rate": 2.7955082742316787e-06,
+      "loss": 0.111,
+      "step": 1597
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.28804004192352295,
+      "learning_rate": 2.79403073286052e-06,
+      "loss": 0.1179,
+      "step": 1598
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.28490689396858215,
+      "learning_rate": 2.7925531914893617e-06,
+      "loss": 0.1156,
+      "step": 1599
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.29740217328071594,
+      "learning_rate": 2.7910756501182033e-06,
+      "loss": 0.113,
+      "step": 1600
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2880135774612427,
+      "learning_rate": 2.7895981087470452e-06,
+      "loss": 0.1185,
+      "step": 1601
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3100912570953369,
+      "learning_rate": 2.7881205673758868e-06,
+      "loss": 0.1193,
+      "step": 1602
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2878260612487793,
+      "learning_rate": 2.7866430260047283e-06,
+      "loss": 0.1111,
+      "step": 1603
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3192022144794464,
+      "learning_rate": 2.7851654846335702e-06,
+      "loss": 0.119,
+      "step": 1604
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2892776429653168,
+      "learning_rate": 2.7836879432624114e-06,
+      "loss": 0.1146,
+      "step": 1605
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2957354187965393,
+      "learning_rate": 2.7822104018912533e-06,
+      "loss": 0.1145,
+      "step": 1606
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2983720898628235,
+      "learning_rate": 2.7807328605200953e-06,
+      "loss": 0.1177,
+      "step": 1607
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2891678810119629,
+      "learning_rate": 2.7792553191489364e-06,
+      "loss": 0.1168,
+      "step": 1608
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.29278263449668884,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.1181,
+      "step": 1609
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.31028038263320923,
+      "learning_rate": 2.7763002364066194e-06,
+      "loss": 0.1188,
+      "step": 1610
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.30459192395210266,
+      "learning_rate": 2.7748226950354614e-06,
+      "loss": 0.1181,
+      "step": 1611
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3110455870628357,
+      "learning_rate": 2.773345153664303e-06,
+      "loss": 0.1185,
+      "step": 1612
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.28712430596351624,
+      "learning_rate": 2.7718676122931445e-06,
+      "loss": 0.1177,
+      "step": 1613
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.31011962890625,
+      "learning_rate": 2.770390070921986e-06,
+      "loss": 0.1128,
+      "step": 1614
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.29422610998153687,
+      "learning_rate": 2.768912529550828e-06,
+      "loss": 0.1169,
+      "step": 1615
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3019821047782898,
+      "learning_rate": 2.767434988179669e-06,
+      "loss": 0.1158,
+      "step": 1616
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.30289024114608765,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 0.1153,
+      "step": 1617
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.28817352652549744,
+      "learning_rate": 2.764479905437352e-06,
+      "loss": 0.1138,
+      "step": 1618
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2857765853404999,
+      "learning_rate": 2.763002364066194e-06,
+      "loss": 0.1166,
+      "step": 1619
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2848632335662842,
+      "learning_rate": 2.761524822695036e-06,
+      "loss": 0.1109,
+      "step": 1620
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.29019591212272644,
+      "learning_rate": 2.760047281323877e-06,
+      "loss": 0.1126,
+      "step": 1621
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2927004098892212,
+      "learning_rate": 2.758569739952719e-06,
+      "loss": 0.1119,
+      "step": 1622
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.31021648645401,
+      "learning_rate": 2.7570921985815607e-06,
+      "loss": 0.1185,
+      "step": 1623
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2918512523174286,
+      "learning_rate": 2.755614657210402e-06,
+      "loss": 0.1133,
+      "step": 1624
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2973274290561676,
+      "learning_rate": 2.7541371158392437e-06,
+      "loss": 0.1167,
+      "step": 1625
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2955513000488281,
+      "learning_rate": 2.7526595744680853e-06,
+      "loss": 0.1169,
+      "step": 1626
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.28820544481277466,
+      "learning_rate": 2.7511820330969268e-06,
+      "loss": 0.1134,
+      "step": 1627
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3079513907432556,
+      "learning_rate": 2.7497044917257687e-06,
+      "loss": 0.1178,
+      "step": 1628
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.29799094796180725,
+      "learning_rate": 2.74822695035461e-06,
+      "loss": 0.116,
+      "step": 1629
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2845512926578522,
+      "learning_rate": 2.746749408983452e-06,
+      "loss": 0.116,
+      "step": 1630
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.29911503195762634,
+      "learning_rate": 2.7452718676122938e-06,
+      "loss": 0.1167,
+      "step": 1631
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.29364052414894104,
+      "learning_rate": 2.743794326241135e-06,
+      "loss": 0.1151,
+      "step": 1632
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.30255818367004395,
+      "learning_rate": 2.742316784869977e-06,
+      "loss": 0.1144,
+      "step": 1633
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3053136169910431,
+      "learning_rate": 2.740839243498818e-06,
+      "loss": 0.1175,
+      "step": 1634
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.29486554861068726,
+      "learning_rate": 2.73936170212766e-06,
+      "loss": 0.12,
+      "step": 1635
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.28754115104675293,
+      "learning_rate": 2.7378841607565014e-06,
+      "loss": 0.1135,
+      "step": 1636
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.30982092022895813,
+      "learning_rate": 2.736406619385343e-06,
+      "loss": 0.1194,
+      "step": 1637
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3011505901813507,
+      "learning_rate": 2.7349290780141845e-06,
+      "loss": 0.1188,
+      "step": 1638
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.28660574555397034,
+      "learning_rate": 2.7334515366430265e-06,
+      "loss": 0.1126,
+      "step": 1639
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2968364953994751,
+      "learning_rate": 2.7319739952718676e-06,
+      "loss": 0.1126,
+      "step": 1640
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2886878252029419,
+      "learning_rate": 2.7304964539007095e-06,
+      "loss": 0.1108,
+      "step": 1641
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.30457553267478943,
+      "learning_rate": 2.7290189125295515e-06,
+      "loss": 0.1171,
+      "step": 1642
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.30036142468452454,
+      "learning_rate": 2.7275413711583926e-06,
+      "loss": 0.1153,
+      "step": 1643
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2911832332611084,
+      "learning_rate": 2.7260638297872346e-06,
+      "loss": 0.1194,
+      "step": 1644
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2887232303619385,
+      "learning_rate": 2.7245862884160757e-06,
+      "loss": 0.1144,
+      "step": 1645
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2982259690761566,
+      "learning_rate": 2.7231087470449176e-06,
+      "loss": 0.1151,
+      "step": 1646
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.30018508434295654,
+      "learning_rate": 2.721631205673759e-06,
+      "loss": 0.1184,
+      "step": 1647
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.30180326104164124,
+      "learning_rate": 2.7201536643026007e-06,
+      "loss": 0.1171,
+      "step": 1648
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2872734069824219,
+      "learning_rate": 2.7186761229314422e-06,
+      "loss": 0.111,
+      "step": 1649
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2746340036392212,
+      "learning_rate": 2.717198581560284e-06,
+      "loss": 0.1123,
+      "step": 1650
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.28895196318626404,
+      "learning_rate": 2.7157210401891253e-06,
+      "loss": 0.116,
+      "step": 1651
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.31148502230644226,
+      "learning_rate": 2.7142434988179672e-06,
+      "loss": 0.1149,
+      "step": 1652
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2928013503551483,
+      "learning_rate": 2.7127659574468084e-06,
+      "loss": 0.1197,
+      "step": 1653
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2865014672279358,
+      "learning_rate": 2.7112884160756503e-06,
+      "loss": 0.1168,
+      "step": 1654
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.27421095967292786,
+      "learning_rate": 2.7098108747044923e-06,
+      "loss": 0.1136,
+      "step": 1655
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.27798932790756226,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 0.1124,
+      "step": 1656
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3020572364330292,
+      "learning_rate": 2.7068557919621753e-06,
+      "loss": 0.1165,
+      "step": 1657
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.295608252286911,
+      "learning_rate": 2.705378250591017e-06,
+      "loss": 0.1144,
+      "step": 1658
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2956421673297882,
+      "learning_rate": 2.7039007092198584e-06,
+      "loss": 0.1161,
+      "step": 1659
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.28097352385520935,
+      "learning_rate": 2.7024231678487e-06,
+      "loss": 0.1124,
+      "step": 1660
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2871864438056946,
+      "learning_rate": 2.7009456264775415e-06,
+      "loss": 0.1153,
+      "step": 1661
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2675824761390686,
+      "learning_rate": 2.699468085106383e-06,
+      "loss": 0.1139,
+      "step": 1662
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2944943606853485,
+      "learning_rate": 2.697990543735225e-06,
+      "loss": 0.1165,
+      "step": 1663
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.28529927134513855,
+      "learning_rate": 2.696513002364066e-06,
+      "loss": 0.1194,
+      "step": 1664
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2922259271144867,
+      "learning_rate": 2.695035460992908e-06,
+      "loss": 0.1171,
+      "step": 1665
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.33257216215133667,
+      "learning_rate": 2.69355791962175e-06,
+      "loss": 0.1182,
+      "step": 1666
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2891588807106018,
+      "learning_rate": 2.692080378250591e-06,
+      "loss": 0.1177,
+      "step": 1667
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2805059850215912,
+      "learning_rate": 2.690602836879433e-06,
+      "loss": 0.117,
+      "step": 1668
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2833464443683624,
+      "learning_rate": 2.6891252955082746e-06,
+      "loss": 0.1177,
+      "step": 1669
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2865210771560669,
+      "learning_rate": 2.687647754137116e-06,
+      "loss": 0.1189,
+      "step": 1670
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3043903708457947,
+      "learning_rate": 2.6861702127659577e-06,
+      "loss": 0.1183,
+      "step": 1671
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2894761562347412,
+      "learning_rate": 2.684692671394799e-06,
+      "loss": 0.115,
+      "step": 1672
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2813999354839325,
+      "learning_rate": 2.6832151300236407e-06,
+      "loss": 0.1143,
+      "step": 1673
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.28010547161102295,
+      "learning_rate": 2.6817375886524827e-06,
+      "loss": 0.1139,
+      "step": 1674
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2791883051395416,
+      "learning_rate": 2.6802600472813238e-06,
+      "loss": 0.1115,
+      "step": 1675
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.296926349401474,
+      "learning_rate": 2.6787825059101657e-06,
+      "loss": 0.1174,
+      "step": 1676
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2935212552547455,
+      "learning_rate": 2.6773049645390077e-06,
+      "loss": 0.1158,
+      "step": 1677
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2720209062099457,
+      "learning_rate": 2.675827423167849e-06,
+      "loss": 0.1083,
+      "step": 1678
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2924852967262268,
+      "learning_rate": 2.6743498817966908e-06,
+      "loss": 0.1153,
+      "step": 1679
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2844301164150238,
+      "learning_rate": 2.672872340425532e-06,
+      "loss": 0.1186,
+      "step": 1680
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3042744994163513,
+      "learning_rate": 2.671394799054374e-06,
+      "loss": 0.117,
+      "step": 1681
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.28683459758758545,
+      "learning_rate": 2.6699172576832154e-06,
+      "loss": 0.119,
+      "step": 1682
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.28836214542388916,
+      "learning_rate": 2.668439716312057e-06,
+      "loss": 0.1157,
+      "step": 1683
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.27478861808776855,
+      "learning_rate": 2.6669621749408984e-06,
+      "loss": 0.111,
+      "step": 1684
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.29842665791511536,
+      "learning_rate": 2.6654846335697404e-06,
+      "loss": 0.1121,
+      "step": 1685
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3033268451690674,
+      "learning_rate": 2.6640070921985815e-06,
+      "loss": 0.1157,
+      "step": 1686
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.31670793890953064,
+      "learning_rate": 2.6625295508274235e-06,
+      "loss": 0.1194,
+      "step": 1687
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.29516011476516724,
+      "learning_rate": 2.6610520094562646e-06,
+      "loss": 0.115,
+      "step": 1688
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2845121920108795,
+      "learning_rate": 2.6595744680851065e-06,
+      "loss": 0.115,
+      "step": 1689
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.29508548974990845,
+      "learning_rate": 2.6580969267139485e-06,
+      "loss": 0.1187,
+      "step": 1690
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2860552668571472,
+      "learning_rate": 2.6566193853427896e-06,
+      "loss": 0.1141,
+      "step": 1691
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.28281691670417786,
+      "learning_rate": 2.6551418439716316e-06,
+      "loss": 0.117,
+      "step": 1692
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.27992716431617737,
+      "learning_rate": 2.653664302600473e-06,
+      "loss": 0.1196,
+      "step": 1693
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2983381748199463,
+      "learning_rate": 2.6521867612293146e-06,
+      "loss": 0.1153,
+      "step": 1694
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.29381421208381653,
+      "learning_rate": 2.650709219858156e-06,
+      "loss": 0.1158,
+      "step": 1695
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2974252700805664,
+      "learning_rate": 2.6492316784869977e-06,
+      "loss": 0.1165,
+      "step": 1696
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2835535705089569,
+      "learning_rate": 2.6477541371158392e-06,
+      "loss": 0.1163,
+      "step": 1697
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.290768563747406,
+      "learning_rate": 2.646276595744681e-06,
+      "loss": 0.1146,
+      "step": 1698
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.28099343180656433,
+      "learning_rate": 2.6447990543735223e-06,
+      "loss": 0.1158,
+      "step": 1699
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.28536152839660645,
+      "learning_rate": 2.6433215130023642e-06,
+      "loss": 0.1161,
+      "step": 1700
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2696172893047333,
+      "learning_rate": 2.641843971631206e-06,
+      "loss": 0.1075,
+      "step": 1701
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.28655049204826355,
+      "learning_rate": 2.6403664302600473e-06,
+      "loss": 0.1166,
+      "step": 1702
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2982497215270996,
+      "learning_rate": 2.6388888888888893e-06,
+      "loss": 0.118,
+      "step": 1703
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.28479912877082825,
+      "learning_rate": 2.637411347517731e-06,
+      "loss": 0.117,
+      "step": 1704
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2885390520095825,
+      "learning_rate": 2.6359338061465723e-06,
+      "loss": 0.112,
+      "step": 1705
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.29115843772888184,
+      "learning_rate": 2.634456264775414e-06,
+      "loss": 0.1135,
+      "step": 1706
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.29853296279907227,
+      "learning_rate": 2.6329787234042554e-06,
+      "loss": 0.1142,
+      "step": 1707
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2896803915500641,
+      "learning_rate": 2.631501182033097e-06,
+      "loss": 0.1155,
+      "step": 1708
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2919914424419403,
+      "learning_rate": 2.630023640661939e-06,
+      "loss": 0.1127,
+      "step": 1709
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2995082139968872,
+      "learning_rate": 2.62854609929078e-06,
+      "loss": 0.1177,
+      "step": 1710
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3067772090435028,
+      "learning_rate": 2.627068557919622e-06,
+      "loss": 0.1153,
+      "step": 1711
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.27984389662742615,
+      "learning_rate": 2.625591016548464e-06,
+      "loss": 0.113,
+      "step": 1712
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2880931496620178,
+      "learning_rate": 2.624113475177305e-06,
+      "loss": 0.114,
+      "step": 1713
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.29385045170783997,
+      "learning_rate": 2.622635933806147e-06,
+      "loss": 0.1139,
+      "step": 1714
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.29723939299583435,
+      "learning_rate": 2.621158392434988e-06,
+      "loss": 0.1157,
+      "step": 1715
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2978956997394562,
+      "learning_rate": 2.61968085106383e-06,
+      "loss": 0.121,
+      "step": 1716
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.28889361023902893,
+      "learning_rate": 2.6182033096926716e-06,
+      "loss": 0.1124,
+      "step": 1717
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.28417134284973145,
+      "learning_rate": 2.616725768321513e-06,
+      "loss": 0.1173,
+      "step": 1718
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.29344823956489563,
+      "learning_rate": 2.6152482269503547e-06,
+      "loss": 0.1194,
+      "step": 1719
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3020975589752197,
+      "learning_rate": 2.6137706855791966e-06,
+      "loss": 0.1156,
+      "step": 1720
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.28589826822280884,
+      "learning_rate": 2.6122931442080377e-06,
+      "loss": 0.113,
+      "step": 1721
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.282889723777771,
+      "learning_rate": 2.6108156028368797e-06,
+      "loss": 0.1143,
+      "step": 1722
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2712370753288269,
+      "learning_rate": 2.609338061465721e-06,
+      "loss": 0.1123,
+      "step": 1723
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2843400239944458,
+      "learning_rate": 2.6078605200945627e-06,
+      "loss": 0.1141,
+      "step": 1724
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.29576143622398376,
+      "learning_rate": 2.6063829787234047e-06,
+      "loss": 0.1218,
+      "step": 1725
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.31621894240379333,
+      "learning_rate": 2.604905437352246e-06,
+      "loss": 0.1231,
+      "step": 1726
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2816074788570404,
+      "learning_rate": 2.6034278959810878e-06,
+      "loss": 0.1136,
+      "step": 1727
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.306791752576828,
+      "learning_rate": 2.6019503546099293e-06,
+      "loss": 0.1163,
+      "step": 1728
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.29944977164268494,
+      "learning_rate": 2.600472813238771e-06,
+      "loss": 0.1184,
+      "step": 1729
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.31420302391052246,
+      "learning_rate": 2.5989952718676124e-06,
+      "loss": 0.1207,
+      "step": 1730
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.28508710861206055,
+      "learning_rate": 2.597517730496454e-06,
+      "loss": 0.1143,
+      "step": 1731
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2918282151222229,
+      "learning_rate": 2.596040189125296e-06,
+      "loss": 0.1157,
+      "step": 1732
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.30676260590553284,
+      "learning_rate": 2.5945626477541374e-06,
+      "loss": 0.1146,
+      "step": 1733
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.29596662521362305,
+      "learning_rate": 2.593085106382979e-06,
+      "loss": 0.1136,
+      "step": 1734
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2966725528240204,
+      "learning_rate": 2.5916075650118205e-06,
+      "loss": 0.1155,
+      "step": 1735
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2881450057029724,
+      "learning_rate": 2.5901300236406624e-06,
+      "loss": 0.1142,
+      "step": 1736
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2984902858734131,
+      "learning_rate": 2.5886524822695035e-06,
+      "loss": 0.1183,
+      "step": 1737
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.28547605872154236,
+      "learning_rate": 2.5871749408983455e-06,
+      "loss": 0.1181,
+      "step": 1738
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.29572775959968567,
+      "learning_rate": 2.5856973995271874e-06,
+      "loss": 0.1153,
+      "step": 1739
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.31525272130966187,
+      "learning_rate": 2.5842198581560286e-06,
+      "loss": 0.1174,
+      "step": 1740
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.29193875193595886,
+      "learning_rate": 2.5827423167848705e-06,
+      "loss": 0.115,
+      "step": 1741
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.28948694467544556,
+      "learning_rate": 2.5812647754137116e-06,
+      "loss": 0.1157,
+      "step": 1742
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.28234297037124634,
+      "learning_rate": 2.5797872340425536e-06,
+      "loss": 0.1174,
+      "step": 1743
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3005118668079376,
+      "learning_rate": 2.578309692671395e-06,
+      "loss": 0.1155,
+      "step": 1744
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.320356547832489,
+      "learning_rate": 2.5768321513002366e-06,
+      "loss": 0.116,
+      "step": 1745
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3094577491283417,
+      "learning_rate": 2.575354609929078e-06,
+      "loss": 0.1176,
+      "step": 1746
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.28986263275146484,
+      "learning_rate": 2.57387706855792e-06,
+      "loss": 0.1157,
+      "step": 1747
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2967703938484192,
+      "learning_rate": 2.5723995271867612e-06,
+      "loss": 0.1173,
+      "step": 1748
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2787644863128662,
+      "learning_rate": 2.570921985815603e-06,
+      "loss": 0.1084,
+      "step": 1749
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.30899956822395325,
+      "learning_rate": 2.5694444444444443e-06,
+      "loss": 0.111,
+      "step": 1750
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.29072219133377075,
+      "learning_rate": 2.5679669030732863e-06,
+      "loss": 0.1109,
+      "step": 1751
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3030107915401459,
+      "learning_rate": 2.5664893617021282e-06,
+      "loss": 0.1163,
+      "step": 1752
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.308183491230011,
+      "learning_rate": 2.5650118203309693e-06,
+      "loss": 0.1197,
+      "step": 1753
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.29148057103157043,
+      "learning_rate": 2.5635342789598113e-06,
+      "loss": 0.1132,
+      "step": 1754
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.28605496883392334,
+      "learning_rate": 2.562056737588653e-06,
+      "loss": 0.1132,
+      "step": 1755
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.30162206292152405,
+      "learning_rate": 2.5605791962174944e-06,
+      "loss": 0.1147,
+      "step": 1756
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2982182800769806,
+      "learning_rate": 2.559101654846336e-06,
+      "loss": 0.1191,
+      "step": 1757
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.27997496724128723,
+      "learning_rate": 2.5576241134751774e-06,
+      "loss": 0.1158,
+      "step": 1758
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.30838632583618164,
+      "learning_rate": 2.556146572104019e-06,
+      "loss": 0.1133,
+      "step": 1759
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2779483497142792,
+      "learning_rate": 2.554669030732861e-06,
+      "loss": 0.1103,
+      "step": 1760
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.28585749864578247,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 0.1155,
+      "step": 1761
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2692381739616394,
+      "learning_rate": 2.551713947990544e-06,
+      "loss": 0.1129,
+      "step": 1762
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2781953811645508,
+      "learning_rate": 2.550236406619386e-06,
+      "loss": 0.1144,
+      "step": 1763
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2877580225467682,
+      "learning_rate": 2.548758865248227e-06,
+      "loss": 0.1143,
+      "step": 1764
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2844424247741699,
+      "learning_rate": 2.547281323877069e-06,
+      "loss": 0.1106,
+      "step": 1765
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.28612884879112244,
+      "learning_rate": 2.5458037825059105e-06,
+      "loss": 0.1147,
+      "step": 1766
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2891106903553009,
+      "learning_rate": 2.544326241134752e-06,
+      "loss": 0.1111,
+      "step": 1767
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2920876741409302,
+      "learning_rate": 2.5428486997635936e-06,
+      "loss": 0.1157,
+      "step": 1768
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2988300621509552,
+      "learning_rate": 2.541371158392435e-06,
+      "loss": 0.119,
+      "step": 1769
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.29913681745529175,
+      "learning_rate": 2.5398936170212767e-06,
+      "loss": 0.1175,
+      "step": 1770
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.28512001037597656,
+      "learning_rate": 2.5384160756501186e-06,
+      "loss": 0.1126,
+      "step": 1771
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.28800663352012634,
+      "learning_rate": 2.5369385342789597e-06,
+      "loss": 0.1166,
+      "step": 1772
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.28620225191116333,
+      "learning_rate": 2.5354609929078017e-06,
+      "loss": 0.1128,
+      "step": 1773
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.28027376532554626,
+      "learning_rate": 2.5339834515366437e-06,
+      "loss": 0.1148,
+      "step": 1774
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2905709743499756,
+      "learning_rate": 2.5325059101654848e-06,
+      "loss": 0.1151,
+      "step": 1775
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2807314693927765,
+      "learning_rate": 2.5310283687943267e-06,
+      "loss": 0.1129,
+      "step": 1776
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.28683072328567505,
+      "learning_rate": 2.529550827423168e-06,
+      "loss": 0.1137,
+      "step": 1777
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2988109290599823,
+      "learning_rate": 2.52807328605201e-06,
+      "loss": 0.1205,
+      "step": 1778
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.28264206647872925,
+      "learning_rate": 2.5265957446808513e-06,
+      "loss": 0.1128,
+      "step": 1779
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3278323709964752,
+      "learning_rate": 2.525118203309693e-06,
+      "loss": 0.1189,
+      "step": 1780
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3016388714313507,
+      "learning_rate": 2.5236406619385344e-06,
+      "loss": 0.1171,
+      "step": 1781
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.30753955245018005,
+      "learning_rate": 2.5221631205673764e-06,
+      "loss": 0.1206,
+      "step": 1782
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2904035747051239,
+      "learning_rate": 2.5206855791962175e-06,
+      "loss": 0.1136,
+      "step": 1783
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2791059613227844,
+      "learning_rate": 2.5192080378250594e-06,
+      "loss": 0.1133,
+      "step": 1784
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3024890124797821,
+      "learning_rate": 2.5177304964539005e-06,
+      "loss": 0.1222,
+      "step": 1785
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2849273383617401,
+      "learning_rate": 2.5162529550827425e-06,
+      "loss": 0.1172,
+      "step": 1786
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.30256059765815735,
+      "learning_rate": 2.5147754137115844e-06,
+      "loss": 0.1155,
+      "step": 1787
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.29325371980667114,
+      "learning_rate": 2.5132978723404256e-06,
+      "loss": 0.1191,
+      "step": 1788
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3001633882522583,
+      "learning_rate": 2.5118203309692675e-06,
+      "loss": 0.1176,
+      "step": 1789
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.29316699504852295,
+      "learning_rate": 2.510342789598109e-06,
+      "loss": 0.1138,
+      "step": 1790
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.28215616941452026,
+      "learning_rate": 2.5088652482269506e-06,
+      "loss": 0.1152,
+      "step": 1791
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.28712812066078186,
+      "learning_rate": 2.507387706855792e-06,
+      "loss": 0.1148,
+      "step": 1792
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.29103708267211914,
+      "learning_rate": 2.5059101654846336e-06,
+      "loss": 0.1153,
+      "step": 1793
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.29552629590034485,
+      "learning_rate": 2.504432624113475e-06,
+      "loss": 0.1199,
+      "step": 1794
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3137848377227783,
+      "learning_rate": 2.502955082742317e-06,
+      "loss": 0.1194,
+      "step": 1795
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.30231961607933044,
+      "learning_rate": 2.5014775413711582e-06,
+      "loss": 0.1145,
+      "step": 1796
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2974165678024292,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1178,
+      "step": 1797
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.29884132742881775,
+      "learning_rate": 2.4985224586288417e-06,
+      "loss": 0.1192,
+      "step": 1798
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.28818535804748535,
+      "learning_rate": 2.4970449172576837e-06,
+      "loss": 0.113,
+      "step": 1799
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2760612368583679,
+      "learning_rate": 2.4955673758865252e-06,
+      "loss": 0.1102,
+      "step": 1800
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.30093708634376526,
+      "learning_rate": 2.4940898345153668e-06,
+      "loss": 0.1149,
+      "step": 1801
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2966955006122589,
+      "learning_rate": 2.4926122931442083e-06,
+      "loss": 0.1153,
+      "step": 1802
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2819238603115082,
+      "learning_rate": 2.49113475177305e-06,
+      "loss": 0.1089,
+      "step": 1803
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2951497733592987,
+      "learning_rate": 2.4896572104018914e-06,
+      "loss": 0.1128,
+      "step": 1804
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2875305414199829,
+      "learning_rate": 2.488179669030733e-06,
+      "loss": 0.1141,
+      "step": 1805
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.28138718008995056,
+      "learning_rate": 2.4867021276595744e-06,
+      "loss": 0.113,
+      "step": 1806
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2869209945201874,
+      "learning_rate": 2.4852245862884164e-06,
+      "loss": 0.1124,
+      "step": 1807
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3309995234012604,
+      "learning_rate": 2.483747044917258e-06,
+      "loss": 0.114,
+      "step": 1808
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.28361091017723083,
+      "learning_rate": 2.4822695035460995e-06,
+      "loss": 0.1153,
+      "step": 1809
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.32110026478767395,
+      "learning_rate": 2.480791962174941e-06,
+      "loss": 0.1108,
+      "step": 1810
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.31635600328445435,
+      "learning_rate": 2.479314420803783e-06,
+      "loss": 0.1131,
+      "step": 1811
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2870709300041199,
+      "learning_rate": 2.4778368794326245e-06,
+      "loss": 0.1082,
+      "step": 1812
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2982943654060364,
+      "learning_rate": 2.476359338061466e-06,
+      "loss": 0.117,
+      "step": 1813
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.30215537548065186,
+      "learning_rate": 2.4748817966903075e-06,
+      "loss": 0.1171,
+      "step": 1814
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3099404573440552,
+      "learning_rate": 2.473404255319149e-06,
+      "loss": 0.1146,
+      "step": 1815
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3013956546783447,
+      "learning_rate": 2.4719267139479906e-06,
+      "loss": 0.1171,
+      "step": 1816
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.32759395241737366,
+      "learning_rate": 2.470449172576832e-06,
+      "loss": 0.1161,
+      "step": 1817
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2977280914783478,
+      "learning_rate": 2.4689716312056737e-06,
+      "loss": 0.1167,
+      "step": 1818
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3042850196361542,
+      "learning_rate": 2.4674940898345156e-06,
+      "loss": 0.1124,
+      "step": 1819
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2889341711997986,
+      "learning_rate": 2.466016548463357e-06,
+      "loss": 0.1153,
+      "step": 1820
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2940898835659027,
+      "learning_rate": 2.4645390070921987e-06,
+      "loss": 0.1127,
+      "step": 1821
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.29065415263175964,
+      "learning_rate": 2.4630614657210402e-06,
+      "loss": 0.1181,
+      "step": 1822
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.30052974820137024,
+      "learning_rate": 2.461583924349882e-06,
+      "loss": 0.1145,
+      "step": 1823
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2908446192741394,
+      "learning_rate": 2.4601063829787237e-06,
+      "loss": 0.1152,
+      "step": 1824
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.30466192960739136,
+      "learning_rate": 2.4586288416075653e-06,
+      "loss": 0.1201,
+      "step": 1825
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.27887600660324097,
+      "learning_rate": 2.457151300236407e-06,
+      "loss": 0.1091,
+      "step": 1826
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.28884103894233704,
+      "learning_rate": 2.4556737588652483e-06,
+      "loss": 0.1121,
+      "step": 1827
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2831375002861023,
+      "learning_rate": 2.45419621749409e-06,
+      "loss": 0.117,
+      "step": 1828
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2932526171207428,
+      "learning_rate": 2.4527186761229314e-06,
+      "loss": 0.1142,
+      "step": 1829
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.288620263338089,
+      "learning_rate": 2.4512411347517734e-06,
+      "loss": 0.1136,
+      "step": 1830
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2911244332790375,
+      "learning_rate": 2.449763593380615e-06,
+      "loss": 0.1181,
+      "step": 1831
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.27385595440864563,
+      "learning_rate": 2.4482860520094564e-06,
+      "loss": 0.1088,
+      "step": 1832
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.28321006894111633,
+      "learning_rate": 2.446808510638298e-06,
+      "loss": 0.1124,
+      "step": 1833
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.28905636072158813,
+      "learning_rate": 2.44533096926714e-06,
+      "loss": 0.1226,
+      "step": 1834
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.28809159994125366,
+      "learning_rate": 2.4438534278959814e-06,
+      "loss": 0.1131,
+      "step": 1835
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3023412823677063,
+      "learning_rate": 2.442375886524823e-06,
+      "loss": 0.1167,
+      "step": 1836
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2827005088329315,
+      "learning_rate": 2.4408983451536645e-06,
+      "loss": 0.1151,
+      "step": 1837
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2977384328842163,
+      "learning_rate": 2.439420803782506e-06,
+      "loss": 0.119,
+      "step": 1838
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3086833953857422,
+      "learning_rate": 2.4379432624113476e-06,
+      "loss": 0.121,
+      "step": 1839
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2868455946445465,
+      "learning_rate": 2.436465721040189e-06,
+      "loss": 0.1176,
+      "step": 1840
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2831099331378937,
+      "learning_rate": 2.4349881796690306e-06,
+      "loss": 0.1141,
+      "step": 1841
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.28741350769996643,
+      "learning_rate": 2.4335106382978726e-06,
+      "loss": 0.1178,
+      "step": 1842
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3091892600059509,
+      "learning_rate": 2.432033096926714e-06,
+      "loss": 0.118,
+      "step": 1843
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2843966782093048,
+      "learning_rate": 2.4305555555555557e-06,
+      "loss": 0.1152,
+      "step": 1844
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.29012230038642883,
+      "learning_rate": 2.429078014184397e-06,
+      "loss": 0.1152,
+      "step": 1845
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2987143397331238,
+      "learning_rate": 2.427600472813239e-06,
+      "loss": 0.1188,
+      "step": 1846
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.29670450091362,
+      "learning_rate": 2.4261229314420807e-06,
+      "loss": 0.1161,
+      "step": 1847
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.31162017583847046,
+      "learning_rate": 2.4246453900709222e-06,
+      "loss": 0.1192,
+      "step": 1848
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.29248949885368347,
+      "learning_rate": 2.4231678486997638e-06,
+      "loss": 0.1159,
+      "step": 1849
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.28830134868621826,
+      "learning_rate": 2.4216903073286053e-06,
+      "loss": 0.1178,
+      "step": 1850
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2896536886692047,
+      "learning_rate": 2.420212765957447e-06,
+      "loss": 0.1151,
+      "step": 1851
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2764412462711334,
+      "learning_rate": 2.4187352245862884e-06,
+      "loss": 0.1142,
+      "step": 1852
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.28309017419815063,
+      "learning_rate": 2.41725768321513e-06,
+      "loss": 0.1105,
+      "step": 1853
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2870856523513794,
+      "learning_rate": 2.415780141843972e-06,
+      "loss": 0.1115,
+      "step": 1854
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2972900867462158,
+      "learning_rate": 2.4143026004728134e-06,
+      "loss": 0.1152,
+      "step": 1855
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2948080599308014,
+      "learning_rate": 2.412825059101655e-06,
+      "loss": 0.1186,
+      "step": 1856
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2792036235332489,
+      "learning_rate": 2.4113475177304965e-06,
+      "loss": 0.1141,
+      "step": 1857
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3004268705844879,
+      "learning_rate": 2.4098699763593384e-06,
+      "loss": 0.1199,
+      "step": 1858
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2848747968673706,
+      "learning_rate": 2.40839243498818e-06,
+      "loss": 0.113,
+      "step": 1859
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.29183247685432434,
+      "learning_rate": 2.4069148936170215e-06,
+      "loss": 0.1131,
+      "step": 1860
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2791503667831421,
+      "learning_rate": 2.405437352245863e-06,
+      "loss": 0.113,
+      "step": 1861
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.28676122426986694,
+      "learning_rate": 2.4039598108747045e-06,
+      "loss": 0.1129,
+      "step": 1862
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2972956597805023,
+      "learning_rate": 2.4024822695035465e-06,
+      "loss": 0.1159,
+      "step": 1863
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2995658218860626,
+      "learning_rate": 2.401004728132388e-06,
+      "loss": 0.116,
+      "step": 1864
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.317130982875824,
+      "learning_rate": 2.3995271867612296e-06,
+      "loss": 0.1189,
+      "step": 1865
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2801668643951416,
+      "learning_rate": 2.398049645390071e-06,
+      "loss": 0.1174,
+      "step": 1866
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.29104942083358765,
+      "learning_rate": 2.3965721040189126e-06,
+      "loss": 0.1159,
+      "step": 1867
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2970658242702484,
+      "learning_rate": 2.395094562647754e-06,
+      "loss": 0.1136,
+      "step": 1868
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.28111767768859863,
+      "learning_rate": 2.393617021276596e-06,
+      "loss": 0.1143,
+      "step": 1869
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2792731821537018,
+      "learning_rate": 2.3921394799054377e-06,
+      "loss": 0.1159,
+      "step": 1870
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2778155207633972,
+      "learning_rate": 2.390661938534279e-06,
+      "loss": 0.1129,
+      "step": 1871
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.28632640838623047,
+      "learning_rate": 2.3891843971631207e-06,
+      "loss": 0.1112,
+      "step": 1872
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.28223931789398193,
+      "learning_rate": 2.3877068557919627e-06,
+      "loss": 0.1169,
+      "step": 1873
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2890729010105133,
+      "learning_rate": 2.3862293144208042e-06,
+      "loss": 0.1125,
+      "step": 1874
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3000413179397583,
+      "learning_rate": 2.3847517730496458e-06,
+      "loss": 0.1161,
+      "step": 1875
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2754424810409546,
+      "learning_rate": 2.3832742316784873e-06,
+      "loss": 0.1116,
+      "step": 1876
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2870715856552124,
+      "learning_rate": 2.381796690307329e-06,
+      "loss": 0.1148,
+      "step": 1877
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2972899377346039,
+      "learning_rate": 2.3803191489361704e-06,
+      "loss": 0.1135,
+      "step": 1878
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2899673581123352,
+      "learning_rate": 2.378841607565012e-06,
+      "loss": 0.1173,
+      "step": 1879
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2824571430683136,
+      "learning_rate": 2.3773640661938534e-06,
+      "loss": 0.1156,
+      "step": 1880
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2759453058242798,
+      "learning_rate": 2.3758865248226954e-06,
+      "loss": 0.1097,
+      "step": 1881
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.268184632062912,
+      "learning_rate": 2.374408983451537e-06,
+      "loss": 0.1115,
+      "step": 1882
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.27184000611305237,
+      "learning_rate": 2.3729314420803784e-06,
+      "loss": 0.1098,
+      "step": 1883
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2911868393421173,
+      "learning_rate": 2.37145390070922e-06,
+      "loss": 0.1138,
+      "step": 1884
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.28168785572052,
+      "learning_rate": 2.369976359338062e-06,
+      "loss": 0.1167,
+      "step": 1885
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2968481779098511,
+      "learning_rate": 2.3684988179669035e-06,
+      "loss": 0.1171,
+      "step": 1886
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.29399606585502625,
+      "learning_rate": 2.367021276595745e-06,
+      "loss": 0.1125,
+      "step": 1887
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.29483121633529663,
+      "learning_rate": 2.3655437352245865e-06,
+      "loss": 0.1122,
+      "step": 1888
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.27542752027511597,
+      "learning_rate": 2.364066193853428e-06,
+      "loss": 0.1129,
+      "step": 1889
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.28775015473365784,
+      "learning_rate": 2.3625886524822696e-06,
+      "loss": 0.1148,
+      "step": 1890
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.27529633045196533,
+      "learning_rate": 2.361111111111111e-06,
+      "loss": 0.1115,
+      "step": 1891
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2927680015563965,
+      "learning_rate": 2.3596335697399527e-06,
+      "loss": 0.1186,
+      "step": 1892
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.29718396067619324,
+      "learning_rate": 2.3581560283687946e-06,
+      "loss": 0.1124,
+      "step": 1893
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.288520485162735,
+      "learning_rate": 2.356678486997636e-06,
+      "loss": 0.1142,
+      "step": 1894
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.27882853150367737,
+      "learning_rate": 2.3552009456264777e-06,
+      "loss": 0.1133,
+      "step": 1895
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.28009846806526184,
+      "learning_rate": 2.3537234042553197e-06,
+      "loss": 0.1174,
+      "step": 1896
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2859962582588196,
+      "learning_rate": 2.352245862884161e-06,
+      "loss": 0.1146,
+      "step": 1897
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2797001004219055,
+      "learning_rate": 2.3507683215130027e-06,
+      "loss": 0.1101,
+      "step": 1898
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.27599966526031494,
+      "learning_rate": 2.3492907801418443e-06,
+      "loss": 0.1093,
+      "step": 1899
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.29249414801597595,
+      "learning_rate": 2.347813238770686e-06,
+      "loss": 0.1173,
+      "step": 1900
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.29532450437545776,
+      "learning_rate": 2.3463356973995273e-06,
+      "loss": 0.1151,
+      "step": 1901
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.28679659962654114,
+      "learning_rate": 2.344858156028369e-06,
+      "loss": 0.1161,
+      "step": 1902
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.28649190068244934,
+      "learning_rate": 2.3433806146572104e-06,
+      "loss": 0.1113,
+      "step": 1903
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3122852146625519,
+      "learning_rate": 2.3419030732860523e-06,
+      "loss": 0.1165,
+      "step": 1904
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.29286447167396545,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 0.1126,
+      "step": 1905
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.31665632128715515,
+      "learning_rate": 2.3389479905437354e-06,
+      "loss": 0.1211,
+      "step": 1906
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.28001415729522705,
+      "learning_rate": 2.337470449172577e-06,
+      "loss": 0.1129,
+      "step": 1907
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2789599299430847,
+      "learning_rate": 2.335992907801419e-06,
+      "loss": 0.1127,
+      "step": 1908
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34256547689437866,
+      "learning_rate": 2.3345153664302604e-06,
+      "loss": 0.1119,
+      "step": 1909
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.30474674701690674,
+      "learning_rate": 2.333037825059102e-06,
+      "loss": 0.1134,
+      "step": 1910
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.28600630164146423,
+      "learning_rate": 2.3315602836879435e-06,
+      "loss": 0.1162,
+      "step": 1911
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.28959837555885315,
+      "learning_rate": 2.330082742316785e-06,
+      "loss": 0.1162,
+      "step": 1912
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2844424843788147,
+      "learning_rate": 2.3286052009456266e-06,
+      "loss": 0.1106,
+      "step": 1913
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2918698787689209,
+      "learning_rate": 2.327127659574468e-06,
+      "loss": 0.116,
+      "step": 1914
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3119930922985077,
+      "learning_rate": 2.3256501182033096e-06,
+      "loss": 0.1155,
+      "step": 1915
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3059247136116028,
+      "learning_rate": 2.3241725768321516e-06,
+      "loss": 0.1247,
+      "step": 1916
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.29329150915145874,
+      "learning_rate": 2.322695035460993e-06,
+      "loss": 0.113,
+      "step": 1917
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.30189597606658936,
+      "learning_rate": 2.3212174940898347e-06,
+      "loss": 0.1192,
+      "step": 1918
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3058381974697113,
+      "learning_rate": 2.319739952718676e-06,
+      "loss": 0.1152,
+      "step": 1919
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2795242369174957,
+      "learning_rate": 2.318262411347518e-06,
+      "loss": 0.1077,
+      "step": 1920
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.27935799956321716,
+      "learning_rate": 2.3167848699763597e-06,
+      "loss": 0.1133,
+      "step": 1921
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.31461602449417114,
+      "learning_rate": 2.3153073286052012e-06,
+      "loss": 0.1207,
+      "step": 1922
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2753186821937561,
+      "learning_rate": 2.3138297872340428e-06,
+      "loss": 0.1066,
+      "step": 1923
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.280995637178421,
+      "learning_rate": 2.3123522458628843e-06,
+      "loss": 0.1158,
+      "step": 1924
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.29846620559692383,
+      "learning_rate": 2.310874704491726e-06,
+      "loss": 0.1174,
+      "step": 1925
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2984260022640228,
+      "learning_rate": 2.3093971631205674e-06,
+      "loss": 0.1141,
+      "step": 1926
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.30352696776390076,
+      "learning_rate": 2.307919621749409e-06,
+      "loss": 0.1165,
+      "step": 1927
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3152119517326355,
+      "learning_rate": 2.306442080378251e-06,
+      "loss": 0.117,
+      "step": 1928
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3000238835811615,
+      "learning_rate": 2.3049645390070924e-06,
+      "loss": 0.1187,
+      "step": 1929
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.30057454109191895,
+      "learning_rate": 2.303486997635934e-06,
+      "loss": 0.1154,
+      "step": 1930
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2817342281341553,
+      "learning_rate": 2.302009456264776e-06,
+      "loss": 0.1106,
+      "step": 1931
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.30073145031929016,
+      "learning_rate": 2.3005319148936174e-06,
+      "loss": 0.1123,
+      "step": 1932
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.31137219071388245,
+      "learning_rate": 2.299054373522459e-06,
+      "loss": 0.1168,
+      "step": 1933
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.29414427280426025,
+      "learning_rate": 2.2975768321513005e-06,
+      "loss": 0.1158,
+      "step": 1934
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2942590117454529,
+      "learning_rate": 2.296099290780142e-06,
+      "loss": 0.1181,
+      "step": 1935
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2882036566734314,
+      "learning_rate": 2.2946217494089835e-06,
+      "loss": 0.1179,
+      "step": 1936
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.28187453746795654,
+      "learning_rate": 2.293144208037825e-06,
+      "loss": 0.1118,
+      "step": 1937
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3011631667613983,
+      "learning_rate": 2.2916666666666666e-06,
+      "loss": 0.1167,
+      "step": 1938
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3146383464336395,
+      "learning_rate": 2.2901891252955086e-06,
+      "loss": 0.1157,
+      "step": 1939
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.307541161775589,
+      "learning_rate": 2.28871158392435e-06,
+      "loss": 0.1138,
+      "step": 1940
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2959631085395813,
+      "learning_rate": 2.2872340425531916e-06,
+      "loss": 0.1118,
+      "step": 1941
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2847319543361664,
+      "learning_rate": 2.285756501182033e-06,
+      "loss": 0.1132,
+      "step": 1942
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2837541103363037,
+      "learning_rate": 2.284278959810875e-06,
+      "loss": 0.1154,
+      "step": 1943
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2996666431427002,
+      "learning_rate": 2.2828014184397167e-06,
+      "loss": 0.1136,
+      "step": 1944
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2826857566833496,
+      "learning_rate": 2.281323877068558e-06,
+      "loss": 0.1149,
+      "step": 1945
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2977179288864136,
+      "learning_rate": 2.2798463356973997e-06,
+      "loss": 0.117,
+      "step": 1946
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.30585452914237976,
+      "learning_rate": 2.2783687943262413e-06,
+      "loss": 0.1168,
+      "step": 1947
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3090185523033142,
+      "learning_rate": 2.276891252955083e-06,
+      "loss": 0.1179,
+      "step": 1948
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2846519947052002,
+      "learning_rate": 2.2754137115839243e-06,
+      "loss": 0.1169,
+      "step": 1949
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2838442027568817,
+      "learning_rate": 2.273936170212766e-06,
+      "loss": 0.1143,
+      "step": 1950
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2834759056568146,
+      "learning_rate": 2.272458628841608e-06,
+      "loss": 0.114,
+      "step": 1951
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.29758554697036743,
+      "learning_rate": 2.2709810874704493e-06,
+      "loss": 0.1167,
+      "step": 1952
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2895980477333069,
+      "learning_rate": 2.269503546099291e-06,
+      "loss": 0.1161,
+      "step": 1953
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.28893157839775085,
+      "learning_rate": 2.2680260047281324e-06,
+      "loss": 0.114,
+      "step": 1954
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.29953494668006897,
+      "learning_rate": 2.2665484633569744e-06,
+      "loss": 0.1127,
+      "step": 1955
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.28809982538223267,
+      "learning_rate": 2.265070921985816e-06,
+      "loss": 0.1145,
+      "step": 1956
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.277861624956131,
+      "learning_rate": 2.2635933806146574e-06,
+      "loss": 0.1106,
+      "step": 1957
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2984499931335449,
+      "learning_rate": 2.262115839243499e-06,
+      "loss": 0.1125,
+      "step": 1958
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.28857162594795227,
+      "learning_rate": 2.2606382978723405e-06,
+      "loss": 0.1114,
+      "step": 1959
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.29593178629875183,
+      "learning_rate": 2.259160756501182e-06,
+      "loss": 0.1117,
+      "step": 1960
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.27818724513053894,
+      "learning_rate": 2.2576832151300236e-06,
+      "loss": 0.115,
+      "step": 1961
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2963353395462036,
+      "learning_rate": 2.2562056737588655e-06,
+      "loss": 0.1169,
+      "step": 1962
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.28198954463005066,
+      "learning_rate": 2.254728132387707e-06,
+      "loss": 0.1126,
+      "step": 1963
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2884441316127777,
+      "learning_rate": 2.2532505910165486e-06,
+      "loss": 0.1182,
+      "step": 1964
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2884747087955475,
+      "learning_rate": 2.25177304964539e-06,
+      "loss": 0.1187,
+      "step": 1965
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.35377413034439087,
+      "learning_rate": 2.250295508274232e-06,
+      "loss": 0.1142,
+      "step": 1966
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.30866456031799316,
+      "learning_rate": 2.2488179669030736e-06,
+      "loss": 0.1181,
+      "step": 1967
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2987980842590332,
+      "learning_rate": 2.247340425531915e-06,
+      "loss": 0.1153,
+      "step": 1968
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3014574348926544,
+      "learning_rate": 2.2458628841607567e-06,
+      "loss": 0.1159,
+      "step": 1969
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.28160780668258667,
+      "learning_rate": 2.2443853427895982e-06,
+      "loss": 0.1125,
+      "step": 1970
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2957876920700073,
+      "learning_rate": 2.2429078014184398e-06,
+      "loss": 0.1168,
+      "step": 1971
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2959586977958679,
+      "learning_rate": 2.2414302600472813e-06,
+      "loss": 0.1153,
+      "step": 1972
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.29878827929496765,
+      "learning_rate": 2.239952718676123e-06,
+      "loss": 0.1161,
+      "step": 1973
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.29964855313301086,
+      "learning_rate": 2.2384751773049648e-06,
+      "loss": 0.1224,
+      "step": 1974
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.29898881912231445,
+      "learning_rate": 2.2369976359338063e-06,
+      "loss": 0.1144,
+      "step": 1975
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.31031155586242676,
+      "learning_rate": 2.235520094562648e-06,
+      "loss": 0.1165,
+      "step": 1976
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2992332875728607,
+      "learning_rate": 2.2340425531914894e-06,
+      "loss": 0.1194,
+      "step": 1977
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.30491992831230164,
+      "learning_rate": 2.2325650118203313e-06,
+      "loss": 0.114,
+      "step": 1978
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2839789092540741,
+      "learning_rate": 2.231087470449173e-06,
+      "loss": 0.1088,
+      "step": 1979
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2687525749206543,
+      "learning_rate": 2.2296099290780144e-06,
+      "loss": 0.1109,
+      "step": 1980
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.28788238763809204,
+      "learning_rate": 2.228132387706856e-06,
+      "loss": 0.1121,
+      "step": 1981
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2904152274131775,
+      "learning_rate": 2.2266548463356975e-06,
+      "loss": 0.1146,
+      "step": 1982
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3062341511249542,
+      "learning_rate": 2.225177304964539e-06,
+      "loss": 0.1135,
+      "step": 1983
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2829311788082123,
+      "learning_rate": 2.2236997635933805e-06,
+      "loss": 0.1147,
+      "step": 1984
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2977240979671478,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.1138,
+      "step": 1985
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3079695999622345,
+      "learning_rate": 2.220744680851064e-06,
+      "loss": 0.1156,
+      "step": 1986
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3145463168621063,
+      "learning_rate": 2.2192671394799056e-06,
+      "loss": 0.1147,
+      "step": 1987
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3046582341194153,
+      "learning_rate": 2.217789598108747e-06,
+      "loss": 0.1186,
+      "step": 1988
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.28766992688179016,
+      "learning_rate": 2.2163120567375886e-06,
+      "loss": 0.1157,
+      "step": 1989
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2931135892868042,
+      "learning_rate": 2.2148345153664306e-06,
+      "loss": 0.1157,
+      "step": 1990
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2860803008079529,
+      "learning_rate": 2.213356973995272e-06,
+      "loss": 0.1131,
+      "step": 1991
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.29874131083488464,
+      "learning_rate": 2.2118794326241137e-06,
+      "loss": 0.1154,
+      "step": 1992
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.28098973631858826,
+      "learning_rate": 2.210401891252955e-06,
+      "loss": 0.113,
+      "step": 1993
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2930745780467987,
+      "learning_rate": 2.208924349881797e-06,
+      "loss": 0.1164,
+      "step": 1994
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.29550185799598694,
+      "learning_rate": 2.2074468085106387e-06,
+      "loss": 0.1137,
+      "step": 1995
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3049972951412201,
+      "learning_rate": 2.2059692671394802e-06,
+      "loss": 0.1174,
+      "step": 1996
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.29784199595451355,
+      "learning_rate": 2.2044917257683217e-06,
+      "loss": 0.115,
+      "step": 1997
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2954116463661194,
+      "learning_rate": 2.2030141843971633e-06,
+      "loss": 0.1145,
+      "step": 1998
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3001748025417328,
+      "learning_rate": 2.201536643026005e-06,
+      "loss": 0.1151,
+      "step": 1999
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.28016290068626404,
+      "learning_rate": 2.2000591016548463e-06,
+      "loss": 0.1119,
+      "step": 2000
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2870945632457733,
+      "learning_rate": 2.1985815602836883e-06,
+      "loss": 0.112,
+      "step": 2001
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.29480886459350586,
+      "learning_rate": 2.19710401891253e-06,
+      "loss": 0.1146,
+      "step": 2002
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.28637635707855225,
+      "learning_rate": 2.1956264775413714e-06,
+      "loss": 0.113,
+      "step": 2003
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2886885106563568,
+      "learning_rate": 2.194148936170213e-06,
+      "loss": 0.1132,
+      "step": 2004
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.28722813725471497,
+      "learning_rate": 2.192671394799055e-06,
+      "loss": 0.1153,
+      "step": 2005
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.29913485050201416,
+      "learning_rate": 2.1911938534278964e-06,
+      "loss": 0.1141,
+      "step": 2006
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.29377543926239014,
+      "learning_rate": 2.189716312056738e-06,
+      "loss": 0.1143,
+      "step": 2007
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.30245083570480347,
+      "learning_rate": 2.1882387706855795e-06,
+      "loss": 0.1148,
+      "step": 2008
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.29203829169273376,
+      "learning_rate": 2.186761229314421e-06,
+      "loss": 0.1161,
+      "step": 2009
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2919359803199768,
+      "learning_rate": 2.1852836879432625e-06,
+      "loss": 0.1137,
+      "step": 2010
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3017304539680481,
+      "learning_rate": 2.183806146572104e-06,
+      "loss": 0.118,
+      "step": 2011
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.272099107503891,
+      "learning_rate": 2.1823286052009456e-06,
+      "loss": 0.1052,
+      "step": 2012
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.28036901354789734,
+      "learning_rate": 2.1808510638297876e-06,
+      "loss": 0.1161,
+      "step": 2013
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2775558531284332,
+      "learning_rate": 2.179373522458629e-06,
+      "loss": 0.1057,
+      "step": 2014
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.29233577847480774,
+      "learning_rate": 2.1778959810874706e-06,
+      "loss": 0.1136,
+      "step": 2015
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2917049825191498,
+      "learning_rate": 2.176418439716312e-06,
+      "loss": 0.1179,
+      "step": 2016
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.27122223377227783,
+      "learning_rate": 2.174940898345154e-06,
+      "loss": 0.1083,
+      "step": 2017
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2984669804573059,
+      "learning_rate": 2.1734633569739956e-06,
+      "loss": 0.1159,
+      "step": 2018
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.29905250668525696,
+      "learning_rate": 2.171985815602837e-06,
+      "loss": 0.1157,
+      "step": 2019
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2953692078590393,
+      "learning_rate": 2.1705082742316787e-06,
+      "loss": 0.113,
+      "step": 2020
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2967214286327362,
+      "learning_rate": 2.1690307328605202e-06,
+      "loss": 0.1141,
+      "step": 2021
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.30087465047836304,
+      "learning_rate": 2.1675531914893618e-06,
+      "loss": 0.1129,
+      "step": 2022
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2705608308315277,
+      "learning_rate": 2.1660756501182033e-06,
+      "loss": 0.1102,
+      "step": 2023
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2867546081542969,
+      "learning_rate": 2.164598108747045e-06,
+      "loss": 0.115,
+      "step": 2024
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.30212005972862244,
+      "learning_rate": 2.163120567375887e-06,
+      "loss": 0.1173,
+      "step": 2025
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2930554747581482,
+      "learning_rate": 2.1616430260047283e-06,
+      "loss": 0.1134,
+      "step": 2026
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2826298475265503,
+      "learning_rate": 2.16016548463357e-06,
+      "loss": 0.1126,
+      "step": 2027
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.27257969975471497,
+      "learning_rate": 2.158687943262412e-06,
+      "loss": 0.1131,
+      "step": 2028
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2848852574825287,
+      "learning_rate": 2.1572104018912534e-06,
+      "loss": 0.1114,
+      "step": 2029
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2891388237476349,
+      "learning_rate": 2.155732860520095e-06,
+      "loss": 0.1157,
+      "step": 2030
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.31753766536712646,
+      "learning_rate": 2.1542553191489364e-06,
+      "loss": 0.1135,
+      "step": 2031
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.28999537229537964,
+      "learning_rate": 2.152777777777778e-06,
+      "loss": 0.117,
+      "step": 2032
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.31126829981803894,
+      "learning_rate": 2.1513002364066195e-06,
+      "loss": 0.1109,
+      "step": 2033
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2811659574508667,
+      "learning_rate": 2.149822695035461e-06,
+      "loss": 0.1121,
+      "step": 2034
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.289581298828125,
+      "learning_rate": 2.1483451536643026e-06,
+      "loss": 0.1153,
+      "step": 2035
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2851566970348358,
+      "learning_rate": 2.1468676122931445e-06,
+      "loss": 0.1108,
+      "step": 2036
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.29975175857543945,
+      "learning_rate": 2.145390070921986e-06,
+      "loss": 0.11,
+      "step": 2037
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3011217415332794,
+      "learning_rate": 2.1439125295508276e-06,
+      "loss": 0.1148,
+      "step": 2038
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.28776800632476807,
+      "learning_rate": 2.142434988179669e-06,
+      "loss": 0.1158,
+      "step": 2039
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.27878338098526,
+      "learning_rate": 2.140957446808511e-06,
+      "loss": 0.1108,
+      "step": 2040
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.27684286236763,
+      "learning_rate": 2.1394799054373526e-06,
+      "loss": 0.1119,
+      "step": 2041
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2992793023586273,
+      "learning_rate": 2.138002364066194e-06,
+      "loss": 0.1152,
+      "step": 2042
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.28693822026252747,
+      "learning_rate": 2.1365248226950357e-06,
+      "loss": 0.1117,
+      "step": 2043
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2983400225639343,
+      "learning_rate": 2.1350472813238772e-06,
+      "loss": 0.116,
+      "step": 2044
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.28364527225494385,
+      "learning_rate": 2.1335697399527187e-06,
+      "loss": 0.1142,
+      "step": 2045
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2928532063961029,
+      "learning_rate": 2.1320921985815603e-06,
+      "loss": 0.1136,
+      "step": 2046
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.28670182824134827,
+      "learning_rate": 2.130614657210402e-06,
+      "loss": 0.1164,
+      "step": 2047
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.30965456366539,
+      "learning_rate": 2.1291371158392438e-06,
+      "loss": 0.1203,
+      "step": 2048
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2876374125480652,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.1153,
+      "step": 2049
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2678670883178711,
+      "learning_rate": 2.126182033096927e-06,
+      "loss": 0.1128,
+      "step": 2050
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.30057376623153687,
+      "learning_rate": 2.1247044917257684e-06,
+      "loss": 0.1179,
+      "step": 2051
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.293866366147995,
+      "learning_rate": 2.1232269503546103e-06,
+      "loss": 0.1162,
+      "step": 2052
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.28677693009376526,
+      "learning_rate": 2.121749408983452e-06,
+      "loss": 0.1153,
+      "step": 2053
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2867712080478668,
+      "learning_rate": 2.1202718676122934e-06,
+      "loss": 0.1176,
+      "step": 2054
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2877953350543976,
+      "learning_rate": 2.118794326241135e-06,
+      "loss": 0.1156,
+      "step": 2055
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.297740638256073,
+      "learning_rate": 2.1173167848699765e-06,
+      "loss": 0.1167,
+      "step": 2056
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.29130303859710693,
+      "learning_rate": 2.115839243498818e-06,
+      "loss": 0.1136,
+      "step": 2057
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3029724955558777,
+      "learning_rate": 2.1143617021276595e-06,
+      "loss": 0.116,
+      "step": 2058
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2867678701877594,
+      "learning_rate": 2.1128841607565015e-06,
+      "loss": 0.1145,
+      "step": 2059
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.28140348196029663,
+      "learning_rate": 2.111406619385343e-06,
+      "loss": 0.1096,
+      "step": 2060
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2789646089076996,
+      "learning_rate": 2.1099290780141846e-06,
+      "loss": 0.1149,
+      "step": 2061
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.30273374915122986,
+      "learning_rate": 2.108451536643026e-06,
+      "loss": 0.1161,
+      "step": 2062
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.30514198541641235,
+      "learning_rate": 2.106973995271868e-06,
+      "loss": 0.1151,
+      "step": 2063
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.29088276624679565,
+      "learning_rate": 2.1054964539007096e-06,
+      "loss": 0.1137,
+      "step": 2064
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.27893608808517456,
+      "learning_rate": 2.104018912529551e-06,
+      "loss": 0.1134,
+      "step": 2065
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2762652337551117,
+      "learning_rate": 2.1025413711583926e-06,
+      "loss": 0.105,
+      "step": 2066
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.29913902282714844,
+      "learning_rate": 2.101063829787234e-06,
+      "loss": 0.1093,
+      "step": 2067
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.28490713238716125,
+      "learning_rate": 2.0995862884160757e-06,
+      "loss": 0.1162,
+      "step": 2068
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2719292938709259,
+      "learning_rate": 2.0981087470449173e-06,
+      "loss": 0.1058,
+      "step": 2069
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.30745750665664673,
+      "learning_rate": 2.0966312056737588e-06,
+      "loss": 0.1154,
+      "step": 2070
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.29631879925727844,
+      "learning_rate": 2.0951536643026007e-06,
+      "loss": 0.1165,
+      "step": 2071
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2982489764690399,
+      "learning_rate": 2.0936761229314423e-06,
+      "loss": 0.1153,
+      "step": 2072
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2918132543563843,
+      "learning_rate": 2.092198581560284e-06,
+      "loss": 0.1132,
+      "step": 2073
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2875410318374634,
+      "learning_rate": 2.0907210401891253e-06,
+      "loss": 0.1115,
+      "step": 2074
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.30258411169052124,
+      "learning_rate": 2.0892434988179673e-06,
+      "loss": 0.1174,
+      "step": 2075
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.28294894099235535,
+      "learning_rate": 2.087765957446809e-06,
+      "loss": 0.1132,
+      "step": 2076
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.29632046818733215,
+      "learning_rate": 2.0862884160756504e-06,
+      "loss": 0.1156,
+      "step": 2077
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2820342183113098,
+      "learning_rate": 2.084810874704492e-06,
+      "loss": 0.1097,
+      "step": 2078
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.285055011510849,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.1109,
+      "step": 2079
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2943456172943115,
+      "learning_rate": 2.081855791962175e-06,
+      "loss": 0.113,
+      "step": 2080
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.3023259937763214,
+      "learning_rate": 2.0803782505910165e-06,
+      "loss": 0.1199,
+      "step": 2081
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.29377272725105286,
+      "learning_rate": 2.078900709219858e-06,
+      "loss": 0.1176,
+      "step": 2082
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.3019704520702362,
+      "learning_rate": 2.0774231678487e-06,
+      "loss": 0.1176,
+      "step": 2083
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2972950339317322,
+      "learning_rate": 2.0759456264775415e-06,
+      "loss": 0.1148,
+      "step": 2084
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2761232256889343,
+      "learning_rate": 2.074468085106383e-06,
+      "loss": 0.1132,
+      "step": 2085
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.29306912422180176,
+      "learning_rate": 2.0729905437352246e-06,
+      "loss": 0.1142,
+      "step": 2086
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.3006981909275055,
+      "learning_rate": 2.0715130023640665e-06,
+      "loss": 0.116,
+      "step": 2087
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2954864501953125,
+      "learning_rate": 2.070035460992908e-06,
+      "loss": 0.1145,
+      "step": 2088
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.28227224946022034,
+      "learning_rate": 2.0685579196217496e-06,
+      "loss": 0.1151,
+      "step": 2089
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.29292428493499756,
+      "learning_rate": 2.067080378250591e-06,
+      "loss": 0.1129,
+      "step": 2090
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.27747640013694763,
+      "learning_rate": 2.0656028368794327e-06,
+      "loss": 0.114,
+      "step": 2091
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2906121611595154,
+      "learning_rate": 2.0641252955082742e-06,
+      "loss": 0.1129,
+      "step": 2092
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.29417839646339417,
+      "learning_rate": 2.0626477541371158e-06,
+      "loss": 0.1163,
+      "step": 2093
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.28785890340805054,
+      "learning_rate": 2.0611702127659577e-06,
+      "loss": 0.1146,
+      "step": 2094
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2887338399887085,
+      "learning_rate": 2.0596926713947992e-06,
+      "loss": 0.1121,
+      "step": 2095
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2889114320278168,
+      "learning_rate": 2.0582151300236408e-06,
+      "loss": 0.1144,
+      "step": 2096
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.29300054907798767,
+      "learning_rate": 2.0567375886524823e-06,
+      "loss": 0.1162,
+      "step": 2097
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.3002152442932129,
+      "learning_rate": 2.0552600472813243e-06,
+      "loss": 0.1102,
+      "step": 2098
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2941839098930359,
+      "learning_rate": 2.053782505910166e-06,
+      "loss": 0.1183,
+      "step": 2099
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.30438682436943054,
+      "learning_rate": 2.0523049645390073e-06,
+      "loss": 0.1188,
+      "step": 2100
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2886125147342682,
+      "learning_rate": 2.050827423167849e-06,
+      "loss": 0.1161,
+      "step": 2101
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2728198766708374,
+      "learning_rate": 2.0493498817966904e-06,
+      "loss": 0.1082,
+      "step": 2102
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.30808454751968384,
+      "learning_rate": 2.047872340425532e-06,
+      "loss": 0.117,
+      "step": 2103
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2848914563655853,
+      "learning_rate": 2.0463947990543735e-06,
+      "loss": 0.1127,
+      "step": 2104
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.3023051917552948,
+      "learning_rate": 2.044917257683215e-06,
+      "loss": 0.1126,
+      "step": 2105
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2926468551158905,
+      "learning_rate": 2.043439716312057e-06,
+      "loss": 0.116,
+      "step": 2106
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2840218245983124,
+      "learning_rate": 2.0419621749408985e-06,
+      "loss": 0.1088,
+      "step": 2107
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.27374690771102905,
+      "learning_rate": 2.04048463356974e-06,
+      "loss": 0.1088,
+      "step": 2108
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.27914440631866455,
+      "learning_rate": 2.0390070921985816e-06,
+      "loss": 0.1123,
+      "step": 2109
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.29767754673957825,
+      "learning_rate": 2.0375295508274235e-06,
+      "loss": 0.1152,
+      "step": 2110
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2942991256713867,
+      "learning_rate": 2.036052009456265e-06,
+      "loss": 0.1177,
+      "step": 2111
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2809903621673584,
+      "learning_rate": 2.0345744680851066e-06,
+      "loss": 0.1139,
+      "step": 2112
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.30496326088905334,
+      "learning_rate": 2.033096926713948e-06,
+      "loss": 0.1156,
+      "step": 2113
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.30311182141304016,
+      "learning_rate": 2.0316193853427897e-06,
+      "loss": 0.1181,
+      "step": 2114
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2825095057487488,
+      "learning_rate": 2.030141843971631e-06,
+      "loss": 0.1105,
+      "step": 2115
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.29001256823539734,
+      "learning_rate": 2.0286643026004727e-06,
+      "loss": 0.1169,
+      "step": 2116
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.30685001611709595,
+      "learning_rate": 2.0271867612293143e-06,
+      "loss": 0.1177,
+      "step": 2117
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.29082387685775757,
+      "learning_rate": 2.025709219858156e-06,
+      "loss": 0.1188,
+      "step": 2118
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.285357266664505,
+      "learning_rate": 2.0242316784869977e-06,
+      "loss": 0.112,
+      "step": 2119
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.31027311086654663,
+      "learning_rate": 2.0227541371158393e-06,
+      "loss": 0.1212,
+      "step": 2120
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.3086315095424652,
+      "learning_rate": 2.021276595744681e-06,
+      "loss": 0.1123,
+      "step": 2121
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.28137117624282837,
+      "learning_rate": 2.0197990543735228e-06,
+      "loss": 0.1109,
+      "step": 2122
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.281871497631073,
+      "learning_rate": 2.0183215130023643e-06,
+      "loss": 0.1124,
+      "step": 2123
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2840617597103119,
+      "learning_rate": 2.016843971631206e-06,
+      "loss": 0.112,
+      "step": 2124
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.28953295946121216,
+      "learning_rate": 2.015366430260048e-06,
+      "loss": 0.1149,
+      "step": 2125
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.29407796263694763,
+      "learning_rate": 2.0138888888888893e-06,
+      "loss": 0.1157,
+      "step": 2126
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2891089916229248,
+      "learning_rate": 2.012411347517731e-06,
+      "loss": 0.1114,
+      "step": 2127
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2966189980506897,
+      "learning_rate": 2.0109338061465724e-06,
+      "loss": 0.1106,
+      "step": 2128
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.288862019777298,
+      "learning_rate": 2.009456264775414e-06,
+      "loss": 0.1107,
+      "step": 2129
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.3019852042198181,
+      "learning_rate": 2.0079787234042555e-06,
+      "loss": 0.1131,
+      "step": 2130
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2896791100502014,
+      "learning_rate": 2.006501182033097e-06,
+      "loss": 0.1138,
+      "step": 2131
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.29517096281051636,
+      "learning_rate": 2.0050236406619385e-06,
+      "loss": 0.112,
+      "step": 2132
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2918021082878113,
+      "learning_rate": 2.0035460992907805e-06,
+      "loss": 0.116,
+      "step": 2133
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.31408873200416565,
+      "learning_rate": 2.002068557919622e-06,
+      "loss": 0.1122,
+      "step": 2134
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2871900498867035,
+      "learning_rate": 2.0005910165484636e-06,
+      "loss": 0.1082,
+      "step": 2135
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.29072195291519165,
+      "learning_rate": 1.999113475177305e-06,
+      "loss": 0.1148,
+      "step": 2136
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2969684302806854,
+      "learning_rate": 1.997635933806147e-06,
+      "loss": 0.1133,
+      "step": 2137
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.28289756178855896,
+      "learning_rate": 1.9961583924349886e-06,
+      "loss": 0.1168,
+      "step": 2138
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.3048178255558014,
+      "learning_rate": 1.99468085106383e-06,
+      "loss": 0.1186,
+      "step": 2139
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2915510833263397,
+      "learning_rate": 1.9932033096926716e-06,
+      "loss": 0.114,
+      "step": 2140
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2859833538532257,
+      "learning_rate": 1.991725768321513e-06,
+      "loss": 0.1213,
+      "step": 2141
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.30284062027931213,
+      "learning_rate": 1.9902482269503547e-06,
+      "loss": 0.1122,
+      "step": 2142
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2962600588798523,
+      "learning_rate": 1.9887706855791962e-06,
+      "loss": 0.1185,
+      "step": 2143
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2962925136089325,
+      "learning_rate": 1.9872931442080378e-06,
+      "loss": 0.1164,
+      "step": 2144
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.28574785590171814,
+      "learning_rate": 1.9858156028368797e-06,
+      "loss": 0.1147,
+      "step": 2145
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.313293993473053,
+      "learning_rate": 1.9843380614657213e-06,
+      "loss": 0.1162,
+      "step": 2146
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.29243847727775574,
+      "learning_rate": 1.982860520094563e-06,
+      "loss": 0.112,
+      "step": 2147
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.27792489528656006,
+      "learning_rate": 1.9813829787234043e-06,
+      "loss": 0.1096,
+      "step": 2148
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.27734920382499695,
+      "learning_rate": 1.9799054373522463e-06,
+      "loss": 0.112,
+      "step": 2149
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2952900528907776,
+      "learning_rate": 1.978427895981088e-06,
+      "loss": 0.1191,
+      "step": 2150
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.28991320729255676,
+      "learning_rate": 1.9769503546099294e-06,
+      "loss": 0.1169,
+      "step": 2151
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2789981961250305,
+      "learning_rate": 1.975472813238771e-06,
+      "loss": 0.1098,
+      "step": 2152
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.28141218423843384,
+      "learning_rate": 1.9739952718676124e-06,
+      "loss": 0.1087,
+      "step": 2153
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2851007580757141,
+      "learning_rate": 1.972517730496454e-06,
+      "loss": 0.1162,
+      "step": 2154
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2707831859588623,
+      "learning_rate": 1.9710401891252955e-06,
+      "loss": 0.1087,
+      "step": 2155
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2749985158443451,
+      "learning_rate": 1.9695626477541374e-06,
+      "loss": 0.1108,
+      "step": 2156
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.27598991990089417,
+      "learning_rate": 1.968085106382979e-06,
+      "loss": 0.1168,
+      "step": 2157
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.29300349950790405,
+      "learning_rate": 1.9666075650118205e-06,
+      "loss": 0.1176,
+      "step": 2158
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.28490304946899414,
+      "learning_rate": 1.965130023640662e-06,
+      "loss": 0.1154,
+      "step": 2159
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.3090299367904663,
+      "learning_rate": 1.963652482269504e-06,
+      "loss": 0.1154,
+      "step": 2160
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2762569785118103,
+      "learning_rate": 1.9621749408983455e-06,
+      "loss": 0.1163,
+      "step": 2161
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.28700074553489685,
+      "learning_rate": 1.960697399527187e-06,
+      "loss": 0.1145,
+      "step": 2162
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.29782989621162415,
+      "learning_rate": 1.9592198581560286e-06,
+      "loss": 0.1188,
+      "step": 2163
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2949253022670746,
+      "learning_rate": 1.95774231678487e-06,
+      "loss": 0.1133,
+      "step": 2164
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.29020193219184875,
+      "learning_rate": 1.9562647754137117e-06,
+      "loss": 0.1146,
+      "step": 2165
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2980537414550781,
+      "learning_rate": 1.954787234042553e-06,
+      "loss": 0.1149,
+      "step": 2166
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2975660562515259,
+      "learning_rate": 1.9533096926713947e-06,
+      "loss": 0.1105,
+      "step": 2167
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2719798684120178,
+      "learning_rate": 1.9518321513002367e-06,
+      "loss": 0.11,
+      "step": 2168
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.29912689328193665,
+      "learning_rate": 1.9503546099290782e-06,
+      "loss": 0.1149,
+      "step": 2169
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2853172719478607,
+      "learning_rate": 1.9488770685579198e-06,
+      "loss": 0.1176,
+      "step": 2170
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.28210878372192383,
+      "learning_rate": 1.9473995271867613e-06,
+      "loss": 0.1186,
+      "step": 2171
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2807007133960724,
+      "learning_rate": 1.9459219858156033e-06,
+      "loss": 0.1113,
+      "step": 2172
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.300578773021698,
+      "learning_rate": 1.944444444444445e-06,
+      "loss": 0.1095,
+      "step": 2173
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2712564170360565,
+      "learning_rate": 1.9429669030732863e-06,
+      "loss": 0.1048,
+      "step": 2174
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2865838408470154,
+      "learning_rate": 1.941489361702128e-06,
+      "loss": 0.1134,
+      "step": 2175
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2857939600944519,
+      "learning_rate": 1.9400118203309694e-06,
+      "loss": 0.1105,
+      "step": 2176
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.30354204773902893,
+      "learning_rate": 1.938534278959811e-06,
+      "loss": 0.114,
+      "step": 2177
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.28888222575187683,
+      "learning_rate": 1.9370567375886525e-06,
+      "loss": 0.1143,
+      "step": 2178
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.3023698627948761,
+      "learning_rate": 1.935579196217494e-06,
+      "loss": 0.1149,
+      "step": 2179
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.31252068281173706,
+      "learning_rate": 1.934101654846336e-06,
+      "loss": 0.1188,
+      "step": 2180
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2918255031108856,
+      "learning_rate": 1.9326241134751775e-06,
+      "loss": 0.1143,
+      "step": 2181
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2880353629589081,
+      "learning_rate": 1.931146572104019e-06,
+      "loss": 0.1153,
+      "step": 2182
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2894575297832489,
+      "learning_rate": 1.9296690307328606e-06,
+      "loss": 0.1141,
+      "step": 2183
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.29142406582832336,
+      "learning_rate": 1.9281914893617025e-06,
+      "loss": 0.1137,
+      "step": 2184
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.27166828513145447,
+      "learning_rate": 1.926713947990544e-06,
+      "loss": 0.1137,
+      "step": 2185
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2828565835952759,
+      "learning_rate": 1.9252364066193856e-06,
+      "loss": 0.1134,
+      "step": 2186
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.28398793935775757,
+      "learning_rate": 1.923758865248227e-06,
+      "loss": 0.1168,
+      "step": 2187
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.28126394748687744,
+      "learning_rate": 1.9222813238770686e-06,
+      "loss": 0.1097,
+      "step": 2188
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.30196845531463623,
+      "learning_rate": 1.92080378250591e-06,
+      "loss": 0.1178,
+      "step": 2189
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.30255094170570374,
+      "learning_rate": 1.9193262411347517e-06,
+      "loss": 0.1129,
+      "step": 2190
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3082413375377655,
+      "learning_rate": 1.9178486997635937e-06,
+      "loss": 0.1181,
+      "step": 2191
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.29044589400291443,
+      "learning_rate": 1.916371158392435e-06,
+      "loss": 0.1155,
+      "step": 2192
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2762584984302521,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 0.1134,
+      "step": 2193
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2921905219554901,
+      "learning_rate": 1.9134160756501183e-06,
+      "loss": 0.1137,
+      "step": 2194
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.27861788868904114,
+      "learning_rate": 1.9119385342789602e-06,
+      "loss": 0.1152,
+      "step": 2195
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2868460416793823,
+      "learning_rate": 1.9104609929078018e-06,
+      "loss": 0.1125,
+      "step": 2196
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.30000385642051697,
+      "learning_rate": 1.9089834515366433e-06,
+      "loss": 0.1144,
+      "step": 2197
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2853168845176697,
+      "learning_rate": 1.907505910165485e-06,
+      "loss": 0.1148,
+      "step": 2198
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2853049337863922,
+      "learning_rate": 1.9060283687943266e-06,
+      "loss": 0.1154,
+      "step": 2199
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.30132025480270386,
+      "learning_rate": 1.904550827423168e-06,
+      "loss": 0.1185,
+      "step": 2200
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.27519527077674866,
+      "learning_rate": 1.9030732860520096e-06,
+      "loss": 0.1119,
+      "step": 2201
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2958602011203766,
+      "learning_rate": 1.9015957446808512e-06,
+      "loss": 0.119,
+      "step": 2202
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2835947275161743,
+      "learning_rate": 1.900118203309693e-06,
+      "loss": 0.1126,
+      "step": 2203
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.305183082818985,
+      "learning_rate": 1.8986406619385345e-06,
+      "loss": 0.1138,
+      "step": 2204
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2993698716163635,
+      "learning_rate": 1.897163120567376e-06,
+      "loss": 0.1183,
+      "step": 2205
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.27937713265419006,
+      "learning_rate": 1.8956855791962175e-06,
+      "loss": 0.1135,
+      "step": 2206
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.3778619170188904,
+      "learning_rate": 1.8942080378250593e-06,
+      "loss": 0.119,
+      "step": 2207
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2794092297554016,
+      "learning_rate": 1.8927304964539008e-06,
+      "loss": 0.1152,
+      "step": 2208
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.28786709904670715,
+      "learning_rate": 1.8912529550827423e-06,
+      "loss": 0.1106,
+      "step": 2209
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2804581820964813,
+      "learning_rate": 1.8897754137115839e-06,
+      "loss": 0.1158,
+      "step": 2210
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2878835201263428,
+      "learning_rate": 1.8882978723404258e-06,
+      "loss": 0.1141,
+      "step": 2211
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.290455162525177,
+      "learning_rate": 1.8868203309692674e-06,
+      "loss": 0.1173,
+      "step": 2212
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2919381260871887,
+      "learning_rate": 1.8853427895981089e-06,
+      "loss": 0.1138,
+      "step": 2213
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.28889450430870056,
+      "learning_rate": 1.8838652482269504e-06,
+      "loss": 0.1152,
+      "step": 2214
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.28930047154426575,
+      "learning_rate": 1.8823877068557922e-06,
+      "loss": 0.1153,
+      "step": 2215
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2951924502849579,
+      "learning_rate": 1.8809101654846337e-06,
+      "loss": 0.1186,
+      "step": 2216
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.284796804189682,
+      "learning_rate": 1.8794326241134752e-06,
+      "loss": 0.1091,
+      "step": 2217
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.28856009244918823,
+      "learning_rate": 1.8779550827423168e-06,
+      "loss": 0.1143,
+      "step": 2218
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2872783839702606,
+      "learning_rate": 1.8764775413711585e-06,
+      "loss": 0.1165,
+      "step": 2219
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2758755683898926,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.1109,
+      "step": 2220
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2860746681690216,
+      "learning_rate": 1.8735224586288418e-06,
+      "loss": 0.1114,
+      "step": 2221
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.27539557218551636,
+      "learning_rate": 1.8720449172576835e-06,
+      "loss": 0.1155,
+      "step": 2222
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.29609543085098267,
+      "learning_rate": 1.870567375886525e-06,
+      "loss": 0.1143,
+      "step": 2223
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.31600242853164673,
+      "learning_rate": 1.8690898345153666e-06,
+      "loss": 0.1155,
+      "step": 2224
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.28252488374710083,
+      "learning_rate": 1.8676122931442081e-06,
+      "loss": 0.115,
+      "step": 2225
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.285282701253891,
+      "learning_rate": 1.8661347517730499e-06,
+      "loss": 0.1125,
+      "step": 2226
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.3068080246448517,
+      "learning_rate": 1.8646572104018914e-06,
+      "loss": 0.1146,
+      "step": 2227
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.28032100200653076,
+      "learning_rate": 1.863179669030733e-06,
+      "loss": 0.1082,
+      "step": 2228
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2798764407634735,
+      "learning_rate": 1.8617021276595745e-06,
+      "loss": 0.1114,
+      "step": 2229
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.28508803248405457,
+      "learning_rate": 1.8602245862884164e-06,
+      "loss": 0.1138,
+      "step": 2230
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.29431527853012085,
+      "learning_rate": 1.858747044917258e-06,
+      "loss": 0.1154,
+      "step": 2231
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.30342376232147217,
+      "learning_rate": 1.8572695035460995e-06,
+      "loss": 0.1147,
+      "step": 2232
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2833671271800995,
+      "learning_rate": 1.855791962174941e-06,
+      "loss": 0.11,
+      "step": 2233
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2891712784767151,
+      "learning_rate": 1.8543144208037828e-06,
+      "loss": 0.1162,
+      "step": 2234
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2837274670600891,
+      "learning_rate": 1.8528368794326243e-06,
+      "loss": 0.1113,
+      "step": 2235
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2900477945804596,
+      "learning_rate": 1.8513593380614659e-06,
+      "loss": 0.1158,
+      "step": 2236
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2959243357181549,
+      "learning_rate": 1.8498817966903074e-06,
+      "loss": 0.1127,
+      "step": 2237
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2977566421031952,
+      "learning_rate": 1.8484042553191491e-06,
+      "loss": 0.1168,
+      "step": 2238
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2897810935974121,
+      "learning_rate": 1.8469267139479907e-06,
+      "loss": 0.11,
+      "step": 2239
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2881297171115875,
+      "learning_rate": 1.8454491725768322e-06,
+      "loss": 0.1133,
+      "step": 2240
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.298189252614975,
+      "learning_rate": 1.8439716312056737e-06,
+      "loss": 0.1169,
+      "step": 2241
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.29187700152397156,
+      "learning_rate": 1.8424940898345157e-06,
+      "loss": 0.1129,
+      "step": 2242
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2757470905780792,
+      "learning_rate": 1.8410165484633572e-06,
+      "loss": 0.1122,
+      "step": 2243
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.29930731654167175,
+      "learning_rate": 1.8395390070921988e-06,
+      "loss": 0.1136,
+      "step": 2244
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.28852495551109314,
+      "learning_rate": 1.8380614657210403e-06,
+      "loss": 0.1158,
+      "step": 2245
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2868964374065399,
+      "learning_rate": 1.836583924349882e-06,
+      "loss": 0.1164,
+      "step": 2246
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2984849810600281,
+      "learning_rate": 1.8351063829787236e-06,
+      "loss": 0.1128,
+      "step": 2247
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.28786176443099976,
+      "learning_rate": 1.8336288416075651e-06,
+      "loss": 0.1131,
+      "step": 2248
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2871016561985016,
+      "learning_rate": 1.8321513002364066e-06,
+      "loss": 0.114,
+      "step": 2249
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.29241591691970825,
+      "learning_rate": 1.8306737588652484e-06,
+      "loss": 0.1155,
+      "step": 2250
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2939310371875763,
+      "learning_rate": 1.82919621749409e-06,
+      "loss": 0.1145,
+      "step": 2251
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2967120110988617,
+      "learning_rate": 1.8277186761229315e-06,
+      "loss": 0.1141,
+      "step": 2252
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2916867136955261,
+      "learning_rate": 1.8262411347517734e-06,
+      "loss": 0.1136,
+      "step": 2253
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3120230734348297,
+      "learning_rate": 1.824763593380615e-06,
+      "loss": 0.1179,
+      "step": 2254
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3026760518550873,
+      "learning_rate": 1.8232860520094565e-06,
+      "loss": 0.1176,
+      "step": 2255
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.29826703667640686,
+      "learning_rate": 1.821808510638298e-06,
+      "loss": 0.1143,
+      "step": 2256
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.29274311661720276,
+      "learning_rate": 1.8203309692671398e-06,
+      "loss": 0.1146,
+      "step": 2257
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2847159206867218,
+      "learning_rate": 1.8188534278959813e-06,
+      "loss": 0.1125,
+      "step": 2258
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.29881909489631653,
+      "learning_rate": 1.8173758865248228e-06,
+      "loss": 0.1198,
+      "step": 2259
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.30652856826782227,
+      "learning_rate": 1.8158983451536644e-06,
+      "loss": 0.1205,
+      "step": 2260
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2915562391281128,
+      "learning_rate": 1.814420803782506e-06,
+      "loss": 0.1149,
+      "step": 2261
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3001241683959961,
+      "learning_rate": 1.8129432624113476e-06,
+      "loss": 0.1142,
+      "step": 2262
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.3032921254634857,
+      "learning_rate": 1.8114657210401892e-06,
+      "loss": 0.1136,
+      "step": 2263
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2904949188232422,
+      "learning_rate": 1.8099881796690307e-06,
+      "loss": 0.1125,
+      "step": 2264
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.29200151562690735,
+      "learning_rate": 1.8085106382978727e-06,
+      "loss": 0.1146,
+      "step": 2265
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.28894492983818054,
+      "learning_rate": 1.8070330969267142e-06,
+      "loss": 0.1166,
+      "step": 2266
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.28607892990112305,
+      "learning_rate": 1.8055555555555557e-06,
+      "loss": 0.1122,
+      "step": 2267
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.30980342626571655,
+      "learning_rate": 1.8040780141843973e-06,
+      "loss": 0.1188,
+      "step": 2268
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2985472083091736,
+      "learning_rate": 1.802600472813239e-06,
+      "loss": 0.1205,
+      "step": 2269
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2962723970413208,
+      "learning_rate": 1.8011229314420805e-06,
+      "loss": 0.1163,
+      "step": 2270
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.291146457195282,
+      "learning_rate": 1.799645390070922e-06,
+      "loss": 0.1126,
+      "step": 2271
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.29331836104393005,
+      "learning_rate": 1.7981678486997636e-06,
+      "loss": 0.1117,
+      "step": 2272
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.28177857398986816,
+      "learning_rate": 1.7966903073286054e-06,
+      "loss": 0.1118,
+      "step": 2273
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.28288203477859497,
+      "learning_rate": 1.7952127659574469e-06,
+      "loss": 0.1114,
+      "step": 2274
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.291878879070282,
+      "learning_rate": 1.7937352245862884e-06,
+      "loss": 0.1136,
+      "step": 2275
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.29085269570350647,
+      "learning_rate": 1.79225768321513e-06,
+      "loss": 0.1142,
+      "step": 2276
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2798583507537842,
+      "learning_rate": 1.790780141843972e-06,
+      "loss": 0.1114,
+      "step": 2277
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.29668503999710083,
+      "learning_rate": 1.7893026004728134e-06,
+      "loss": 0.117,
+      "step": 2278
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2943309247493744,
+      "learning_rate": 1.787825059101655e-06,
+      "loss": 0.1103,
+      "step": 2279
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.3087671399116516,
+      "learning_rate": 1.7863475177304965e-06,
+      "loss": 0.1185,
+      "step": 2280
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.29699620604515076,
+      "learning_rate": 1.7848699763593383e-06,
+      "loss": 0.1138,
+      "step": 2281
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.3048483431339264,
+      "learning_rate": 1.7833924349881798e-06,
+      "loss": 0.1179,
+      "step": 2282
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2899823486804962,
+      "learning_rate": 1.7819148936170213e-06,
+      "loss": 0.1149,
+      "step": 2283
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2938592731952667,
+      "learning_rate": 1.7804373522458629e-06,
+      "loss": 0.1094,
+      "step": 2284
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2925514876842499,
+      "learning_rate": 1.7789598108747046e-06,
+      "loss": 0.1147,
+      "step": 2285
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.27580395340919495,
+      "learning_rate": 1.7774822695035463e-06,
+      "loss": 0.1082,
+      "step": 2286
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.29947608709335327,
+      "learning_rate": 1.7760047281323879e-06,
+      "loss": 0.115,
+      "step": 2287
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.3095090091228485,
+      "learning_rate": 1.7745271867612296e-06,
+      "loss": 0.1148,
+      "step": 2288
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.30156761407852173,
+      "learning_rate": 1.7730496453900712e-06,
+      "loss": 0.1178,
+      "step": 2289
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.29009100794792175,
+      "learning_rate": 1.7715721040189127e-06,
+      "loss": 0.1146,
+      "step": 2290
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2717399597167969,
+      "learning_rate": 1.7700945626477542e-06,
+      "loss": 0.1117,
+      "step": 2291
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.30193158984184265,
+      "learning_rate": 1.768617021276596e-06,
+      "loss": 0.1065,
+      "step": 2292
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.3018586337566376,
+      "learning_rate": 1.7671394799054375e-06,
+      "loss": 0.1195,
+      "step": 2293
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2862337529659271,
+      "learning_rate": 1.765661938534279e-06,
+      "loss": 0.1119,
+      "step": 2294
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.29999321699142456,
+      "learning_rate": 1.7641843971631206e-06,
+      "loss": 0.114,
+      "step": 2295
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.27612584829330444,
+      "learning_rate": 1.7627068557919625e-06,
+      "loss": 0.1164,
+      "step": 2296
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.28371861577033997,
+      "learning_rate": 1.761229314420804e-06,
+      "loss": 0.1111,
+      "step": 2297
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2856762707233429,
+      "learning_rate": 1.7597517730496456e-06,
+      "loss": 0.1114,
+      "step": 2298
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.29717156291007996,
+      "learning_rate": 1.7582742316784871e-06,
+      "loss": 0.1113,
+      "step": 2299
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2878396511077881,
+      "learning_rate": 1.7567966903073289e-06,
+      "loss": 0.1161,
+      "step": 2300
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.29438477754592896,
+      "learning_rate": 1.7553191489361704e-06,
+      "loss": 0.1141,
+      "step": 2301
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.30244189500808716,
+      "learning_rate": 1.753841607565012e-06,
+      "loss": 0.1108,
+      "step": 2302
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.30174821615219116,
+      "learning_rate": 1.7523640661938535e-06,
+      "loss": 0.1147,
+      "step": 2303
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2939281165599823,
+      "learning_rate": 1.7508865248226952e-06,
+      "loss": 0.1108,
+      "step": 2304
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2996091842651367,
+      "learning_rate": 1.7494089834515368e-06,
+      "loss": 0.1108,
+      "step": 2305
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.27365562319755554,
+      "learning_rate": 1.7479314420803783e-06,
+      "loss": 0.1135,
+      "step": 2306
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.29914361238479614,
+      "learning_rate": 1.7464539007092198e-06,
+      "loss": 0.1187,
+      "step": 2307
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2949599623680115,
+      "learning_rate": 1.7449763593380618e-06,
+      "loss": 0.1148,
+      "step": 2308
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2832069396972656,
+      "learning_rate": 1.7434988179669033e-06,
+      "loss": 0.1132,
+      "step": 2309
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.28514814376831055,
+      "learning_rate": 1.7420212765957448e-06,
+      "loss": 0.1134,
+      "step": 2310
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2938980460166931,
+      "learning_rate": 1.7405437352245864e-06,
+      "loss": 0.1145,
+      "step": 2311
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2806869447231293,
+      "learning_rate": 1.7390661938534281e-06,
+      "loss": 0.1124,
+      "step": 2312
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2887020409107208,
+      "learning_rate": 1.7375886524822697e-06,
+      "loss": 0.1131,
+      "step": 2313
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.29470404982566833,
+      "learning_rate": 1.7361111111111112e-06,
+      "loss": 0.1128,
+      "step": 2314
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.28729313611984253,
+      "learning_rate": 1.7346335697399527e-06,
+      "loss": 0.1105,
+      "step": 2315
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.30053889751434326,
+      "learning_rate": 1.7331560283687945e-06,
+      "loss": 0.1184,
+      "step": 2316
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2935018241405487,
+      "learning_rate": 1.731678486997636e-06,
+      "loss": 0.1158,
+      "step": 2317
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.28204721212387085,
+      "learning_rate": 1.7302009456264775e-06,
+      "loss": 0.1086,
+      "step": 2318
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2948352098464966,
+      "learning_rate": 1.7287234042553195e-06,
+      "loss": 0.1079,
+      "step": 2319
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.28264257311820984,
+      "learning_rate": 1.727245862884161e-06,
+      "loss": 0.1114,
+      "step": 2320
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.31276652216911316,
+      "learning_rate": 1.7257683215130026e-06,
+      "loss": 0.1185,
+      "step": 2321
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.313482403755188,
+      "learning_rate": 1.724290780141844e-06,
+      "loss": 0.1194,
+      "step": 2322
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2782033383846283,
+      "learning_rate": 1.7228132387706858e-06,
+      "loss": 0.1124,
+      "step": 2323
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2881643772125244,
+      "learning_rate": 1.7213356973995274e-06,
+      "loss": 0.1136,
+      "step": 2324
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2864632308483124,
+      "learning_rate": 1.719858156028369e-06,
+      "loss": 0.1128,
+      "step": 2325
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.29543817043304443,
+      "learning_rate": 1.7183806146572104e-06,
+      "loss": 0.1143,
+      "step": 2326
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.27833855152130127,
+      "learning_rate": 1.7169030732860522e-06,
+      "loss": 0.1027,
+      "step": 2327
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2832610011100769,
+      "learning_rate": 1.7154255319148937e-06,
+      "loss": 0.1032,
+      "step": 2328
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.27698007225990295,
+      "learning_rate": 1.7139479905437353e-06,
+      "loss": 0.1026,
+      "step": 2329
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2640492618083954,
+      "learning_rate": 1.7124704491725768e-06,
+      "loss": 0.1041,
+      "step": 2330
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2630833387374878,
+      "learning_rate": 1.7109929078014187e-06,
+      "loss": 0.1002,
+      "step": 2331
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.279683381319046,
+      "learning_rate": 1.7095153664302603e-06,
+      "loss": 0.1041,
+      "step": 2332
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.27962374687194824,
+      "learning_rate": 1.7080378250591018e-06,
+      "loss": 0.0963,
+      "step": 2333
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.26919347047805786,
+      "learning_rate": 1.7065602836879433e-06,
+      "loss": 0.0974,
+      "step": 2334
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.27701398730278015,
+      "learning_rate": 1.705082742316785e-06,
+      "loss": 0.0977,
+      "step": 2335
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2857283353805542,
+      "learning_rate": 1.7036052009456266e-06,
+      "loss": 0.1035,
+      "step": 2336
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.27304279804229736,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 0.1011,
+      "step": 2337
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.29119873046875,
+      "learning_rate": 1.7006501182033097e-06,
+      "loss": 0.0975,
+      "step": 2338
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.28418490290641785,
+      "learning_rate": 1.6991725768321514e-06,
+      "loss": 0.1016,
+      "step": 2339
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2920931875705719,
+      "learning_rate": 1.697695035460993e-06,
+      "loss": 0.1055,
+      "step": 2340
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.30196309089660645,
+      "learning_rate": 1.6962174940898345e-06,
+      "loss": 0.1026,
+      "step": 2341
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.3092053532600403,
+      "learning_rate": 1.694739952718676e-06,
+      "loss": 0.0992,
+      "step": 2342
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.27946844696998596,
+      "learning_rate": 1.693262411347518e-06,
+      "loss": 0.1016,
+      "step": 2343
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.292696475982666,
+      "learning_rate": 1.6917848699763595e-06,
+      "loss": 0.1024,
+      "step": 2344
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.27167221903800964,
+      "learning_rate": 1.690307328605201e-06,
+      "loss": 0.0967,
+      "step": 2345
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2841625213623047,
+      "learning_rate": 1.6888297872340426e-06,
+      "loss": 0.1006,
+      "step": 2346
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2752596437931061,
+      "learning_rate": 1.6873522458628843e-06,
+      "loss": 0.1004,
+      "step": 2347
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.28646716475486755,
+      "learning_rate": 1.6858747044917259e-06,
+      "loss": 0.1031,
+      "step": 2348
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2770024240016937,
+      "learning_rate": 1.6843971631205674e-06,
+      "loss": 0.1006,
+      "step": 2349
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2720467448234558,
+      "learning_rate": 1.682919621749409e-06,
+      "loss": 0.0981,
+      "step": 2350
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2653568387031555,
+      "learning_rate": 1.6814420803782507e-06,
+      "loss": 0.0977,
+      "step": 2351
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2717125713825226,
+      "learning_rate": 1.6799645390070924e-06,
+      "loss": 0.0976,
+      "step": 2352
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.27147597074508667,
+      "learning_rate": 1.678486997635934e-06,
+      "loss": 0.0982,
+      "step": 2353
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2934548556804657,
+      "learning_rate": 1.6770094562647757e-06,
+      "loss": 0.0972,
+      "step": 2354
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.28466281294822693,
+      "learning_rate": 1.6755319148936172e-06,
+      "loss": 0.1001,
+      "step": 2355
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.27623334527015686,
+      "learning_rate": 1.6740543735224588e-06,
+      "loss": 0.0984,
+      "step": 2356
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.27629122138023376,
+      "learning_rate": 1.6725768321513003e-06,
+      "loss": 0.1022,
+      "step": 2357
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2733404040336609,
+      "learning_rate": 1.671099290780142e-06,
+      "loss": 0.0991,
+      "step": 2358
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2766924798488617,
+      "learning_rate": 1.6696217494089836e-06,
+      "loss": 0.1016,
+      "step": 2359
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.297608345746994,
+      "learning_rate": 1.6681442080378251e-06,
+      "loss": 0.1039,
+      "step": 2360
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.28820985555648804,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.1,
+      "step": 2361
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.28066518902778625,
+      "learning_rate": 1.6651891252955086e-06,
+      "loss": 0.1001,
+      "step": 2362
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2893994152545929,
+      "learning_rate": 1.6637115839243502e-06,
+      "loss": 0.1049,
+      "step": 2363
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2885749936103821,
+      "learning_rate": 1.6622340425531917e-06,
+      "loss": 0.0984,
+      "step": 2364
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.28250664472579956,
+      "learning_rate": 1.6607565011820332e-06,
+      "loss": 0.1021,
+      "step": 2365
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2792646586894989,
+      "learning_rate": 1.659278959810875e-06,
+      "loss": 0.098,
+      "step": 2366
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2783917784690857,
+      "learning_rate": 1.6578014184397165e-06,
+      "loss": 0.1029,
+      "step": 2367
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.287564218044281,
+      "learning_rate": 1.656323877068558e-06,
+      "loss": 0.1003,
+      "step": 2368
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.28596606850624084,
+      "learning_rate": 1.6548463356973996e-06,
+      "loss": 0.1049,
+      "step": 2369
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.27297908067703247,
+      "learning_rate": 1.6533687943262413e-06,
+      "loss": 0.1017,
+      "step": 2370
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.28268593549728394,
+      "learning_rate": 1.6518912529550828e-06,
+      "loss": 0.1009,
+      "step": 2371
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2934612035751343,
+      "learning_rate": 1.6504137115839244e-06,
+      "loss": 0.107,
+      "step": 2372
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2998588979244232,
+      "learning_rate": 1.648936170212766e-06,
+      "loss": 0.1077,
+      "step": 2373
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.28457850217819214,
+      "learning_rate": 1.6474586288416079e-06,
+      "loss": 0.1013,
+      "step": 2374
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2709556221961975,
+      "learning_rate": 1.6459810874704494e-06,
+      "loss": 0.1015,
+      "step": 2375
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.28184080123901367,
+      "learning_rate": 1.644503546099291e-06,
+      "loss": 0.1023,
+      "step": 2376
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2929559051990509,
+      "learning_rate": 1.6430260047281325e-06,
+      "loss": 0.1009,
+      "step": 2377
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.26676449179649353,
+      "learning_rate": 1.6415484633569742e-06,
+      "loss": 0.1022,
+      "step": 2378
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2755708396434784,
+      "learning_rate": 1.6400709219858157e-06,
+      "loss": 0.1037,
+      "step": 2379
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.27607542276382446,
+      "learning_rate": 1.6385933806146573e-06,
+      "loss": 0.1004,
+      "step": 2380
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2789907455444336,
+      "learning_rate": 1.6371158392434988e-06,
+      "loss": 0.1009,
+      "step": 2381
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.294049471616745,
+      "learning_rate": 1.6356382978723406e-06,
+      "loss": 0.1011,
+      "step": 2382
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.27792078256607056,
+      "learning_rate": 1.634160756501182e-06,
+      "loss": 0.1032,
+      "step": 2383
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2859567701816559,
+      "learning_rate": 1.6326832151300236e-06,
+      "loss": 0.1016,
+      "step": 2384
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2963901460170746,
+      "learning_rate": 1.6312056737588656e-06,
+      "loss": 0.1019,
+      "step": 2385
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2874448597431183,
+      "learning_rate": 1.6297281323877071e-06,
+      "loss": 0.102,
+      "step": 2386
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.27752822637557983,
+      "learning_rate": 1.6282505910165487e-06,
+      "loss": 0.1021,
+      "step": 2387
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2879457175731659,
+      "learning_rate": 1.6267730496453902e-06,
+      "loss": 0.1055,
+      "step": 2388
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.26426151394844055,
+      "learning_rate": 1.625295508274232e-06,
+      "loss": 0.101,
+      "step": 2389
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.26926976442337036,
+      "learning_rate": 1.6238179669030735e-06,
+      "loss": 0.1006,
+      "step": 2390
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.30479148030281067,
+      "learning_rate": 1.622340425531915e-06,
+      "loss": 0.1013,
+      "step": 2391
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.28178951144218445,
+      "learning_rate": 1.6208628841607565e-06,
+      "loss": 0.1008,
+      "step": 2392
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.3022918701171875,
+      "learning_rate": 1.6193853427895983e-06,
+      "loss": 0.1026,
+      "step": 2393
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.2791891396045685,
+      "learning_rate": 1.6179078014184398e-06,
+      "loss": 0.0961,
+      "step": 2394
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.27909085154533386,
+      "learning_rate": 1.6164302600472813e-06,
+      "loss": 0.0995,
+      "step": 2395
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.27849242091178894,
+      "learning_rate": 1.6149527186761229e-06,
+      "loss": 0.0978,
+      "step": 2396
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.29277774691581726,
+      "learning_rate": 1.6134751773049648e-06,
+      "loss": 0.1047,
+      "step": 2397
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.28572532534599304,
+      "learning_rate": 1.6119976359338064e-06,
+      "loss": 0.1026,
+      "step": 2398
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.285919189453125,
+      "learning_rate": 1.610520094562648e-06,
+      "loss": 0.1009,
+      "step": 2399
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.29157981276512146,
+      "learning_rate": 1.6090425531914894e-06,
+      "loss": 0.1045,
+      "step": 2400
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.28791433572769165,
+      "learning_rate": 1.6075650118203312e-06,
+      "loss": 0.1017,
+      "step": 2401
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2839343249797821,
+      "learning_rate": 1.6060874704491727e-06,
+      "loss": 0.1015,
+      "step": 2402
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2876354157924652,
+      "learning_rate": 1.6046099290780142e-06,
+      "loss": 0.1004,
+      "step": 2403
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.28077203035354614,
+      "learning_rate": 1.6031323877068558e-06,
+      "loss": 0.0953,
+      "step": 2404
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2791903018951416,
+      "learning_rate": 1.6016548463356975e-06,
+      "loss": 0.1,
+      "step": 2405
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2938844561576843,
+      "learning_rate": 1.600177304964539e-06,
+      "loss": 0.1033,
+      "step": 2406
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2853633165359497,
+      "learning_rate": 1.5986997635933806e-06,
+      "loss": 0.1061,
+      "step": 2407
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.29488539695739746,
+      "learning_rate": 1.5972222222222221e-06,
+      "loss": 0.1022,
+      "step": 2408
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2931897044181824,
+      "learning_rate": 1.595744680851064e-06,
+      "loss": 0.1033,
+      "step": 2409
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.29492074251174927,
+      "learning_rate": 1.5942671394799056e-06,
+      "loss": 0.1017,
+      "step": 2410
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.28838834166526794,
+      "learning_rate": 1.5927895981087472e-06,
+      "loss": 0.0979,
+      "step": 2411
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.26292821764945984,
+      "learning_rate": 1.5913120567375887e-06,
+      "loss": 0.101,
+      "step": 2412
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.27453941106796265,
+      "learning_rate": 1.5898345153664304e-06,
+      "loss": 0.0988,
+      "step": 2413
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2912222743034363,
+      "learning_rate": 1.588356973995272e-06,
+      "loss": 0.104,
+      "step": 2414
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.29044854640960693,
+      "learning_rate": 1.5868794326241135e-06,
+      "loss": 0.0996,
+      "step": 2415
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2874346375465393,
+      "learning_rate": 1.5854018912529552e-06,
+      "loss": 0.0977,
+      "step": 2416
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.3096331059932709,
+      "learning_rate": 1.583924349881797e-06,
+      "loss": 0.0996,
+      "step": 2417
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2928108870983124,
+      "learning_rate": 1.5824468085106385e-06,
+      "loss": 0.102,
+      "step": 2418
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2843978703022003,
+      "learning_rate": 1.58096926713948e-06,
+      "loss": 0.0983,
+      "step": 2419
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.29167672991752625,
+      "learning_rate": 1.5794917257683218e-06,
+      "loss": 0.104,
+      "step": 2420
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2938622534275055,
+      "learning_rate": 1.5780141843971633e-06,
+      "loss": 0.1011,
+      "step": 2421
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.28277653455734253,
+      "learning_rate": 1.5765366430260049e-06,
+      "loss": 0.0982,
+      "step": 2422
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.28260311484336853,
+      "learning_rate": 1.5750591016548464e-06,
+      "loss": 0.1044,
+      "step": 2423
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.29971638321876526,
+      "learning_rate": 1.5735815602836881e-06,
+      "loss": 0.0973,
+      "step": 2424
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.3148963749408722,
+      "learning_rate": 1.5721040189125297e-06,
+      "loss": 0.1043,
+      "step": 2425
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.28889045119285583,
+      "learning_rate": 1.5706264775413712e-06,
+      "loss": 0.1035,
+      "step": 2426
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.29094818234443665,
+      "learning_rate": 1.5691489361702128e-06,
+      "loss": 0.0993,
+      "step": 2427
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2852533757686615,
+      "learning_rate": 1.5676713947990547e-06,
+      "loss": 0.098,
+      "step": 2428
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.27466657757759094,
+      "learning_rate": 1.5661938534278962e-06,
+      "loss": 0.0974,
+      "step": 2429
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2786017954349518,
+      "learning_rate": 1.5647163120567378e-06,
+      "loss": 0.0973,
+      "step": 2430
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2868878245353699,
+      "learning_rate": 1.5632387706855793e-06,
+      "loss": 0.0991,
+      "step": 2431
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2835158407688141,
+      "learning_rate": 1.561761229314421e-06,
+      "loss": 0.0998,
+      "step": 2432
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2792942225933075,
+      "learning_rate": 1.5602836879432626e-06,
+      "loss": 0.1007,
+      "step": 2433
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.284585565328598,
+      "learning_rate": 1.5588061465721041e-06,
+      "loss": 0.1023,
+      "step": 2434
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.28843092918395996,
+      "learning_rate": 1.5573286052009457e-06,
+      "loss": 0.0992,
+      "step": 2435
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.31233447790145874,
+      "learning_rate": 1.5558510638297874e-06,
+      "loss": 0.0976,
+      "step": 2436
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2842386066913605,
+      "learning_rate": 1.554373522458629e-06,
+      "loss": 0.1029,
+      "step": 2437
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.29805105924606323,
+      "learning_rate": 1.5528959810874705e-06,
+      "loss": 0.0961,
+      "step": 2438
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.3055539131164551,
+      "learning_rate": 1.551418439716312e-06,
+      "loss": 0.1033,
+      "step": 2439
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2998804748058319,
+      "learning_rate": 1.549940898345154e-06,
+      "loss": 0.1007,
+      "step": 2440
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2983750104904175,
+      "learning_rate": 1.5484633569739955e-06,
+      "loss": 0.1019,
+      "step": 2441
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2895788848400116,
+      "learning_rate": 1.546985815602837e-06,
+      "loss": 0.1008,
+      "step": 2442
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2977020740509033,
+      "learning_rate": 1.5455082742316786e-06,
+      "loss": 0.1023,
+      "step": 2443
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.29555702209472656,
+      "learning_rate": 1.5440307328605203e-06,
+      "loss": 0.0987,
+      "step": 2444
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2802613079547882,
+      "learning_rate": 1.5425531914893618e-06,
+      "loss": 0.0963,
+      "step": 2445
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2977522611618042,
+      "learning_rate": 1.5410756501182034e-06,
+      "loss": 0.1009,
+      "step": 2446
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2892318367958069,
+      "learning_rate": 1.539598108747045e-06,
+      "loss": 0.1009,
+      "step": 2447
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2982611060142517,
+      "learning_rate": 1.5381205673758866e-06,
+      "loss": 0.0985,
+      "step": 2448
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2954762578010559,
+      "learning_rate": 1.5366430260047282e-06,
+      "loss": 0.1043,
+      "step": 2449
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.28837716579437256,
+      "learning_rate": 1.5351654846335697e-06,
+      "loss": 0.1012,
+      "step": 2450
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.306397944688797,
+      "learning_rate": 1.5336879432624117e-06,
+      "loss": 0.1006,
+      "step": 2451
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2779596447944641,
+      "learning_rate": 1.5322104018912532e-06,
+      "loss": 0.0984,
+      "step": 2452
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2985957860946655,
+      "learning_rate": 1.5307328605200947e-06,
+      "loss": 0.1085,
+      "step": 2453
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.3000498116016388,
+      "learning_rate": 1.5292553191489363e-06,
+      "loss": 0.1051,
+      "step": 2454
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.3055376410484314,
+      "learning_rate": 1.527777777777778e-06,
+      "loss": 0.1037,
+      "step": 2455
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2855037748813629,
+      "learning_rate": 1.5263002364066196e-06,
+      "loss": 0.1002,
+      "step": 2456
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2804585099220276,
+      "learning_rate": 1.524822695035461e-06,
+      "loss": 0.0996,
+      "step": 2457
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2828192114830017,
+      "learning_rate": 1.5233451536643026e-06,
+      "loss": 0.1016,
+      "step": 2458
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2757384181022644,
+      "learning_rate": 1.5218676122931444e-06,
+      "loss": 0.1018,
+      "step": 2459
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.28859207034111023,
+      "learning_rate": 1.520390070921986e-06,
+      "loss": 0.102,
+      "step": 2460
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.282955139875412,
+      "learning_rate": 1.5189125295508274e-06,
+      "loss": 0.0982,
+      "step": 2461
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2767374813556671,
+      "learning_rate": 1.517434988179669e-06,
+      "loss": 0.1003,
+      "step": 2462
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2955954372882843,
+      "learning_rate": 1.515957446808511e-06,
+      "loss": 0.1001,
+      "step": 2463
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.28480204939842224,
+      "learning_rate": 1.5144799054373525e-06,
+      "loss": 0.099,
+      "step": 2464
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.28144410252571106,
+      "learning_rate": 1.513002364066194e-06,
+      "loss": 0.0977,
+      "step": 2465
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.3027268946170807,
+      "learning_rate": 1.5115248226950355e-06,
+      "loss": 0.1035,
+      "step": 2466
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.29462218284606934,
+      "learning_rate": 1.5100472813238773e-06,
+      "loss": 0.1017,
+      "step": 2467
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2944556474685669,
+      "learning_rate": 1.5085697399527188e-06,
+      "loss": 0.1047,
+      "step": 2468
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2802707850933075,
+      "learning_rate": 1.5070921985815603e-06,
+      "loss": 0.0977,
+      "step": 2469
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2949898838996887,
+      "learning_rate": 1.5056146572104019e-06,
+      "loss": 0.099,
+      "step": 2470
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2897087335586548,
+      "learning_rate": 1.5041371158392436e-06,
+      "loss": 0.1024,
+      "step": 2471
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.2748096287250519,
+      "learning_rate": 1.5026595744680852e-06,
+      "loss": 0.096,
+      "step": 2472
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.2901870310306549,
+      "learning_rate": 1.5011820330969267e-06,
+      "loss": 0.101,
+      "step": 2473
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.2855484187602997,
+      "learning_rate": 1.4997044917257682e-06,
+      "loss": 0.1018,
+      "step": 2474
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.326937735080719,
+      "learning_rate": 1.4982269503546102e-06,
+      "loss": 0.1022,
+      "step": 2475
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.3026944398880005,
+      "learning_rate": 1.4967494089834517e-06,
+      "loss": 0.0997,
+      "step": 2476
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.3010585904121399,
+      "learning_rate": 1.4952718676122932e-06,
+      "loss": 0.1036,
+      "step": 2477
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.3869937062263489,
+      "learning_rate": 1.4937943262411348e-06,
+      "loss": 0.1053,
+      "step": 2478
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.28908172249794006,
+      "learning_rate": 1.4923167848699765e-06,
+      "loss": 0.099,
+      "step": 2479
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.2949136197566986,
+      "learning_rate": 1.490839243498818e-06,
+      "loss": 0.1021,
+      "step": 2480
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.29135873913764954,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 0.1022,
+      "step": 2481
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.296333372592926,
+      "learning_rate": 1.4878841607565013e-06,
+      "loss": 0.1023,
+      "step": 2482
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.28624671697616577,
+      "learning_rate": 1.486406619385343e-06,
+      "loss": 0.1008,
+      "step": 2483
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2879456877708435,
+      "learning_rate": 1.4849290780141846e-06,
+      "loss": 0.1039,
+      "step": 2484
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.28483426570892334,
+      "learning_rate": 1.4834515366430261e-06,
+      "loss": 0.0995,
+      "step": 2485
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.28305456042289734,
+      "learning_rate": 1.4819739952718679e-06,
+      "loss": 0.0986,
+      "step": 2486
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.301544725894928,
+      "learning_rate": 1.4804964539007094e-06,
+      "loss": 0.1003,
+      "step": 2487
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.29170653223991394,
+      "learning_rate": 1.479018912529551e-06,
+      "loss": 0.1027,
+      "step": 2488
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2962697148323059,
+      "learning_rate": 1.4775413711583925e-06,
+      "loss": 0.1008,
+      "step": 2489
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.28422003984451294,
+      "learning_rate": 1.4760638297872342e-06,
+      "loss": 0.1011,
+      "step": 2490
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.29257017374038696,
+      "learning_rate": 1.4745862884160758e-06,
+      "loss": 0.0983,
+      "step": 2491
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.29132208228111267,
+      "learning_rate": 1.4731087470449173e-06,
+      "loss": 0.0989,
+      "step": 2492
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2927258014678955,
+      "learning_rate": 1.4716312056737588e-06,
+      "loss": 0.0997,
+      "step": 2493
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2784199118614197,
+      "learning_rate": 1.4701536643026008e-06,
+      "loss": 0.1012,
+      "step": 2494
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.281755268573761,
+      "learning_rate": 1.4686761229314423e-06,
+      "loss": 0.1003,
+      "step": 2495
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.3046222925186157,
+      "learning_rate": 1.4671985815602839e-06,
+      "loss": 0.101,
+      "step": 2496
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2778962254524231,
+      "learning_rate": 1.4657210401891254e-06,
+      "loss": 0.1013,
+      "step": 2497
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.28655120730400085,
+      "learning_rate": 1.4642434988179671e-06,
+      "loss": 0.1012,
+      "step": 2498
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.28885889053344727,
+      "learning_rate": 1.4627659574468087e-06,
+      "loss": 0.102,
+      "step": 2499
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.27904340624809265,
+      "learning_rate": 1.4612884160756502e-06,
+      "loss": 0.0972,
+      "step": 2500
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2867487370967865,
+      "learning_rate": 1.4598108747044917e-06,
+      "loss": 0.099,
+      "step": 2501
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2888018488883972,
+      "learning_rate": 1.4583333333333335e-06,
+      "loss": 0.101,
+      "step": 2502
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2878059148788452,
+      "learning_rate": 1.456855791962175e-06,
+      "loss": 0.0975,
+      "step": 2503
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.28317585587501526,
+      "learning_rate": 1.4553782505910166e-06,
+      "loss": 0.0998,
+      "step": 2504
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2976023852825165,
+      "learning_rate": 1.453900709219858e-06,
+      "loss": 0.1062,
+      "step": 2505
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.28345340490341187,
+      "learning_rate": 1.4524231678487e-06,
+      "loss": 0.0997,
+      "step": 2506
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2787289321422577,
+      "learning_rate": 1.4509456264775416e-06,
+      "loss": 0.0975,
+      "step": 2507
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.3053015172481537,
+      "learning_rate": 1.4494680851063831e-06,
+      "loss": 0.1073,
+      "step": 2508
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.304909348487854,
+      "learning_rate": 1.4479905437352246e-06,
+      "loss": 0.1011,
+      "step": 2509
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.29397496581077576,
+      "learning_rate": 1.4465130023640664e-06,
+      "loss": 0.0997,
+      "step": 2510
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2873506546020508,
+      "learning_rate": 1.445035460992908e-06,
+      "loss": 0.1033,
+      "step": 2511
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.294401079416275,
+      "learning_rate": 1.4435579196217495e-06,
+      "loss": 0.0982,
+      "step": 2512
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.3069527745246887,
+      "learning_rate": 1.4420803782505912e-06,
+      "loss": 0.1048,
+      "step": 2513
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2794750928878784,
+      "learning_rate": 1.4406028368794327e-06,
+      "loss": 0.0962,
+      "step": 2514
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.29790565371513367,
+      "learning_rate": 1.4391252955082743e-06,
+      "loss": 0.0964,
+      "step": 2515
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2879948616027832,
+      "learning_rate": 1.4376477541371158e-06,
+      "loss": 0.1,
+      "step": 2516
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2881058156490326,
+      "learning_rate": 1.4361702127659578e-06,
+      "loss": 0.1041,
+      "step": 2517
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.305267333984375,
+      "learning_rate": 1.4346926713947993e-06,
+      "loss": 0.1008,
+      "step": 2518
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.29523733258247375,
+      "learning_rate": 1.4332151300236408e-06,
+      "loss": 0.0995,
+      "step": 2519
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.29262807965278625,
+      "learning_rate": 1.4317375886524824e-06,
+      "loss": 0.0985,
+      "step": 2520
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.27941974997520447,
+      "learning_rate": 1.4302600472813241e-06,
+      "loss": 0.0983,
+      "step": 2521
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.28749069571495056,
+      "learning_rate": 1.4287825059101656e-06,
+      "loss": 0.0999,
+      "step": 2522
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.27913329005241394,
+      "learning_rate": 1.4273049645390072e-06,
+      "loss": 0.0988,
+      "step": 2523
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.28391385078430176,
+      "learning_rate": 1.4258274231678487e-06,
+      "loss": 0.0989,
+      "step": 2524
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.29209619760513306,
+      "learning_rate": 1.4243498817966905e-06,
+      "loss": 0.0991,
+      "step": 2525
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2947562336921692,
+      "learning_rate": 1.422872340425532e-06,
+      "loss": 0.1024,
+      "step": 2526
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.27451324462890625,
+      "learning_rate": 1.4213947990543735e-06,
+      "loss": 0.1005,
+      "step": 2527
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2845952808856964,
+      "learning_rate": 1.419917257683215e-06,
+      "loss": 0.0995,
+      "step": 2528
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2810964286327362,
+      "learning_rate": 1.418439716312057e-06,
+      "loss": 0.0985,
+      "step": 2529
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.29670843482017517,
+      "learning_rate": 1.4169621749408985e-06,
+      "loss": 0.1002,
+      "step": 2530
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.28005945682525635,
+      "learning_rate": 1.41548463356974e-06,
+      "loss": 0.1031,
+      "step": 2531
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.29916125535964966,
+      "learning_rate": 1.4140070921985816e-06,
+      "loss": 0.0976,
+      "step": 2532
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2828112244606018,
+      "learning_rate": 1.4125295508274234e-06,
+      "loss": 0.1003,
+      "step": 2533
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2964456081390381,
+      "learning_rate": 1.4110520094562649e-06,
+      "loss": 0.1009,
+      "step": 2534
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2872307598590851,
+      "learning_rate": 1.4095744680851064e-06,
+      "loss": 0.1032,
+      "step": 2535
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2993559241294861,
+      "learning_rate": 1.408096926713948e-06,
+      "loss": 0.1043,
+      "step": 2536
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.3101213574409485,
+      "learning_rate": 1.4066193853427897e-06,
+      "loss": 0.0994,
+      "step": 2537
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2871704697608948,
+      "learning_rate": 1.4051418439716312e-06,
+      "loss": 0.098,
+      "step": 2538
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.27517789602279663,
+      "learning_rate": 1.4036643026004728e-06,
+      "loss": 0.0988,
+      "step": 2539
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2929779291152954,
+      "learning_rate": 1.4021867612293143e-06,
+      "loss": 0.1019,
+      "step": 2540
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2799249291419983,
+      "learning_rate": 1.4007092198581563e-06,
+      "loss": 0.0985,
+      "step": 2541
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.29473015666007996,
+      "learning_rate": 1.3992316784869978e-06,
+      "loss": 0.0995,
+      "step": 2542
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2734103798866272,
+      "learning_rate": 1.3977541371158393e-06,
+      "loss": 0.1003,
+      "step": 2543
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.29323965311050415,
+      "learning_rate": 1.3962765957446809e-06,
+      "loss": 0.0977,
+      "step": 2544
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.28401991724967957,
+      "learning_rate": 1.3947990543735226e-06,
+      "loss": 0.1009,
+      "step": 2545
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2964780628681183,
+      "learning_rate": 1.3933215130023641e-06,
+      "loss": 0.1009,
+      "step": 2546
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.28310421109199524,
+      "learning_rate": 1.3918439716312057e-06,
+      "loss": 0.1022,
+      "step": 2547
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.3037421703338623,
+      "learning_rate": 1.3903664302600476e-06,
+      "loss": 0.1056,
+      "step": 2548
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2831652760505676,
+      "learning_rate": 1.3888888888888892e-06,
+      "loss": 0.0982,
+      "step": 2549
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2927112281322479,
+      "learning_rate": 1.3874113475177307e-06,
+      "loss": 0.1038,
+      "step": 2550
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.29487597942352295,
+      "learning_rate": 1.3859338061465722e-06,
+      "loss": 0.1037,
+      "step": 2551
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.29078540205955505,
+      "learning_rate": 1.384456264775414e-06,
+      "loss": 0.1049,
+      "step": 2552
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2882029116153717,
+      "learning_rate": 1.3829787234042555e-06,
+      "loss": 0.0996,
+      "step": 2553
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.28395238518714905,
+      "learning_rate": 1.381501182033097e-06,
+      "loss": 0.0972,
+      "step": 2554
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2835884094238281,
+      "learning_rate": 1.3800236406619386e-06,
+      "loss": 0.0998,
+      "step": 2555
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2692258656024933,
+      "learning_rate": 1.3785460992907803e-06,
+      "loss": 0.0968,
+      "step": 2556
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2895731031894684,
+      "learning_rate": 1.3770685579196219e-06,
+      "loss": 0.0997,
+      "step": 2557
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.28379395604133606,
+      "learning_rate": 1.3755910165484634e-06,
+      "loss": 0.0977,
+      "step": 2558
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2734008729457855,
+      "learning_rate": 1.374113475177305e-06,
+      "loss": 0.0972,
+      "step": 2559
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.28540322184562683,
+      "learning_rate": 1.3726359338061469e-06,
+      "loss": 0.0991,
+      "step": 2560
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2855961322784424,
+      "learning_rate": 1.3711583924349884e-06,
+      "loss": 0.105,
+      "step": 2561
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.29528793692588806,
+      "learning_rate": 1.36968085106383e-06,
+      "loss": 0.1012,
+      "step": 2562
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2916938364505768,
+      "learning_rate": 1.3682033096926715e-06,
+      "loss": 0.1006,
+      "step": 2563
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2831311821937561,
+      "learning_rate": 1.3667257683215132e-06,
+      "loss": 0.1007,
+      "step": 2564
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.29988333582878113,
+      "learning_rate": 1.3652482269503548e-06,
+      "loss": 0.105,
+      "step": 2565
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2824275493621826,
+      "learning_rate": 1.3637706855791963e-06,
+      "loss": 0.0966,
+      "step": 2566
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2922384738922119,
+      "learning_rate": 1.3622931442080378e-06,
+      "loss": 0.1015,
+      "step": 2567
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2818516790866852,
+      "learning_rate": 1.3608156028368796e-06,
+      "loss": 0.1005,
+      "step": 2568
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2792305052280426,
+      "learning_rate": 1.3593380614657211e-06,
+      "loss": 0.1011,
+      "step": 2569
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2940114438533783,
+      "learning_rate": 1.3578605200945626e-06,
+      "loss": 0.1025,
+      "step": 2570
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2990158200263977,
+      "learning_rate": 1.3563829787234042e-06,
+      "loss": 0.0999,
+      "step": 2571
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.29137930274009705,
+      "learning_rate": 1.3549054373522461e-06,
+      "loss": 0.1016,
+      "step": 2572
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.31098029017448425,
+      "learning_rate": 1.3534278959810877e-06,
+      "loss": 0.1017,
+      "step": 2573
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2917681932449341,
+      "learning_rate": 1.3519503546099292e-06,
+      "loss": 0.0988,
+      "step": 2574
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.30096524953842163,
+      "learning_rate": 1.3504728132387707e-06,
+      "loss": 0.1021,
+      "step": 2575
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2981046140193939,
+      "learning_rate": 1.3489952718676125e-06,
+      "loss": 0.1027,
+      "step": 2576
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.29524290561676025,
+      "learning_rate": 1.347517730496454e-06,
+      "loss": 0.1007,
+      "step": 2577
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.28835827112197876,
+      "learning_rate": 1.3460401891252955e-06,
+      "loss": 0.1033,
+      "step": 2578
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2963818907737732,
+      "learning_rate": 1.3445626477541373e-06,
+      "loss": 0.1,
+      "step": 2579
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.3014226257801056,
+      "learning_rate": 1.3430851063829788e-06,
+      "loss": 0.1004,
+      "step": 2580
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.30180823802948,
+      "learning_rate": 1.3416075650118204e-06,
+      "loss": 0.102,
+      "step": 2581
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2911633849143982,
+      "learning_rate": 1.3401300236406619e-06,
+      "loss": 0.102,
+      "step": 2582
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.29165199398994446,
+      "learning_rate": 1.3386524822695039e-06,
+      "loss": 0.1032,
+      "step": 2583
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.28109994530677795,
+      "learning_rate": 1.3371749408983454e-06,
+      "loss": 0.0996,
+      "step": 2584
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.28780633211135864,
+      "learning_rate": 1.335697399527187e-06,
+      "loss": 0.0997,
+      "step": 2585
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2797524631023407,
+      "learning_rate": 1.3342198581560285e-06,
+      "loss": 0.0982,
+      "step": 2586
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.29814624786376953,
+      "learning_rate": 1.3327423167848702e-06,
+      "loss": 0.1012,
+      "step": 2587
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.3046199679374695,
+      "learning_rate": 1.3312647754137117e-06,
+      "loss": 0.1062,
+      "step": 2588
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.27812275290489197,
+      "learning_rate": 1.3297872340425533e-06,
+      "loss": 0.0995,
+      "step": 2589
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2765676975250244,
+      "learning_rate": 1.3283096926713948e-06,
+      "loss": 0.0991,
+      "step": 2590
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.27323317527770996,
+      "learning_rate": 1.3268321513002365e-06,
+      "loss": 0.095,
+      "step": 2591
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2854834496974945,
+      "learning_rate": 1.325354609929078e-06,
+      "loss": 0.1008,
+      "step": 2592
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.30844664573669434,
+      "learning_rate": 1.3238770685579196e-06,
+      "loss": 0.1013,
+      "step": 2593
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.28207823634147644,
+      "learning_rate": 1.3223995271867611e-06,
+      "loss": 0.0982,
+      "step": 2594
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.29778367280960083,
+      "learning_rate": 1.320921985815603e-06,
+      "loss": 0.1031,
+      "step": 2595
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.3009220361709595,
+      "learning_rate": 1.3194444444444446e-06,
+      "loss": 0.1029,
+      "step": 2596
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.28126275539398193,
+      "learning_rate": 1.3179669030732862e-06,
+      "loss": 0.1004,
+      "step": 2597
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.3123842179775238,
+      "learning_rate": 1.3164893617021277e-06,
+      "loss": 0.1003,
+      "step": 2598
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.28384894132614136,
+      "learning_rate": 1.3150118203309694e-06,
+      "loss": 0.1008,
+      "step": 2599
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.28906840085983276,
+      "learning_rate": 1.313534278959811e-06,
+      "loss": 0.1019,
+      "step": 2600
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2816874384880066,
+      "learning_rate": 1.3120567375886525e-06,
+      "loss": 0.0987,
+      "step": 2601
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.29358232021331787,
+      "learning_rate": 1.310579196217494e-06,
+      "loss": 0.0985,
+      "step": 2602
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.31373322010040283,
+      "learning_rate": 1.3091016548463358e-06,
+      "loss": 0.0994,
+      "step": 2603
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.29676130414009094,
+      "learning_rate": 1.3076241134751773e-06,
+      "loss": 0.1006,
+      "step": 2604
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2956346869468689,
+      "learning_rate": 1.3061465721040189e-06,
+      "loss": 0.1011,
+      "step": 2605
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.29391372203826904,
+      "learning_rate": 1.3046690307328604e-06,
+      "loss": 0.1038,
+      "step": 2606
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.29173383116722107,
+      "learning_rate": 1.3031914893617024e-06,
+      "loss": 0.0995,
+      "step": 2607
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.28803884983062744,
+      "learning_rate": 1.3017139479905439e-06,
+      "loss": 0.1019,
+      "step": 2608
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.27921730279922485,
+      "learning_rate": 1.3002364066193854e-06,
+      "loss": 0.0943,
+      "step": 2609
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.27652543783187866,
+      "learning_rate": 1.298758865248227e-06,
+      "loss": 0.1005,
+      "step": 2610
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2916536331176758,
+      "learning_rate": 1.2972813238770687e-06,
+      "loss": 0.0995,
+      "step": 2611
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.28120341897010803,
+      "learning_rate": 1.2958037825059102e-06,
+      "loss": 0.1008,
+      "step": 2612
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2926959693431854,
+      "learning_rate": 1.2943262411347518e-06,
+      "loss": 0.0973,
+      "step": 2613
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.28686630725860596,
+      "learning_rate": 1.2928486997635937e-06,
+      "loss": 0.0989,
+      "step": 2614
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2951143682003021,
+      "learning_rate": 1.2913711583924353e-06,
+      "loss": 0.1042,
+      "step": 2615
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2878260016441345,
+      "learning_rate": 1.2898936170212768e-06,
+      "loss": 0.0975,
+      "step": 2616
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.27820301055908203,
+      "learning_rate": 1.2884160756501183e-06,
+      "loss": 0.0982,
+      "step": 2617
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2839721441268921,
+      "learning_rate": 1.28693853427896e-06,
+      "loss": 0.0982,
+      "step": 2618
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.29253509640693665,
+      "learning_rate": 1.2854609929078016e-06,
+      "loss": 0.1031,
+      "step": 2619
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.29048773646354675,
+      "learning_rate": 1.2839834515366431e-06,
+      "loss": 0.1018,
+      "step": 2620
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.3143782913684845,
+      "learning_rate": 1.2825059101654847e-06,
+      "loss": 0.1028,
+      "step": 2621
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.29310721158981323,
+      "learning_rate": 1.2810283687943264e-06,
+      "loss": 0.1047,
+      "step": 2622
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.28151217103004456,
+      "learning_rate": 1.279550827423168e-06,
+      "loss": 0.1002,
+      "step": 2623
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2948727607727051,
+      "learning_rate": 1.2780732860520095e-06,
+      "loss": 0.1022,
+      "step": 2624
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2815762460231781,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 0.0969,
+      "step": 2625
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2971389591693878,
+      "learning_rate": 1.275118203309693e-06,
+      "loss": 0.1031,
+      "step": 2626
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.29214245080947876,
+      "learning_rate": 1.2736406619385345e-06,
+      "loss": 0.1002,
+      "step": 2627
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2950294613838196,
+      "learning_rate": 1.272163120567376e-06,
+      "loss": 0.1017,
+      "step": 2628
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.3087725043296814,
+      "learning_rate": 1.2706855791962176e-06,
+      "loss": 0.0996,
+      "step": 2629
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.29620638489723206,
+      "learning_rate": 1.2692080378250593e-06,
+      "loss": 0.1015,
+      "step": 2630
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2882344126701355,
+      "learning_rate": 1.2677304964539009e-06,
+      "loss": 0.1001,
+      "step": 2631
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2925781011581421,
+      "learning_rate": 1.2662529550827424e-06,
+      "loss": 0.1012,
+      "step": 2632
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2959561049938202,
+      "learning_rate": 1.264775413711584e-06,
+      "loss": 0.0969,
+      "step": 2633
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2817199230194092,
+      "learning_rate": 1.2632978723404257e-06,
+      "loss": 0.0956,
+      "step": 2634
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.31056445837020874,
+      "learning_rate": 1.2618203309692672e-06,
+      "loss": 0.1019,
+      "step": 2635
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.30532965064048767,
+      "learning_rate": 1.2603427895981087e-06,
+      "loss": 0.1011,
+      "step": 2636
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2935227155685425,
+      "learning_rate": 1.2588652482269503e-06,
+      "loss": 0.1015,
+      "step": 2637
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.28477272391319275,
+      "learning_rate": 1.2573877068557922e-06,
+      "loss": 0.1019,
+      "step": 2638
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2968229055404663,
+      "learning_rate": 1.2559101654846338e-06,
+      "loss": 0.104,
+      "step": 2639
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.29241105914115906,
+      "learning_rate": 1.2544326241134753e-06,
+      "loss": 0.0995,
+      "step": 2640
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.3011491894721985,
+      "learning_rate": 1.2529550827423168e-06,
+      "loss": 0.1037,
+      "step": 2641
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.28209346532821655,
+      "learning_rate": 1.2514775413711586e-06,
+      "loss": 0.1004,
+      "step": 2642
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.3160993456840515,
+      "learning_rate": 1.25e-06,
+      "loss": 0.1054,
+      "step": 2643
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.28129342198371887,
+      "learning_rate": 1.2485224586288418e-06,
+      "loss": 0.098,
+      "step": 2644
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2937127351760864,
+      "learning_rate": 1.2470449172576834e-06,
+      "loss": 0.0982,
+      "step": 2645
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.2814466655254364,
+      "learning_rate": 1.245567375886525e-06,
+      "loss": 0.097,
+      "step": 2646
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.31185996532440186,
+      "learning_rate": 1.2440898345153664e-06,
+      "loss": 0.1003,
+      "step": 2647
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.30120548605918884,
+      "learning_rate": 1.2426122931442082e-06,
+      "loss": 0.0992,
+      "step": 2648
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3021382987499237,
+      "learning_rate": 1.2411347517730497e-06,
+      "loss": 0.0991,
+      "step": 2649
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.29702848196029663,
+      "learning_rate": 1.2396572104018915e-06,
+      "loss": 0.1017,
+      "step": 2650
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.2970198094844818,
+      "learning_rate": 1.238179669030733e-06,
+      "loss": 0.1003,
+      "step": 2651
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.2858641743659973,
+      "learning_rate": 1.2367021276595745e-06,
+      "loss": 0.097,
+      "step": 2652
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3146287202835083,
+      "learning_rate": 1.235224586288416e-06,
+      "loss": 0.1034,
+      "step": 2653
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.30120575428009033,
+      "learning_rate": 1.2337470449172578e-06,
+      "loss": 0.1006,
+      "step": 2654
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.29078367352485657,
+      "learning_rate": 1.2322695035460994e-06,
+      "loss": 0.107,
+      "step": 2655
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3000693917274475,
+      "learning_rate": 1.230791962174941e-06,
+      "loss": 0.1031,
+      "step": 2656
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.2897482216358185,
+      "learning_rate": 1.2293144208037826e-06,
+      "loss": 0.0978,
+      "step": 2657
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.3052290976047516,
+      "learning_rate": 1.2278368794326242e-06,
+      "loss": 0.1025,
+      "step": 2658
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2962765693664551,
+      "learning_rate": 1.2263593380614657e-06,
+      "loss": 0.0993,
+      "step": 2659
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2958114743232727,
+      "learning_rate": 1.2248817966903074e-06,
+      "loss": 0.1015,
+      "step": 2660
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.3029790222644806,
+      "learning_rate": 1.223404255319149e-06,
+      "loss": 0.0985,
+      "step": 2661
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2966248691082001,
+      "learning_rate": 1.2219267139479907e-06,
+      "loss": 0.1028,
+      "step": 2662
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2948656976222992,
+      "learning_rate": 1.2204491725768323e-06,
+      "loss": 0.1014,
+      "step": 2663
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.3018767535686493,
+      "learning_rate": 1.2189716312056738e-06,
+      "loss": 0.0982,
+      "step": 2664
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.282370924949646,
+      "learning_rate": 1.2174940898345153e-06,
+      "loss": 0.1008,
+      "step": 2665
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.3088902235031128,
+      "learning_rate": 1.216016548463357e-06,
+      "loss": 0.1024,
+      "step": 2666
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.28619828820228577,
+      "learning_rate": 1.2145390070921986e-06,
+      "loss": 0.0996,
+      "step": 2667
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2833898365497589,
+      "learning_rate": 1.2130614657210403e-06,
+      "loss": 0.0974,
+      "step": 2668
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.289203017950058,
+      "learning_rate": 1.2115839243498819e-06,
+      "loss": 0.1,
+      "step": 2669
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2880306541919708,
+      "learning_rate": 1.2101063829787234e-06,
+      "loss": 0.0995,
+      "step": 2670
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2951520085334778,
+      "learning_rate": 1.208628841607565e-06,
+      "loss": 0.101,
+      "step": 2671
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2928524911403656,
+      "learning_rate": 1.2071513002364067e-06,
+      "loss": 0.1018,
+      "step": 2672
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.28666532039642334,
+      "learning_rate": 1.2056737588652482e-06,
+      "loss": 0.1007,
+      "step": 2673
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.29114654660224915,
+      "learning_rate": 1.20419621749409e-06,
+      "loss": 0.1039,
+      "step": 2674
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.28709572553634644,
+      "learning_rate": 1.2027186761229315e-06,
+      "loss": 0.0978,
+      "step": 2675
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.28625017404556274,
+      "learning_rate": 1.2012411347517733e-06,
+      "loss": 0.1023,
+      "step": 2676
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2860890030860901,
+      "learning_rate": 1.1997635933806148e-06,
+      "loss": 0.1005,
+      "step": 2677
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2895687222480774,
+      "learning_rate": 1.1982860520094563e-06,
+      "loss": 0.0969,
+      "step": 2678
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2909514605998993,
+      "learning_rate": 1.196808510638298e-06,
+      "loss": 0.099,
+      "step": 2679
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.27398183941841125,
+      "learning_rate": 1.1953309692671396e-06,
+      "loss": 0.1001,
+      "step": 2680
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.294829785823822,
+      "learning_rate": 1.1938534278959813e-06,
+      "loss": 0.0997,
+      "step": 2681
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.30538126826286316,
+      "learning_rate": 1.1923758865248229e-06,
+      "loss": 0.1012,
+      "step": 2682
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2880515456199646,
+      "learning_rate": 1.1908983451536644e-06,
+      "loss": 0.1014,
+      "step": 2683
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.29294338822364807,
+      "learning_rate": 1.189420803782506e-06,
+      "loss": 0.098,
+      "step": 2684
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2915271520614624,
+      "learning_rate": 1.1879432624113477e-06,
+      "loss": 0.1005,
+      "step": 2685
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.3017413914203644,
+      "learning_rate": 1.1864657210401892e-06,
+      "loss": 0.1018,
+      "step": 2686
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.29048147797584534,
+      "learning_rate": 1.184988179669031e-06,
+      "loss": 0.1018,
+      "step": 2687
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.293397456407547,
+      "learning_rate": 1.1835106382978725e-06,
+      "loss": 0.1026,
+      "step": 2688
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.27730804681777954,
+      "learning_rate": 1.182033096926714e-06,
+      "loss": 0.0962,
+      "step": 2689
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.28659364581108093,
+      "learning_rate": 1.1805555555555556e-06,
+      "loss": 0.0966,
+      "step": 2690
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.29424241185188293,
+      "learning_rate": 1.1790780141843973e-06,
+      "loss": 0.1004,
+      "step": 2691
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.316184937953949,
+      "learning_rate": 1.1776004728132388e-06,
+      "loss": 0.1071,
+      "step": 2692
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.29108530282974243,
+      "learning_rate": 1.1761229314420806e-06,
+      "loss": 0.1022,
+      "step": 2693
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.30403319001197815,
+      "learning_rate": 1.1746453900709221e-06,
+      "loss": 0.1069,
+      "step": 2694
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2905270755290985,
+      "learning_rate": 1.1731678486997637e-06,
+      "loss": 0.0991,
+      "step": 2695
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.28768104314804077,
+      "learning_rate": 1.1716903073286052e-06,
+      "loss": 0.0998,
+      "step": 2696
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.301690936088562,
+      "learning_rate": 1.170212765957447e-06,
+      "loss": 0.103,
+      "step": 2697
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.29621148109436035,
+      "learning_rate": 1.1687352245862885e-06,
+      "loss": 0.1033,
+      "step": 2698
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.29729101061820984,
+      "learning_rate": 1.1672576832151302e-06,
+      "loss": 0.1005,
+      "step": 2699
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.27758485078811646,
+      "learning_rate": 1.1657801418439718e-06,
+      "loss": 0.0977,
+      "step": 2700
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2879749536514282,
+      "learning_rate": 1.1643026004728133e-06,
+      "loss": 0.0988,
+      "step": 2701
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2713543772697449,
+      "learning_rate": 1.1628250591016548e-06,
+      "loss": 0.0998,
+      "step": 2702
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2882237434387207,
+      "learning_rate": 1.1613475177304966e-06,
+      "loss": 0.0997,
+      "step": 2703
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.2874648869037628,
+      "learning_rate": 1.159869976359338e-06,
+      "loss": 0.1031,
+      "step": 2704
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.30636066198349,
+      "learning_rate": 1.1583924349881798e-06,
+      "loss": 0.1045,
+      "step": 2705
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.2844293415546417,
+      "learning_rate": 1.1569148936170214e-06,
+      "loss": 0.1008,
+      "step": 2706
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.30712318420410156,
+      "learning_rate": 1.155437352245863e-06,
+      "loss": 0.0987,
+      "step": 2707
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.29389747977256775,
+      "learning_rate": 1.1539598108747044e-06,
+      "loss": 0.0989,
+      "step": 2708
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.285461962223053,
+      "learning_rate": 1.1524822695035462e-06,
+      "loss": 0.0988,
+      "step": 2709
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.3036527931690216,
+      "learning_rate": 1.151004728132388e-06,
+      "loss": 0.1013,
+      "step": 2710
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.2744309604167938,
+      "learning_rate": 1.1495271867612295e-06,
+      "loss": 0.0992,
+      "step": 2711
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.28376033902168274,
+      "learning_rate": 1.148049645390071e-06,
+      "loss": 0.0997,
+      "step": 2712
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.2899521589279175,
+      "learning_rate": 1.1465721040189125e-06,
+      "loss": 0.1052,
+      "step": 2713
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.297088086605072,
+      "learning_rate": 1.1450945626477543e-06,
+      "loss": 0.1027,
+      "step": 2714
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.289885014295578,
+      "learning_rate": 1.1436170212765958e-06,
+      "loss": 0.1018,
+      "step": 2715
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3001596927642822,
+      "learning_rate": 1.1421394799054376e-06,
+      "loss": 0.1005,
+      "step": 2716
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2943061888217926,
+      "learning_rate": 1.140661938534279e-06,
+      "loss": 0.1042,
+      "step": 2717
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.30126798152923584,
+      "learning_rate": 1.1391843971631206e-06,
+      "loss": 0.0999,
+      "step": 2718
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2938699722290039,
+      "learning_rate": 1.1377068557919622e-06,
+      "loss": 0.1043,
+      "step": 2719
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.28688114881515503,
+      "learning_rate": 1.136229314420804e-06,
+      "loss": 0.0998,
+      "step": 2720
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2972783148288727,
+      "learning_rate": 1.1347517730496454e-06,
+      "loss": 0.1035,
+      "step": 2721
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2821040153503418,
+      "learning_rate": 1.1332742316784872e-06,
+      "loss": 0.096,
+      "step": 2722
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.28433874249458313,
+      "learning_rate": 1.1317966903073287e-06,
+      "loss": 0.1016,
+      "step": 2723
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2975385785102844,
+      "learning_rate": 1.1303191489361703e-06,
+      "loss": 0.1008,
+      "step": 2724
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.27922743558883667,
+      "learning_rate": 1.1288416075650118e-06,
+      "loss": 0.1,
+      "step": 2725
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2874366044998169,
+      "learning_rate": 1.1273640661938535e-06,
+      "loss": 0.1021,
+      "step": 2726
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3065992295742035,
+      "learning_rate": 1.125886524822695e-06,
+      "loss": 0.1047,
+      "step": 2727
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.293060839176178,
+      "learning_rate": 1.1244089834515368e-06,
+      "loss": 0.0974,
+      "step": 2728
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.289791464805603,
+      "learning_rate": 1.1229314420803783e-06,
+      "loss": 0.1021,
+      "step": 2729
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2886219918727875,
+      "learning_rate": 1.1214539007092199e-06,
+      "loss": 0.1026,
+      "step": 2730
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.29779472947120667,
+      "learning_rate": 1.1199763593380614e-06,
+      "loss": 0.1028,
+      "step": 2731
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.30056414008140564,
+      "learning_rate": 1.1184988179669032e-06,
+      "loss": 0.1015,
+      "step": 2732
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.28907081484794617,
+      "learning_rate": 1.1170212765957447e-06,
+      "loss": 0.1001,
+      "step": 2733
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.29127663373947144,
+      "learning_rate": 1.1155437352245864e-06,
+      "loss": 0.0987,
+      "step": 2734
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.3074101507663727,
+      "learning_rate": 1.114066193853428e-06,
+      "loss": 0.0992,
+      "step": 2735
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2988634407520294,
+      "learning_rate": 1.1125886524822695e-06,
+      "loss": 0.1051,
+      "step": 2736
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2810848355293274,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.0981,
+      "step": 2737
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.3018767535686493,
+      "learning_rate": 1.1096335697399528e-06,
+      "loss": 0.1049,
+      "step": 2738
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.28153783082962036,
+      "learning_rate": 1.1081560283687943e-06,
+      "loss": 0.0988,
+      "step": 2739
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2854481637477875,
+      "learning_rate": 1.106678486997636e-06,
+      "loss": 0.1025,
+      "step": 2740
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.29290053248405457,
+      "learning_rate": 1.1052009456264776e-06,
+      "loss": 0.1033,
+      "step": 2741
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.29502326250076294,
+      "learning_rate": 1.1037234042553193e-06,
+      "loss": 0.1018,
+      "step": 2742
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.27135977149009705,
+      "learning_rate": 1.1022458628841609e-06,
+      "loss": 0.0977,
+      "step": 2743
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.28091344237327576,
+      "learning_rate": 1.1007683215130024e-06,
+      "loss": 0.099,
+      "step": 2744
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2980334758758545,
+      "learning_rate": 1.0992907801418442e-06,
+      "loss": 0.1017,
+      "step": 2745
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.3068128526210785,
+      "learning_rate": 1.0978132387706857e-06,
+      "loss": 0.1015,
+      "step": 2746
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2793194651603699,
+      "learning_rate": 1.0963356973995274e-06,
+      "loss": 0.1023,
+      "step": 2747
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.30465635657310486,
+      "learning_rate": 1.094858156028369e-06,
+      "loss": 0.1011,
+      "step": 2748
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.28814515471458435,
+      "learning_rate": 1.0933806146572105e-06,
+      "loss": 0.1022,
+      "step": 2749
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.29599475860595703,
+      "learning_rate": 1.091903073286052e-06,
+      "loss": 0.1017,
+      "step": 2750
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2938927412033081,
+      "learning_rate": 1.0904255319148938e-06,
+      "loss": 0.0992,
+      "step": 2751
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.3008365035057068,
+      "learning_rate": 1.0889479905437353e-06,
+      "loss": 0.1023,
+      "step": 2752
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2860044836997986,
+      "learning_rate": 1.087470449172577e-06,
+      "loss": 0.0985,
+      "step": 2753
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.29876452684402466,
+      "learning_rate": 1.0859929078014186e-06,
+      "loss": 0.1051,
+      "step": 2754
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2892434298992157,
+      "learning_rate": 1.0845153664302601e-06,
+      "loss": 0.1012,
+      "step": 2755
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.3051392138004303,
+      "learning_rate": 1.0830378250591017e-06,
+      "loss": 0.1027,
+      "step": 2756
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.29516929388046265,
+      "learning_rate": 1.0815602836879434e-06,
+      "loss": 0.1016,
+      "step": 2757
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.3024262487888336,
+      "learning_rate": 1.080082742316785e-06,
+      "loss": 0.1034,
+      "step": 2758
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2915303409099579,
+      "learning_rate": 1.0786052009456267e-06,
+      "loss": 0.1015,
+      "step": 2759
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2958263158798218,
+      "learning_rate": 1.0771276595744682e-06,
+      "loss": 0.1022,
+      "step": 2760
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.31569913029670715,
+      "learning_rate": 1.0756501182033097e-06,
+      "loss": 0.0991,
+      "step": 2761
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.29078567028045654,
+      "learning_rate": 1.0741725768321513e-06,
+      "loss": 0.1006,
+      "step": 2762
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.29640620946884155,
+      "learning_rate": 1.072695035460993e-06,
+      "loss": 0.1001,
+      "step": 2763
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.32687878608703613,
+      "learning_rate": 1.0712174940898346e-06,
+      "loss": 0.0999,
+      "step": 2764
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2732146978378296,
+      "learning_rate": 1.0697399527186763e-06,
+      "loss": 0.097,
+      "step": 2765
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2881612777709961,
+      "learning_rate": 1.0682624113475178e-06,
+      "loss": 0.0983,
+      "step": 2766
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.3095841109752655,
+      "learning_rate": 1.0667848699763594e-06,
+      "loss": 0.1038,
+      "step": 2767
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.3025471270084381,
+      "learning_rate": 1.065307328605201e-06,
+      "loss": 0.1019,
+      "step": 2768
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2799895107746124,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 0.0964,
+      "step": 2769
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.29736506938934326,
+      "learning_rate": 1.0623522458628842e-06,
+      "loss": 0.1029,
+      "step": 2770
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2896970510482788,
+      "learning_rate": 1.060874704491726e-06,
+      "loss": 0.1002,
+      "step": 2771
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.31091010570526123,
+      "learning_rate": 1.0593971631205675e-06,
+      "loss": 0.1016,
+      "step": 2772
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.31489723920822144,
+      "learning_rate": 1.057919621749409e-06,
+      "loss": 0.1014,
+      "step": 2773
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.3119453489780426,
+      "learning_rate": 1.0564420803782507e-06,
+      "loss": 0.1032,
+      "step": 2774
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2809605300426483,
+      "learning_rate": 1.0549645390070923e-06,
+      "loss": 0.1,
+      "step": 2775
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.308035671710968,
+      "learning_rate": 1.053486997635934e-06,
+      "loss": 0.1011,
+      "step": 2776
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.28076261281967163,
+      "learning_rate": 1.0520094562647756e-06,
+      "loss": 0.0967,
+      "step": 2777
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.283022403717041,
+      "learning_rate": 1.050531914893617e-06,
+      "loss": 0.0986,
+      "step": 2778
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2869526147842407,
+      "learning_rate": 1.0490543735224586e-06,
+      "loss": 0.1014,
+      "step": 2779
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.28648924827575684,
+      "learning_rate": 1.0475768321513004e-06,
+      "loss": 0.1029,
+      "step": 2780
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2810591161251068,
+      "learning_rate": 1.046099290780142e-06,
+      "loss": 0.0989,
+      "step": 2781
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2824496626853943,
+      "learning_rate": 1.0446217494089836e-06,
+      "loss": 0.099,
+      "step": 2782
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.29349786043167114,
+      "learning_rate": 1.0431442080378252e-06,
+      "loss": 0.0982,
+      "step": 2783
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.27743223309516907,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 0.0963,
+      "step": 2784
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.3050210177898407,
+      "learning_rate": 1.0401891252955083e-06,
+      "loss": 0.1008,
+      "step": 2785
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2933866083621979,
+      "learning_rate": 1.03871158392435e-06,
+      "loss": 0.1037,
+      "step": 2786
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.28359368443489075,
+      "learning_rate": 1.0372340425531915e-06,
+      "loss": 0.0993,
+      "step": 2787
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3085693120956421,
+      "learning_rate": 1.0357565011820333e-06,
+      "loss": 0.1021,
+      "step": 2788
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2862800061702728,
+      "learning_rate": 1.0342789598108748e-06,
+      "loss": 0.1007,
+      "step": 2789
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2946535646915436,
+      "learning_rate": 1.0328014184397163e-06,
+      "loss": 0.1033,
+      "step": 2790
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2901286780834198,
+      "learning_rate": 1.0313238770685579e-06,
+      "loss": 0.0994,
+      "step": 2791
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.29477375745773315,
+      "learning_rate": 1.0298463356973996e-06,
+      "loss": 0.1026,
+      "step": 2792
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2964722812175751,
+      "learning_rate": 1.0283687943262412e-06,
+      "loss": 0.104,
+      "step": 2793
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2761469781398773,
+      "learning_rate": 1.026891252955083e-06,
+      "loss": 0.0953,
+      "step": 2794
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.290297269821167,
+      "learning_rate": 1.0254137115839244e-06,
+      "loss": 0.1008,
+      "step": 2795
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3044756054878235,
+      "learning_rate": 1.023936170212766e-06,
+      "loss": 0.1062,
+      "step": 2796
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2901773750782013,
+      "learning_rate": 1.0224586288416075e-06,
+      "loss": 0.1008,
+      "step": 2797
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.291571706533432,
+      "learning_rate": 1.0209810874704492e-06,
+      "loss": 0.1038,
+      "step": 2798
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.3010854423046112,
+      "learning_rate": 1.0195035460992908e-06,
+      "loss": 0.0977,
+      "step": 2799
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.30066177248954773,
+      "learning_rate": 1.0180260047281325e-06,
+      "loss": 0.0998,
+      "step": 2800
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.3019755780696869,
+      "learning_rate": 1.016548463356974e-06,
+      "loss": 0.1011,
+      "step": 2801
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2892332375049591,
+      "learning_rate": 1.0150709219858156e-06,
+      "loss": 0.1006,
+      "step": 2802
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.30072087049484253,
+      "learning_rate": 1.0135933806146571e-06,
+      "loss": 0.1032,
+      "step": 2803
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2851409912109375,
+      "learning_rate": 1.0121158392434989e-06,
+      "loss": 0.1005,
+      "step": 2804
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.30061519145965576,
+      "learning_rate": 1.0106382978723404e-06,
+      "loss": 0.0987,
+      "step": 2805
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.29518571496009827,
+      "learning_rate": 1.0091607565011821e-06,
+      "loss": 0.103,
+      "step": 2806
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.3077875077724457,
+      "learning_rate": 1.007683215130024e-06,
+      "loss": 0.1034,
+      "step": 2807
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2864879369735718,
+      "learning_rate": 1.0062056737588654e-06,
+      "loss": 0.1017,
+      "step": 2808
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2855875492095947,
+      "learning_rate": 1.004728132387707e-06,
+      "loss": 0.094,
+      "step": 2809
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2906632125377655,
+      "learning_rate": 1.0032505910165485e-06,
+      "loss": 0.0993,
+      "step": 2810
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.29428011178970337,
+      "learning_rate": 1.0017730496453902e-06,
+      "loss": 0.0992,
+      "step": 2811
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2958046495914459,
+      "learning_rate": 1.0002955082742318e-06,
+      "loss": 0.0991,
+      "step": 2812
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.298820436000824,
+      "learning_rate": 9.988179669030735e-07,
+      "loss": 0.1041,
+      "step": 2813
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.29265356063842773,
+      "learning_rate": 9.97340425531915e-07,
+      "loss": 0.0975,
+      "step": 2814
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.28962936997413635,
+      "learning_rate": 9.958628841607566e-07,
+      "loss": 0.0995,
+      "step": 2815
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.29972410202026367,
+      "learning_rate": 9.943853427895981e-07,
+      "loss": 0.1025,
+      "step": 2816
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.26854565739631653,
+      "learning_rate": 9.929078014184399e-07,
+      "loss": 0.0948,
+      "step": 2817
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2878590524196625,
+      "learning_rate": 9.914302600472814e-07,
+      "loss": 0.1017,
+      "step": 2818
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.28923988342285156,
+      "learning_rate": 9.899527186761231e-07,
+      "loss": 0.1011,
+      "step": 2819
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.3198642432689667,
+      "learning_rate": 9.884751773049647e-07,
+      "loss": 0.1026,
+      "step": 2820
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2762257158756256,
+      "learning_rate": 9.869976359338062e-07,
+      "loss": 0.0964,
+      "step": 2821
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.28085795044898987,
+      "learning_rate": 9.855200945626477e-07,
+      "loss": 0.0982,
+      "step": 2822
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2961018681526184,
+      "learning_rate": 9.840425531914895e-07,
+      "loss": 0.1007,
+      "step": 2823
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.28563687205314636,
+      "learning_rate": 9.82565011820331e-07,
+      "loss": 0.0995,
+      "step": 2824
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3008669316768646,
+      "learning_rate": 9.810874704491728e-07,
+      "loss": 0.0974,
+      "step": 2825
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.30179286003112793,
+      "learning_rate": 9.796099290780143e-07,
+      "loss": 0.099,
+      "step": 2826
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.28248944878578186,
+      "learning_rate": 9.781323877068558e-07,
+      "loss": 0.0996,
+      "step": 2827
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.31358087062835693,
+      "learning_rate": 9.766548463356974e-07,
+      "loss": 0.0989,
+      "step": 2828
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2905551493167877,
+      "learning_rate": 9.751773049645391e-07,
+      "loss": 0.1014,
+      "step": 2829
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3012705147266388,
+      "learning_rate": 9.736997635933807e-07,
+      "loss": 0.1027,
+      "step": 2830
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2929949164390564,
+      "learning_rate": 9.722222222222224e-07,
+      "loss": 0.0991,
+      "step": 2831
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.38551101088523865,
+      "learning_rate": 9.70744680851064e-07,
+      "loss": 0.1005,
+      "step": 2832
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2964586317539215,
+      "learning_rate": 9.692671394799055e-07,
+      "loss": 0.106,
+      "step": 2833
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.28564804792404175,
+      "learning_rate": 9.67789598108747e-07,
+      "loss": 0.1011,
+      "step": 2834
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.28093159198760986,
+      "learning_rate": 9.663120567375887e-07,
+      "loss": 0.0983,
+      "step": 2835
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.299272358417511,
+      "learning_rate": 9.648345153664303e-07,
+      "loss": 0.1031,
+      "step": 2836
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.300737589597702,
+      "learning_rate": 9.63356973995272e-07,
+      "loss": 0.103,
+      "step": 2837
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2836374342441559,
+      "learning_rate": 9.618794326241136e-07,
+      "loss": 0.0959,
+      "step": 2838
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2918663024902344,
+      "learning_rate": 9.60401891252955e-07,
+      "loss": 0.1003,
+      "step": 2839
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2785128653049469,
+      "learning_rate": 9.589243498817968e-07,
+      "loss": 0.0987,
+      "step": 2840
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.28116706013679504,
+      "learning_rate": 9.574468085106384e-07,
+      "loss": 0.0962,
+      "step": 2841
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.28128284215927124,
+      "learning_rate": 9.559692671394801e-07,
+      "loss": 0.1007,
+      "step": 2842
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.30037355422973633,
+      "learning_rate": 9.544917257683216e-07,
+      "loss": 0.0995,
+      "step": 2843
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2907056510448456,
+      "learning_rate": 9.530141843971633e-07,
+      "loss": 0.1004,
+      "step": 2844
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.28128039836883545,
+      "learning_rate": 9.515366430260048e-07,
+      "loss": 0.0968,
+      "step": 2845
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3154534697532654,
+      "learning_rate": 9.500591016548465e-07,
+      "loss": 0.1006,
+      "step": 2846
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.353351354598999,
+      "learning_rate": 9.48581560283688e-07,
+      "loss": 0.0981,
+      "step": 2847
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3033483624458313,
+      "learning_rate": 9.471040189125296e-07,
+      "loss": 0.1072,
+      "step": 2848
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2815799415111542,
+      "learning_rate": 9.456264775413712e-07,
+      "loss": 0.0972,
+      "step": 2849
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2860986888408661,
+      "learning_rate": 9.441489361702129e-07,
+      "loss": 0.102,
+      "step": 2850
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.29191893339157104,
+      "learning_rate": 9.426713947990544e-07,
+      "loss": 0.0993,
+      "step": 2851
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.28981664776802063,
+      "learning_rate": 9.411938534278961e-07,
+      "loss": 0.1006,
+      "step": 2852
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.29072147607803345,
+      "learning_rate": 9.397163120567376e-07,
+      "loss": 0.0977,
+      "step": 2853
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.30578887462615967,
+      "learning_rate": 9.382387706855793e-07,
+      "loss": 0.0988,
+      "step": 2854
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.29915106296539307,
+      "learning_rate": 9.367612293144209e-07,
+      "loss": 0.0993,
+      "step": 2855
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.28647637367248535,
+      "learning_rate": 9.352836879432625e-07,
+      "loss": 0.1041,
+      "step": 2856
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2972520887851715,
+      "learning_rate": 9.338061465721041e-07,
+      "loss": 0.1056,
+      "step": 2857
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.3061383366584778,
+      "learning_rate": 9.323286052009457e-07,
+      "loss": 0.1038,
+      "step": 2858
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.29745516180992126,
+      "learning_rate": 9.308510638297872e-07,
+      "loss": 0.1017,
+      "step": 2859
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.29421311616897583,
+      "learning_rate": 9.29373522458629e-07,
+      "loss": 0.1026,
+      "step": 2860
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2981680929660797,
+      "learning_rate": 9.278959810874705e-07,
+      "loss": 0.1035,
+      "step": 2861
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.30175986886024475,
+      "learning_rate": 9.264184397163122e-07,
+      "loss": 0.1022,
+      "step": 2862
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2928481996059418,
+      "learning_rate": 9.249408983451537e-07,
+      "loss": 0.0974,
+      "step": 2863
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.30058106780052185,
+      "learning_rate": 9.234633569739953e-07,
+      "loss": 0.0985,
+      "step": 2864
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2931995987892151,
+      "learning_rate": 9.219858156028369e-07,
+      "loss": 0.1009,
+      "step": 2865
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.29406002163887024,
+      "learning_rate": 9.205082742316786e-07,
+      "loss": 0.101,
+      "step": 2866
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.3006937503814697,
+      "learning_rate": 9.190307328605201e-07,
+      "loss": 0.1017,
+      "step": 2867
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.3065166771411896,
+      "learning_rate": 9.175531914893618e-07,
+      "loss": 0.1,
+      "step": 2868
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.3141879439353943,
+      "learning_rate": 9.160756501182033e-07,
+      "loss": 0.0993,
+      "step": 2869
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.29548919200897217,
+      "learning_rate": 9.14598108747045e-07,
+      "loss": 0.102,
+      "step": 2870
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.29190176725387573,
+      "learning_rate": 9.131205673758867e-07,
+      "loss": 0.103,
+      "step": 2871
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.30462419986724854,
+      "learning_rate": 9.116430260047282e-07,
+      "loss": 0.1033,
+      "step": 2872
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.291686087846756,
+      "learning_rate": 9.101654846335699e-07,
+      "loss": 0.1013,
+      "step": 2873
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.2886969745159149,
+      "learning_rate": 9.086879432624114e-07,
+      "loss": 0.0985,
+      "step": 2874
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.31121140718460083,
+      "learning_rate": 9.07210401891253e-07,
+      "loss": 0.1038,
+      "step": 2875
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.29642900824546814,
+      "learning_rate": 9.057328605200946e-07,
+      "loss": 0.0999,
+      "step": 2876
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.3031664192676544,
+      "learning_rate": 9.042553191489363e-07,
+      "loss": 0.1007,
+      "step": 2877
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.28943681716918945,
+      "learning_rate": 9.027777777777779e-07,
+      "loss": 0.0974,
+      "step": 2878
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.2956269383430481,
+      "learning_rate": 9.013002364066195e-07,
+      "loss": 0.1013,
+      "step": 2879
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.28794893622398376,
+      "learning_rate": 8.99822695035461e-07,
+      "loss": 0.0998,
+      "step": 2880
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.27833840250968933,
+      "learning_rate": 8.983451536643027e-07,
+      "loss": 0.0991,
+      "step": 2881
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.29869675636291504,
+      "learning_rate": 8.968676122931442e-07,
+      "loss": 0.1015,
+      "step": 2882
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.28746286034584045,
+      "learning_rate": 8.95390070921986e-07,
+      "loss": 0.0966,
+      "step": 2883
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3120495676994324,
+      "learning_rate": 8.939125295508275e-07,
+      "loss": 0.1032,
+      "step": 2884
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3064665198326111,
+      "learning_rate": 8.924349881796691e-07,
+      "loss": 0.1027,
+      "step": 2885
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.28617846965789795,
+      "learning_rate": 8.909574468085107e-07,
+      "loss": 0.1059,
+      "step": 2886
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3061176836490631,
+      "learning_rate": 8.894799054373523e-07,
+      "loss": 0.1009,
+      "step": 2887
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.2989843189716339,
+      "learning_rate": 8.880023640661939e-07,
+      "loss": 0.1007,
+      "step": 2888
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.29698726534843445,
+      "learning_rate": 8.865248226950356e-07,
+      "loss": 0.1018,
+      "step": 2889
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.28415513038635254,
+      "learning_rate": 8.850472813238771e-07,
+      "loss": 0.1014,
+      "step": 2890
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2988301217556,
+      "learning_rate": 8.835697399527188e-07,
+      "loss": 0.1021,
+      "step": 2891
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.29267290234565735,
+      "learning_rate": 8.820921985815603e-07,
+      "loss": 0.1013,
+      "step": 2892
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2903957664966583,
+      "learning_rate": 8.80614657210402e-07,
+      "loss": 0.0967,
+      "step": 2893
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.308270126581192,
+      "learning_rate": 8.791371158392436e-07,
+      "loss": 0.1032,
+      "step": 2894
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.295505553483963,
+      "learning_rate": 8.776595744680852e-07,
+      "loss": 0.1009,
+      "step": 2895
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2870788276195526,
+      "learning_rate": 8.761820330969267e-07,
+      "loss": 0.0963,
+      "step": 2896
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.29682108759880066,
+      "learning_rate": 8.747044917257684e-07,
+      "loss": 0.1005,
+      "step": 2897
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.29423806071281433,
+      "learning_rate": 8.732269503546099e-07,
+      "loss": 0.1021,
+      "step": 2898
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.28941142559051514,
+      "learning_rate": 8.717494089834517e-07,
+      "loss": 0.1015,
+      "step": 2899
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2867008447647095,
+      "learning_rate": 8.702718676122932e-07,
+      "loss": 0.1016,
+      "step": 2900
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2877733111381531,
+      "learning_rate": 8.687943262411348e-07,
+      "loss": 0.0998,
+      "step": 2901
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2984996736049652,
+      "learning_rate": 8.673167848699764e-07,
+      "loss": 0.1034,
+      "step": 2902
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2939130365848541,
+      "learning_rate": 8.65839243498818e-07,
+      "loss": 0.0971,
+      "step": 2903
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2885243594646454,
+      "learning_rate": 8.643617021276597e-07,
+      "loss": 0.1029,
+      "step": 2904
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.28547102212905884,
+      "learning_rate": 8.628841607565013e-07,
+      "loss": 0.1042,
+      "step": 2905
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.29665929079055786,
+      "learning_rate": 8.614066193853429e-07,
+      "loss": 0.1023,
+      "step": 2906
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.29190170764923096,
+      "learning_rate": 8.599290780141845e-07,
+      "loss": 0.0975,
+      "step": 2907
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.27988186478614807,
+      "learning_rate": 8.584515366430261e-07,
+      "loss": 0.1013,
+      "step": 2908
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.27806246280670166,
+      "learning_rate": 8.569739952718676e-07,
+      "loss": 0.0993,
+      "step": 2909
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.28322136402130127,
+      "learning_rate": 8.554964539007094e-07,
+      "loss": 0.0994,
+      "step": 2910
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.292763352394104,
+      "learning_rate": 8.540189125295509e-07,
+      "loss": 0.0984,
+      "step": 2911
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2851690649986267,
+      "learning_rate": 8.525413711583925e-07,
+      "loss": 0.0977,
+      "step": 2912
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.30315104126930237,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 0.1011,
+      "step": 2913
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2887428104877472,
+      "learning_rate": 8.495862884160757e-07,
+      "loss": 0.0991,
+      "step": 2914
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2720642685890198,
+      "learning_rate": 8.481087470449173e-07,
+      "loss": 0.0943,
+      "step": 2915
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.300809770822525,
+      "learning_rate": 8.46631205673759e-07,
+      "loss": 0.1023,
+      "step": 2916
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.30130618810653687,
+      "learning_rate": 8.451536643026005e-07,
+      "loss": 0.1052,
+      "step": 2917
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.3137718439102173,
+      "learning_rate": 8.436761229314422e-07,
+      "loss": 0.1051,
+      "step": 2918
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2853814959526062,
+      "learning_rate": 8.421985815602837e-07,
+      "loss": 0.0992,
+      "step": 2919
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2999713718891144,
+      "learning_rate": 8.407210401891253e-07,
+      "loss": 0.1061,
+      "step": 2920
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.29466015100479126,
+      "learning_rate": 8.39243498817967e-07,
+      "loss": 0.1042,
+      "step": 2921
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2791386842727661,
+      "learning_rate": 8.377659574468086e-07,
+      "loss": 0.0985,
+      "step": 2922
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.30222687125205994,
+      "learning_rate": 8.362884160756502e-07,
+      "loss": 0.1034,
+      "step": 2923
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.3009599447250366,
+      "learning_rate": 8.348108747044918e-07,
+      "loss": 0.0977,
+      "step": 2924
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.279634028673172,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.0997,
+      "step": 2925
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.29688337445259094,
+      "learning_rate": 8.318557919621751e-07,
+      "loss": 0.1015,
+      "step": 2926
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2865789532661438,
+      "learning_rate": 8.303782505910166e-07,
+      "loss": 0.1027,
+      "step": 2927
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.28429552912712097,
+      "learning_rate": 8.289007092198582e-07,
+      "loss": 0.0977,
+      "step": 2928
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.29092469811439514,
+      "learning_rate": 8.274231678486998e-07,
+      "loss": 0.097,
+      "step": 2929
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.30970582365989685,
+      "learning_rate": 8.259456264775414e-07,
+      "loss": 0.103,
+      "step": 2930
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2910176217556,
+      "learning_rate": 8.24468085106383e-07,
+      "loss": 0.0969,
+      "step": 2931
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2959626317024231,
+      "learning_rate": 8.229905437352247e-07,
+      "loss": 0.1034,
+      "step": 2932
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.28358665108680725,
+      "learning_rate": 8.215130023640662e-07,
+      "loss": 0.0994,
+      "step": 2933
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.29149356484413147,
+      "learning_rate": 8.200354609929079e-07,
+      "loss": 0.1002,
+      "step": 2934
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2894674837589264,
+      "learning_rate": 8.185579196217494e-07,
+      "loss": 0.0954,
+      "step": 2935
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.30106693506240845,
+      "learning_rate": 8.17080378250591e-07,
+      "loss": 0.1032,
+      "step": 2936
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2927505671977997,
+      "learning_rate": 8.156028368794328e-07,
+      "loss": 0.0966,
+      "step": 2937
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2981710135936737,
+      "learning_rate": 8.141252955082743e-07,
+      "loss": 0.103,
+      "step": 2938
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.31989383697509766,
+      "learning_rate": 8.12647754137116e-07,
+      "loss": 0.1046,
+      "step": 2939
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.31127825379371643,
+      "learning_rate": 8.111702127659575e-07,
+      "loss": 0.0988,
+      "step": 2940
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.30211105942726135,
+      "learning_rate": 8.096926713947991e-07,
+      "loss": 0.103,
+      "step": 2941
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2955443859100342,
+      "learning_rate": 8.082151300236407e-07,
+      "loss": 0.0963,
+      "step": 2942
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.3013782501220703,
+      "learning_rate": 8.067375886524824e-07,
+      "loss": 0.1021,
+      "step": 2943
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.29456835985183716,
+      "learning_rate": 8.05260047281324e-07,
+      "loss": 0.1016,
+      "step": 2944
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.30215513706207275,
+      "learning_rate": 8.037825059101656e-07,
+      "loss": 0.1001,
+      "step": 2945
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2833058834075928,
+      "learning_rate": 8.023049645390071e-07,
+      "loss": 0.0999,
+      "step": 2946
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.29116910696029663,
+      "learning_rate": 8.008274231678488e-07,
+      "loss": 0.0983,
+      "step": 2947
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.297916978597641,
+      "learning_rate": 7.993498817966903e-07,
+      "loss": 0.1011,
+      "step": 2948
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.29713013768196106,
+      "learning_rate": 7.97872340425532e-07,
+      "loss": 0.1021,
+      "step": 2949
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.28667354583740234,
+      "learning_rate": 7.963947990543736e-07,
+      "loss": 0.0986,
+      "step": 2950
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.28612220287323,
+      "learning_rate": 7.949172576832152e-07,
+      "loss": 0.1027,
+      "step": 2951
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.303057461977005,
+      "learning_rate": 7.934397163120567e-07,
+      "loss": 0.1011,
+      "step": 2952
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.29398488998413086,
+      "learning_rate": 7.919621749408985e-07,
+      "loss": 0.1014,
+      "step": 2953
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3039552867412567,
+      "learning_rate": 7.9048463356974e-07,
+      "loss": 0.103,
+      "step": 2954
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.2897822856903076,
+      "learning_rate": 7.890070921985817e-07,
+      "loss": 0.1,
+      "step": 2955
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.30103448033332825,
+      "learning_rate": 7.875295508274232e-07,
+      "loss": 0.1028,
+      "step": 2956
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.29206156730651855,
+      "learning_rate": 7.860520094562648e-07,
+      "loss": 0.1001,
+      "step": 2957
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.29797887802124023,
+      "learning_rate": 7.845744680851064e-07,
+      "loss": 0.1,
+      "step": 2958
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.28990575671195984,
+      "learning_rate": 7.830969267139481e-07,
+      "loss": 0.1013,
+      "step": 2959
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.29574206471443176,
+      "learning_rate": 7.816193853427897e-07,
+      "loss": 0.1005,
+      "step": 2960
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2913941740989685,
+      "learning_rate": 7.801418439716313e-07,
+      "loss": 0.1015,
+      "step": 2961
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.29508161544799805,
+      "learning_rate": 7.786643026004728e-07,
+      "loss": 0.0995,
+      "step": 2962
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2888253927230835,
+      "learning_rate": 7.771867612293145e-07,
+      "loss": 0.0975,
+      "step": 2963
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.30230095982551575,
+      "learning_rate": 7.75709219858156e-07,
+      "loss": 0.0973,
+      "step": 2964
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2799548804759979,
+      "learning_rate": 7.742316784869977e-07,
+      "loss": 0.0995,
+      "step": 2965
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2934662997722626,
+      "learning_rate": 7.727541371158393e-07,
+      "loss": 0.1004,
+      "step": 2966
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.27563154697418213,
+      "learning_rate": 7.712765957446809e-07,
+      "loss": 0.0977,
+      "step": 2967
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.30211102962493896,
+      "learning_rate": 7.697990543735225e-07,
+      "loss": 0.1035,
+      "step": 2968
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2879369854927063,
+      "learning_rate": 7.683215130023641e-07,
+      "loss": 0.0984,
+      "step": 2969
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.28050658106803894,
+      "learning_rate": 7.668439716312058e-07,
+      "loss": 0.0993,
+      "step": 2970
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2956594228744507,
+      "learning_rate": 7.653664302600474e-07,
+      "loss": 0.0982,
+      "step": 2971
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.285321444272995,
+      "learning_rate": 7.63888888888889e-07,
+      "loss": 0.1,
+      "step": 2972
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3093894422054291,
+      "learning_rate": 7.624113475177305e-07,
+      "loss": 0.1033,
+      "step": 2973
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.29528602957725525,
+      "learning_rate": 7.609338061465722e-07,
+      "loss": 0.1033,
+      "step": 2974
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2800599932670593,
+      "learning_rate": 7.594562647754137e-07,
+      "loss": 0.0978,
+      "step": 2975
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.301089346408844,
+      "learning_rate": 7.579787234042555e-07,
+      "loss": 0.1028,
+      "step": 2976
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2935054898262024,
+      "learning_rate": 7.56501182033097e-07,
+      "loss": 0.1038,
+      "step": 2977
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3007417917251587,
+      "learning_rate": 7.550236406619386e-07,
+      "loss": 0.0978,
+      "step": 2978
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2877776622772217,
+      "learning_rate": 7.535460992907802e-07,
+      "loss": 0.1029,
+      "step": 2979
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.283892959356308,
+      "learning_rate": 7.520685579196218e-07,
+      "loss": 0.0978,
+      "step": 2980
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2816719114780426,
+      "learning_rate": 7.505910165484633e-07,
+      "loss": 0.0945,
+      "step": 2981
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.29446297883987427,
+      "learning_rate": 7.491134751773051e-07,
+      "loss": 0.1012,
+      "step": 2982
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2932320535182953,
+      "learning_rate": 7.476359338061466e-07,
+      "loss": 0.0992,
+      "step": 2983
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.3051176369190216,
+      "learning_rate": 7.461583924349883e-07,
+      "loss": 0.1047,
+      "step": 2984
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2896937429904938,
+      "learning_rate": 7.446808510638298e-07,
+      "loss": 0.103,
+      "step": 2985
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2901441156864166,
+      "learning_rate": 7.432033096926715e-07,
+      "loss": 0.0968,
+      "step": 2986
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2830792963504791,
+      "learning_rate": 7.417257683215131e-07,
+      "loss": 0.0993,
+      "step": 2987
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.31729981303215027,
+      "learning_rate": 7.402482269503547e-07,
+      "loss": 0.1073,
+      "step": 2988
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2926541268825531,
+      "learning_rate": 7.387706855791962e-07,
+      "loss": 0.0964,
+      "step": 2989
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.30170807242393494,
+      "learning_rate": 7.372931442080379e-07,
+      "loss": 0.0989,
+      "step": 2990
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.3067688047885895,
+      "learning_rate": 7.358156028368794e-07,
+      "loss": 0.1053,
+      "step": 2991
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.29287537932395935,
+      "learning_rate": 7.343380614657212e-07,
+      "loss": 0.0973,
+      "step": 2992
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.28648877143859863,
+      "learning_rate": 7.328605200945627e-07,
+      "loss": 0.0988,
+      "step": 2993
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.28888073563575745,
+      "learning_rate": 7.313829787234043e-07,
+      "loss": 0.0971,
+      "step": 2994
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.28870508074760437,
+      "learning_rate": 7.299054373522459e-07,
+      "loss": 0.0979,
+      "step": 2995
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.30402082204818726,
+      "learning_rate": 7.284278959810875e-07,
+      "loss": 0.1036,
+      "step": 2996
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.28851741552352905,
+      "learning_rate": 7.26950354609929e-07,
+      "loss": 0.0964,
+      "step": 2997
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2824156880378723,
+      "learning_rate": 7.254728132387708e-07,
+      "loss": 0.0993,
+      "step": 2998
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.27733680605888367,
+      "learning_rate": 7.239952718676123e-07,
+      "loss": 0.0997,
+      "step": 2999
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2903364300727844,
+      "learning_rate": 7.22517730496454e-07,
+      "loss": 0.1009,
+      "step": 3000
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.29385676980018616,
+      "learning_rate": 7.210401891252956e-07,
+      "loss": 0.1023,
+      "step": 3001
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.28179335594177246,
+      "learning_rate": 7.195626477541371e-07,
+      "loss": 0.1008,
+      "step": 3002
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2913713753223419,
+      "learning_rate": 7.180851063829789e-07,
+      "loss": 0.0987,
+      "step": 3003
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.29837390780448914,
+      "learning_rate": 7.166075650118204e-07,
+      "loss": 0.1039,
+      "step": 3004
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.3011203706264496,
+      "learning_rate": 7.151300236406621e-07,
+      "loss": 0.1053,
+      "step": 3005
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2935463786125183,
+      "learning_rate": 7.136524822695036e-07,
+      "loss": 0.1009,
+      "step": 3006
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.28958970308303833,
+      "learning_rate": 7.121749408983452e-07,
+      "loss": 0.1032,
+      "step": 3007
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.308200478553772,
+      "learning_rate": 7.106973995271868e-07,
+      "loss": 0.1035,
+      "step": 3008
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2873825430870056,
+      "learning_rate": 7.092198581560285e-07,
+      "loss": 0.0994,
+      "step": 3009
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2968955636024475,
+      "learning_rate": 7.0774231678487e-07,
+      "loss": 0.1006,
+      "step": 3010
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2786247730255127,
+      "learning_rate": 7.062647754137117e-07,
+      "loss": 0.0973,
+      "step": 3011
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.3046036660671234,
+      "learning_rate": 7.047872340425532e-07,
+      "loss": 0.1005,
+      "step": 3012
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2819902002811432,
+      "learning_rate": 7.033096926713949e-07,
+      "loss": 0.0974,
+      "step": 3013
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.30306100845336914,
+      "learning_rate": 7.018321513002364e-07,
+      "loss": 0.1018,
+      "step": 3014
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2917092442512512,
+      "learning_rate": 7.003546099290781e-07,
+      "loss": 0.0997,
+      "step": 3015
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.3216758966445923,
+      "learning_rate": 6.988770685579197e-07,
+      "loss": 0.0995,
+      "step": 3016
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2969210743904114,
+      "learning_rate": 6.973995271867613e-07,
+      "loss": 0.0974,
+      "step": 3017
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.2986419200897217,
+      "learning_rate": 6.959219858156028e-07,
+      "loss": 0.1004,
+      "step": 3018
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.27990296483039856,
+      "learning_rate": 6.944444444444446e-07,
+      "loss": 0.0989,
+      "step": 3019
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.29205796122550964,
+      "learning_rate": 6.929669030732861e-07,
+      "loss": 0.1032,
+      "step": 3020
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.30241334438323975,
+      "learning_rate": 6.914893617021278e-07,
+      "loss": 0.1011,
+      "step": 3021
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3049878180027008,
+      "learning_rate": 6.900118203309693e-07,
+      "loss": 0.1049,
+      "step": 3022
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.28830352425575256,
+      "learning_rate": 6.885342789598109e-07,
+      "loss": 0.1013,
+      "step": 3023
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.28061676025390625,
+      "learning_rate": 6.870567375886525e-07,
+      "loss": 0.0977,
+      "step": 3024
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.302949994802475,
+      "learning_rate": 6.855791962174942e-07,
+      "loss": 0.1048,
+      "step": 3025
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.280875027179718,
+      "learning_rate": 6.841016548463357e-07,
+      "loss": 0.097,
+      "step": 3026
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.2862735688686371,
+      "learning_rate": 6.826241134751774e-07,
+      "loss": 0.1031,
+      "step": 3027
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.2974378168582916,
+      "learning_rate": 6.811465721040189e-07,
+      "loss": 0.1039,
+      "step": 3028
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3044671416282654,
+      "learning_rate": 6.796690307328606e-07,
+      "loss": 0.1036,
+      "step": 3029
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2969146966934204,
+      "learning_rate": 6.781914893617021e-07,
+      "loss": 0.0999,
+      "step": 3030
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2826891243457794,
+      "learning_rate": 6.767139479905438e-07,
+      "loss": 0.0987,
+      "step": 3031
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2863085865974426,
+      "learning_rate": 6.752364066193854e-07,
+      "loss": 0.0994,
+      "step": 3032
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.29002559185028076,
+      "learning_rate": 6.73758865248227e-07,
+      "loss": 0.1039,
+      "step": 3033
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.285457044839859,
+      "learning_rate": 6.722813238770686e-07,
+      "loss": 0.1008,
+      "step": 3034
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.29994022846221924,
+      "learning_rate": 6.708037825059102e-07,
+      "loss": 0.104,
+      "step": 3035
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.28790104389190674,
+      "learning_rate": 6.693262411347519e-07,
+      "loss": 0.1017,
+      "step": 3036
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2845805585384369,
+      "learning_rate": 6.678486997635935e-07,
+      "loss": 0.0978,
+      "step": 3037
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2992018461227417,
+      "learning_rate": 6.663711583924351e-07,
+      "loss": 0.1019,
+      "step": 3038
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.28522050380706787,
+      "learning_rate": 6.648936170212766e-07,
+      "loss": 0.1007,
+      "step": 3039
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.29484012722969055,
+      "learning_rate": 6.634160756501183e-07,
+      "loss": 0.1005,
+      "step": 3040
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2922971248626709,
+      "learning_rate": 6.619385342789598e-07,
+      "loss": 0.0997,
+      "step": 3041
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.28032514452934265,
+      "learning_rate": 6.604609929078016e-07,
+      "loss": 0.0982,
+      "step": 3042
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2958720624446869,
+      "learning_rate": 6.589834515366431e-07,
+      "loss": 0.0997,
+      "step": 3043
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2862323224544525,
+      "learning_rate": 6.575059101654847e-07,
+      "loss": 0.0994,
+      "step": 3044
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.30775293707847595,
+      "learning_rate": 6.560283687943263e-07,
+      "loss": 0.0981,
+      "step": 3045
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2850129008293152,
+      "learning_rate": 6.545508274231679e-07,
+      "loss": 0.1003,
+      "step": 3046
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2836865186691284,
+      "learning_rate": 6.530732860520094e-07,
+      "loss": 0.1025,
+      "step": 3047
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.29065871238708496,
+      "learning_rate": 6.515957446808512e-07,
+      "loss": 0.0974,
+      "step": 3048
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.27818062901496887,
+      "learning_rate": 6.501182033096927e-07,
+      "loss": 0.0996,
+      "step": 3049
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2964308559894562,
+      "learning_rate": 6.486406619385343e-07,
+      "loss": 0.1031,
+      "step": 3050
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2839726209640503,
+      "learning_rate": 6.471631205673759e-07,
+      "loss": 0.0995,
+      "step": 3051
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.32063916325569153,
+      "learning_rate": 6.456855791962176e-07,
+      "loss": 0.1004,
+      "step": 3052
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3052508533000946,
+      "learning_rate": 6.442080378250592e-07,
+      "loss": 0.1034,
+      "step": 3053
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3084821105003357,
+      "learning_rate": 6.427304964539008e-07,
+      "loss": 0.1007,
+      "step": 3054
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.28898391127586365,
+      "learning_rate": 6.412529550827423e-07,
+      "loss": 0.1027,
+      "step": 3055
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.29600197076797485,
+      "learning_rate": 6.39775413711584e-07,
+      "loss": 0.0991,
+      "step": 3056
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.2969059944152832,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 0.0988,
+      "step": 3057
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.2989344298839569,
+      "learning_rate": 6.368203309692673e-07,
+      "loss": 0.0998,
+      "step": 3058
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.29680922627449036,
+      "learning_rate": 6.353427895981088e-07,
+      "loss": 0.1002,
+      "step": 3059
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.30277103185653687,
+      "learning_rate": 6.338652482269504e-07,
+      "loss": 0.1028,
+      "step": 3060
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.29336634278297424,
+      "learning_rate": 6.32387706855792e-07,
+      "loss": 0.1026,
+      "step": 3061
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3019260764122009,
+      "learning_rate": 6.309101654846336e-07,
+      "loss": 0.1013,
+      "step": 3062
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3115624785423279,
+      "learning_rate": 6.294326241134751e-07,
+      "loss": 0.1023,
+      "step": 3063
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3000103831291199,
+      "learning_rate": 6.279550827423169e-07,
+      "loss": 0.0976,
+      "step": 3064
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2888752222061157,
+      "learning_rate": 6.264775413711584e-07,
+      "loss": 0.1005,
+      "step": 3065
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29305657744407654,
+      "learning_rate": 6.25e-07,
+      "loss": 0.0982,
+      "step": 3066
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2918168604373932,
+      "learning_rate": 6.235224586288417e-07,
+      "loss": 0.1014,
+      "step": 3067
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2931625247001648,
+      "learning_rate": 6.220449172576832e-07,
+      "loss": 0.1009,
+      "step": 3068
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29553675651550293,
+      "learning_rate": 6.205673758865249e-07,
+      "loss": 0.1001,
+      "step": 3069
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29019299149513245,
+      "learning_rate": 6.190898345153665e-07,
+      "loss": 0.1016,
+      "step": 3070
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29510360956192017,
+      "learning_rate": 6.17612293144208e-07,
+      "loss": 0.0999,
+      "step": 3071
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2989548444747925,
+      "learning_rate": 6.161347517730497e-07,
+      "loss": 0.1017,
+      "step": 3072
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29093608260154724,
+      "learning_rate": 6.146572104018913e-07,
+      "loss": 0.0985,
+      "step": 3073
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29467877745628357,
+      "learning_rate": 6.131796690307328e-07,
+      "loss": 0.1034,
+      "step": 3074
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.3069037199020386,
+      "learning_rate": 6.117021276595745e-07,
+      "loss": 0.101,
+      "step": 3075
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2977922558784485,
+      "learning_rate": 6.102245862884161e-07,
+      "loss": 0.1009,
+      "step": 3076
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2957204580307007,
+      "learning_rate": 6.087470449172577e-07,
+      "loss": 0.0974,
+      "step": 3077
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.28425145149230957,
+      "learning_rate": 6.072695035460993e-07,
+      "loss": 0.0964,
+      "step": 3078
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2853310704231262,
+      "learning_rate": 6.057919621749409e-07,
+      "loss": 0.0984,
+      "step": 3079
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.30463528633117676,
+      "learning_rate": 6.043144208037825e-07,
+      "loss": 0.0978,
+      "step": 3080
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.3013131320476532,
+      "learning_rate": 6.028368794326241e-07,
+      "loss": 0.1,
+      "step": 3081
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2993744909763336,
+      "learning_rate": 6.013593380614658e-07,
+      "loss": 0.1016,
+      "step": 3082
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2832348644733429,
+      "learning_rate": 5.998817966903074e-07,
+      "loss": 0.1002,
+      "step": 3083
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.29118210077285767,
+      "learning_rate": 5.98404255319149e-07,
+      "loss": 0.1022,
+      "step": 3084
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.3079187273979187,
+      "learning_rate": 5.969267139479907e-07,
+      "loss": 0.1027,
+      "step": 3085
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.29184913635253906,
+      "learning_rate": 5.954491725768322e-07,
+      "loss": 0.102,
+      "step": 3086
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.289700448513031,
+      "learning_rate": 5.939716312056738e-07,
+      "loss": 0.1013,
+      "step": 3087
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.29641756415367126,
+      "learning_rate": 5.924940898345155e-07,
+      "loss": 0.1021,
+      "step": 3088
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.29890719056129456,
+      "learning_rate": 5.91016548463357e-07,
+      "loss": 0.1021,
+      "step": 3089
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2940806448459625,
+      "learning_rate": 5.895390070921987e-07,
+      "loss": 0.1028,
+      "step": 3090
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.3055453598499298,
+      "learning_rate": 5.880614657210403e-07,
+      "loss": 0.1021,
+      "step": 3091
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2881205677986145,
+      "learning_rate": 5.865839243498818e-07,
+      "loss": 0.1013,
+      "step": 3092
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2925069034099579,
+      "learning_rate": 5.851063829787235e-07,
+      "loss": 0.1022,
+      "step": 3093
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.30679821968078613,
+      "learning_rate": 5.836288416075651e-07,
+      "loss": 0.1014,
+      "step": 3094
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2992866337299347,
+      "learning_rate": 5.821513002364066e-07,
+      "loss": 0.0986,
+      "step": 3095
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.29628366231918335,
+      "learning_rate": 5.806737588652483e-07,
+      "loss": 0.1033,
+      "step": 3096
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.31407296657562256,
+      "learning_rate": 5.791962174940899e-07,
+      "loss": 0.1054,
+      "step": 3097
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.28960010409355164,
+      "learning_rate": 5.777186761229315e-07,
+      "loss": 0.0996,
+      "step": 3098
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2831580638885498,
+      "learning_rate": 5.762411347517731e-07,
+      "loss": 0.0993,
+      "step": 3099
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.28401294350624084,
+      "learning_rate": 5.747635933806147e-07,
+      "loss": 0.1015,
+      "step": 3100
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.3234121799468994,
+      "learning_rate": 5.732860520094563e-07,
+      "loss": 0.1051,
+      "step": 3101
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2908365726470947,
+      "learning_rate": 5.718085106382979e-07,
+      "loss": 0.0973,
+      "step": 3102
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2927663028240204,
+      "learning_rate": 5.703309692671395e-07,
+      "loss": 0.0993,
+      "step": 3103
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.31612542271614075,
+      "learning_rate": 5.688534278959811e-07,
+      "loss": 0.1028,
+      "step": 3104
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2908136546611786,
+      "learning_rate": 5.673758865248227e-07,
+      "loss": 0.1012,
+      "step": 3105
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.28986722230911255,
+      "learning_rate": 5.658983451536644e-07,
+      "loss": 0.1,
+      "step": 3106
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2795732021331787,
+      "learning_rate": 5.644208037825059e-07,
+      "loss": 0.0998,
+      "step": 3107
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.28216084837913513,
+      "learning_rate": 5.629432624113475e-07,
+      "loss": 0.0994,
+      "step": 3108
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2967695891857147,
+      "learning_rate": 5.614657210401892e-07,
+      "loss": 0.099,
+      "step": 3109
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.29375335574150085,
+      "learning_rate": 5.599881796690307e-07,
+      "loss": 0.0987,
+      "step": 3110
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.30044302344322205,
+      "learning_rate": 5.585106382978723e-07,
+      "loss": 0.0974,
+      "step": 3111
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.29533493518829346,
+      "learning_rate": 5.57033096926714e-07,
+      "loss": 0.1008,
+      "step": 3112
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.30284395813941956,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.0986,
+      "step": 3113
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.286884605884552,
+      "learning_rate": 5.540780141843972e-07,
+      "loss": 0.098,
+      "step": 3114
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.3027244806289673,
+      "learning_rate": 5.526004728132388e-07,
+      "loss": 0.1003,
+      "step": 3115
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.30712053179740906,
+      "learning_rate": 5.511229314420804e-07,
+      "loss": 0.103,
+      "step": 3116
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.301013708114624,
+      "learning_rate": 5.496453900709221e-07,
+      "loss": 0.1,
+      "step": 3117
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.29280179738998413,
+      "learning_rate": 5.481678486997637e-07,
+      "loss": 0.1007,
+      "step": 3118
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2868701219558716,
+      "learning_rate": 5.466903073286052e-07,
+      "loss": 0.0972,
+      "step": 3119
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2882191836833954,
+      "learning_rate": 5.452127659574469e-07,
+      "loss": 0.1013,
+      "step": 3120
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.30170050263404846,
+      "learning_rate": 5.437352245862885e-07,
+      "loss": 0.1008,
+      "step": 3121
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2974720299243927,
+      "learning_rate": 5.422576832151301e-07,
+      "loss": 0.1015,
+      "step": 3122
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.3039587140083313,
+      "learning_rate": 5.407801418439717e-07,
+      "loss": 0.1029,
+      "step": 3123
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.3278217017650604,
+      "learning_rate": 5.393026004728133e-07,
+      "loss": 0.1002,
+      "step": 3124
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.30176177620887756,
+      "learning_rate": 5.378250591016549e-07,
+      "loss": 0.1003,
+      "step": 3125
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2856448292732239,
+      "learning_rate": 5.363475177304965e-07,
+      "loss": 0.0982,
+      "step": 3126
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.27896958589553833,
+      "learning_rate": 5.348699763593382e-07,
+      "loss": 0.0958,
+      "step": 3127
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2881554067134857,
+      "learning_rate": 5.333924349881797e-07,
+      "loss": 0.0995,
+      "step": 3128
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.29578450322151184,
+      "learning_rate": 5.319148936170213e-07,
+      "loss": 0.1023,
+      "step": 3129
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2835855484008789,
+      "learning_rate": 5.30437352245863e-07,
+      "loss": 0.0974,
+      "step": 3130
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.29111170768737793,
+      "learning_rate": 5.289598108747045e-07,
+      "loss": 0.0968,
+      "step": 3131
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2896515429019928,
+      "learning_rate": 5.274822695035461e-07,
+      "loss": 0.0977,
+      "step": 3132
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.28907546401023865,
+      "learning_rate": 5.260047281323878e-07,
+      "loss": 0.1006,
+      "step": 3133
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.30252498388290405,
+      "learning_rate": 5.245271867612293e-07,
+      "loss": 0.1044,
+      "step": 3134
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2873133420944214,
+      "learning_rate": 5.23049645390071e-07,
+      "loss": 0.1002,
+      "step": 3135
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2922936975955963,
+      "learning_rate": 5.215721040189126e-07,
+      "loss": 0.1025,
+      "step": 3136
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.29615217447280884,
+      "learning_rate": 5.200945626477541e-07,
+      "loss": 0.1049,
+      "step": 3137
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.3067047894001007,
+      "learning_rate": 5.186170212765958e-07,
+      "loss": 0.1049,
+      "step": 3138
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.30281513929367065,
+      "learning_rate": 5.171394799054374e-07,
+      "loss": 0.0995,
+      "step": 3139
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.29096826910972595,
+      "learning_rate": 5.156619385342789e-07,
+      "loss": 0.1015,
+      "step": 3140
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.308479905128479,
+      "learning_rate": 5.141843971631206e-07,
+      "loss": 0.1043,
+      "step": 3141
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.29051145911216736,
+      "learning_rate": 5.127068557919622e-07,
+      "loss": 0.1007,
+      "step": 3142
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2863350212574005,
+      "learning_rate": 5.112293144208038e-07,
+      "loss": 0.101,
+      "step": 3143
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.3130321800708771,
+      "learning_rate": 5.097517730496454e-07,
+      "loss": 0.1001,
+      "step": 3144
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.28423893451690674,
+      "learning_rate": 5.08274231678487e-07,
+      "loss": 0.0987,
+      "step": 3145
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.28704968094825745,
+      "learning_rate": 5.067966903073286e-07,
+      "loss": 0.1024,
+      "step": 3146
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.3076588809490204,
+      "learning_rate": 5.053191489361702e-07,
+      "loss": 0.0993,
+      "step": 3147
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.3118914067745209,
+      "learning_rate": 5.03841607565012e-07,
+      "loss": 0.1021,
+      "step": 3148
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2936742603778839,
+      "learning_rate": 5.023640661938535e-07,
+      "loss": 0.1002,
+      "step": 3149
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.282441109418869,
+      "learning_rate": 5.008865248226951e-07,
+      "loss": 0.1005,
+      "step": 3150
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.29804307222366333,
+      "learning_rate": 4.994089834515368e-07,
+      "loss": 0.1007,
+      "step": 3151
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.29237258434295654,
+      "learning_rate": 4.979314420803783e-07,
+      "loss": 0.0982,
+      "step": 3152
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2982313632965088,
+      "learning_rate": 4.964539007092199e-07,
+      "loss": 0.0979,
+      "step": 3153
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.29577890038490295,
+      "learning_rate": 4.949763593380616e-07,
+      "loss": 0.1018,
+      "step": 3154
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2712281346321106,
+      "learning_rate": 4.934988179669031e-07,
+      "loss": 0.1006,
+      "step": 3155
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.30041977763175964,
+      "learning_rate": 4.920212765957447e-07,
+      "loss": 0.0995,
+      "step": 3156
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.28760164976119995,
+      "learning_rate": 4.905437352245864e-07,
+      "loss": 0.0994,
+      "step": 3157
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.31253767013549805,
+      "learning_rate": 4.890661938534279e-07,
+      "loss": 0.0988,
+      "step": 3158
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.28944775462150574,
+      "learning_rate": 4.875886524822696e-07,
+      "loss": 0.102,
+      "step": 3159
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.30013445019721985,
+      "learning_rate": 4.861111111111112e-07,
+      "loss": 0.0961,
+      "step": 3160
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.292417973279953,
+      "learning_rate": 4.846335697399527e-07,
+      "loss": 0.0992,
+      "step": 3161
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2782425880432129,
+      "learning_rate": 4.831560283687944e-07,
+      "loss": 0.095,
+      "step": 3162
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2884424030780792,
+      "learning_rate": 4.81678486997636e-07,
+      "loss": 0.0955,
+      "step": 3163
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2914351224899292,
+      "learning_rate": 4.802009456264775e-07,
+      "loss": 0.1012,
+      "step": 3164
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.29791468381881714,
+      "learning_rate": 4.787234042553192e-07,
+      "loss": 0.1033,
+      "step": 3165
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.28530699014663696,
+      "learning_rate": 4.772458628841608e-07,
+      "loss": 0.0997,
+      "step": 3166
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2834513783454895,
+      "learning_rate": 4.757683215130024e-07,
+      "loss": 0.0964,
+      "step": 3167
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.3007907271385193,
+      "learning_rate": 4.74290780141844e-07,
+      "loss": 0.0969,
+      "step": 3168
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.3055894672870636,
+      "learning_rate": 4.728132387706856e-07,
+      "loss": 0.1018,
+      "step": 3169
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.30420035123825073,
+      "learning_rate": 4.713356973995272e-07,
+      "loss": 0.101,
+      "step": 3170
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.28663694858551025,
+      "learning_rate": 4.698581560283688e-07,
+      "loss": 0.0979,
+      "step": 3171
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.29340553283691406,
+      "learning_rate": 4.6838061465721045e-07,
+      "loss": 0.0984,
+      "step": 3172
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.29488351941108704,
+      "learning_rate": 4.6690307328605203e-07,
+      "loss": 0.1028,
+      "step": 3173
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.2917100787162781,
+      "learning_rate": 4.654255319148936e-07,
+      "loss": 0.1012,
+      "step": 3174
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.2944326102733612,
+      "learning_rate": 4.6394799054373526e-07,
+      "loss": 0.1029,
+      "step": 3175
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.29237106442451477,
+      "learning_rate": 4.6247044917257685e-07,
+      "loss": 0.0956,
+      "step": 3176
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.29847726225852966,
+      "learning_rate": 4.6099290780141843e-07,
+      "loss": 0.0999,
+      "step": 3177
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.28914281725883484,
+      "learning_rate": 4.5951536643026007e-07,
+      "loss": 0.0993,
+      "step": 3178
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.28048333525657654,
+      "learning_rate": 4.5803782505910166e-07,
+      "loss": 0.1002,
+      "step": 3179
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3109818696975708,
+      "learning_rate": 4.5656028368794335e-07,
+      "loss": 0.0993,
+      "step": 3180
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3007102608680725,
+      "learning_rate": 4.5508274231678494e-07,
+      "loss": 0.1005,
+      "step": 3181
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2797643840312958,
+      "learning_rate": 4.536052009456265e-07,
+      "loss": 0.096,
+      "step": 3182
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.28448039293289185,
+      "learning_rate": 4.5212765957446816e-07,
+      "loss": 0.0996,
+      "step": 3183
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2921052873134613,
+      "learning_rate": 4.5065011820330975e-07,
+      "loss": 0.0996,
+      "step": 3184
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.29880839586257935,
+      "learning_rate": 4.4917257683215134e-07,
+      "loss": 0.0995,
+      "step": 3185
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2864736318588257,
+      "learning_rate": 4.47695035460993e-07,
+      "loss": 0.1018,
+      "step": 3186
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3114159405231476,
+      "learning_rate": 4.4621749408983456e-07,
+      "loss": 0.1001,
+      "step": 3187
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2995733618736267,
+      "learning_rate": 4.4473995271867615e-07,
+      "loss": 0.1042,
+      "step": 3188
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.29584336280822754,
+      "learning_rate": 4.432624113475178e-07,
+      "loss": 0.1008,
+      "step": 3189
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.29479894042015076,
+      "learning_rate": 4.417848699763594e-07,
+      "loss": 0.1026,
+      "step": 3190
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.285696804523468,
+      "learning_rate": 4.40307328605201e-07,
+      "loss": 0.097,
+      "step": 3191
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2800580859184265,
+      "learning_rate": 4.388297872340426e-07,
+      "loss": 0.0987,
+      "step": 3192
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.28899091482162476,
+      "learning_rate": 4.373522458628842e-07,
+      "loss": 0.1021,
+      "step": 3193
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2926108241081238,
+      "learning_rate": 4.3587470449172583e-07,
+      "loss": 0.0983,
+      "step": 3194
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.27967798709869385,
+      "learning_rate": 4.343971631205674e-07,
+      "loss": 0.097,
+      "step": 3195
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.28893759846687317,
+      "learning_rate": 4.32919621749409e-07,
+      "loss": 0.0994,
+      "step": 3196
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.292180597782135,
+      "learning_rate": 4.3144208037825064e-07,
+      "loss": 0.1029,
+      "step": 3197
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3064509332180023,
+      "learning_rate": 4.2996453900709223e-07,
+      "loss": 0.0989,
+      "step": 3198
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.29233554005622864,
+      "learning_rate": 4.284869976359338e-07,
+      "loss": 0.0997,
+      "step": 3199
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2989186644554138,
+      "learning_rate": 4.2700945626477545e-07,
+      "loss": 0.0976,
+      "step": 3200
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.29185956716537476,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 0.1002,
+      "step": 3201
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.27969053387641907,
+      "learning_rate": 4.2405437352245863e-07,
+      "loss": 0.0979,
+      "step": 3202
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2914850115776062,
+      "learning_rate": 4.2257683215130027e-07,
+      "loss": 0.0974,
+      "step": 3203
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3051094710826874,
+      "learning_rate": 4.2109929078014185e-07,
+      "loss": 0.1019,
+      "step": 3204
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.3007698357105255,
+      "learning_rate": 4.196217494089835e-07,
+      "loss": 0.1023,
+      "step": 3205
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2982529103755951,
+      "learning_rate": 4.181442080378251e-07,
+      "loss": 0.0987,
+      "step": 3206
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2981809973716736,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 0.1007,
+      "step": 3207
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.30896785855293274,
+      "learning_rate": 4.151891252955083e-07,
+      "loss": 0.1016,
+      "step": 3208
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2968890964984894,
+      "learning_rate": 4.137115839243499e-07,
+      "loss": 0.0977,
+      "step": 3209
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.29388725757598877,
+      "learning_rate": 4.122340425531915e-07,
+      "loss": 0.1002,
+      "step": 3210
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.28996309638023376,
+      "learning_rate": 4.107565011820331e-07,
+      "loss": 0.1013,
+      "step": 3211
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.3018192946910858,
+      "learning_rate": 4.092789598108747e-07,
+      "loss": 0.0979,
+      "step": 3212
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2953970432281494,
+      "learning_rate": 4.078014184397164e-07,
+      "loss": 0.1021,
+      "step": 3213
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.29449236392974854,
+      "learning_rate": 4.06323877068558e-07,
+      "loss": 0.1004,
+      "step": 3214
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.28408554196357727,
+      "learning_rate": 4.0484633569739957e-07,
+      "loss": 0.1005,
+      "step": 3215
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2912413775920868,
+      "learning_rate": 4.033687943262412e-07,
+      "loss": 0.0975,
+      "step": 3216
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.287748783826828,
+      "learning_rate": 4.018912529550828e-07,
+      "loss": 0.0979,
+      "step": 3217
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.30039867758750916,
+      "learning_rate": 4.004137115839244e-07,
+      "loss": 0.1037,
+      "step": 3218
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2908068597316742,
+      "learning_rate": 3.98936170212766e-07,
+      "loss": 0.1008,
+      "step": 3219
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.29198575019836426,
+      "learning_rate": 3.974586288416076e-07,
+      "loss": 0.098,
+      "step": 3220
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.29601743817329407,
+      "learning_rate": 3.9598108747044925e-07,
+      "loss": 0.1002,
+      "step": 3221
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.295124888420105,
+      "learning_rate": 3.9450354609929083e-07,
+      "loss": 0.0989,
+      "step": 3222
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.29923877120018005,
+      "learning_rate": 3.930260047281324e-07,
+      "loss": 0.1008,
+      "step": 3223
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.30230453610420227,
+      "learning_rate": 3.9154846335697406e-07,
+      "loss": 0.104,
+      "step": 3224
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.28801625967025757,
+      "learning_rate": 3.9007092198581565e-07,
+      "loss": 0.1016,
+      "step": 3225
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.291517972946167,
+      "learning_rate": 3.8859338061465723e-07,
+      "loss": 0.1005,
+      "step": 3226
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2919277250766754,
+      "learning_rate": 3.8711583924349887e-07,
+      "loss": 0.1002,
+      "step": 3227
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.30763471126556396,
+      "learning_rate": 3.8563829787234046e-07,
+      "loss": 0.0981,
+      "step": 3228
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2947371006011963,
+      "learning_rate": 3.8416075650118205e-07,
+      "loss": 0.1004,
+      "step": 3229
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2985070049762726,
+      "learning_rate": 3.826832151300237e-07,
+      "loss": 0.1002,
+      "step": 3230
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2864234149456024,
+      "learning_rate": 3.8120567375886527e-07,
+      "loss": 0.0974,
+      "step": 3231
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.277732253074646,
+      "learning_rate": 3.7972813238770686e-07,
+      "loss": 0.0982,
+      "step": 3232
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2844245135784149,
+      "learning_rate": 3.782505910165485e-07,
+      "loss": 0.1003,
+      "step": 3233
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.28954553604125977,
+      "learning_rate": 3.767730496453901e-07,
+      "loss": 0.0987,
+      "step": 3234
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3000849187374115,
+      "learning_rate": 3.7529550827423167e-07,
+      "loss": 0.097,
+      "step": 3235
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2906058132648468,
+      "learning_rate": 3.738179669030733e-07,
+      "loss": 0.0994,
+      "step": 3236
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3089418113231659,
+      "learning_rate": 3.723404255319149e-07,
+      "loss": 0.1012,
+      "step": 3237
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.30087733268737793,
+      "learning_rate": 3.7086288416075654e-07,
+      "loss": 0.1014,
+      "step": 3238
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3096245229244232,
+      "learning_rate": 3.693853427895981e-07,
+      "loss": 0.1046,
+      "step": 3239
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3011225461959839,
+      "learning_rate": 3.679078014184397e-07,
+      "loss": 0.1019,
+      "step": 3240
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2994038164615631,
+      "learning_rate": 3.6643026004728135e-07,
+      "loss": 0.1048,
+      "step": 3241
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.28039488196372986,
+      "learning_rate": 3.6495271867612294e-07,
+      "loss": 0.0983,
+      "step": 3242
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3006604313850403,
+      "learning_rate": 3.634751773049645e-07,
+      "loss": 0.1017,
+      "step": 3243
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.29698190093040466,
+      "learning_rate": 3.6199763593380616e-07,
+      "loss": 0.1017,
+      "step": 3244
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.30535686016082764,
+      "learning_rate": 3.605200945626478e-07,
+      "loss": 0.104,
+      "step": 3245
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.28788578510284424,
+      "learning_rate": 3.5904255319148944e-07,
+      "loss": 0.1018,
+      "step": 3246
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.29092541337013245,
+      "learning_rate": 3.5756501182033103e-07,
+      "loss": 0.0993,
+      "step": 3247
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2982770800590515,
+      "learning_rate": 3.560874704491726e-07,
+      "loss": 0.1013,
+      "step": 3248
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2855170965194702,
+      "learning_rate": 3.5460992907801425e-07,
+      "loss": 0.0995,
+      "step": 3249
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3044281005859375,
+      "learning_rate": 3.5313238770685584e-07,
+      "loss": 0.1029,
+      "step": 3250
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2925157845020294,
+      "learning_rate": 3.516548463356974e-07,
+      "loss": 0.0999,
+      "step": 3251
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.29421383142471313,
+      "learning_rate": 3.5017730496453907e-07,
+      "loss": 0.0997,
+      "step": 3252
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2912918031215668,
+      "learning_rate": 3.4869976359338065e-07,
+      "loss": 0.097,
+      "step": 3253
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.3049495220184326,
+      "learning_rate": 3.472222222222223e-07,
+      "loss": 0.1003,
+      "step": 3254
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2882755398750305,
+      "learning_rate": 3.457446808510639e-07,
+      "loss": 0.1005,
+      "step": 3255
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.30853840708732605,
+      "learning_rate": 3.4426713947990546e-07,
+      "loss": 0.1018,
+      "step": 3256
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.29855167865753174,
+      "learning_rate": 3.427895981087471e-07,
+      "loss": 0.1023,
+      "step": 3257
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.283933162689209,
+      "learning_rate": 3.413120567375887e-07,
+      "loss": 0.0997,
+      "step": 3258
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2944255471229553,
+      "learning_rate": 3.398345153664303e-07,
+      "loss": 0.0993,
+      "step": 3259
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.29534801840782166,
+      "learning_rate": 3.383569739952719e-07,
+      "loss": 0.1034,
+      "step": 3260
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.29837891459465027,
+      "learning_rate": 3.368794326241135e-07,
+      "loss": 0.1021,
+      "step": 3261
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.27940067648887634,
+      "learning_rate": 3.354018912529551e-07,
+      "loss": 0.1034,
+      "step": 3262
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.29715830087661743,
+      "learning_rate": 3.3392434988179673e-07,
+      "loss": 0.1037,
+      "step": 3263
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.29531028866767883,
+      "learning_rate": 3.324468085106383e-07,
+      "loss": 0.1023,
+      "step": 3264
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.28897911310195923,
+      "learning_rate": 3.309692671394799e-07,
+      "loss": 0.0992,
+      "step": 3265
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2843620777130127,
+      "learning_rate": 3.2949172576832154e-07,
+      "loss": 0.0973,
+      "step": 3266
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2797899842262268,
+      "learning_rate": 3.2801418439716313e-07,
+      "loss": 0.0964,
+      "step": 3267
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.28645768761634827,
+      "learning_rate": 3.265366430260047e-07,
+      "loss": 0.1,
+      "step": 3268
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2838907241821289,
+      "learning_rate": 3.2505910165484635e-07,
+      "loss": 0.0983,
+      "step": 3269
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2955852746963501,
+      "learning_rate": 3.2358156028368794e-07,
+      "loss": 0.0994,
+      "step": 3270
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2828039228916168,
+      "learning_rate": 3.221040189125296e-07,
+      "loss": 0.1,
+      "step": 3271
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2960423231124878,
+      "learning_rate": 3.2062647754137117e-07,
+      "loss": 0.1006,
+      "step": 3272
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.287566214799881,
+      "learning_rate": 3.1914893617021275e-07,
+      "loss": 0.1007,
+      "step": 3273
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.27013954520225525,
+      "learning_rate": 3.176713947990544e-07,
+      "loss": 0.0995,
+      "step": 3274
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.27947941422462463,
+      "learning_rate": 3.16193853427896e-07,
+      "loss": 0.0971,
+      "step": 3275
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.3190213441848755,
+      "learning_rate": 3.1471631205673757e-07,
+      "loss": 0.1043,
+      "step": 3276
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.30662423372268677,
+      "learning_rate": 3.132387706855792e-07,
+      "loss": 0.1017,
+      "step": 3277
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.30271029472351074,
+      "learning_rate": 3.1176122931442085e-07,
+      "loss": 0.106,
+      "step": 3278
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2864120602607727,
+      "learning_rate": 3.1028368794326243e-07,
+      "loss": 0.0991,
+      "step": 3279
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2892053425312042,
+      "learning_rate": 3.08806146572104e-07,
+      "loss": 0.1026,
+      "step": 3280
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.28111135959625244,
+      "learning_rate": 3.0732860520094566e-07,
+      "loss": 0.1001,
+      "step": 3281
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2661232650279999,
+      "learning_rate": 3.0585106382978724e-07,
+      "loss": 0.0944,
+      "step": 3282
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2917569577693939,
+      "learning_rate": 3.0437352245862883e-07,
+      "loss": 0.1024,
+      "step": 3283
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2828235328197479,
+      "learning_rate": 3.0289598108747047e-07,
+      "loss": 0.0975,
+      "step": 3284
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.28836554288864136,
+      "learning_rate": 3.0141843971631206e-07,
+      "loss": 0.1002,
+      "step": 3285
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.29511120915412903,
+      "learning_rate": 2.999408983451537e-07,
+      "loss": 0.1026,
+      "step": 3286
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.29372188448905945,
+      "learning_rate": 2.9846335697399534e-07,
+      "loss": 0.1013,
+      "step": 3287
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.28520238399505615,
+      "learning_rate": 2.969858156028369e-07,
+      "loss": 0.0998,
+      "step": 3288
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.30671435594558716,
+      "learning_rate": 2.955082742316785e-07,
+      "loss": 0.1038,
+      "step": 3289
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.32998016476631165,
+      "learning_rate": 2.9403073286052015e-07,
+      "loss": 0.0987,
+      "step": 3290
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.3044099509716034,
+      "learning_rate": 2.9255319148936174e-07,
+      "loss": 0.105,
+      "step": 3291
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.28645607829093933,
+      "learning_rate": 2.910756501182033e-07,
+      "loss": 0.1012,
+      "step": 3292
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2924303114414215,
+      "learning_rate": 2.8959810874704496e-07,
+      "loss": 0.0999,
+      "step": 3293
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2927074730396271,
+      "learning_rate": 2.8812056737588655e-07,
+      "loss": 0.1013,
+      "step": 3294
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2848796546459198,
+      "learning_rate": 2.8664302600472813e-07,
+      "loss": 0.0992,
+      "step": 3295
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2813242971897125,
+      "learning_rate": 2.851654846335698e-07,
+      "loss": 0.0955,
+      "step": 3296
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.3030376136302948,
+      "learning_rate": 2.8368794326241136e-07,
+      "loss": 0.1011,
+      "step": 3297
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.28521743416786194,
+      "learning_rate": 2.8221040189125295e-07,
+      "loss": 0.1042,
+      "step": 3298
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2935108244419098,
+      "learning_rate": 2.807328605200946e-07,
+      "loss": 0.0989,
+      "step": 3299
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.30593711137771606,
+      "learning_rate": 2.7925531914893617e-07,
+      "loss": 0.1001,
+      "step": 3300
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2947891056537628,
+      "learning_rate": 2.7777777777777776e-07,
+      "loss": 0.1009,
+      "step": 3301
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2941044270992279,
+      "learning_rate": 2.763002364066194e-07,
+      "loss": 0.1011,
+      "step": 3302
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2926792800426483,
+      "learning_rate": 2.7482269503546104e-07,
+      "loss": 0.0987,
+      "step": 3303
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2926498353481293,
+      "learning_rate": 2.733451536643026e-07,
+      "loss": 0.1034,
+      "step": 3304
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.28817710280418396,
+      "learning_rate": 2.7186761229314426e-07,
+      "loss": 0.0999,
+      "step": 3305
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.30679962038993835,
+      "learning_rate": 2.7039007092198585e-07,
+      "loss": 0.1015,
+      "step": 3306
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.27825838327407837,
+      "learning_rate": 2.6891252955082744e-07,
+      "loss": 0.0967,
+      "step": 3307
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2853482961654663,
+      "learning_rate": 2.674349881796691e-07,
+      "loss": 0.1012,
+      "step": 3308
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.3037569224834442,
+      "learning_rate": 2.6595744680851066e-07,
+      "loss": 0.1037,
+      "step": 3309
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.297299325466156,
+      "learning_rate": 2.6447990543735225e-07,
+      "loss": 0.0992,
+      "step": 3310
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2857692837715149,
+      "learning_rate": 2.630023640661939e-07,
+      "loss": 0.1023,
+      "step": 3311
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2810015082359314,
+      "learning_rate": 2.615248226950355e-07,
+      "loss": 0.0971,
+      "step": 3312
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.29854997992515564,
+      "learning_rate": 2.6004728132387706e-07,
+      "loss": 0.0999,
+      "step": 3313
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2918824851512909,
+      "learning_rate": 2.585697399527187e-07,
+      "loss": 0.0978,
+      "step": 3314
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2950204014778137,
+      "learning_rate": 2.570921985815603e-07,
+      "loss": 0.0969,
+      "step": 3315
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.29736363887786865,
+      "learning_rate": 2.556146572104019e-07,
+      "loss": 0.1007,
+      "step": 3316
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.30513665080070496,
+      "learning_rate": 2.541371158392435e-07,
+      "loss": 0.1005,
+      "step": 3317
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.28804028034210205,
+      "learning_rate": 2.526595744680851e-07,
+      "loss": 0.1006,
+      "step": 3318
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2950001060962677,
+      "learning_rate": 2.5118203309692674e-07,
+      "loss": 0.0993,
+      "step": 3319
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2907906472682953,
+      "learning_rate": 2.497044917257684e-07,
+      "loss": 0.1023,
+      "step": 3320
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2923189103603363,
+      "learning_rate": 2.4822695035460997e-07,
+      "loss": 0.1003,
+      "step": 3321
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.30383339524269104,
+      "learning_rate": 2.4674940898345155e-07,
+      "loss": 0.0975,
+      "step": 3322
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.30165350437164307,
+      "learning_rate": 2.452718676122932e-07,
+      "loss": 0.1047,
+      "step": 3323
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.295097678899765,
+      "learning_rate": 2.437943262411348e-07,
+      "loss": 0.1028,
+      "step": 3324
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2958454489707947,
+      "learning_rate": 2.4231678486997637e-07,
+      "loss": 0.1041,
+      "step": 3325
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.29658475518226624,
+      "learning_rate": 2.40839243498818e-07,
+      "loss": 0.0995,
+      "step": 3326
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.28040602803230286,
+      "learning_rate": 2.393617021276596e-07,
+      "loss": 0.0969,
+      "step": 3327
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2849106192588806,
+      "learning_rate": 2.378841607565012e-07,
+      "loss": 0.0983,
+      "step": 3328
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.27472877502441406,
+      "learning_rate": 2.364066193853428e-07,
+      "loss": 0.099,
+      "step": 3329
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.28400999307632446,
+      "learning_rate": 2.349290780141844e-07,
+      "loss": 0.0971,
+      "step": 3330
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2841739058494568,
+      "learning_rate": 2.3345153664302602e-07,
+      "loss": 0.0987,
+      "step": 3331
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2831608057022095,
+      "learning_rate": 2.3197399527186763e-07,
+      "loss": 0.098,
+      "step": 3332
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2966896891593933,
+      "learning_rate": 2.3049645390070922e-07,
+      "loss": 0.1055,
+      "step": 3333
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2822111248970032,
+      "learning_rate": 2.2901891252955083e-07,
+      "loss": 0.0982,
+      "step": 3334
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.30328476428985596,
+      "learning_rate": 2.2754137115839247e-07,
+      "loss": 0.1007,
+      "step": 3335
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.28947389125823975,
+      "learning_rate": 2.2606382978723408e-07,
+      "loss": 0.0967,
+      "step": 3336
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.3034358322620392,
+      "learning_rate": 2.2458628841607567e-07,
+      "loss": 0.1033,
+      "step": 3337
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.28458371758461,
+      "learning_rate": 2.2310874704491728e-07,
+      "loss": 0.099,
+      "step": 3338
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.29691368341445923,
+      "learning_rate": 2.216312056737589e-07,
+      "loss": 0.1022,
+      "step": 3339
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2907390296459198,
+      "learning_rate": 2.201536643026005e-07,
+      "loss": 0.0995,
+      "step": 3340
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.30089688301086426,
+      "learning_rate": 2.186761229314421e-07,
+      "loss": 0.0991,
+      "step": 3341
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.29386812448501587,
+      "learning_rate": 2.171985815602837e-07,
+      "loss": 0.0992,
+      "step": 3342
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.30215704441070557,
+      "learning_rate": 2.1572104018912532e-07,
+      "loss": 0.1022,
+      "step": 3343
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3089015483856201,
+      "learning_rate": 2.142434988179669e-07,
+      "loss": 0.1068,
+      "step": 3344
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.29736074805259705,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 0.1015,
+      "step": 3345
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.298833429813385,
+      "learning_rate": 2.1128841607565013e-07,
+      "loss": 0.1005,
+      "step": 3346
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.290835976600647,
+      "learning_rate": 2.0981087470449175e-07,
+      "loss": 0.1007,
+      "step": 3347
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.2878603935241699,
+      "learning_rate": 2.0833333333333333e-07,
+      "loss": 0.1024,
+      "step": 3348
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.29022616147994995,
+      "learning_rate": 2.0685579196217495e-07,
+      "loss": 0.0999,
+      "step": 3349
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.2884187400341034,
+      "learning_rate": 2.0537825059101656e-07,
+      "loss": 0.0976,
+      "step": 3350
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.29454505443573,
+      "learning_rate": 2.039007092198582e-07,
+      "loss": 0.1017,
+      "step": 3351
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3062115013599396,
+      "learning_rate": 2.0242316784869978e-07,
+      "loss": 0.1028,
+      "step": 3352
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.28808677196502686,
+      "learning_rate": 2.009456264775414e-07,
+      "loss": 0.0985,
+      "step": 3353
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.299273818731308,
+      "learning_rate": 1.99468085106383e-07,
+      "loss": 0.1016,
+      "step": 3354
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3078691065311432,
+      "learning_rate": 1.9799054373522462e-07,
+      "loss": 0.1012,
+      "step": 3355
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.28446710109710693,
+      "learning_rate": 1.965130023640662e-07,
+      "loss": 0.1,
+      "step": 3356
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.30727890133857727,
+      "learning_rate": 1.9503546099290782e-07,
+      "loss": 0.0982,
+      "step": 3357
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.3024684190750122,
+      "learning_rate": 1.9355791962174944e-07,
+      "loss": 0.1022,
+      "step": 3358
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.30678850412368774,
+      "learning_rate": 1.9208037825059102e-07,
+      "loss": 0.1022,
+      "step": 3359
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.27412235736846924,
+      "learning_rate": 1.9060283687943264e-07,
+      "loss": 0.0989,
+      "step": 3360
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2876332402229309,
+      "learning_rate": 1.8912529550827425e-07,
+      "loss": 0.1012,
+      "step": 3361
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.3078194558620453,
+      "learning_rate": 1.8764775413711584e-07,
+      "loss": 0.1024,
+      "step": 3362
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.293345183134079,
+      "learning_rate": 1.8617021276595745e-07,
+      "loss": 0.0945,
+      "step": 3363
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2867828607559204,
+      "learning_rate": 1.8469267139479906e-07,
+      "loss": 0.0984,
+      "step": 3364
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.29020947217941284,
+      "learning_rate": 1.8321513002364067e-07,
+      "loss": 0.1035,
+      "step": 3365
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.28689587116241455,
+      "learning_rate": 1.8173758865248226e-07,
+      "loss": 0.0994,
+      "step": 3366
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.28980886936187744,
+      "learning_rate": 1.802600472813239e-07,
+      "loss": 0.0991,
+      "step": 3367
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2807579040527344,
+      "learning_rate": 1.7878250591016551e-07,
+      "loss": 0.0957,
+      "step": 3368
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2856517732143402,
+      "learning_rate": 1.7730496453900713e-07,
+      "loss": 0.1002,
+      "step": 3369
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.28456971049308777,
+      "learning_rate": 1.758274231678487e-07,
+      "loss": 0.0994,
+      "step": 3370
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.3095431625843048,
+      "learning_rate": 1.7434988179669033e-07,
+      "loss": 0.1021,
+      "step": 3371
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.3035014271736145,
+      "learning_rate": 1.7287234042553194e-07,
+      "loss": 0.1011,
+      "step": 3372
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2828143835067749,
+      "learning_rate": 1.7139479905437355e-07,
+      "loss": 0.1008,
+      "step": 3373
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.29041945934295654,
+      "learning_rate": 1.6991725768321514e-07,
+      "loss": 0.0993,
+      "step": 3374
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.30088818073272705,
+      "learning_rate": 1.6843971631205675e-07,
+      "loss": 0.1012,
+      "step": 3375
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.29245930910110474,
+      "learning_rate": 1.6696217494089836e-07,
+      "loss": 0.0987,
+      "step": 3376
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.29735520482063293,
+      "learning_rate": 1.6548463356973995e-07,
+      "loss": 0.1008,
+      "step": 3377
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.27912163734436035,
+      "learning_rate": 1.6400709219858156e-07,
+      "loss": 0.098,
+      "step": 3378
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.29138869047164917,
+      "learning_rate": 1.6252955082742318e-07,
+      "loss": 0.0991,
+      "step": 3379
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2952117621898651,
+      "learning_rate": 1.610520094562648e-07,
+      "loss": 0.1032,
+      "step": 3380
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.27844762802124023,
+      "learning_rate": 1.5957446808510638e-07,
+      "loss": 0.0975,
+      "step": 3381
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.28062328696250916,
+      "learning_rate": 1.58096926713948e-07,
+      "loss": 0.1008,
+      "step": 3382
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2944590151309967,
+      "learning_rate": 1.566193853427896e-07,
+      "loss": 0.1021,
+      "step": 3383
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.3003998100757599,
+      "learning_rate": 1.5514184397163122e-07,
+      "loss": 0.1008,
+      "step": 3384
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.29515713453292847,
+      "learning_rate": 1.5366430260047283e-07,
+      "loss": 0.1031,
+      "step": 3385
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.27879828214645386,
+      "learning_rate": 1.5218676122931442e-07,
+      "loss": 0.1008,
+      "step": 3386
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.28016576170921326,
+      "learning_rate": 1.5070921985815603e-07,
+      "loss": 0.098,
+      "step": 3387
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2855539619922638,
+      "learning_rate": 1.4923167848699767e-07,
+      "loss": 0.0965,
+      "step": 3388
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.29991909861564636,
+      "learning_rate": 1.4775413711583925e-07,
+      "loss": 0.1018,
+      "step": 3389
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.30334773659706116,
+      "learning_rate": 1.4627659574468087e-07,
+      "loss": 0.099,
+      "step": 3390
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2943590581417084,
+      "learning_rate": 1.4479905437352248e-07,
+      "loss": 0.1039,
+      "step": 3391
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.29932430386543274,
+      "learning_rate": 1.4332151300236407e-07,
+      "loss": 0.1023,
+      "step": 3392
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.26667240262031555,
+      "learning_rate": 1.4184397163120568e-07,
+      "loss": 0.0921,
+      "step": 3393
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2913690507411957,
+      "learning_rate": 1.403664302600473e-07,
+      "loss": 0.1019,
+      "step": 3394
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2920913100242615,
+      "learning_rate": 1.3888888888888888e-07,
+      "loss": 0.1006,
+      "step": 3395
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.27679651975631714,
+      "learning_rate": 1.3741134751773052e-07,
+      "loss": 0.0979,
+      "step": 3396
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2970958650112152,
+      "learning_rate": 1.3593380614657213e-07,
+      "loss": 0.1008,
+      "step": 3397
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2868645191192627,
+      "learning_rate": 1.3445626477541372e-07,
+      "loss": 0.103,
+      "step": 3398
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.29753348231315613,
+      "learning_rate": 1.3297872340425533e-07,
+      "loss": 0.1043,
+      "step": 3399
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2839721143245697,
+      "learning_rate": 1.3150118203309694e-07,
+      "loss": 0.0954,
+      "step": 3400
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.3019815981388092,
+      "learning_rate": 1.3002364066193853e-07,
+      "loss": 0.1,
+      "step": 3401
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.300241619348526,
+      "learning_rate": 1.2854609929078014e-07,
+      "loss": 0.1007,
+      "step": 3402
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2881091833114624,
+      "learning_rate": 1.2706855791962176e-07,
+      "loss": 0.1005,
+      "step": 3403
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.29561030864715576,
+      "learning_rate": 1.2559101654846337e-07,
+      "loss": 0.1017,
+      "step": 3404
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.30144816637039185,
+      "learning_rate": 1.2411347517730498e-07,
+      "loss": 0.099,
+      "step": 3405
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2764350175857544,
+      "learning_rate": 1.226359338061466e-07,
+      "loss": 0.0972,
+      "step": 3406
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3121400475502014,
+      "learning_rate": 1.2115839243498818e-07,
+      "loss": 0.104,
+      "step": 3407
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2967803180217743,
+      "learning_rate": 1.196808510638298e-07,
+      "loss": 0.1037,
+      "step": 3408
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2955235242843628,
+      "learning_rate": 1.182033096926714e-07,
+      "loss": 0.0997,
+      "step": 3409
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.292117178440094,
+      "learning_rate": 1.1672576832151301e-07,
+      "loss": 0.1008,
+      "step": 3410
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.30653539299964905,
+      "learning_rate": 1.1524822695035461e-07,
+      "loss": 0.1013,
+      "step": 3411
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.29063141345977783,
+      "learning_rate": 1.1377068557919623e-07,
+      "loss": 0.0992,
+      "step": 3412
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2725222110748291,
+      "learning_rate": 1.1229314420803783e-07,
+      "loss": 0.0976,
+      "step": 3413
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.3076043128967285,
+      "learning_rate": 1.1081560283687945e-07,
+      "loss": 0.1022,
+      "step": 3414
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.3010319769382477,
+      "learning_rate": 1.0933806146572105e-07,
+      "loss": 0.1013,
+      "step": 3415
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2830990254878998,
+      "learning_rate": 1.0786052009456266e-07,
+      "loss": 0.0992,
+      "step": 3416
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.31569141149520874,
+      "learning_rate": 1.0638297872340426e-07,
+      "loss": 0.1057,
+      "step": 3417
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.27784407138824463,
+      "learning_rate": 1.0490543735224587e-07,
+      "loss": 0.1002,
+      "step": 3418
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.30672913789749146,
+      "learning_rate": 1.0342789598108747e-07,
+      "loss": 0.0981,
+      "step": 3419
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2799972593784332,
+      "learning_rate": 1.019503546099291e-07,
+      "loss": 0.0969,
+      "step": 3420
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2792622745037079,
+      "learning_rate": 1.004728132387707e-07,
+      "loss": 0.0987,
+      "step": 3421
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2826020121574402,
+      "learning_rate": 9.899527186761231e-08,
+      "loss": 0.0988,
+      "step": 3422
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.29048794507980347,
+      "learning_rate": 9.751773049645391e-08,
+      "loss": 0.0969,
+      "step": 3423
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2992827594280243,
+      "learning_rate": 9.604018912529551e-08,
+      "loss": 0.1026,
+      "step": 3424
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2988862097263336,
+      "learning_rate": 9.456264775413712e-08,
+      "loss": 0.1032,
+      "step": 3425
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.29507583379745483,
+      "learning_rate": 9.308510638297872e-08,
+      "loss": 0.0986,
+      "step": 3426
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.30378058552742004,
+      "learning_rate": 9.160756501182034e-08,
+      "loss": 0.1008,
+      "step": 3427
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.3084165155887604,
+      "learning_rate": 9.013002364066195e-08,
+      "loss": 0.1021,
+      "step": 3428
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.2972557246685028,
+      "learning_rate": 8.865248226950356e-08,
+      "loss": 0.1009,
+      "step": 3429
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.2884296476840973,
+      "learning_rate": 8.717494089834516e-08,
+      "loss": 0.0957,
+      "step": 3430
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.28628695011138916,
+      "learning_rate": 8.569739952718678e-08,
+      "loss": 0.0988,
+      "step": 3431
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.28985264897346497,
+      "learning_rate": 8.421985815602838e-08,
+      "loss": 0.0993,
+      "step": 3432
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.2933429479598999,
+      "learning_rate": 8.274231678486998e-08,
+      "loss": 0.0996,
+      "step": 3433
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.29415765404701233,
+      "learning_rate": 8.126477541371159e-08,
+      "loss": 0.1018,
+      "step": 3434
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.30490729212760925,
+      "learning_rate": 7.978723404255319e-08,
+      "loss": 0.0994,
+      "step": 3435
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.3100927472114563,
+      "learning_rate": 7.83096926713948e-08,
+      "loss": 0.103,
+      "step": 3436
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2980824112892151,
+      "learning_rate": 7.683215130023641e-08,
+      "loss": 0.1006,
+      "step": 3437
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.30111047625541687,
+      "learning_rate": 7.535460992907801e-08,
+      "loss": 0.1011,
+      "step": 3438
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2920394837856293,
+      "learning_rate": 7.387706855791963e-08,
+      "loss": 0.1034,
+      "step": 3439
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2894846498966217,
+      "learning_rate": 7.239952718676124e-08,
+      "loss": 0.1012,
+      "step": 3440
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2791222035884857,
+      "learning_rate": 7.092198581560284e-08,
+      "loss": 0.0992,
+      "step": 3441
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.28812167048454285,
+      "learning_rate": 6.944444444444444e-08,
+      "loss": 0.1,
+      "step": 3442
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.29871422052383423,
+      "learning_rate": 6.796690307328607e-08,
+      "loss": 0.1033,
+      "step": 3443
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2987375855445862,
+      "learning_rate": 6.648936170212767e-08,
+      "loss": 0.1027,
+      "step": 3444
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2827816307544708,
+      "learning_rate": 6.501182033096927e-08,
+      "loss": 0.0965,
+      "step": 3445
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2952755093574524,
+      "learning_rate": 6.353427895981088e-08,
+      "loss": 0.0991,
+      "step": 3446
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2988056242465973,
+      "learning_rate": 6.205673758865249e-08,
+      "loss": 0.098,
+      "step": 3447
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.3044067919254303,
+      "learning_rate": 6.057919621749409e-08,
+      "loss": 0.1006,
+      "step": 3448
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2961951792240143,
+      "learning_rate": 5.91016548463357e-08,
+      "loss": 0.1008,
+      "step": 3449
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.29294413328170776,
+      "learning_rate": 5.7624113475177304e-08,
+      "loss": 0.103,
+      "step": 3450
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2907746136188507,
+      "learning_rate": 5.614657210401892e-08,
+      "loss": 0.1013,
+      "step": 3451
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2909153699874878,
+      "learning_rate": 5.4669030732860524e-08,
+      "loss": 0.098,
+      "step": 3452
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2770833969116211,
+      "learning_rate": 5.319148936170213e-08,
+      "loss": 0.0965,
+      "step": 3453
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2828409671783447,
+      "learning_rate": 5.1713947990543736e-08,
+      "loss": 0.0973,
+      "step": 3454
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.29271915555000305,
+      "learning_rate": 5.023640661938535e-08,
+      "loss": 0.0975,
+      "step": 3455
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2918560802936554,
+      "learning_rate": 4.8758865248226956e-08,
+      "loss": 0.1015,
+      "step": 3456
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.29700955748558044,
+      "learning_rate": 4.728132387706856e-08,
+      "loss": 0.0986,
+      "step": 3457
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.29509779810905457,
+      "learning_rate": 4.580378250591017e-08,
+      "loss": 0.1023,
+      "step": 3458
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.295709490776062,
+      "learning_rate": 4.432624113475178e-08,
+      "loss": 0.102,
+      "step": 3459
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2859545648097992,
+      "learning_rate": 4.284869976359339e-08,
+      "loss": 0.1011,
+      "step": 3460
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.3015943765640259,
+      "learning_rate": 4.137115839243499e-08,
+      "loss": 0.1056,
+      "step": 3461
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.27578216791152954,
+      "learning_rate": 3.9893617021276594e-08,
+      "loss": 0.1023,
+      "step": 3462
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2955705225467682,
+      "learning_rate": 3.841607565011821e-08,
+      "loss": 0.0976,
+      "step": 3463
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2894452214241028,
+      "learning_rate": 3.6938534278959814e-08,
+      "loss": 0.0994,
+      "step": 3464
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.29981255531311035,
+      "learning_rate": 3.546099290780142e-08,
+      "loss": 0.0992,
+      "step": 3465
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.324081152677536,
+      "learning_rate": 3.398345153664303e-08,
+      "loss": 0.1012,
+      "step": 3466
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.27969592809677124,
+      "learning_rate": 3.250591016548463e-08,
+      "loss": 0.0982,
+      "step": 3467
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.27975597977638245,
+      "learning_rate": 3.1028368794326246e-08,
+      "loss": 0.0949,
+      "step": 3468
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.29814818501472473,
+      "learning_rate": 2.955082742316785e-08,
+      "loss": 0.1045,
+      "step": 3469
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.28994208574295044,
+      "learning_rate": 2.807328605200946e-08,
+      "loss": 0.0993,
+      "step": 3470
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.28579577803611755,
+      "learning_rate": 2.6595744680851065e-08,
+      "loss": 0.1002,
+      "step": 3471
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2850336730480194,
+      "learning_rate": 2.5118203309692675e-08,
+      "loss": 0.0983,
+      "step": 3472
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.29380032420158386,
+      "learning_rate": 2.364066193853428e-08,
+      "loss": 0.0982,
+      "step": 3473
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.29162782430648804,
+      "learning_rate": 2.216312056737589e-08,
+      "loss": 0.1003,
+      "step": 3474
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.29252490401268005,
+      "learning_rate": 2.0685579196217494e-08,
+      "loss": 0.1007,
+      "step": 3475
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.3083917498588562,
+      "learning_rate": 1.9208037825059104e-08,
+      "loss": 0.0994,
+      "step": 3476
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.28565075993537903,
+      "learning_rate": 1.773049645390071e-08,
+      "loss": 0.0987,
+      "step": 3477
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2891337275505066,
+      "learning_rate": 1.6252955082742316e-08,
+      "loss": 0.101,
+      "step": 3478
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.28994470834732056,
+      "learning_rate": 1.4775413711583924e-08,
+      "loss": 0.0981,
+      "step": 3479
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.28105318546295166,
+      "learning_rate": 1.3297872340425533e-08,
+      "loss": 0.0976,
+      "step": 3480
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.29527804255485535,
+      "learning_rate": 1.182033096926714e-08,
+      "loss": 0.0964,
+      "step": 3481
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.28097090125083923,
+      "learning_rate": 1.0342789598108747e-08,
+      "loss": 0.0987,
+      "step": 3482
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.3093743920326233,
+      "learning_rate": 8.865248226950355e-09,
+      "loss": 0.1027,
+      "step": 3483
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2985726296901703,
+      "learning_rate": 7.387706855791962e-09,
+      "loss": 0.1021,
+      "step": 3484
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2969771921634674,
+      "learning_rate": 5.91016548463357e-09,
+      "loss": 0.0995,
+      "step": 3485
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2804962694644928,
+      "learning_rate": 4.4326241134751775e-09,
+      "loss": 0.0968,
+      "step": 3486
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2925645112991333,
+      "learning_rate": 2.955082742316785e-09,
+      "loss": 0.1013,
+      "step": 3487
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.28656816482543945,
+      "learning_rate": 1.4775413711583926e-09,
+      "loss": 0.1008,
+      "step": 3488
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2871009409427643,
+      "learning_rate": 0.0,
+      "loss": 0.0981,
+      "step": 3489
+    },
+    {
+      "epoch": 3.0,
+      "step": 3489,
+      "total_flos": 1.9763607241333146e+18,
+      "train_loss": 0.12524562087321084,
+      "train_runtime": 16413.9429,
+      "train_samples_per_second": 54.406,
+      "train_steps_per_second": 0.213
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3489,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 3000,
+  "total_flos": 1.9763607241333146e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}