diff --git "a/Baseline + Grid/trainer_state.json" "b/Baseline + Grid/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/Baseline + Grid/trainer_state.json"
@@ -0,0 +1,53440 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 7630,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9762861728668213,
+      "learning_rate": 4.366812227074236e-08,
+      "loss": 1.3072,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.928385019302368,
+      "learning_rate": 8.733624454148472e-08,
+      "loss": 1.2834,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.775402069091797,
+      "learning_rate": 1.3100436681222707e-07,
+      "loss": 1.3959,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.31943416595459,
+      "learning_rate": 1.7467248908296944e-07,
+      "loss": 1.4208,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.607449531555176,
+      "learning_rate": 2.183406113537118e-07,
+      "loss": 1.3888,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.493844032287598,
+      "learning_rate": 2.6200873362445414e-07,
+      "loss": 1.3983,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.53220272064209,
+      "learning_rate": 3.056768558951965e-07,
+      "loss": 1.4656,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.078151226043701,
+      "learning_rate": 3.4934497816593887e-07,
+      "loss": 1.2953,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.353097915649414,
+      "learning_rate": 3.930131004366813e-07,
+      "loss": 1.3304,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.830784797668457,
+      "learning_rate": 4.366812227074236e-07,
+      "loss": 1.3677,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9784624576568604,
+      "learning_rate": 4.80349344978166e-07,
+      "loss": 1.3001,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.703521728515625,
+      "learning_rate": 5.240174672489083e-07,
+      "loss": 1.4074,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.004316329956055,
+      "learning_rate": 5.676855895196507e-07,
+      "loss": 1.3956,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.621962547302246,
+      "learning_rate": 6.11353711790393e-07,
+      "loss": 1.3968,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.088995933532715,
+      "learning_rate": 6.550218340611354e-07,
+      "loss": 1.4246,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9195985794067383,
+      "learning_rate": 6.986899563318777e-07,
+      "loss": 1.2995,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.311702728271484,
+      "learning_rate": 7.423580786026202e-07,
+      "loss": 1.3695,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.357166767120361,
+      "learning_rate": 7.860262008733626e-07,
+      "loss": 1.4061,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.77955436706543,
+      "learning_rate": 8.296943231441049e-07,
+      "loss": 1.318,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.139078140258789,
+      "learning_rate": 8.733624454148472e-07,
+      "loss": 1.3356,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.619923114776611,
+      "learning_rate": 9.170305676855896e-07,
+      "loss": 1.3937,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5499277114868164,
+      "learning_rate": 9.60698689956332e-07,
+      "loss": 1.2707,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.90549898147583,
+      "learning_rate": 1.0043668122270742e-06,
+      "loss": 1.3205,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.9104132652282715,
+      "learning_rate": 1.0480349344978166e-06,
+      "loss": 1.3233,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.417962551116943,
+      "learning_rate": 1.091703056768559e-06,
+      "loss": 1.3047,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4479217529296875,
+      "learning_rate": 1.1353711790393014e-06,
+      "loss": 1.26,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.029177188873291,
+      "learning_rate": 1.1790393013100437e-06,
+      "loss": 1.328,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.560856342315674,
+      "learning_rate": 1.222707423580786e-06,
+      "loss": 1.237,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.368199586868286,
+      "learning_rate": 1.2663755458515283e-06,
+      "loss": 1.2749,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9247214794158936,
+      "learning_rate": 1.3100436681222709e-06,
+      "loss": 1.2915,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.679678201675415,
+      "learning_rate": 1.3537117903930134e-06,
+      "loss": 1.2649,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5161683559417725,
+      "learning_rate": 1.3973799126637555e-06,
+      "loss": 1.2133,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8029417991638184,
+      "learning_rate": 1.441048034934498e-06,
+      "loss": 1.2347,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.292768955230713,
+      "learning_rate": 1.4847161572052403e-06,
+      "loss": 1.2573,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.25894832611084,
+      "learning_rate": 1.5283842794759826e-06,
+      "loss": 1.2882,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.507974624633789,
+      "learning_rate": 1.5720524017467252e-06,
+      "loss": 1.2508,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.151370048522949,
+      "learning_rate": 1.6157205240174673e-06,
+      "loss": 1.2426,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.437225103378296,
+      "learning_rate": 1.6593886462882098e-06,
+      "loss": 1.2701,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.00418758392334,
+      "learning_rate": 1.703056768558952e-06,
+      "loss": 1.2608,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8030459880828857,
+      "learning_rate": 1.7467248908296944e-06,
+      "loss": 1.2405,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2376179695129395,
+      "learning_rate": 1.790393013100437e-06,
+      "loss": 1.2641,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.188955545425415,
+      "learning_rate": 1.8340611353711792e-06,
+      "loss": 1.2448,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2253551483154297,
+      "learning_rate": 1.8777292576419216e-06,
+      "loss": 1.261,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.187809705734253,
+      "learning_rate": 1.921397379912664e-06,
+      "loss": 1.2108,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0644752979278564,
+      "learning_rate": 1.965065502183406e-06,
+      "loss": 1.1871,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4131577014923096,
+      "learning_rate": 2.0087336244541485e-06,
+      "loss": 1.2484,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.983105182647705,
+      "learning_rate": 2.052401746724891e-06,
+      "loss": 1.1817,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.319720983505249,
+      "learning_rate": 2.096069868995633e-06,
+      "loss": 1.2697,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9813400506973267,
+      "learning_rate": 2.1397379912663756e-06,
+      "loss": 1.2087,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9288873672485352,
+      "learning_rate": 2.183406113537118e-06,
+      "loss": 1.1757,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9189386367797852,
+      "learning_rate": 2.2270742358078603e-06,
+      "loss": 1.1995,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7903482913970947,
+      "learning_rate": 2.270742358078603e-06,
+      "loss": 1.1791,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7450220584869385,
+      "learning_rate": 2.3144104803493453e-06,
+      "loss": 1.1537,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8957092761993408,
+      "learning_rate": 2.3580786026200874e-06,
+      "loss": 1.2431,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6985628604888916,
+      "learning_rate": 2.40174672489083e-06,
+      "loss": 1.2095,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6078908443450928,
+      "learning_rate": 2.445414847161572e-06,
+      "loss": 1.2007,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.646765947341919,
+      "learning_rate": 2.4890829694323146e-06,
+      "loss": 1.1651,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7008386850357056,
+      "learning_rate": 2.5327510917030567e-06,
+      "loss": 1.1238,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.592303991317749,
+      "learning_rate": 2.576419213973799e-06,
+      "loss": 1.1572,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3370763063430786,
+      "learning_rate": 2.6200873362445417e-06,
+      "loss": 1.2389,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5988514423370361,
+      "learning_rate": 2.6637554585152842e-06,
+      "loss": 1.14,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2982804775238037,
+      "learning_rate": 2.7074235807860268e-06,
+      "loss": 1.2657,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6043944358825684,
+      "learning_rate": 2.7510917030567684e-06,
+      "loss": 1.1623,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4861419200897217,
+      "learning_rate": 2.794759825327511e-06,
+      "loss": 1.1524,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4339959621429443,
+      "learning_rate": 2.8384279475982535e-06,
+      "loss": 1.0953,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4928590059280396,
+      "learning_rate": 2.882096069868996e-06,
+      "loss": 1.1422,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4441134929656982,
+      "learning_rate": 2.9257641921397385e-06,
+      "loss": 1.1116,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4727232456207275,
+      "learning_rate": 2.9694323144104806e-06,
+      "loss": 1.1517,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4340986013412476,
+      "learning_rate": 3.0131004366812227e-06,
+      "loss": 1.1792,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5223362445831299,
+      "learning_rate": 3.0567685589519653e-06,
+      "loss": 1.0746,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3745989799499512,
+      "learning_rate": 3.100436681222708e-06,
+      "loss": 1.1206,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.343002200126648,
+      "learning_rate": 3.1441048034934503e-06,
+      "loss": 1.0916,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4382805824279785,
+      "learning_rate": 3.1877729257641924e-06,
+      "loss": 1.0965,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5478129386901855,
+      "learning_rate": 3.2314410480349345e-06,
+      "loss": 1.0883,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2844624519348145,
+      "learning_rate": 3.275109170305677e-06,
+      "loss": 1.1144,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3805232048034668,
+      "learning_rate": 3.3187772925764196e-06,
+      "loss": 1.0719,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.429984211921692,
+      "learning_rate": 3.362445414847162e-06,
+      "loss": 1.0415,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1679091453552246,
+      "learning_rate": 3.406113537117904e-06,
+      "loss": 1.2457,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3966615200042725,
+      "learning_rate": 3.4497816593886467e-06,
+      "loss": 1.1062,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3488271236419678,
+      "learning_rate": 3.493449781659389e-06,
+      "loss": 1.1267,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.092294692993164,
+      "learning_rate": 3.5371179039301313e-06,
+      "loss": 1.244,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0650908946990967,
+      "learning_rate": 3.580786026200874e-06,
+      "loss": 1.2474,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9599583745002747,
+      "learning_rate": 3.624454148471616e-06,
+      "loss": 1.2461,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5420949459075928,
+      "learning_rate": 3.6681222707423585e-06,
+      "loss": 1.0605,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4207594394683838,
+      "learning_rate": 3.7117903930131006e-06,
+      "loss": 1.1012,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4431995153427124,
+      "learning_rate": 3.755458515283843e-06,
+      "loss": 1.1063,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4478545188903809,
+      "learning_rate": 3.7991266375545856e-06,
+      "loss": 1.1121,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4228595495224,
+      "learning_rate": 3.842794759825328e-06,
+      "loss": 1.0681,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8995394110679626,
+      "learning_rate": 3.88646288209607e-06,
+      "loss": 1.2622,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4481686353683472,
+      "learning_rate": 3.930131004366812e-06,
+      "loss": 1.0924,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8588674068450928,
+      "learning_rate": 3.9737991266375545e-06,
+      "loss": 1.2573,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.382635474205017,
+      "learning_rate": 4.017467248908297e-06,
+      "loss": 0.9936,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4126375913619995,
+      "learning_rate": 4.0611353711790395e-06,
+      "loss": 1.1012,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3713363409042358,
+      "learning_rate": 4.104803493449782e-06,
+      "loss": 1.1338,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3677173852920532,
+      "learning_rate": 4.1484716157205246e-06,
+      "loss": 1.1283,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4161489009857178,
+      "learning_rate": 4.192139737991266e-06,
+      "loss": 1.0206,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8825482726097107,
+      "learning_rate": 4.235807860262009e-06,
+      "loss": 1.2173,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8655318021774292,
+      "learning_rate": 4.279475982532751e-06,
+      "loss": 1.2311,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4151586294174194,
+      "learning_rate": 4.323144104803494e-06,
+      "loss": 1.0621,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7219696044921875,
+      "learning_rate": 4.366812227074236e-06,
+      "loss": 1.2159,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7416970133781433,
+      "learning_rate": 4.410480349344978e-06,
+      "loss": 1.2412,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5807145833969116,
+      "learning_rate": 4.4541484716157205e-06,
+      "loss": 1.0211,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5095086097717285,
+      "learning_rate": 4.497816593886463e-06,
+      "loss": 1.058,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.430260181427002,
+      "learning_rate": 4.541484716157206e-06,
+      "loss": 1.0727,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8102776408195496,
+      "learning_rate": 4.585152838427948e-06,
+      "loss": 1.2255,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5629417896270752,
+      "learning_rate": 4.628820960698691e-06,
+      "loss": 1.0714,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.399341106414795,
+      "learning_rate": 4.672489082969432e-06,
+      "loss": 1.0665,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4884237051010132,
+      "learning_rate": 4.716157205240175e-06,
+      "loss": 1.0879,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3632824420928955,
+      "learning_rate": 4.759825327510917e-06,
+      "loss": 1.0531,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3362115621566772,
+      "learning_rate": 4.80349344978166e-06,
+      "loss": 1.0609,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3452637195587158,
+      "learning_rate": 4.847161572052402e-06,
+      "loss": 1.0505,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.347224473953247,
+      "learning_rate": 4.890829694323144e-06,
+      "loss": 1.0359,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4778406620025635,
+      "learning_rate": 4.934497816593887e-06,
+      "loss": 1.055,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3513973951339722,
+      "learning_rate": 4.978165938864629e-06,
+      "loss": 1.0534,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4141337871551514,
+      "learning_rate": 5.021834061135371e-06,
+      "loss": 1.0657,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4096206426620483,
+      "learning_rate": 5.065502183406113e-06,
+      "loss": 1.0462,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3665956258773804,
+      "learning_rate": 5.109170305676856e-06,
+      "loss": 1.039,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5227785110473633,
+      "learning_rate": 5.152838427947598e-06,
+      "loss": 1.0266,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5189164876937866,
+      "learning_rate": 5.196506550218341e-06,
+      "loss": 1.0489,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9776817560195923,
+      "learning_rate": 5.2401746724890834e-06,
+      "loss": 1.2244,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3285025358200073,
+      "learning_rate": 5.283842794759826e-06,
+      "loss": 1.0407,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7842618227005005,
+      "learning_rate": 5.3275109170305685e-06,
+      "loss": 1.2022,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4452645778656006,
+      "learning_rate": 5.371179039301311e-06,
+      "loss": 1.0338,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3620645999908447,
+      "learning_rate": 5.4148471615720535e-06,
+      "loss": 1.0548,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.422279715538025,
+      "learning_rate": 5.458515283842796e-06,
+      "loss": 1.0879,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.462520956993103,
+      "learning_rate": 5.502183406113537e-06,
+      "loss": 1.0766,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9695852398872375,
+      "learning_rate": 5.545851528384279e-06,
+      "loss": 1.2056,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4943437576293945,
+      "learning_rate": 5.589519650655022e-06,
+      "loss": 1.03,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8334753513336182,
+      "learning_rate": 5.6331877729257645e-06,
+      "loss": 1.2159,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.424392819404602,
+      "learning_rate": 5.676855895196507e-06,
+      "loss": 1.0004,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3460170030593872,
+      "learning_rate": 5.7205240174672495e-06,
+      "loss": 1.0228,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3665870428085327,
+      "learning_rate": 5.764192139737992e-06,
+      "loss": 1.0094,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4244742393493652,
+      "learning_rate": 5.8078602620087346e-06,
+      "loss": 1.0207,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4888854026794434,
+      "learning_rate": 5.851528384279477e-06,
+      "loss": 1.0885,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0823116302490234,
+      "learning_rate": 5.89519650655022e-06,
+      "loss": 1.204,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.505212426185608,
+      "learning_rate": 5.938864628820961e-06,
+      "loss": 1.0087,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.417958378791809,
+      "learning_rate": 5.982532751091703e-06,
+      "loss": 1.0387,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3787816762924194,
+      "learning_rate": 6.0262008733624455e-06,
+      "loss": 0.9739,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4174026250839233,
+      "learning_rate": 6.069868995633188e-06,
+      "loss": 1.0243,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.440824031829834,
+      "learning_rate": 6.1135371179039305e-06,
+      "loss": 1.0182,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8154327869415283,
+      "learning_rate": 6.157205240174673e-06,
+      "loss": 1.2115,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8038000464439392,
+      "learning_rate": 6.200873362445416e-06,
+      "loss": 1.2577,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6156922578811646,
+      "learning_rate": 6.244541484716158e-06,
+      "loss": 1.028,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6841146945953369,
+      "learning_rate": 6.288209606986901e-06,
+      "loss": 1.2034,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5045748949050903,
+      "learning_rate": 6.331877729257643e-06,
+      "loss": 1.0562,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8079639673233032,
+      "learning_rate": 6.375545851528385e-06,
+      "loss": 1.1918,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4941452741622925,
+      "learning_rate": 6.419213973799127e-06,
+      "loss": 1.021,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5231505632400513,
+      "learning_rate": 6.462882096069869e-06,
+      "loss": 1.0227,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3450863361358643,
+      "learning_rate": 6.5065502183406116e-06,
+      "loss": 0.9768,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7153328061103821,
+      "learning_rate": 6.550218340611354e-06,
+      "loss": 1.2175,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4517680406570435,
+      "learning_rate": 6.593886462882097e-06,
+      "loss": 0.9887,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.397452473640442,
+      "learning_rate": 6.637554585152839e-06,
+      "loss": 1.0124,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7740655541419983,
+      "learning_rate": 6.681222707423582e-06,
+      "loss": 1.2164,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3531179428100586,
+      "learning_rate": 6.724890829694324e-06,
+      "loss": 1.0296,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3414291143417358,
+      "learning_rate": 6.768558951965066e-06,
+      "loss": 0.9966,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.406220555305481,
+      "learning_rate": 6.812227074235808e-06,
+      "loss": 1.0734,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6642008423805237,
+      "learning_rate": 6.855895196506551e-06,
+      "loss": 1.1811,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3642995357513428,
+      "learning_rate": 6.8995633187772934e-06,
+      "loss": 0.998,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3557102680206299,
+      "learning_rate": 6.943231441048035e-06,
+      "loss": 1.0239,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7368056178092957,
+      "learning_rate": 6.986899563318778e-06,
+      "loss": 1.2004,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6497008204460144,
+      "learning_rate": 7.03056768558952e-06,
+      "loss": 1.1764,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.445486068725586,
+      "learning_rate": 7.074235807860263e-06,
+      "loss": 0.9697,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7167004346847534,
+      "learning_rate": 7.117903930131005e-06,
+      "loss": 1.2071,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4675872325897217,
+      "learning_rate": 7.161572052401748e-06,
+      "loss": 0.9834,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.411812663078308,
+      "learning_rate": 7.205240174672489e-06,
+      "loss": 1.0549,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.813985288143158,
+      "learning_rate": 7.248908296943232e-06,
+      "loss": 1.2071,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4216095209121704,
+      "learning_rate": 7.2925764192139745e-06,
+      "loss": 1.0343,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4082406759262085,
+      "learning_rate": 7.336244541484717e-06,
+      "loss": 0.9477,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3392952680587769,
+      "learning_rate": 7.3799126637554595e-06,
+      "loss": 0.9617,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4519870281219482,
+      "learning_rate": 7.423580786026201e-06,
+      "loss": 1.0677,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4016577005386353,
+      "learning_rate": 7.467248908296944e-06,
+      "loss": 0.9874,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.471930980682373,
+      "learning_rate": 7.510917030567686e-06,
+      "loss": 1.01,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9688252806663513,
+      "learning_rate": 7.554585152838429e-06,
+      "loss": 1.2218,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7991207242012024,
+      "learning_rate": 7.598253275109171e-06,
+      "loss": 1.1844,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4857248067855835,
+      "learning_rate": 7.641921397379913e-06,
+      "loss": 1.0273,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3977161645889282,
+      "learning_rate": 7.685589519650655e-06,
+      "loss": 1.0036,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.50105619430542,
+      "learning_rate": 7.729257641921398e-06,
+      "loss": 0.999,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5781958103179932,
+      "learning_rate": 7.77292576419214e-06,
+      "loss": 0.9736,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4932281970977783,
+      "learning_rate": 7.816593886462883e-06,
+      "loss": 0.9966,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1842398643493652,
+      "learning_rate": 7.860262008733624e-06,
+      "loss": 1.2013,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4550504684448242,
+      "learning_rate": 7.903930131004366e-06,
+      "loss": 0.8988,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7951863408088684,
+      "learning_rate": 7.947598253275109e-06,
+      "loss": 1.1957,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5487887859344482,
+      "learning_rate": 7.991266375545851e-06,
+      "loss": 1.025,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8496047258377075,
+      "learning_rate": 8.034934497816594e-06,
+      "loss": 1.1756,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8684196472167969,
+      "learning_rate": 8.078602620087337e-06,
+      "loss": 1.211,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.552699089050293,
+      "learning_rate": 8.122270742358079e-06,
+      "loss": 1.0063,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4625284671783447,
+      "learning_rate": 8.165938864628822e-06,
+      "loss": 1.0637,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4318928718566895,
+      "learning_rate": 8.209606986899564e-06,
+      "loss": 0.9807,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5218669176101685,
+      "learning_rate": 8.253275109170307e-06,
+      "loss": 0.9982,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5467698574066162,
+      "learning_rate": 8.296943231441049e-06,
+      "loss": 0.965,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.563277244567871,
+      "learning_rate": 8.34061135371179e-06,
+      "loss": 1.1843,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4803440570831299,
+      "learning_rate": 8.384279475982532e-06,
+      "loss": 1.0306,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5081746578216553,
+      "learning_rate": 8.427947598253275e-06,
+      "loss": 1.0158,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.477040410041809,
+      "learning_rate": 8.471615720524018e-06,
+      "loss": 1.0196,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3506752252578735,
+      "learning_rate": 8.51528384279476e-06,
+      "loss": 0.9705,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9146580696105957,
+      "learning_rate": 8.558951965065503e-06,
+      "loss": 1.1934,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5538229942321777,
+      "learning_rate": 8.602620087336245e-06,
+      "loss": 0.9665,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4603041410446167,
+      "learning_rate": 8.646288209606988e-06,
+      "loss": 0.9655,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7806007862091064,
+      "learning_rate": 8.68995633187773e-06,
+      "loss": 1.1812,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.471705436706543,
+      "learning_rate": 8.733624454148473e-06,
+      "loss": 0.9646,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5152798891067505,
+      "learning_rate": 8.777292576419215e-06,
+      "loss": 0.9699,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3325138092041016,
+      "learning_rate": 8.820960698689956e-06,
+      "loss": 1.0274,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.779484748840332,
+      "learning_rate": 8.864628820960699e-06,
+      "loss": 1.159,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7416744828224182,
+      "learning_rate": 8.908296943231441e-06,
+      "loss": 1.1805,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.675710916519165,
+      "learning_rate": 8.951965065502184e-06,
+      "loss": 0.9922,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7615010142326355,
+      "learning_rate": 8.995633187772926e-06,
+      "loss": 1.1655,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5184879302978516,
+      "learning_rate": 9.039301310043669e-06,
+      "loss": 0.9916,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4446998834609985,
+      "learning_rate": 9.082969432314411e-06,
+      "loss": 0.9714,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.319512963294983,
+      "learning_rate": 9.126637554585154e-06,
+      "loss": 0.9861,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8355891108512878,
+      "learning_rate": 9.170305676855896e-06,
+      "loss": 1.2058,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7832462191581726,
+      "learning_rate": 9.213973799126639e-06,
+      "loss": 1.1851,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6658129692077637,
+      "learning_rate": 9.257641921397381e-06,
+      "loss": 1.1694,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.637667179107666,
+      "learning_rate": 9.301310043668122e-06,
+      "loss": 0.9749,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5358294248580933,
+      "learning_rate": 9.344978165938865e-06,
+      "loss": 0.9467,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5597167015075684,
+      "learning_rate": 9.388646288209607e-06,
+      "loss": 1.0266,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5246849060058594,
+      "learning_rate": 9.43231441048035e-06,
+      "loss": 0.9364,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4327346086502075,
+      "learning_rate": 9.475982532751092e-06,
+      "loss": 1.0005,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4170622825622559,
+      "learning_rate": 9.519650655021835e-06,
+      "loss": 0.9711,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4605021476745605,
+      "learning_rate": 9.563318777292577e-06,
+      "loss": 0.9509,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3937458992004395,
+      "learning_rate": 9.60698689956332e-06,
+      "loss": 0.9308,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4637393951416016,
+      "learning_rate": 9.650655021834062e-06,
+      "loss": 0.9696,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6090011596679688,
+      "learning_rate": 9.694323144104805e-06,
+      "loss": 0.9713,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3515111207962036,
+      "learning_rate": 9.737991266375547e-06,
+      "loss": 0.9793,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4589847326278687,
+      "learning_rate": 9.781659388646288e-06,
+      "loss": 0.9644,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.565245270729065,
+      "learning_rate": 9.82532751091703e-06,
+      "loss": 1.1691,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5389862060546875,
+      "learning_rate": 9.868995633187773e-06,
+      "loss": 0.9764,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5852655172348022,
+      "learning_rate": 9.912663755458516e-06,
+      "loss": 1.0116,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4474867582321167,
+      "learning_rate": 9.956331877729258e-06,
+      "loss": 0.9913,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.521257758140564,
+      "learning_rate": 1e-05,
+      "loss": 0.9971,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4365307092666626,
+      "learning_rate": 9.999999549537193e-06,
+      "loss": 1.0358,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9128078818321228,
+      "learning_rate": 9.999998198148852e-06,
+      "loss": 1.1727,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.787290632724762,
+      "learning_rate": 9.999995945835221e-06,
+      "loss": 1.1673,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6633785963058472,
+      "learning_rate": 9.999992792596707e-06,
+      "loss": 0.9412,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6636563539505005,
+      "learning_rate": 9.999988738433875e-06,
+      "loss": 1.0054,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5141115188598633,
+      "learning_rate": 9.999983783347458e-06,
+      "loss": 0.963,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4434840679168701,
+      "learning_rate": 9.999977927338348e-06,
+      "loss": 1.0072,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5513427257537842,
+      "learning_rate": 9.999971170407601e-06,
+      "loss": 1.0202,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.303257703781128,
+      "learning_rate": 9.999963512556433e-06,
+      "loss": 1.2045,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5431580543518066,
+      "learning_rate": 9.999954953786226e-06,
+      "loss": 0.9491,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5449200868606567,
+      "learning_rate": 9.99994549409852e-06,
+      "loss": 0.9617,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9710614085197449,
+      "learning_rate": 9.99993513349502e-06,
+      "loss": 1.1754,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4396934509277344,
+      "learning_rate": 9.999923871977593e-06,
+      "loss": 0.9471,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3730570077896118,
+      "learning_rate": 9.999911709548269e-06,
+      "loss": 0.9182,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4151428937911987,
+      "learning_rate": 9.99989864620924e-06,
+      "loss": 1.0047,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3808568716049194,
+      "learning_rate": 9.999884681962857e-06,
+      "loss": 0.9775,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5857722759246826,
+      "learning_rate": 9.999869816811638e-06,
+      "loss": 1.1721,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5388017892837524,
+      "learning_rate": 9.999854050758263e-06,
+      "loss": 0.9695,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2519099712371826,
+      "learning_rate": 9.99983738380557e-06,
+      "loss": 1.1651,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5330753326416016,
+      "learning_rate": 9.999819815956561e-06,
+      "loss": 0.9465,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4193384647369385,
+      "learning_rate": 9.999801347214407e-06,
+      "loss": 1.0162,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9960240125656128,
+      "learning_rate": 9.999781977582432e-06,
+      "loss": 1.1661,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.436417818069458,
+      "learning_rate": 9.999761707064126e-06,
+      "loss": 0.9945,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5325478315353394,
+      "learning_rate": 9.999740535663142e-06,
+      "loss": 0.9975,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2886042594909668,
+      "learning_rate": 9.999718463383295e-06,
+      "loss": 1.1897,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4714573621749878,
+      "learning_rate": 9.999695490228563e-06,
+      "loss": 0.9627,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.433445692062378,
+      "learning_rate": 9.999671616203082e-06,
+      "loss": 0.942,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4018750190734863,
+      "learning_rate": 9.999646841311157e-06,
+      "loss": 0.9132,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8541656136512756,
+      "learning_rate": 9.999621165557252e-06,
+      "loss": 1.1494,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.476057767868042,
+      "learning_rate": 9.999594588945993e-06,
+      "loss": 0.9203,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7959000468254089,
+      "learning_rate": 9.999567111482166e-06,
+      "loss": 1.158,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5034557580947876,
+      "learning_rate": 9.999538733170727e-06,
+      "loss": 0.92,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8090248107910156,
+      "learning_rate": 9.999509454016784e-06,
+      "loss": 1.1842,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7873092293739319,
+      "learning_rate": 9.999479274025618e-06,
+      "loss": 1.1782,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6767832636833191,
+      "learning_rate": 9.999448193202663e-06,
+      "loss": 1.1502,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6455926895141602,
+      "learning_rate": 9.99941621155352e-06,
+      "loss": 0.9347,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4017856121063232,
+      "learning_rate": 9.999383329083953e-06,
+      "loss": 1.0301,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1533317565917969,
+      "learning_rate": 9.999349545799886e-06,
+      "loss": 1.1695,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3936187028884888,
+      "learning_rate": 9.999314861707406e-06,
+      "loss": 0.9692,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3313649892807007,
+      "learning_rate": 9.999279276812763e-06,
+      "loss": 0.9268,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3454363346099854,
+      "learning_rate": 9.99924279112237e-06,
+      "loss": 0.9698,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9323685169219971,
+      "learning_rate": 9.9992054046428e-06,
+      "loss": 1.1572,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4118221998214722,
+      "learning_rate": 9.999167117380788e-06,
+      "loss": 0.9788,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7146197557449341,
+      "learning_rate": 9.999127929343236e-06,
+      "loss": 1.1794,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4928665161132812,
+      "learning_rate": 9.999087840537202e-06,
+      "loss": 0.9764,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4540326595306396,
+      "learning_rate": 9.999046850969912e-06,
+      "loss": 1.0026,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9282329678535461,
+      "learning_rate": 9.999004960648749e-06,
+      "loss": 1.1789,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.405978798866272,
+      "learning_rate": 9.998962169581263e-06,
+      "loss": 0.9603,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3647347688674927,
+      "learning_rate": 9.998918477775164e-06,
+      "loss": 0.9931,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3796380758285522,
+      "learning_rate": 9.998873885238326e-06,
+      "loss": 0.9467,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4442263841629028,
+      "learning_rate": 9.998828391978779e-06,
+      "loss": 1.0017,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.32794988155365,
+      "learning_rate": 9.998781998004725e-06,
+      "loss": 0.9915,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.416075348854065,
+      "learning_rate": 9.998734703324524e-06,
+      "loss": 0.9582,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4231287240982056,
+      "learning_rate": 9.998686507946693e-06,
+      "loss": 0.9628,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.491829752922058,
+      "learning_rate": 9.99863741187992e-06,
+      "loss": 0.9755,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3802821636199951,
+      "learning_rate": 9.998587415133052e-06,
+      "loss": 0.947,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4962403774261475,
+      "learning_rate": 9.998536517715093e-06,
+      "loss": 0.9752,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4559823274612427,
+      "learning_rate": 9.99848471963522e-06,
+      "loss": 0.9716,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.420289397239685,
+      "learning_rate": 9.998432020902762e-06,
+      "loss": 0.9943,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3070297241210938,
+      "learning_rate": 9.998378421527216e-06,
+      "loss": 0.9428,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3490428924560547,
+      "learning_rate": 9.99832392151824e-06,
+      "loss": 1.0346,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4025980234146118,
+      "learning_rate": 9.998268520885652e-06,
+      "loss": 1.1945,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5228384733200073,
+      "learning_rate": 9.998212219639438e-06,
+      "loss": 0.914,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.759404718875885,
+      "learning_rate": 9.998155017789739e-06,
+      "loss": 1.1959,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4503040313720703,
+      "learning_rate": 9.998096915346865e-06,
+      "loss": 0.9401,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9316030740737915,
+      "learning_rate": 9.99803791232128e-06,
+      "loss": 1.1535,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8257102966308594,
+      "learning_rate": 9.997978008723624e-06,
+      "loss": 1.1416,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7212945818901062,
+      "learning_rate": 9.997917204564684e-06,
+      "loss": 1.1705,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6444920301437378,
+      "learning_rate": 9.997855499855418e-06,
+      "loss": 0.9789,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.503499984741211,
+      "learning_rate": 9.997792894606945e-06,
+      "loss": 1.0333,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3783987760543823,
+      "learning_rate": 9.997729388830544e-06,
+      "loss": 0.9781,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4366778135299683,
+      "learning_rate": 9.997664982537659e-06,
+      "loss": 1.0025,
+      "step": 301
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6682184934616089,
+      "learning_rate": 9.997599675739895e-06,
+      "loss": 1.1543,
+      "step": 302
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4257774353027344,
+      "learning_rate": 9.997533468449016e-06,
+      "loss": 1.161,
+      "step": 303
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.583132266998291,
+      "learning_rate": 9.997466360676958e-06,
+      "loss": 0.9153,
+      "step": 304
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.565984845161438,
+      "learning_rate": 9.997398352435808e-06,
+      "loss": 0.9872,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5014756917953491,
+      "learning_rate": 9.997329443737823e-06,
+      "loss": 0.9925,
+      "step": 306
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5646944046020508,
+      "learning_rate": 9.997259634595415e-06,
+      "loss": 0.9704,
+      "step": 307
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4402631521224976,
+      "learning_rate": 9.997188925021167e-06,
+      "loss": 0.9094,
+      "step": 308
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3597862720489502,
+      "learning_rate": 9.997117315027817e-06,
+      "loss": 0.9903,
+      "step": 309
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3809672594070435,
+      "learning_rate": 9.99704480462827e-06,
+      "loss": 0.8879,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.773985743522644,
+      "learning_rate": 9.996971393835592e-06,
+      "loss": 1.1507,
+      "step": 311
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5496139526367188,
+      "learning_rate": 9.996897082663007e-06,
+      "loss": 0.9787,
+      "step": 312
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6554973125457764,
+      "learning_rate": 9.996821871123907e-06,
+      "loss": 0.9924,
+      "step": 313
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5883963108062744,
+      "learning_rate": 9.996745759231845e-06,
+      "loss": 1.0368,
+      "step": 314
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.427638053894043,
+      "learning_rate": 9.996668747000533e-06,
+      "loss": 0.9794,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5523178577423096,
+      "learning_rate": 9.996590834443848e-06,
+      "loss": 1.0043,
+      "step": 316
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.529066801071167,
+      "learning_rate": 9.996512021575831e-06,
+      "loss": 0.955,
+      "step": 317
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4080655574798584,
+      "learning_rate": 9.99643230841068e-06,
+      "loss": 0.8986,
+      "step": 318
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5633721351623535,
+      "learning_rate": 9.996351694962758e-06,
+      "loss": 0.9788,
+      "step": 319
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.934578537940979,
+      "learning_rate": 9.996270181246593e-06,
+      "loss": 1.1521,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5036811828613281,
+      "learning_rate": 9.99618776727687e-06,
+      "loss": 0.9438,
+      "step": 321
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4979288578033447,
+      "learning_rate": 9.99610445306844e-06,
+      "loss": 0.9339,
+      "step": 322
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5303502082824707,
+      "learning_rate": 9.996020238636315e-06,
+      "loss": 0.931,
+      "step": 323
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4902194738388062,
+      "learning_rate": 9.99593512399567e-06,
+      "loss": 1.0296,
+      "step": 324
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3262659311294556,
+      "learning_rate": 9.99584910916184e-06,
+      "loss": 1.1764,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.341464877128601,
+      "learning_rate": 9.995762194150324e-06,
+      "loss": 0.9143,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7401795387268066,
+      "learning_rate": 9.995674378976783e-06,
+      "loss": 1.1655,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4292312860488892,
+      "learning_rate": 9.99558566365704e-06,
+      "loss": 0.9418,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9038122296333313,
+      "learning_rate": 9.995496048207078e-06,
+      "loss": 1.1666,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.376439094543457,
+      "learning_rate": 9.995405532643048e-06,
+      "loss": 0.9639,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.387746810913086,
+      "learning_rate": 9.995314116981256e-06,
+      "loss": 0.9898,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3900483846664429,
+      "learning_rate": 9.995221801238177e-06,
+      "loss": 1.0081,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4389894008636475,
+      "learning_rate": 9.995128585430445e-06,
+      "loss": 0.9393,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.519869089126587,
+      "learning_rate": 9.995034469574854e-06,
+      "loss": 0.8989,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4383097887039185,
+      "learning_rate": 9.994939453688361e-06,
+      "loss": 0.9873,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1208152770996094,
+      "learning_rate": 9.99484353778809e-06,
+      "loss": 1.178,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3579117059707642,
+      "learning_rate": 9.994746721891322e-06,
+      "loss": 0.9679,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7335168719291687,
+      "learning_rate": 9.994649006015502e-06,
+      "loss": 1.1427,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4384958744049072,
+      "learning_rate": 9.994550390178237e-06,
+      "loss": 0.9307,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7981409430503845,
+      "learning_rate": 9.994450874397294e-06,
+      "loss": 1.1617,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7658079862594604,
+      "learning_rate": 9.994350458690606e-06,
+      "loss": 1.1395,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.373764157295227,
+      "learning_rate": 9.99424914307627e-06,
+      "loss": 0.8793,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3759604692459106,
+      "learning_rate": 9.994146927572535e-06,
+      "loss": 0.9381,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.964495837688446,
+      "learning_rate": 9.994043812197823e-06,
+      "loss": 1.1355,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3795030117034912,
+      "learning_rate": 9.993939796970713e-06,
+      "loss": 0.9854,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7948021292686462,
+      "learning_rate": 9.993834881909944e-06,
+      "loss": 1.1489,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.413388729095459,
+      "learning_rate": 9.993729067034426e-06,
+      "loss": 0.9658,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3056944608688354,
+      "learning_rate": 9.993622352363221e-06,
+      "loss": 0.958,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8675932288169861,
+      "learning_rate": 9.99351473791556e-06,
+      "loss": 1.1675,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7293811440467834,
+      "learning_rate": 9.99340622371083e-06,
+      "loss": 1.1619,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7171947956085205,
+      "learning_rate": 9.993296809768585e-06,
+      "loss": 1.154,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4400566816329956,
+      "learning_rate": 9.993186496108542e-06,
+      "loss": 0.9786,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7656499743461609,
+      "learning_rate": 9.993075282750575e-06,
+      "loss": 1.1497,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6883093118667603,
+      "learning_rate": 9.992963169714726e-06,
+      "loss": 1.1424,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.326298713684082,
+      "learning_rate": 9.992850157021193e-06,
+      "loss": 0.9445,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3720530271530151,
+      "learning_rate": 9.992736244690341e-06,
+      "loss": 0.9669,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3761656284332275,
+      "learning_rate": 9.992621432742696e-06,
+      "loss": 0.997,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.427048683166504,
+      "learning_rate": 9.992505721198942e-06,
+      "loss": 0.9303,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4590128660202026,
+      "learning_rate": 9.992389110079932e-06,
+      "loss": 0.8992,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1679272651672363,
+      "learning_rate": 9.992271599406675e-06,
+      "loss": 1.1533,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4149800539016724,
+      "learning_rate": 9.992153189200349e-06,
+      "loss": 0.9171,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.362635850906372,
+      "learning_rate": 9.992033879482284e-06,
+      "loss": 0.8517,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5442136526107788,
+      "learning_rate": 9.991913670273982e-06,
+      "loss": 0.9926,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3346638679504395,
+      "learning_rate": 9.9917925615971e-06,
+      "loss": 0.8995,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9814811944961548,
+      "learning_rate": 9.991670553473463e-06,
+      "loss": 1.188,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.489993691444397,
+      "learning_rate": 9.991547645925055e-06,
+      "loss": 0.9451,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4442644119262695,
+      "learning_rate": 9.991423838974018e-06,
+      "loss": 0.9621,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9175399541854858,
+      "learning_rate": 9.991299132642665e-06,
+      "loss": 1.1531,
+      "step": 368
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4872815608978271,
+      "learning_rate": 9.991173526953462e-06,
+      "loss": 0.9255,
+      "step": 369
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6996262073516846,
+      "learning_rate": 9.991047021929044e-06,
+      "loss": 1.1727,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7504169940948486,
+      "learning_rate": 9.990919617592205e-06,
+      "loss": 1.1431,
+      "step": 371
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3642596006393433,
+      "learning_rate": 9.990791313965902e-06,
+      "loss": 0.9159,
+      "step": 372
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4238793849945068,
+      "learning_rate": 9.990662111073252e-06,
+      "loss": 0.9388,
+      "step": 373
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.01309072971344,
+      "learning_rate": 9.990532008937535e-06,
+      "loss": 1.1511,
+      "step": 374
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.376466155052185,
+      "learning_rate": 9.990401007582196e-06,
+      "loss": 0.9852,
+      "step": 375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4529038667678833,
+      "learning_rate": 9.990269107030836e-06,
+      "loss": 0.8656,
+      "step": 376
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9543941020965576,
+      "learning_rate": 9.990136307307224e-06,
+      "loss": 1.1223,
+      "step": 377
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2876418828964233,
+      "learning_rate": 9.990002608435287e-06,
+      "loss": 0.9057,
+      "step": 378
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3643189668655396,
+      "learning_rate": 9.989868010439118e-06,
+      "loss": 0.926,
+      "step": 379
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7492734789848328,
+      "learning_rate": 9.989732513342969e-06,
+      "loss": 1.131,
+      "step": 380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.384361743927002,
+      "learning_rate": 9.98959611717125e-06,
+      "loss": 0.9074,
+      "step": 381
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3749462366104126,
+      "learning_rate": 9.989458821948544e-06,
+      "loss": 0.9543,
+      "step": 382
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7603498101234436,
+      "learning_rate": 9.989320627699585e-06,
+      "loss": 1.1622,
+      "step": 383
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3727092742919922,
+      "learning_rate": 9.989181534449277e-06,
+      "loss": 0.9647,
+      "step": 384
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3829072713851929,
+      "learning_rate": 9.98904154222268e-06,
+      "loss": 0.947,
+      "step": 385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3873432874679565,
+      "learning_rate": 9.988900651045019e-06,
+      "loss": 0.9577,
+      "step": 386
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.379676103591919,
+      "learning_rate": 9.988758860941682e-06,
+      "loss": 0.965,
+      "step": 387
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4259883165359497,
+      "learning_rate": 9.988616171938215e-06,
+      "loss": 0.9469,
+      "step": 388
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5079751014709473,
+      "learning_rate": 9.98847258406033e-06,
+      "loss": 0.9761,
+      "step": 389
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4035929441452026,
+      "learning_rate": 9.9883280973339e-06,
+      "loss": 0.9,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3511199951171875,
+      "learning_rate": 9.988182711784956e-06,
+      "loss": 0.9835,
+      "step": 391
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.452821969985962,
+      "learning_rate": 9.988036427439698e-06,
+      "loss": 0.9692,
+      "step": 392
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3352254629135132,
+      "learning_rate": 9.987889244324485e-06,
+      "loss": 0.9774,
+      "step": 393
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3556424379348755,
+      "learning_rate": 9.987741162465834e-06,
+      "loss": 0.9605,
+      "step": 394
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3688325881958008,
+      "learning_rate": 9.987592181890426e-06,
+      "loss": 0.9579,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9375126361846924,
+      "learning_rate": 9.98744230262511e-06,
+      "loss": 1.1589,
+      "step": 396
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3631701469421387,
+      "learning_rate": 9.987291524696889e-06,
+      "loss": 0.9333,
+      "step": 397
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3106107711791992,
+      "learning_rate": 9.987139848132931e-06,
+      "loss": 0.9437,
+      "step": 398
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8266355395317078,
+      "learning_rate": 9.986987272960567e-06,
+      "loss": 1.1797,
+      "step": 399
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8118479251861572,
+      "learning_rate": 9.986833799207288e-06,
+      "loss": 1.1435,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3840752840042114,
+      "learning_rate": 9.986679426900748e-06,
+      "loss": 0.8724,
+      "step": 401
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3857775926589966,
+      "learning_rate": 9.986524156068762e-06,
+      "loss": 0.8981,
+      "step": 402
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.362464189529419,
+      "learning_rate": 9.986367986739307e-06,
+      "loss": 1.0302,
+      "step": 403
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4417916536331177,
+      "learning_rate": 9.986210918940524e-06,
+      "loss": 0.9349,
+      "step": 404
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.587675929069519,
+      "learning_rate": 9.986052952700714e-06,
+      "loss": 1.1862,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1709288358688354,
+      "learning_rate": 9.985894088048339e-06,
+      "loss": 1.1703,
+      "step": 406
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4201661348342896,
+      "learning_rate": 9.985734325012023e-06,
+      "loss": 0.9004,
+      "step": 407
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5698002576828003,
+      "learning_rate": 9.985573663620559e-06,
+      "loss": 0.9055,
+      "step": 408
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4792811870574951,
+      "learning_rate": 9.985412103902888e-06,
+      "loss": 0.9428,
+      "step": 409
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4745115041732788,
+      "learning_rate": 9.985249645888124e-06,
+      "loss": 0.9368,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3852499723434448,
+      "learning_rate": 9.98508628960554e-06,
+      "loss": 0.9146,
+      "step": 411
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.411238670349121,
+      "learning_rate": 9.98492203508457e-06,
+      "loss": 0.9618,
+      "step": 412
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3763689994812012,
+      "learning_rate": 9.984756882354812e-06,
+      "loss": 0.958,
+      "step": 413
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4473611116409302,
+      "learning_rate": 9.984590831446022e-06,
+      "loss": 0.926,
+      "step": 414
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4207489490509033,
+      "learning_rate": 9.984423882388118e-06,
+      "loss": 0.9067,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.535839319229126,
+      "learning_rate": 9.984256035211185e-06,
+      "loss": 0.9886,
+      "step": 416
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3661099672317505,
+      "learning_rate": 9.984087289945466e-06,
+      "loss": 0.9396,
+      "step": 417
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.358784556388855,
+      "learning_rate": 9.983917646621365e-06,
+      "loss": 1.0032,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.401901364326477,
+      "learning_rate": 9.98374710526945e-06,
+      "loss": 0.885,
+      "step": 419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4286376237869263,
+      "learning_rate": 9.983575665920451e-06,
+      "loss": 0.968,
+      "step": 420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3755630254745483,
+      "learning_rate": 9.983403328605256e-06,
+      "loss": 0.9619,
+      "step": 421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3962163925170898,
+      "learning_rate": 9.983230093354921e-06,
+      "loss": 0.9803,
+      "step": 422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.309749722480774,
+      "learning_rate": 9.98305596020066e-06,
+      "loss": 0.8903,
+      "step": 423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3325660228729248,
+      "learning_rate": 9.982880929173846e-06,
+      "loss": 0.9546,
+      "step": 424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.446644902229309,
+      "learning_rate": 9.982705000306022e-06,
+      "loss": 0.9281,
+      "step": 425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3128139972686768,
+      "learning_rate": 9.982528173628882e-06,
+      "loss": 1.1882,
+      "step": 426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3814377784729004,
+      "learning_rate": 9.982350449174292e-06,
+      "loss": 0.9309,
+      "step": 427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3711718320846558,
+      "learning_rate": 9.982171826974272e-06,
+      "loss": 0.9094,
+      "step": 428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.392824411392212,
+      "learning_rate": 9.98199230706101e-06,
+      "loss": 0.9528,
+      "step": 429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2860655784606934,
+      "learning_rate": 9.981811889466852e-06,
+      "loss": 1.1384,
+      "step": 430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4406713247299194,
+      "learning_rate": 9.981630574224306e-06,
+      "loss": 0.9343,
+      "step": 431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3696845769882202,
+      "learning_rate": 9.981448361366041e-06,
+      "loss": 0.9022,
+      "step": 432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3962363004684448,
+      "learning_rate": 9.981265250924893e-06,
+      "loss": 0.9126,
+      "step": 433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4280849695205688,
+      "learning_rate": 9.98108124293385e-06,
+      "loss": 0.9956,
+      "step": 434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4079115390777588,
+      "learning_rate": 9.980896337426073e-06,
+      "loss": 0.9851,
+      "step": 435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3094536066055298,
+      "learning_rate": 9.980710534434877e-06,
+      "loss": 0.9805,
+      "step": 436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0166232585906982,
+      "learning_rate": 9.98052383399374e-06,
+      "loss": 1.1728,
+      "step": 437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3525216579437256,
+      "learning_rate": 9.980336236136306e-06,
+      "loss": 0.9063,
+      "step": 438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4866044521331787,
+      "learning_rate": 9.980147740896373e-06,
+      "loss": 0.9256,
+      "step": 439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4255015850067139,
+      "learning_rate": 9.979958348307906e-06,
+      "loss": 0.9945,
+      "step": 440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0617774724960327,
+      "learning_rate": 9.979768058405034e-06,
+      "loss": 1.1695,
+      "step": 441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.392736792564392,
+      "learning_rate": 9.979576871222041e-06,
+      "loss": 0.8937,
+      "step": 442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4055452346801758,
+      "learning_rate": 9.979384786793378e-06,
+      "loss": 0.9718,
+      "step": 443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4223439693450928,
+      "learning_rate": 9.979191805153654e-06,
+      "loss": 0.9037,
+      "step": 444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4126672744750977,
+      "learning_rate": 9.978997926337641e-06,
+      "loss": 0.99,
+      "step": 445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3866801261901855,
+      "learning_rate": 9.978803150380278e-06,
+      "loss": 0.995,
+      "step": 446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.412125825881958,
+      "learning_rate": 9.978607477316656e-06,
+      "loss": 0.8734,
+      "step": 447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3386189937591553,
+      "learning_rate": 9.97841090718203e-06,
+      "loss": 0.9461,
+      "step": 448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7198209762573242,
+      "learning_rate": 9.978213440011826e-06,
+      "loss": 1.1468,
+      "step": 449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3756874799728394,
+      "learning_rate": 9.978015075841622e-06,
+      "loss": 0.9561,
+      "step": 450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1801114082336426,
+      "learning_rate": 9.977815814707157e-06,
+      "loss": 1.1509,
+      "step": 451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8026167154312134,
+      "learning_rate": 9.977615656644339e-06,
+      "loss": 1.1356,
+      "step": 452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4633667469024658,
+      "learning_rate": 9.977414601689231e-06,
+      "loss": 0.958,
+      "step": 453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.497083067893982,
+      "learning_rate": 9.977212649878063e-06,
+      "loss": 0.9408,
+      "step": 454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6588555574417114,
+      "learning_rate": 9.977009801247219e-06,
+      "loss": 1.1526,
+      "step": 455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5464911460876465,
+      "learning_rate": 9.976806055833253e-06,
+      "loss": 1.1546,
+      "step": 456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1092222929000854,
+      "learning_rate": 9.976601413672876e-06,
+      "loss": 1.1466,
+      "step": 457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5068655014038086,
+      "learning_rate": 9.976395874802963e-06,
+      "loss": 0.9726,
+      "step": 458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2124660015106201,
+      "learning_rate": 9.976189439260545e-06,
+      "loss": 1.1199,
+      "step": 459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.339544653892517,
+      "learning_rate": 9.975982107082821e-06,
+      "loss": 1.1578,
+      "step": 460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4358220100402832,
+      "learning_rate": 9.975773878307152e-06,
+      "loss": 0.9344,
+      "step": 461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5225763320922852,
+      "learning_rate": 9.975564752971051e-06,
+      "loss": 0.9394,
+      "step": 462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.454634666442871,
+      "learning_rate": 9.975354731112206e-06,
+      "loss": 0.8831,
+      "step": 463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5660711526870728,
+      "learning_rate": 9.975143812768457e-06,
+      "loss": 0.9667,
+      "step": 464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2920292615890503,
+      "learning_rate": 9.974931997977808e-06,
+      "loss": 1.1412,
+      "step": 465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4268141984939575,
+      "learning_rate": 9.974719286778425e-06,
+      "loss": 0.9085,
+      "step": 466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5291410684585571,
+      "learning_rate": 9.974505679208637e-06,
+      "loss": 0.9079,
+      "step": 467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5313109159469604,
+      "learning_rate": 9.974291175306929e-06,
+      "loss": 0.985,
+      "step": 468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8482195138931274,
+      "learning_rate": 9.974075775111956e-06,
+      "loss": 1.1209,
+      "step": 469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8127344846725464,
+      "learning_rate": 9.973859478662527e-06,
+      "loss": 1.1226,
+      "step": 470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6147809028625488,
+      "learning_rate": 9.973642285997617e-06,
+      "loss": 0.9578,
+      "step": 471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7350525856018066,
+      "learning_rate": 9.973424197156359e-06,
+      "loss": 1.1172,
+      "step": 472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4260021448135376,
+      "learning_rate": 9.973205212178052e-06,
+      "loss": 0.9805,
+      "step": 473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.754204273223877,
+      "learning_rate": 9.97298533110215e-06,
+      "loss": 1.1603,
+      "step": 474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3530151844024658,
+      "learning_rate": 9.972764553968277e-06,
+      "loss": 0.9397,
+      "step": 475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3757436275482178,
+      "learning_rate": 9.97254288081621e-06,
+      "loss": 0.8892,
+      "step": 476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7864536046981812,
+      "learning_rate": 9.972320311685894e-06,
+      "loss": 1.158,
+      "step": 477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7867088913917542,
+      "learning_rate": 9.97209684661743e-06,
+      "loss": 1.1476,
+      "step": 478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6807409524917603,
+      "learning_rate": 9.971872485651086e-06,
+      "loss": 1.1329,
+      "step": 479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.39301598072052,
+      "learning_rate": 9.971647228827285e-06,
+      "loss": 0.9711,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7387423515319824,
+      "learning_rate": 9.971421076186619e-06,
+      "loss": 1.1445,
+      "step": 481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3388198614120483,
+      "learning_rate": 9.971194027769833e-06,
+      "loss": 0.9426,
+      "step": 482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7445107698440552,
+      "learning_rate": 9.970966083617841e-06,
+      "loss": 1.1445,
+      "step": 483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.370663046836853,
+      "learning_rate": 9.970737243771714e-06,
+      "loss": 0.8929,
+      "step": 484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3347265720367432,
+      "learning_rate": 9.970507508272684e-06,
+      "loss": 0.8724,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.323834776878357,
+      "learning_rate": 9.970276877162149e-06,
+      "loss": 0.9065,
+      "step": 486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9365085959434509,
+      "learning_rate": 9.970045350481663e-06,
+      "loss": 1.1132,
+      "step": 487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2952345609664917,
+      "learning_rate": 9.969812928272946e-06,
+      "loss": 0.9501,
+      "step": 488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6905801892280579,
+      "learning_rate": 9.969579610577874e-06,
+      "loss": 1.1336,
+      "step": 489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6712061762809753,
+      "learning_rate": 9.969345397438488e-06,
+      "loss": 1.1267,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4091941118240356,
+      "learning_rate": 9.969110288896992e-06,
+      "loss": 0.9356,
+      "step": 491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4133572578430176,
+      "learning_rate": 9.968874284995749e-06,
+      "loss": 0.9071,
+      "step": 492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3780145645141602,
+      "learning_rate": 9.96863738577728e-06,
+      "loss": 0.9611,
+      "step": 493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7852421998977661,
+      "learning_rate": 9.968399591284273e-06,
+      "loss": 1.1612,
+      "step": 494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6964321732521057,
+      "learning_rate": 9.968160901559575e-06,
+      "loss": 1.1209,
+      "step": 495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5055832862854004,
+      "learning_rate": 9.967921316646194e-06,
+      "loss": 0.993,
+      "step": 496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4953628778457642,
+      "learning_rate": 9.9676808365873e-06,
+      "loss": 0.9493,
+      "step": 497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.367328405380249,
+      "learning_rate": 9.967439461426223e-06,
+      "loss": 0.9607,
+      "step": 498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0375980138778687,
+      "learning_rate": 9.967197191206456e-06,
+      "loss": 1.1414,
+      "step": 499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3754279613494873,
+      "learning_rate": 9.966954025971654e-06,
+      "loss": 0.9635,
+      "step": 500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3997414112091064,
+      "learning_rate": 9.966709965765628e-06,
+      "loss": 0.9448,
+      "step": 501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7252254486083984,
+      "learning_rate": 9.966465010632359e-06,
+      "loss": 1.1422,
+      "step": 502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3658297061920166,
+      "learning_rate": 9.966219160615978e-06,
+      "loss": 0.9104,
+      "step": 503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3202418088912964,
+      "learning_rate": 9.965972415760787e-06,
+      "loss": 0.9306,
+      "step": 504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8117262721061707,
+      "learning_rate": 9.965724776111247e-06,
+      "loss": 1.1293,
+      "step": 505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6313713788986206,
+      "learning_rate": 9.965476241711979e-06,
+      "loss": 1.1283,
+      "step": 506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6669440865516663,
+      "learning_rate": 9.965226812607762e-06,
+      "loss": 1.1134,
+      "step": 507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3847283124923706,
+      "learning_rate": 9.964976488843544e-06,
+      "loss": 0.9486,
+      "step": 508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7835860848426819,
+      "learning_rate": 9.964725270464423e-06,
+      "loss": 1.0997,
+      "step": 509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.424512267112732,
+      "learning_rate": 9.96447315751567e-06,
+      "loss": 0.9448,
+      "step": 510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3565753698349,
+      "learning_rate": 9.964220150042711e-06,
+      "loss": 0.9384,
+      "step": 511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7213389277458191,
+      "learning_rate": 9.963966248091134e-06,
+      "loss": 1.1467,
+      "step": 512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3535267114639282,
+      "learning_rate": 9.96371145170669e-06,
+      "loss": 0.9142,
+      "step": 513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.338820219039917,
+      "learning_rate": 9.963455760935286e-06,
+      "loss": 0.9384,
+      "step": 514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4353440999984741,
+      "learning_rate": 9.963199175822995e-06,
+      "loss": 0.9844,
+      "step": 515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3628755807876587,
+      "learning_rate": 9.96294169641605e-06,
+      "loss": 0.9521,
+      "step": 516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3240785598754883,
+      "learning_rate": 9.962683322760846e-06,
+      "loss": 0.9098,
+      "step": 517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2709004878997803,
+      "learning_rate": 9.962424054903939e-06,
+      "loss": 0.9734,
+      "step": 518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0641906261444092,
+      "learning_rate": 9.962163892892041e-06,
+      "loss": 1.1279,
+      "step": 519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3190505504608154,
+      "learning_rate": 9.961902836772033e-06,
+      "loss": 0.9095,
+      "step": 520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4533170461654663,
+      "learning_rate": 9.961640886590952e-06,
+      "loss": 0.9294,
+      "step": 521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7705588340759277,
+      "learning_rate": 9.961378042395997e-06,
+      "loss": 0.9324,
+      "step": 522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5343326330184937,
+      "learning_rate": 9.961114304234531e-06,
+      "loss": 0.9055,
+      "step": 523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1116387844085693,
+      "learning_rate": 9.960849672154073e-06,
+      "loss": 1.1411,
+      "step": 524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.372399091720581,
+      "learning_rate": 9.960584146202309e-06,
+      "loss": 0.955,
+      "step": 525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4383169412612915,
+      "learning_rate": 9.960317726427079e-06,
+      "loss": 0.968,
+      "step": 526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4079073667526245,
+      "learning_rate": 9.960050412876388e-06,
+      "loss": 0.9282,
+      "step": 527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4399936199188232,
+      "learning_rate": 9.959782205598405e-06,
+      "loss": 0.9342,
+      "step": 528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4502651691436768,
+      "learning_rate": 9.959513104641457e-06,
+      "loss": 0.9574,
+      "step": 529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9833511114120483,
+      "learning_rate": 9.959243110054027e-06,
+      "loss": 1.1257,
+      "step": 530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4199950695037842,
+      "learning_rate": 9.95897222188477e-06,
+      "loss": 0.8574,
+      "step": 531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4284460544586182,
+      "learning_rate": 9.958700440182492e-06,
+      "loss": 0.9763,
+      "step": 532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3758163452148438,
+      "learning_rate": 9.958427764996168e-06,
+      "loss": 0.9596,
+      "step": 533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4043824672698975,
+      "learning_rate": 9.958154196374926e-06,
+      "loss": 0.9288,
+      "step": 534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.339060664176941,
+      "learning_rate": 9.957879734368059e-06,
+      "loss": 0.9135,
+      "step": 535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1346993446350098,
+      "learning_rate": 9.957604379025025e-06,
+      "loss": 1.1339,
+      "step": 536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4682562351226807,
+      "learning_rate": 9.957328130395434e-06,
+      "loss": 0.9003,
+      "step": 537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4210519790649414,
+      "learning_rate": 9.957050988529067e-06,
+      "loss": 0.9022,
+      "step": 538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.335793375968933,
+      "learning_rate": 9.956772953475857e-06,
+      "loss": 0.8885,
+      "step": 539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.494789719581604,
+      "learning_rate": 9.9564940252859e-06,
+      "loss": 0.986,
+      "step": 540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4773472547531128,
+      "learning_rate": 9.956214204009462e-06,
+      "loss": 0.9505,
+      "step": 541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3993499279022217,
+      "learning_rate": 9.955933489696956e-06,
+      "loss": 0.9207,
+      "step": 542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.122336506843567,
+      "learning_rate": 9.955651882398966e-06,
+      "loss": 1.1389,
+      "step": 543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3761601448059082,
+      "learning_rate": 9.95536938216623e-06,
+      "loss": 0.901,
+      "step": 544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3746429681777954,
+      "learning_rate": 9.955085989049655e-06,
+      "loss": 0.8781,
+      "step": 545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.350751519203186,
+      "learning_rate": 9.954801703100302e-06,
+      "loss": 0.9222,
+      "step": 546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3852475881576538,
+      "learning_rate": 9.954516524369395e-06,
+      "loss": 0.9175,
+      "step": 547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3916113376617432,
+      "learning_rate": 9.954230452908316e-06,
+      "loss": 0.934,
+      "step": 548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3390744924545288,
+      "learning_rate": 9.953943488768616e-06,
+      "loss": 0.9205,
+      "step": 549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.367653250694275,
+      "learning_rate": 9.953655632002e-06,
+      "loss": 0.9575,
+      "step": 550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0399491786956787,
+      "learning_rate": 9.953366882660334e-06,
+      "loss": 1.1391,
+      "step": 551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7214741110801697,
+      "learning_rate": 9.953077240795646e-06,
+      "loss": 1.1389,
+      "step": 552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5081775188446045,
+      "learning_rate": 9.952786706460127e-06,
+      "loss": 0.9395,
+      "step": 553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6154195070266724,
+      "learning_rate": 9.952495279706127e-06,
+      "loss": 0.9205,
+      "step": 554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3953231573104858,
+      "learning_rate": 9.952202960586157e-06,
+      "loss": 0.8954,
+      "step": 555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3875144720077515,
+      "learning_rate": 9.951909749152886e-06,
+      "loss": 0.9218,
+      "step": 556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5237627029418945,
+      "learning_rate": 9.951615645459149e-06,
+      "loss": 0.9415,
+      "step": 557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4737483263015747,
+      "learning_rate": 9.951320649557938e-06,
+      "loss": 0.9362,
+      "step": 558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4022161960601807,
+      "learning_rate": 9.951024761502405e-06,
+      "loss": 0.9062,
+      "step": 559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3162556886672974,
+      "learning_rate": 9.950727981345869e-06,
+      "loss": 0.9315,
+      "step": 560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.404616355895996,
+      "learning_rate": 9.950430309141804e-06,
+      "loss": 0.9578,
+      "step": 561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3658745288848877,
+      "learning_rate": 9.950131744943842e-06,
+      "loss": 0.9349,
+      "step": 562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3644921779632568,
+      "learning_rate": 9.949832288805784e-06,
+      "loss": 0.9585,
+      "step": 563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3159843683242798,
+      "learning_rate": 9.949531940781587e-06,
+      "loss": 0.8826,
+      "step": 564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.358473300933838,
+      "learning_rate": 9.949230700925368e-06,
+      "loss": 0.9935,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2850130796432495,
+      "learning_rate": 9.948928569291407e-06,
+      "loss": 0.9129,
+      "step": 566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3044729232788086,
+      "learning_rate": 9.948625545934142e-06,
+      "loss": 0.944,
+      "step": 567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4392162561416626,
+      "learning_rate": 9.948321630908178e-06,
+      "loss": 0.9034,
+      "step": 568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3263932466506958,
+      "learning_rate": 9.94801682426827e-06,
+      "loss": 0.9517,
+      "step": 569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.328125238418579,
+      "learning_rate": 9.947711126069341e-06,
+      "loss": 0.9212,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3271721601486206,
+      "learning_rate": 9.947404536366477e-06,
+      "loss": 0.9213,
+      "step": 571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.353461503982544,
+      "learning_rate": 9.947097055214915e-06,
+      "loss": 0.934,
+      "step": 572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2413610219955444,
+      "learning_rate": 9.946788682670066e-06,
+      "loss": 0.8734,
+      "step": 573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.162074089050293,
+      "learning_rate": 9.946479418787488e-06,
+      "loss": 1.121,
+      "step": 574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4177528619766235,
+      "learning_rate": 9.946169263622906e-06,
+      "loss": 0.9416,
+      "step": 575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2661558389663696,
+      "learning_rate": 9.94585821723221e-06,
+      "loss": 1.1482,
+      "step": 576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4106310606002808,
+      "learning_rate": 9.945546279671444e-06,
+      "loss": 0.9529,
+      "step": 577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.199392318725586,
+      "learning_rate": 9.945233450996809e-06,
+      "loss": 1.1493,
+      "step": 578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4077966213226318,
+      "learning_rate": 9.944919731264679e-06,
+      "loss": 0.9368,
+      "step": 579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4410579204559326,
+      "learning_rate": 9.94460512053158e-06,
+      "loss": 0.8954,
+      "step": 580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.316868543624878,
+      "learning_rate": 9.9442896188542e-06,
+      "loss": 0.8547,
+      "step": 581
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3385674953460693,
+      "learning_rate": 9.943973226289385e-06,
+      "loss": 0.9224,
+      "step": 582
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3002310991287231,
+      "learning_rate": 9.943655942894146e-06,
+      "loss": 0.9198,
+      "step": 583
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.352347493171692,
+      "learning_rate": 9.943337768725656e-06,
+      "loss": 0.9032,
+      "step": 584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.865230679512024,
+      "learning_rate": 9.94301870384124e-06,
+      "loss": 1.1497,
+      "step": 585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.45187246799469,
+      "learning_rate": 9.942698748298391e-06,
+      "loss": 0.8957,
+      "step": 586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3785253763198853,
+      "learning_rate": 9.94237790215476e-06,
+      "loss": 0.8849,
+      "step": 587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4125572443008423,
+      "learning_rate": 9.942056165468161e-06,
+      "loss": 0.9162,
+      "step": 588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3438467979431152,
+      "learning_rate": 9.941733538296564e-06,
+      "loss": 0.9219,
+      "step": 589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4063750505447388,
+      "learning_rate": 9.9414100206981e-06,
+      "loss": 0.9837,
+      "step": 590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2916902303695679,
+      "learning_rate": 9.941085612731065e-06,
+      "loss": 0.9489,
+      "step": 591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4685876369476318,
+      "learning_rate": 9.940760314453911e-06,
+      "loss": 0.9245,
+      "step": 592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4118586778640747,
+      "learning_rate": 9.940434125925253e-06,
+      "loss": 0.936,
+      "step": 593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2775986194610596,
+      "learning_rate": 9.940107047203864e-06,
+      "loss": 0.9012,
+      "step": 594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3765825033187866,
+      "learning_rate": 9.93977907834868e-06,
+      "loss": 0.9622,
+      "step": 595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.391733169555664,
+      "learning_rate": 9.939450219418793e-06,
+      "loss": 0.9579,
+      "step": 596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8806833624839783,
+      "learning_rate": 9.939120470473464e-06,
+      "loss": 1.1211,
+      "step": 597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.409346103668213,
+      "learning_rate": 9.938789831572104e-06,
+      "loss": 0.8649,
+      "step": 598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.412896752357483,
+      "learning_rate": 9.93845830277429e-06,
+      "loss": 0.9289,
+      "step": 599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.705312192440033,
+      "learning_rate": 9.938125884139762e-06,
+      "loss": 1.1221,
+      "step": 600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3317503929138184,
+      "learning_rate": 9.937792575728413e-06,
+      "loss": 0.9348,
+      "step": 601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.339111089706421,
+      "learning_rate": 9.937458377600303e-06,
+      "loss": 0.9553,
+      "step": 602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2810546159744263,
+      "learning_rate": 9.937123289815646e-06,
+      "loss": 0.8578,
+      "step": 603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9462709426879883,
+      "learning_rate": 9.936787312434824e-06,
+      "loss": 1.1213,
+      "step": 604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7775394320487976,
+      "learning_rate": 9.93645044551837e-06,
+      "loss": 1.1031,
+      "step": 605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6664707064628601,
+      "learning_rate": 9.936112689126989e-06,
+      "loss": 1.138,
+      "step": 606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3769171237945557,
+      "learning_rate": 9.935774043321535e-06,
+      "loss": 0.9166,
+      "step": 607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0796873569488525,
+      "learning_rate": 9.935434508163028e-06,
+      "loss": 1.1434,
+      "step": 608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.859926700592041,
+      "learning_rate": 9.935094083712647e-06,
+      "loss": 1.1098,
+      "step": 609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.406424880027771,
+      "learning_rate": 9.934752770031732e-06,
+      "loss": 0.9869,
+      "step": 610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.368245005607605,
+      "learning_rate": 9.934410567181783e-06,
+      "loss": 0.9385,
+      "step": 611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4283738136291504,
+      "learning_rate": 9.934067475224459e-06,
+      "loss": 0.9589,
+      "step": 612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4087960720062256,
+      "learning_rate": 9.93372349422158e-06,
+      "loss": 0.8924,
+      "step": 613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.222011685371399,
+      "learning_rate": 9.933378624235125e-06,
+      "loss": 1.1399,
+      "step": 614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3843708038330078,
+      "learning_rate": 9.933032865327238e-06,
+      "loss": 0.917,
+      "step": 615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4562110900878906,
+      "learning_rate": 9.932686217560218e-06,
+      "loss": 0.906,
+      "step": 616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3438234329223633,
+      "learning_rate": 9.932338680996524e-06,
+      "loss": 0.8306,
+      "step": 617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3754431009292603,
+      "learning_rate": 9.931990255698781e-06,
+      "loss": 0.9323,
+      "step": 618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3538984060287476,
+      "learning_rate": 9.931640941729765e-06,
+      "loss": 0.9538,
+      "step": 619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0014146566390991,
+      "learning_rate": 9.931290739152419e-06,
+      "loss": 1.1245,
+      "step": 620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4163830280303955,
+      "learning_rate": 9.930939648029848e-06,
+      "loss": 0.9543,
+      "step": 621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7041628360748291,
+      "learning_rate": 9.930587668425307e-06,
+      "loss": 1.135,
+      "step": 622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7910506129264832,
+      "learning_rate": 9.930234800402221e-06,
+      "loss": 1.114,
+      "step": 623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.568239688873291,
+      "learning_rate": 9.929881044024172e-06,
+      "loss": 0.9018,
+      "step": 624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3587279319763184,
+      "learning_rate": 9.929526399354901e-06,
+      "loss": 0.8993,
+      "step": 625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3951034545898438,
+      "learning_rate": 9.929170866458309e-06,
+      "loss": 0.8833,
+      "step": 626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3297063112258911,
+      "learning_rate": 9.92881444539846e-06,
+      "loss": 0.9021,
+      "step": 627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.389297366142273,
+      "learning_rate": 9.928457136239573e-06,
+      "loss": 0.9332,
+      "step": 628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.309766411781311,
+      "learning_rate": 9.928098939046032e-06,
+      "loss": 0.8864,
+      "step": 629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3815124034881592,
+      "learning_rate": 9.927739853882377e-06,
+      "loss": 0.8827,
+      "step": 630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3248754739761353,
+      "learning_rate": 9.927379880813309e-06,
+      "loss": 0.9098,
+      "step": 631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.412638545036316,
+      "learning_rate": 9.927019019903692e-06,
+      "loss": 0.8825,
+      "step": 632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3837202787399292,
+      "learning_rate": 9.926657271218549e-06,
+      "loss": 0.8931,
+      "step": 633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3559536933898926,
+      "learning_rate": 9.92629463482306e-06,
+      "loss": 0.8953,
+      "step": 634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2405205965042114,
+      "learning_rate": 9.925931110782564e-06,
+      "loss": 1.1316,
+      "step": 635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3649438619613647,
+      "learning_rate": 9.925566699162566e-06,
+      "loss": 0.8937,
+      "step": 636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5147351026535034,
+      "learning_rate": 9.925201400028728e-06,
+      "loss": 0.9217,
+      "step": 637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6635342836380005,
+      "learning_rate": 9.924835213446869e-06,
+      "loss": 0.8997,
+      "step": 638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.353182077407837,
+      "learning_rate": 9.924468139482972e-06,
+      "loss": 0.9024,
+      "step": 639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8075530529022217,
+      "learning_rate": 9.924100178203178e-06,
+      "loss": 1.1048,
+      "step": 640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3564703464508057,
+      "learning_rate": 9.923731329673788e-06,
+      "loss": 0.9455,
+      "step": 641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7167832255363464,
+      "learning_rate": 9.92336159396126e-06,
+      "loss": 1.164,
+      "step": 642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.430503487586975,
+      "learning_rate": 9.922990971132221e-06,
+      "loss": 0.9324,
+      "step": 643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7091605067253113,
+      "learning_rate": 9.922619461253449e-06,
+      "loss": 1.1404,
+      "step": 644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7337227463722229,
+      "learning_rate": 9.922247064391883e-06,
+      "loss": 1.0948,
+      "step": 645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4942271709442139,
+      "learning_rate": 9.921873780614624e-06,
+      "loss": 0.9264,
+      "step": 646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.376370906829834,
+      "learning_rate": 9.921499609988934e-06,
+      "loss": 0.9055,
+      "step": 647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3125874996185303,
+      "learning_rate": 9.921124552582231e-06,
+      "loss": 0.9271,
+      "step": 648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3731166124343872,
+      "learning_rate": 9.920748608462094e-06,
+      "loss": 0.9172,
+      "step": 649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3073210716247559,
+      "learning_rate": 9.920371777696267e-06,
+      "loss": 0.9453,
+      "step": 650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.147413730621338,
+      "learning_rate": 9.919994060352643e-06,
+      "loss": 1.1273,
+      "step": 651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.932919442653656,
+      "learning_rate": 9.919615456499286e-06,
+      "loss": 1.1165,
+      "step": 652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7064737677574158,
+      "learning_rate": 9.919235966204414e-06,
+      "loss": 1.1337,
+      "step": 653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7934197187423706,
+      "learning_rate": 9.918855589536403e-06,
+      "loss": 1.1033,
+      "step": 654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5389935970306396,
+      "learning_rate": 9.918474326563794e-06,
+      "loss": 0.9861,
+      "step": 655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4745523929595947,
+      "learning_rate": 9.918092177355283e-06,
+      "loss": 0.9266,
+      "step": 656
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0311862230300903,
+      "learning_rate": 9.917709141979728e-06,
+      "loss": 1.0976,
+      "step": 657
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3473914861679077,
+      "learning_rate": 9.917325220506147e-06,
+      "loss": 0.9384,
+      "step": 658
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8294597864151001,
+      "learning_rate": 9.916940413003717e-06,
+      "loss": 1.1076,
+      "step": 659
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3246793746948242,
+      "learning_rate": 9.916554719541773e-06,
+      "loss": 0.9015,
+      "step": 660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2983108758926392,
+      "learning_rate": 9.916168140189813e-06,
+      "loss": 0.9109,
+      "step": 661
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3240963220596313,
+      "learning_rate": 9.915780675017492e-06,
+      "loss": 0.9368,
+      "step": 662
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.950232207775116,
+      "learning_rate": 9.915392324094626e-06,
+      "loss": 1.1196,
+      "step": 663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7867540717124939,
+      "learning_rate": 9.91500308749119e-06,
+      "loss": 1.0761,
+      "step": 664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3589928150177002,
+      "learning_rate": 9.914612965277318e-06,
+      "loss": 0.9462,
+      "step": 665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.411209225654602,
+      "learning_rate": 9.914221957523304e-06,
+      "loss": 0.9498,
+      "step": 666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3364570140838623,
+      "learning_rate": 9.913830064299605e-06,
+      "loss": 0.8809,
+      "step": 667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1477309465408325,
+      "learning_rate": 9.913437285676829e-06,
+      "loss": 1.0957,
+      "step": 668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3296834230422974,
+      "learning_rate": 9.913043621725752e-06,
+      "loss": 0.8239,
+      "step": 669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4664157629013062,
+      "learning_rate": 9.912649072517306e-06,
+      "loss": 0.9748,
+      "step": 670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.372475504875183,
+      "learning_rate": 9.912253638122582e-06,
+      "loss": 0.9643,
+      "step": 671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3326736688613892,
+      "learning_rate": 9.911857318612834e-06,
+      "loss": 0.8976,
+      "step": 672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3416414260864258,
+      "learning_rate": 9.91146011405947e-06,
+      "loss": 0.828,
+      "step": 673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8342621326446533,
+      "learning_rate": 9.911062024534062e-06,
+      "loss": 1.1332,
+      "step": 674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3394322395324707,
+      "learning_rate": 9.91066305010834e-06,
+      "loss": 0.951,
+      "step": 675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6696964502334595,
+      "learning_rate": 9.910263190854193e-06,
+      "loss": 1.122,
+      "step": 676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.441505789756775,
+      "learning_rate": 9.909862446843667e-06,
+      "loss": 0.8839,
+      "step": 677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.291011095046997,
+      "learning_rate": 9.909460818148972e-06,
+      "loss": 0.9058,
+      "step": 678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2460265159606934,
+      "learning_rate": 9.909058304842479e-06,
+      "loss": 0.8775,
+      "step": 679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8411524295806885,
+      "learning_rate": 9.908654906996708e-06,
+      "loss": 1.1265,
+      "step": 680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3943681716918945,
+      "learning_rate": 9.908250624684352e-06,
+      "loss": 0.9227,
+      "step": 681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3191968202590942,
+      "learning_rate": 9.907845457978253e-06,
+      "loss": 0.9578,
+      "step": 682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3396016359329224,
+      "learning_rate": 9.907439406951416e-06,
+      "loss": 0.9191,
+      "step": 683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7487990260124207,
+      "learning_rate": 9.907032471677006e-06,
+      "loss": 1.1028,
+      "step": 684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3827826976776123,
+      "learning_rate": 9.906624652228348e-06,
+      "loss": 0.9409,
+      "step": 685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4748131036758423,
+      "learning_rate": 9.906215948678921e-06,
+      "loss": 0.896,
+      "step": 686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6919950246810913,
+      "learning_rate": 9.905806361102374e-06,
+      "loss": 1.1043,
+      "step": 687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.320577621459961,
+      "learning_rate": 9.905395889572502e-06,
+      "loss": 0.9166,
+      "step": 688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6819447875022888,
+      "learning_rate": 9.904984534163268e-06,
+      "loss": 1.1026,
+      "step": 689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.311959147453308,
+      "learning_rate": 9.904572294948795e-06,
+      "loss": 0.9048,
+      "step": 690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2776756286621094,
+      "learning_rate": 9.904159172003357e-06,
+      "loss": 0.9211,
+      "step": 691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3180441856384277,
+      "learning_rate": 9.903745165401399e-06,
+      "loss": 0.8751,
+      "step": 692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.333703637123108,
+      "learning_rate": 9.903330275217512e-06,
+      "loss": 0.9393,
+      "step": 693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8805934190750122,
+      "learning_rate": 9.90291450152646e-06,
+      "loss": 1.1074,
+      "step": 694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7536527514457703,
+      "learning_rate": 9.902497844403155e-06,
+      "loss": 1.0937,
+      "step": 695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6768956184387207,
+      "learning_rate": 9.902080303922672e-06,
+      "loss": 1.1342,
+      "step": 696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7291260957717896,
+      "learning_rate": 9.901661880160245e-06,
+      "loss": 1.114,
+      "step": 697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7519636750221252,
+      "learning_rate": 9.901242573191272e-06,
+      "loss": 1.113,
+      "step": 698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7331888675689697,
+      "learning_rate": 9.900822383091303e-06,
+      "loss": 1.0978,
+      "step": 699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6780292391777039,
+      "learning_rate": 9.900401309936049e-06,
+      "loss": 1.1164,
+      "step": 700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6972759962081909,
+      "learning_rate": 9.899979353801384e-06,
+      "loss": 1.1123,
+      "step": 701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5952041149139404,
+      "learning_rate": 9.899556514763337e-06,
+      "loss": 0.8845,
+      "step": 702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8340471982955933,
+      "learning_rate": 9.899132792898096e-06,
+      "loss": 1.0803,
+      "step": 703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.448333501815796,
+      "learning_rate": 9.89870818828201e-06,
+      "loss": 0.916,
+      "step": 704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2917073965072632,
+      "learning_rate": 9.898282700991588e-06,
+      "loss": 0.9032,
+      "step": 705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3096857070922852,
+      "learning_rate": 9.897856331103495e-06,
+      "loss": 0.9579,
+      "step": 706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3468931913375854,
+      "learning_rate": 9.897429078694555e-06,
+      "loss": 0.9413,
+      "step": 707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2677875757217407,
+      "learning_rate": 9.897000943841756e-06,
+      "loss": 0.9491,
+      "step": 708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0122675895690918,
+      "learning_rate": 9.89657192662224e-06,
+      "loss": 1.1262,
+      "step": 709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4099767208099365,
+      "learning_rate": 9.896142027113308e-06,
+      "loss": 0.897,
+      "step": 710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4676541090011597,
+      "learning_rate": 9.895711245392424e-06,
+      "loss": 0.8896,
+      "step": 711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8682623505592346,
+      "learning_rate": 9.895279581537206e-06,
+      "loss": 1.11,
+      "step": 712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4274992942810059,
+      "learning_rate": 9.894847035625437e-06,
+      "loss": 0.9755,
+      "step": 713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4266479015350342,
+      "learning_rate": 9.89441360773505e-06,
+      "loss": 0.9033,
+      "step": 714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3855398893356323,
+      "learning_rate": 9.893979297944147e-06,
+      "loss": 0.9091,
+      "step": 715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.30911123752594,
+      "learning_rate": 9.893544106330983e-06,
+      "loss": 0.8915,
+      "step": 716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6306880712509155,
+      "learning_rate": 9.893108032973973e-06,
+      "loss": 0.9208,
+      "step": 717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8919384479522705,
+      "learning_rate": 9.892671077951688e-06,
+      "loss": 1.1143,
+      "step": 718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4347028732299805,
+      "learning_rate": 9.892233241342865e-06,
+      "loss": 0.8851,
+      "step": 719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3671742677688599,
+      "learning_rate": 9.891794523226395e-06,
+      "loss": 0.9812,
+      "step": 720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3297487497329712,
+      "learning_rate": 9.891354923681325e-06,
+      "loss": 0.9585,
+      "step": 721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3297545909881592,
+      "learning_rate": 9.890914442786868e-06,
+      "loss": 0.8972,
+      "step": 722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4062058925628662,
+      "learning_rate": 9.890473080622389e-06,
+      "loss": 0.9513,
+      "step": 723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0028488636016846,
+      "learning_rate": 9.890030837267419e-06,
+      "loss": 1.1133,
+      "step": 724
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.285094141960144,
+      "learning_rate": 9.889587712801641e-06,
+      "loss": 0.9125,
+      "step": 725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.425707459449768,
+      "learning_rate": 9.889143707304897e-06,
+      "loss": 0.91,
+      "step": 726
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3327258825302124,
+      "learning_rate": 9.888698820857197e-06,
+      "loss": 0.9397,
+      "step": 727
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.813483715057373,
+      "learning_rate": 9.888253053538697e-06,
+      "loss": 1.1004,
+      "step": 728
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3534005880355835,
+      "learning_rate": 9.88780640542972e-06,
+      "loss": 0.9431,
+      "step": 729
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.315233826637268,
+      "learning_rate": 9.887358876610745e-06,
+      "loss": 0.9173,
+      "step": 730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3832995891571045,
+      "learning_rate": 9.886910467162409e-06,
+      "loss": 0.9222,
+      "step": 731
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3002419471740723,
+      "learning_rate": 9.886461177165509e-06,
+      "loss": 0.9118,
+      "step": 732
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2939592599868774,
+      "learning_rate": 9.886011006701001e-06,
+      "loss": 0.8283,
+      "step": 733
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.391329288482666,
+      "learning_rate": 9.885559955850001e-06,
+      "loss": 0.8979,
+      "step": 734
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.403883934020996,
+      "learning_rate": 9.885108024693778e-06,
+      "loss": 0.9085,
+      "step": 735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3280601501464844,
+      "learning_rate": 9.884655213313766e-06,
+      "loss": 0.9174,
+      "step": 736
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8006996512413025,
+      "learning_rate": 9.884201521791554e-06,
+      "loss": 1.1058,
+      "step": 737
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4678330421447754,
+      "learning_rate": 9.88374695020889e-06,
+      "loss": 0.8929,
+      "step": 738
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.403671145439148,
+      "learning_rate": 9.88329149864768e-06,
+      "loss": 0.9122,
+      "step": 739
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3127541542053223,
+      "learning_rate": 9.882835167189993e-06,
+      "loss": 0.9493,
+      "step": 740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.320647954940796,
+      "learning_rate": 9.88237795591805e-06,
+      "loss": 0.9528,
+      "step": 741
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3584613800048828,
+      "learning_rate": 9.881919864914233e-06,
+      "loss": 0.9471,
+      "step": 742
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4007147550582886,
+      "learning_rate": 9.881460894261088e-06,
+      "loss": 0.9579,
+      "step": 743
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3660777807235718,
+      "learning_rate": 9.881001044041309e-06,
+      "loss": 0.9042,
+      "step": 744
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9862294793128967,
+      "learning_rate": 9.88054031433776e-06,
+      "loss": 1.1097,
+      "step": 745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.307870626449585,
+      "learning_rate": 9.88007870523345e-06,
+      "loss": 0.8884,
+      "step": 746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6971320509910583,
+      "learning_rate": 9.87961621681156e-06,
+      "loss": 1.124,
+      "step": 747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7526207566261292,
+      "learning_rate": 9.879152849155424e-06,
+      "loss": 1.1211,
+      "step": 748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4040582180023193,
+      "learning_rate": 9.87868860234853e-06,
+      "loss": 0.9854,
+      "step": 749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.419785499572754,
+      "learning_rate": 9.878223476474531e-06,
+      "loss": 0.9023,
+      "step": 750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8395545482635498,
+      "learning_rate": 9.877757471617235e-06,
+      "loss": 1.1111,
+      "step": 751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3280115127563477,
+      "learning_rate": 9.87729058786061e-06,
+      "loss": 0.9082,
+      "step": 752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7241206765174866,
+      "learning_rate": 9.87682282528878e-06,
+      "loss": 1.1238,
+      "step": 753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2947766780853271,
+      "learning_rate": 9.87635418398603e-06,
+      "loss": 0.8981,
+      "step": 754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3356424570083618,
+      "learning_rate": 9.875884664036801e-06,
+      "loss": 0.9182,
+      "step": 755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3048721551895142,
+      "learning_rate": 9.875414265525696e-06,
+      "loss": 0.9105,
+      "step": 756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2781139612197876,
+      "learning_rate": 9.874942988537471e-06,
+      "loss": 0.8762,
+      "step": 757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3027689456939697,
+      "learning_rate": 9.874470833157045e-06,
+      "loss": 0.9157,
+      "step": 758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3622400760650635,
+      "learning_rate": 9.87399779946949e-06,
+      "loss": 0.8759,
+      "step": 759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.203501582145691,
+      "learning_rate": 9.873523887560044e-06,
+      "loss": 1.1133,
+      "step": 760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3335654735565186,
+      "learning_rate": 9.873049097514096e-06,
+      "loss": 0.8954,
+      "step": 761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3803139925003052,
+      "learning_rate": 9.872573429417199e-06,
+      "loss": 0.9212,
+      "step": 762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.37216317653656,
+      "learning_rate": 9.872096883355059e-06,
+      "loss": 0.8615,
+      "step": 763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4121893644332886,
+      "learning_rate": 9.871619459413543e-06,
+      "loss": 0.9381,
+      "step": 764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8170282244682312,
+      "learning_rate": 9.871141157678676e-06,
+      "loss": 1.1238,
+      "step": 765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3291423320770264,
+      "learning_rate": 9.87066197823664e-06,
+      "loss": 0.894,
+      "step": 766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7147064805030823,
+      "learning_rate": 9.870181921173776e-06,
+      "loss": 1.1289,
+      "step": 767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3840837478637695,
+      "learning_rate": 9.869700986576585e-06,
+      "loss": 0.9326,
+      "step": 768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7092385292053223,
+      "learning_rate": 9.869219174531724e-06,
+      "loss": 1.1059,
+      "step": 769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4101241827011108,
+      "learning_rate": 9.868736485126006e-06,
+      "loss": 0.9142,
+      "step": 770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.354705572128296,
+      "learning_rate": 9.868252918446404e-06,
+      "loss": 0.8505,
+      "step": 771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3685630559921265,
+      "learning_rate": 9.867768474580053e-06,
+      "loss": 0.9276,
+      "step": 772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2282153367996216,
+      "learning_rate": 9.867283153614241e-06,
+      "loss": 0.9325,
+      "step": 773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2805473804473877,
+      "learning_rate": 9.866796955636416e-06,
+      "loss": 0.9808,
+      "step": 774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.362629771232605,
+      "learning_rate": 9.866309880734181e-06,
+      "loss": 0.9245,
+      "step": 775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3089823722839355,
+      "learning_rate": 9.865821928995305e-06,
+      "loss": 0.9084,
+      "step": 776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3029990196228027,
+      "learning_rate": 9.865333100507704e-06,
+      "loss": 0.9233,
+      "step": 777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3688348531723022,
+      "learning_rate": 9.864843395359462e-06,
+      "loss": 0.9043,
+      "step": 778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2697064876556396,
+      "learning_rate": 9.864352813638813e-06,
+      "loss": 0.9169,
+      "step": 779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3603781461715698,
+      "learning_rate": 9.863861355434155e-06,
+      "loss": 0.9359,
+      "step": 780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8828596472740173,
+      "learning_rate": 9.863369020834039e-06,
+      "loss": 1.1213,
+      "step": 781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7934764623641968,
+      "learning_rate": 9.86287580992718e-06,
+      "loss": 1.1108,
+      "step": 782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6547415256500244,
+      "learning_rate": 9.862381722802445e-06,
+      "loss": 1.0787,
+      "step": 783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3447428941726685,
+      "learning_rate": 9.861886759548862e-06,
+      "loss": 0.8377,
+      "step": 784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3550069332122803,
+      "learning_rate": 9.861390920255613e-06,
+      "loss": 0.9259,
+      "step": 785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.402618646621704,
+      "learning_rate": 9.860894205012046e-06,
+      "loss": 0.9436,
+      "step": 786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3039226531982422,
+      "learning_rate": 9.860396613907658e-06,
+      "loss": 0.9328,
+      "step": 787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3577814102172852,
+      "learning_rate": 9.859898147032107e-06,
+      "loss": 0.881,
+      "step": 788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2764225006103516,
+      "learning_rate": 9.859398804475214e-06,
+      "loss": 0.9233,
+      "step": 789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3595134019851685,
+      "learning_rate": 9.858898586326947e-06,
+      "loss": 1.1173,
+      "step": 790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.398391604423523,
+      "learning_rate": 9.858397492677443e-06,
+      "loss": 0.8756,
+      "step": 791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9541093111038208,
+      "learning_rate": 9.857895523616988e-06,
+      "loss": 1.1135,
+      "step": 792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4161243438720703,
+      "learning_rate": 9.85739267923603e-06,
+      "loss": 0.8959,
+      "step": 793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3684262037277222,
+      "learning_rate": 9.856888959625177e-06,
+      "loss": 0.9194,
+      "step": 794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8969137072563171,
+      "learning_rate": 9.856384364875189e-06,
+      "loss": 1.1082,
+      "step": 795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8056027889251709,
+      "learning_rate": 9.855878895076986e-06,
+      "loss": 1.0945,
+      "step": 796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6842461228370667,
+      "learning_rate": 9.855372550321648e-06,
+      "loss": 1.1188,
+      "step": 797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4950729608535767,
+      "learning_rate": 9.854865330700411e-06,
+      "loss": 0.9329,
+      "step": 798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4502121210098267,
+      "learning_rate": 9.854357236304666e-06,
+      "loss": 0.8751,
+      "step": 799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.37982177734375,
+      "learning_rate": 9.853848267225967e-06,
+      "loss": 0.9072,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4420177936553955,
+      "learning_rate": 9.85333842355602e-06,
+      "loss": 0.8808,
+      "step": 801
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3189878463745117,
+      "learning_rate": 9.852827705386693e-06,
+      "loss": 0.8616,
+      "step": 802
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4199342727661133,
+      "learning_rate": 9.85231611281001e-06,
+      "loss": 0.8734,
+      "step": 803
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2564648389816284,
+      "learning_rate": 9.85180364591815e-06,
+      "loss": 0.8495,
+      "step": 804
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3986817598342896,
+      "learning_rate": 9.851290304803454e-06,
+      "loss": 0.939,
+      "step": 805
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3764325380325317,
+      "learning_rate": 9.850776089558418e-06,
+      "loss": 0.8862,
+      "step": 806
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3923914432525635,
+      "learning_rate": 9.850261000275695e-06,
+      "loss": 0.9212,
+      "step": 807
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3455246686935425,
+      "learning_rate": 9.8497450370481e-06,
+      "loss": 0.8861,
+      "step": 808
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4592523574829102,
+      "learning_rate": 9.849228199968597e-06,
+      "loss": 0.8879,
+      "step": 809
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7521485090255737,
+      "learning_rate": 9.848710489130314e-06,
+      "loss": 1.1242,
+      "step": 810
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4074522256851196,
+      "learning_rate": 9.848191904626536e-06,
+      "loss": 0.9207,
+      "step": 811
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3955937623977661,
+      "learning_rate": 9.847672446550705e-06,
+      "loss": 0.8772,
+      "step": 812
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.331168293952942,
+      "learning_rate": 9.847152114996417e-06,
+      "loss": 0.8136,
+      "step": 813
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3390915393829346,
+      "learning_rate": 9.846630910057428e-06,
+      "loss": 0.8862,
+      "step": 814
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0296097993850708,
+      "learning_rate": 9.846108831827655e-06,
+      "loss": 1.1147,
+      "step": 815
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.321384072303772,
+      "learning_rate": 9.845585880401165e-06,
+      "loss": 0.9138,
+      "step": 816
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.370473027229309,
+      "learning_rate": 9.845062055872188e-06,
+      "loss": 0.9252,
+      "step": 817
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4552267789840698,
+      "learning_rate": 9.844537358335108e-06,
+      "loss": 0.908,
+      "step": 818
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9070262908935547,
+      "learning_rate": 9.844011787884468e-06,
+      "loss": 1.0944,
+      "step": 819
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8189317584037781,
+      "learning_rate": 9.843485344614969e-06,
+      "loss": 1.1372,
+      "step": 820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7088820338249207,
+      "learning_rate": 9.84295802862147e-06,
+      "loss": 1.1296,
+      "step": 821
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.684767484664917,
+      "learning_rate": 9.842429839998979e-06,
+      "loss": 1.0821,
+      "step": 822
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.389959692955017,
+      "learning_rate": 9.841900778842675e-06,
+      "loss": 0.8876,
+      "step": 823
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4564353227615356,
+      "learning_rate": 9.841370845247881e-06,
+      "loss": 0.9371,
+      "step": 824
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4186900854110718,
+      "learning_rate": 9.84084003931009e-06,
+      "loss": 0.8807,
+      "step": 825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.275727391242981,
+      "learning_rate": 9.840308361124938e-06,
+      "loss": 0.8674,
+      "step": 826
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2786340713500977,
+      "learning_rate": 9.83977581078823e-06,
+      "loss": 1.1294,
+      "step": 827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3801414966583252,
+      "learning_rate": 9.839242388395922e-06,
+      "loss": 0.9276,
+      "step": 828
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3899602890014648,
+      "learning_rate": 9.83870809404413e-06,
+      "loss": 0.943,
+      "step": 829
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3898988962173462,
+      "learning_rate": 9.838172927829123e-06,
+      "loss": 0.9065,
+      "step": 830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3858681917190552,
+      "learning_rate": 9.837636889847334e-06,
+      "loss": 0.8618,
+      "step": 831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.334793210029602,
+      "learning_rate": 9.837099980195348e-06,
+      "loss": 0.8751,
+      "step": 832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2837599515914917,
+      "learning_rate": 9.836562198969906e-06,
+      "loss": 0.8372,
+      "step": 833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3155395984649658,
+      "learning_rate": 9.83602354626791e-06,
+      "loss": 0.9551,
+      "step": 834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4000002145767212,
+      "learning_rate": 9.835484022186418e-06,
+      "loss": 0.8988,
+      "step": 835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1411073207855225,
+      "learning_rate": 9.83494362682264e-06,
+      "loss": 1.1051,
+      "step": 836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2459871768951416,
+      "learning_rate": 9.834402360273952e-06,
+      "loss": 0.9045,
+      "step": 837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7277729511260986,
+      "learning_rate": 9.83386022263788e-06,
+      "loss": 1.1405,
+      "step": 838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7660993337631226,
+      "learning_rate": 9.83331721401211e-06,
+      "loss": 1.1147,
+      "step": 839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6828975677490234,
+      "learning_rate": 9.832773334494483e-06,
+      "loss": 0.918,
+      "step": 840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6007356643676758,
+      "learning_rate": 9.832228584182998e-06,
+      "loss": 0.9308,
+      "step": 841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3910608291625977,
+      "learning_rate": 9.831682963175811e-06,
+      "loss": 0.8587,
+      "step": 842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3229790925979614,
+      "learning_rate": 9.831136471571236e-06,
+      "loss": 0.892,
+      "step": 843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.319281816482544,
+      "learning_rate": 9.830589109467743e-06,
+      "loss": 0.8992,
+      "step": 844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0977191925048828,
+      "learning_rate": 9.830040876963956e-06,
+      "loss": 1.1158,
+      "step": 845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4177289009094238,
+      "learning_rate": 9.829491774158659e-06,
+      "loss": 0.9397,
+      "step": 846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4666606187820435,
+      "learning_rate": 9.828941801150796e-06,
+      "loss": 0.8951,
+      "step": 847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2998011112213135,
+      "learning_rate": 9.828390958039458e-06,
+      "loss": 0.8387,
+      "step": 848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.444300889968872,
+      "learning_rate": 9.827839244923903e-06,
+      "loss": 0.92,
+      "step": 849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3749456405639648,
+      "learning_rate": 9.827286661903539e-06,
+      "loss": 0.9266,
+      "step": 850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3582322597503662,
+      "learning_rate": 9.826733209077934e-06,
+      "loss": 0.9174,
+      "step": 851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3099451065063477,
+      "learning_rate": 9.826178886546813e-06,
+      "loss": 0.9515,
+      "step": 852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3631728887557983,
+      "learning_rate": 9.825623694410056e-06,
+      "loss": 0.8852,
+      "step": 853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0469313859939575,
+      "learning_rate": 9.8250676327677e-06,
+      "loss": 1.1157,
+      "step": 854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3978304862976074,
+      "learning_rate": 9.824510701719938e-06,
+      "loss": 0.9242,
+      "step": 855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4113707542419434,
+      "learning_rate": 9.823952901367124e-06,
+      "loss": 0.8896,
+      "step": 856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4157973527908325,
+      "learning_rate": 9.823394231809765e-06,
+      "loss": 0.8912,
+      "step": 857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7479792237281799,
+      "learning_rate": 9.82283469314852e-06,
+      "loss": 1.0996,
+      "step": 858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4014570713043213,
+      "learning_rate": 9.822274285484215e-06,
+      "loss": 0.9234,
+      "step": 859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.700994610786438,
+      "learning_rate": 9.821713008917822e-06,
+      "loss": 1.1121,
+      "step": 860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4814058542251587,
+      "learning_rate": 9.821150863550481e-06,
+      "loss": 0.8712,
+      "step": 861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7227991819381714,
+      "learning_rate": 9.820587849483476e-06,
+      "loss": 1.1134,
+      "step": 862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6746354103088379,
+      "learning_rate": 9.82002396681826e-06,
+      "loss": 1.0885,
+      "step": 863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3871876001358032,
+      "learning_rate": 9.81945921565643e-06,
+      "loss": 0.9191,
+      "step": 864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7334367632865906,
+      "learning_rate": 9.81889359609975e-06,
+      "loss": 1.105,
+      "step": 865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6942464113235474,
+      "learning_rate": 9.818327108250134e-06,
+      "loss": 1.1008,
+      "step": 866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3996092081069946,
+      "learning_rate": 9.817759752209658e-06,
+      "loss": 0.9741,
+      "step": 867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7424430847167969,
+      "learning_rate": 9.817191528080545e-06,
+      "loss": 1.1234,
+      "step": 868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3791247606277466,
+      "learning_rate": 9.816622435965187e-06,
+      "loss": 0.855,
+      "step": 869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7496314644813538,
+      "learning_rate": 9.816052475966122e-06,
+      "loss": 1.1082,
+      "step": 870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2771210670471191,
+      "learning_rate": 9.81548164818605e-06,
+      "loss": 0.8751,
+      "step": 871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2822200059890747,
+      "learning_rate": 9.814909952727825e-06,
+      "loss": 0.8989,
+      "step": 872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3620318174362183,
+      "learning_rate": 9.814337389694458e-06,
+      "loss": 0.919,
+      "step": 873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2557100057601929,
+      "learning_rate": 9.813763959189115e-06,
+      "loss": 0.8745,
+      "step": 874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3286696672439575,
+      "learning_rate": 9.813189661315123e-06,
+      "loss": 0.891,
+      "step": 875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.06153404712677,
+      "learning_rate": 9.812614496175961e-06,
+      "loss": 1.1073,
+      "step": 876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2568786144256592,
+      "learning_rate": 9.812038463875262e-06,
+      "loss": 0.8701,
+      "step": 877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6660780310630798,
+      "learning_rate": 9.811461564516821e-06,
+      "loss": 1.0949,
+      "step": 878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8087523579597473,
+      "learning_rate": 9.810883798204588e-06,
+      "loss": 1.1055,
+      "step": 879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7400908470153809,
+      "learning_rate": 9.810305165042666e-06,
+      "loss": 1.1101,
+      "step": 880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3709179162979126,
+      "learning_rate": 9.809725665135314e-06,
+      "loss": 0.9064,
+      "step": 881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3831565380096436,
+      "learning_rate": 9.809145298586954e-06,
+      "loss": 0.8897,
+      "step": 882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0454497337341309,
+      "learning_rate": 9.808564065502157e-06,
+      "loss": 1.1134,
+      "step": 883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3147666454315186,
+      "learning_rate": 9.807981965985653e-06,
+      "loss": 0.9106,
+      "step": 884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2900927066802979,
+      "learning_rate": 9.807399000142326e-06,
+      "loss": 0.8953,
+      "step": 885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2813183069229126,
+      "learning_rate": 9.806815168077219e-06,
+      "loss": 0.885,
+      "step": 886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3327758312225342,
+      "learning_rate": 9.806230469895532e-06,
+      "loss": 0.8995,
+      "step": 887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3881449699401855,
+      "learning_rate": 9.805644905702615e-06,
+      "loss": 0.9044,
+      "step": 888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3176392316818237,
+      "learning_rate": 9.805058475603979e-06,
+      "loss": 0.8627,
+      "step": 889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.943459689617157,
+      "learning_rate": 9.804471179705293e-06,
+      "loss": 1.0955,
+      "step": 890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2853516340255737,
+      "learning_rate": 9.803883018112376e-06,
+      "loss": 0.8953,
+      "step": 891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7421061396598816,
+      "learning_rate": 9.803293990931206e-06,
+      "loss": 1.124,
+      "step": 892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3171552419662476,
+      "learning_rate": 9.802704098267917e-06,
+      "loss": 0.9098,
+      "step": 893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.30195152759552,
+      "learning_rate": 9.8021133402288e-06,
+      "loss": 0.8559,
+      "step": 894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3548176288604736,
+      "learning_rate": 9.801521716920303e-06,
+      "loss": 0.89,
+      "step": 895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3332535028457642,
+      "learning_rate": 9.800929228449023e-06,
+      "loss": 0.8899,
+      "step": 896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3899776935577393,
+      "learning_rate": 9.80033587492172e-06,
+      "loss": 0.8997,
+      "step": 897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.285434365272522,
+      "learning_rate": 9.799741656445307e-06,
+      "loss": 1.0863,
+      "step": 898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3082239627838135,
+      "learning_rate": 9.799146573126853e-06,
+      "loss": 0.9238,
+      "step": 899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3742952346801758,
+      "learning_rate": 9.798550625073584e-06,
+      "loss": 0.8993,
+      "step": 900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3441476821899414,
+      "learning_rate": 9.79795381239288e-06,
+      "loss": 0.8677,
+      "step": 901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3291044235229492,
+      "learning_rate": 9.79735613519228e-06,
+      "loss": 0.8582,
+      "step": 902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3673014640808105,
+      "learning_rate": 9.796757593579473e-06,
+      "loss": 0.8786,
+      "step": 903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3296339511871338,
+      "learning_rate": 9.79615818766231e-06,
+      "loss": 0.8804,
+      "step": 904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3106472492218018,
+      "learning_rate": 9.795557917548796e-06,
+      "loss": 0.8208,
+      "step": 905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2618303298950195,
+      "learning_rate": 9.794956783347088e-06,
+      "loss": 1.114,
+      "step": 906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9077434539794922,
+      "learning_rate": 9.794354785165501e-06,
+      "loss": 1.1015,
+      "step": 907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6914875507354736,
+      "learning_rate": 9.793751923112508e-06,
+      "loss": 1.1116,
+      "step": 908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8955706357955933,
+      "learning_rate": 9.793148197296737e-06,
+      "loss": 1.0867,
+      "step": 909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.574976921081543,
+      "learning_rate": 9.792543607826968e-06,
+      "loss": 0.8966,
+      "step": 910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9658063650131226,
+      "learning_rate": 9.79193815481214e-06,
+      "loss": 1.1141,
+      "step": 911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4675484895706177,
+      "learning_rate": 9.791331838361345e-06,
+      "loss": 0.969,
+      "step": 912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3620882034301758,
+      "learning_rate": 9.790724658583834e-06,
+      "loss": 0.9241,
+      "step": 913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2954328060150146,
+      "learning_rate": 9.790116615589011e-06,
+      "loss": 0.9415,
+      "step": 914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2754459381103516,
+      "learning_rate": 9.789507709486439e-06,
+      "loss": 0.8316,
+      "step": 915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.363478422164917,
+      "learning_rate": 9.788897940385828e-06,
+      "loss": 0.879,
+      "step": 916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.323106288909912,
+      "learning_rate": 9.788287308397056e-06,
+      "loss": 0.8814,
+      "step": 917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3126457929611206,
+      "learning_rate": 9.787675813630145e-06,
+      "loss": 0.8698,
+      "step": 918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3239237070083618,
+      "learning_rate": 9.787063456195278e-06,
+      "loss": 0.9453,
+      "step": 919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4531527757644653,
+      "learning_rate": 9.786450236202796e-06,
+      "loss": 1.0911,
+      "step": 920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3522270917892456,
+      "learning_rate": 9.785836153763189e-06,
+      "loss": 0.8923,
+      "step": 921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8278368711471558,
+      "learning_rate": 9.785221208987104e-06,
+      "loss": 1.1288,
+      "step": 922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7372597455978394,
+      "learning_rate": 9.78460540198535e-06,
+      "loss": 1.0823,
+      "step": 923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4672778844833374,
+      "learning_rate": 9.783988732868881e-06,
+      "loss": 0.8567,
+      "step": 924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1153934001922607,
+      "learning_rate": 9.783371201748815e-06,
+      "loss": 1.1127,
+      "step": 925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8929865956306458,
+      "learning_rate": 9.782752808736422e-06,
+      "loss": 1.1097,
+      "step": 926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4206846952438354,
+      "learning_rate": 9.782133553943125e-06,
+      "loss": 0.8492,
+      "step": 927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4169299602508545,
+      "learning_rate": 9.781513437480505e-06,
+      "loss": 0.9297,
+      "step": 928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3908072710037231,
+      "learning_rate": 9.780892459460298e-06,
+      "loss": 0.9036,
+      "step": 929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6831588745117188,
+      "learning_rate": 9.780270619994397e-06,
+      "loss": 1.1006,
+      "step": 930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3913823366165161,
+      "learning_rate": 9.779647919194846e-06,
+      "loss": 0.9494,
+      "step": 931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3427947759628296,
+      "learning_rate": 9.779024357173848e-06,
+      "loss": 0.9358,
+      "step": 932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3139034509658813,
+      "learning_rate": 9.778399934043756e-06,
+      "loss": 0.8375,
+      "step": 933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3166587352752686,
+      "learning_rate": 9.777774649917086e-06,
+      "loss": 0.8748,
+      "step": 934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2857178449630737,
+      "learning_rate": 9.777148504906504e-06,
+      "loss": 0.8615,
+      "step": 935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3704177141189575,
+      "learning_rate": 9.77652149912483e-06,
+      "loss": 0.931,
+      "step": 936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3921277523040771,
+      "learning_rate": 9.775893632685043e-06,
+      "loss": 0.9108,
+      "step": 937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3841384649276733,
+      "learning_rate": 9.775264905700275e-06,
+      "loss": 0.8184,
+      "step": 938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3511210680007935,
+      "learning_rate": 9.774635318283811e-06,
+      "loss": 0.9183,
+      "step": 939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3351964950561523,
+      "learning_rate": 9.774004870549097e-06,
+      "loss": 0.8911,
+      "step": 940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1216524839401245,
+      "learning_rate": 9.773373562609729e-06,
+      "loss": 1.1082,
+      "step": 941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8596928119659424,
+      "learning_rate": 9.772741394579457e-06,
+      "loss": 1.0954,
+      "step": 942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4232929944992065,
+      "learning_rate": 9.772108366572191e-06,
+      "loss": 0.8892,
+      "step": 943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3949291706085205,
+      "learning_rate": 9.771474478701992e-06,
+      "loss": 0.8635,
+      "step": 944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3991998434066772,
+      "learning_rate": 9.770839731083076e-06,
+      "loss": 0.8662,
+      "step": 945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3209315538406372,
+      "learning_rate": 9.770204123829818e-06,
+      "loss": 0.8853,
+      "step": 946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2774323225021362,
+      "learning_rate": 9.769567657056742e-06,
+      "loss": 0.8537,
+      "step": 947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3018674850463867,
+      "learning_rate": 9.768930330878533e-06,
+      "loss": 0.92,
+      "step": 948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4322720766067505,
+      "learning_rate": 9.768292145410025e-06,
+      "loss": 0.9177,
+      "step": 949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.070636749267578,
+      "learning_rate": 9.767653100766212e-06,
+      "loss": 1.1002,
+      "step": 950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6578713655471802,
+      "learning_rate": 9.767013197062238e-06,
+      "loss": 1.0833,
+      "step": 951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.921327531337738,
+      "learning_rate": 9.766372434413404e-06,
+      "loss": 1.095,
+      "step": 952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7495396137237549,
+      "learning_rate": 9.765730812935167e-06,
+      "loss": 0.9603,
+      "step": 953
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6756030321121216,
+      "learning_rate": 9.765088332743137e-06,
+      "loss": 0.8357,
+      "step": 954
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5314592123031616,
+      "learning_rate": 9.764444993953082e-06,
+      "loss": 0.9243,
+      "step": 955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.409179449081421,
+      "learning_rate": 9.763800796680919e-06,
+      "loss": 0.8471,
+      "step": 956
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4916013479232788,
+      "learning_rate": 9.763155741042724e-06,
+      "loss": 0.9082,
+      "step": 957
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4974182844161987,
+      "learning_rate": 9.762509827154725e-06,
+      "loss": 0.9658,
+      "step": 958
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4910370111465454,
+      "learning_rate": 9.761863055133308e-06,
+      "loss": 0.8495,
+      "step": 959
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0384650230407715,
+      "learning_rate": 9.761215425095013e-06,
+      "loss": 1.1449,
+      "step": 960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4574644565582275,
+      "learning_rate": 9.760566937156527e-06,
+      "loss": 0.8883,
+      "step": 961
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3986386060714722,
+      "learning_rate": 9.759917591434705e-06,
+      "loss": 0.8968,
+      "step": 962
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5176039934158325,
+      "learning_rate": 9.759267388046545e-06,
+      "loss": 0.9197,
+      "step": 963
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4728986024856567,
+      "learning_rate": 9.758616327109207e-06,
+      "loss": 0.843,
+      "step": 964
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.490685224533081,
+      "learning_rate": 9.75796440874e-06,
+      "loss": 0.8629,
+      "step": 965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.231182336807251,
+      "learning_rate": 9.757311633056393e-06,
+      "loss": 1.1,
+      "step": 966
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3866735696792603,
+      "learning_rate": 9.756658000176001e-06,
+      "loss": 0.8513,
+      "step": 967
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.418514609336853,
+      "learning_rate": 9.756003510216605e-06,
+      "loss": 0.9718,
+      "step": 968
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3869946002960205,
+      "learning_rate": 9.75534816329613e-06,
+      "loss": 0.9411,
+      "step": 969
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3889185190200806,
+      "learning_rate": 9.754691959532664e-06,
+      "loss": 0.8783,
+      "step": 970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4792498350143433,
+      "learning_rate": 9.75403489904444e-06,
+      "loss": 0.9586,
+      "step": 971
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3141164779663086,
+      "learning_rate": 9.753376981949854e-06,
+      "loss": 0.9548,
+      "step": 972
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2880481481552124,
+      "learning_rate": 9.752718208367454e-06,
+      "loss": 0.8497,
+      "step": 973
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3423330783843994,
+      "learning_rate": 9.752058578415938e-06,
+      "loss": 0.9304,
+      "step": 974
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3311506509780884,
+      "learning_rate": 9.751398092214162e-06,
+      "loss": 0.8709,
+      "step": 975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1253836154937744,
+      "learning_rate": 9.750736749881136e-06,
+      "loss": 1.1171,
+      "step": 976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8899155855178833,
+      "learning_rate": 9.750074551536026e-06,
+      "loss": 1.0837,
+      "step": 977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5204510688781738,
+      "learning_rate": 9.749411497298148e-06,
+      "loss": 0.848,
+      "step": 978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8055835962295532,
+      "learning_rate": 9.748747587286976e-06,
+      "loss": 1.1097,
+      "step": 979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4942314624786377,
+      "learning_rate": 9.748082821622137e-06,
+      "loss": 0.901,
+      "step": 980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9159333109855652,
+      "learning_rate": 9.747417200423409e-06,
+      "loss": 1.0757,
+      "step": 981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.338801622390747,
+      "learning_rate": 9.746750723810729e-06,
+      "loss": 0.8844,
+      "step": 982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7683328986167908,
+      "learning_rate": 9.74608339190419e-06,
+      "loss": 1.1114,
+      "step": 983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2722275257110596,
+      "learning_rate": 9.745415204824026e-06,
+      "loss": 0.8798,
+      "step": 984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.402712345123291,
+      "learning_rate": 9.744746162690644e-06,
+      "loss": 0.8857,
+      "step": 985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4178311824798584,
+      "learning_rate": 9.744076265624589e-06,
+      "loss": 0.9161,
+      "step": 986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9016255140304565,
+      "learning_rate": 9.74340551374657e-06,
+      "loss": 1.1046,
+      "step": 987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2652454376220703,
+      "learning_rate": 9.742733907177446e-06,
+      "loss": 0.8729,
+      "step": 988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7513898611068726,
+      "learning_rate": 9.742061446038229e-06,
+      "loss": 1.0904,
+      "step": 989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7182430624961853,
+      "learning_rate": 9.741388130450086e-06,
+      "loss": 1.0675,
+      "step": 990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4111249446868896,
+      "learning_rate": 9.740713960534341e-06,
+      "loss": 0.8449,
+      "step": 991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7964020371437073,
+      "learning_rate": 9.740038936412467e-06,
+      "loss": 1.1206,
+      "step": 992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.41074800491333,
+      "learning_rate": 9.739363058206096e-06,
+      "loss": 0.9422,
+      "step": 993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.679141104221344,
+      "learning_rate": 9.738686326037009e-06,
+      "loss": 1.0953,
+      "step": 994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6632534265518188,
+      "learning_rate": 9.738008740027144e-06,
+      "loss": 1.111,
+      "step": 995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3533493280410767,
+      "learning_rate": 9.73733030029859e-06,
+      "loss": 0.8688,
+      "step": 996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.308886170387268,
+      "learning_rate": 9.736651006973597e-06,
+      "loss": 0.8639,
+      "step": 997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.800689160823822,
+      "learning_rate": 9.735970860174556e-06,
+      "loss": 1.1097,
+      "step": 998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3439810276031494,
+      "learning_rate": 9.735289860024025e-06,
+      "loss": 0.8764,
+      "step": 999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2513384819030762,
+      "learning_rate": 9.734608006644708e-06,
+      "loss": 0.8472,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2657442092895508,
+      "learning_rate": 9.733925300159466e-06,
+      "loss": 0.8748,
+      "step": 1001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3224653005599976,
+      "learning_rate": 9.733241740691312e-06,
+      "loss": 0.8711,
+      "step": 1002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3591641187667847,
+      "learning_rate": 9.732557328363413e-06,
+      "loss": 0.918,
+      "step": 1003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8184081315994263,
+      "learning_rate": 9.731872063299088e-06,
+      "loss": 1.0809,
+      "step": 1004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.287413239479065,
+      "learning_rate": 9.731185945621814e-06,
+      "loss": 0.894,
+      "step": 1005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7455149292945862,
+      "learning_rate": 9.73049897545522e-06,
+      "loss": 1.0744,
+      "step": 1006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.355089545249939,
+      "learning_rate": 9.729811152923086e-06,
+      "loss": 0.8923,
+      "step": 1007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3711113929748535,
+      "learning_rate": 9.729122478149346e-06,
+      "loss": 0.9405,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2691717147827148,
+      "learning_rate": 9.728432951258094e-06,
+      "loss": 0.8476,
+      "step": 1009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.292731523513794,
+      "learning_rate": 9.727742572373566e-06,
+      "loss": 0.8761,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3416204452514648,
+      "learning_rate": 9.727051341620164e-06,
+      "loss": 0.931,
+      "step": 1011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3341429233551025,
+      "learning_rate": 9.726359259122434e-06,
+      "loss": 0.8617,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3206145763397217,
+      "learning_rate": 9.72566632500508e-06,
+      "loss": 0.9442,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.276137351989746,
+      "learning_rate": 9.724972539392957e-06,
+      "loss": 0.8839,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3275524377822876,
+      "learning_rate": 9.724277902411077e-06,
+      "loss": 0.8578,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3516069650650024,
+      "learning_rate": 9.723582414184602e-06,
+      "loss": 0.8979,
+      "step": 1016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3075344562530518,
+      "learning_rate": 9.722886074838849e-06,
+      "loss": 0.887,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3517531156539917,
+      "learning_rate": 9.722188884499288e-06,
+      "loss": 0.8372,
+      "step": 1018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1386748552322388,
+      "learning_rate": 9.721490843291543e-06,
+      "loss": 1.0936,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4362287521362305,
+      "learning_rate": 9.720791951341388e-06,
+      "loss": 0.8815,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8074418902397156,
+      "learning_rate": 9.720092208774756e-06,
+      "loss": 1.0934,
+      "step": 1021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.377079963684082,
+      "learning_rate": 9.719391615717729e-06,
+      "loss": 0.8922,
+      "step": 1022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3184384107589722,
+      "learning_rate": 9.718690172296544e-06,
+      "loss": 0.8646,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3054969310760498,
+      "learning_rate": 9.71798787863759e-06,
+      "loss": 0.8893,
+      "step": 1024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.321425437927246,
+      "learning_rate": 9.71728473486741e-06,
+      "loss": 0.8303,
+      "step": 1025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2818922996520996,
+      "learning_rate": 9.716580741112698e-06,
+      "loss": 0.9033,
+      "step": 1026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3421045541763306,
+      "learning_rate": 9.715875897500307e-06,
+      "loss": 0.9493,
+      "step": 1027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2930256128311157,
+      "learning_rate": 9.715170204157239e-06,
+      "loss": 0.9124,
+      "step": 1028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.352359414100647,
+      "learning_rate": 9.714463661210645e-06,
+      "loss": 1.1128,
+      "step": 1029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.387694239616394,
+      "learning_rate": 9.713756268787837e-06,
+      "loss": 0.9057,
+      "step": 1030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3820563554763794,
+      "learning_rate": 9.713048027016278e-06,
+      "loss": 0.9643,
+      "step": 1031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.341307520866394,
+      "learning_rate": 9.71233893602358e-06,
+      "loss": 0.9236,
+      "step": 1032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2902225255966187,
+      "learning_rate": 9.711628995937513e-06,
+      "loss": 0.9283,
+      "step": 1033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3425657749176025,
+      "learning_rate": 9.710918206885993e-06,
+      "loss": 0.901,
+      "step": 1034
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2869259119033813,
+      "learning_rate": 9.710206568997098e-06,
+      "loss": 0.8795,
+      "step": 1035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3430553674697876,
+      "learning_rate": 9.709494082399053e-06,
+      "loss": 0.9078,
+      "step": 1036
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3185373544692993,
+      "learning_rate": 9.708780747220238e-06,
+      "loss": 0.8582,
+      "step": 1037
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2303787469863892,
+      "learning_rate": 9.708066563589185e-06,
+      "loss": 1.1053,
+      "step": 1038
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.316267490386963,
+      "learning_rate": 9.707351531634579e-06,
+      "loss": 0.8806,
+      "step": 1039
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2983468770980835,
+      "learning_rate": 9.70663565148526e-06,
+      "loss": 0.8647,
+      "step": 1040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7597914934158325,
+      "learning_rate": 9.705918923270216e-06,
+      "loss": 1.0972,
+      "step": 1041
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.769464373588562,
+      "learning_rate": 9.705201347118592e-06,
+      "loss": 1.0976,
+      "step": 1042
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3396140336990356,
+      "learning_rate": 9.704482923159685e-06,
+      "loss": 0.8804,
+      "step": 1043
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4534107446670532,
+      "learning_rate": 9.703763651522944e-06,
+      "loss": 0.8927,
+      "step": 1044
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.427994728088379,
+      "learning_rate": 9.703043532337972e-06,
+      "loss": 0.902,
+      "step": 1045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3250004053115845,
+      "learning_rate": 9.702322565734524e-06,
+      "loss": 0.9025,
+      "step": 1046
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3217374086380005,
+      "learning_rate": 9.701600751842504e-06,
+      "loss": 0.8324,
+      "step": 1047
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1454473733901978,
+      "learning_rate": 9.700878090791975e-06,
+      "loss": 1.1135,
+      "step": 1048
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4145585298538208,
+      "learning_rate": 9.70015458271315e-06,
+      "loss": 0.9549,
+      "step": 1049
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3167780637741089,
+      "learning_rate": 9.699430227736393e-06,
+      "loss": 0.8417,
+      "step": 1050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.286990761756897,
+      "learning_rate": 9.698705025992222e-06,
+      "loss": 0.9032,
+      "step": 1051
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4566078186035156,
+      "learning_rate": 9.697978977611308e-06,
+      "loss": 0.9323,
+      "step": 1052
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9970657825469971,
+      "learning_rate": 9.697252082724475e-06,
+      "loss": 1.0857,
+      "step": 1053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.282199501991272,
+      "learning_rate": 9.696524341462698e-06,
+      "loss": 0.9052,
+      "step": 1054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6728658676147461,
+      "learning_rate": 9.695795753957105e-06,
+      "loss": 1.095,
+      "step": 1055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6614185571670532,
+      "learning_rate": 9.695066320338977e-06,
+      "loss": 1.0959,
+      "step": 1056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7020531296730042,
+      "learning_rate": 9.694336040739747e-06,
+      "loss": 1.119,
+      "step": 1057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4104490280151367,
+      "learning_rate": 9.693604915291e-06,
+      "loss": 0.9178,
+      "step": 1058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7003157734870911,
+      "learning_rate": 9.692872944124473e-06,
+      "loss": 1.0616,
+      "step": 1059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6680921316146851,
+      "learning_rate": 9.69214012737206e-06,
+      "loss": 1.0973,
+      "step": 1060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.38895845413208,
+      "learning_rate": 9.6914064651658e-06,
+      "loss": 0.8852,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8099476099014282,
+      "learning_rate": 9.690671957637889e-06,
+      "loss": 1.1016,
+      "step": 1062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.431815505027771,
+      "learning_rate": 9.689936604920675e-06,
+      "loss": 0.9405,
+      "step": 1063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.293208360671997,
+      "learning_rate": 9.689200407146658e-06,
+      "loss": 0.8475,
+      "step": 1064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2978321313858032,
+      "learning_rate": 9.688463364448486e-06,
+      "loss": 0.9292,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7683027386665344,
+      "learning_rate": 9.68772547695897e-06,
+      "loss": 1.0699,
+      "step": 1066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7235736846923828,
+      "learning_rate": 9.68698674481106e-06,
+      "loss": 1.0802,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3511606454849243,
+      "learning_rate": 9.686247168137867e-06,
+      "loss": 0.8622,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.226919412612915,
+      "learning_rate": 9.685506747072655e-06,
+      "loss": 0.8452,
+      "step": 1069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9355242848396301,
+      "learning_rate": 9.68476548174883e-06,
+      "loss": 1.0826,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3155908584594727,
+      "learning_rate": 9.684023372299963e-06,
+      "loss": 0.9133,
+      "step": 1071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3269561529159546,
+      "learning_rate": 9.683280418859769e-06,
+      "loss": 0.9217,
+      "step": 1072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2730751037597656,
+      "learning_rate": 9.682536621562115e-06,
+      "loss": 0.8728,
+      "step": 1073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3912400007247925,
+      "learning_rate": 9.681791980541025e-06,
+      "loss": 0.8875,
+      "step": 1074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.565213918685913,
+      "learning_rate": 9.68104649593067e-06,
+      "loss": 0.8659,
+      "step": 1075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3085055351257324,
+      "learning_rate": 9.68030016786538e-06,
+      "loss": 0.9076,
+      "step": 1076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3263310194015503,
+      "learning_rate": 9.679552996479625e-06,
+      "loss": 0.8792,
+      "step": 1077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3487540483474731,
+      "learning_rate": 9.67880498190804e-06,
+      "loss": 0.8545,
+      "step": 1078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2758139371871948,
+      "learning_rate": 9.678056124285405e-06,
+      "loss": 0.842,
+      "step": 1079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9471149444580078,
+      "learning_rate": 9.677306423746649e-06,
+      "loss": 1.0987,
+      "step": 1080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.755728542804718,
+      "learning_rate": 9.676555880426862e-06,
+      "loss": 1.1041,
+      "step": 1081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3139748573303223,
+      "learning_rate": 9.675804494461279e-06,
+      "loss": 0.8628,
+      "step": 1082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3995956182479858,
+      "learning_rate": 9.675052265985288e-06,
+      "loss": 0.9157,
+      "step": 1083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3195377588272095,
+      "learning_rate": 9.67429919513443e-06,
+      "loss": 0.9035,
+      "step": 1084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.212686538696289,
+      "learning_rate": 9.673545282044395e-06,
+      "loss": 1.0923,
+      "step": 1085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.375620722770691,
+      "learning_rate": 9.67279052685103e-06,
+      "loss": 0.8767,
+      "step": 1086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2780696153640747,
+      "learning_rate": 9.672034929690328e-06,
+      "loss": 0.8194,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8213459253311157,
+      "learning_rate": 9.67127849069844e-06,
+      "loss": 1.0911,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4527965784072876,
+      "learning_rate": 9.670521210011662e-06,
+      "loss": 0.8757,
+      "step": 1089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8058233857154846,
+      "learning_rate": 9.669763087766444e-06,
+      "loss": 1.1078,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2183420658111572,
+      "learning_rate": 9.669004124099393e-06,
+      "loss": 0.8544,
+      "step": 1091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8058820366859436,
+      "learning_rate": 9.668244319147258e-06,
+      "loss": 1.0883,
+      "step": 1092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.259606122970581,
+      "learning_rate": 9.667483673046946e-06,
+      "loss": 0.8514,
+      "step": 1093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2961606979370117,
+      "learning_rate": 9.666722185935514e-06,
+      "loss": 0.8635,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9213928580284119,
+      "learning_rate": 9.665959857950172e-06,
+      "loss": 1.1035,
+      "step": 1095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3604215383529663,
+      "learning_rate": 9.665196689228281e-06,
+      "loss": 0.8862,
+      "step": 1096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4049075841903687,
+      "learning_rate": 9.664432679907347e-06,
+      "loss": 0.8832,
+      "step": 1097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3904292583465576,
+      "learning_rate": 9.663667830125041e-06,
+      "loss": 0.9507,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0167733430862427,
+      "learning_rate": 9.662902140019172e-06,
+      "loss": 1.0966,
+      "step": 1099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3304858207702637,
+      "learning_rate": 9.662135609727708e-06,
+      "loss": 0.8925,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.311123251914978,
+      "learning_rate": 9.661368239388764e-06,
+      "loss": 0.8619,
+      "step": 1101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.450952410697937,
+      "learning_rate": 9.660600029140612e-06,
+      "loss": 0.9693,
+      "step": 1102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7415351867675781,
+      "learning_rate": 9.659830979121671e-06,
+      "loss": 1.0874,
+      "step": 1103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.371199369430542,
+      "learning_rate": 9.659061089470512e-06,
+      "loss": 0.9012,
+      "step": 1104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6902478337287903,
+      "learning_rate": 9.658290360325859e-06,
+      "loss": 1.0792,
+      "step": 1105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7478721141815186,
+      "learning_rate": 9.657518791826583e-06,
+      "loss": 0.9108,
+      "step": 1106
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3023587465286255,
+      "learning_rate": 9.656746384111711e-06,
+      "loss": 0.8814,
+      "step": 1107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2429667711257935,
+      "learning_rate": 9.655973137320419e-06,
+      "loss": 0.8566,
+      "step": 1108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3905569314956665,
+      "learning_rate": 9.655199051592034e-06,
+      "loss": 0.9015,
+      "step": 1109
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3471429347991943,
+      "learning_rate": 9.654424127066037e-06,
+      "loss": 0.9135,
+      "step": 1110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8538874387741089,
+      "learning_rate": 9.653648363882057e-06,
+      "loss": 1.1,
+      "step": 1111
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.343463659286499,
+      "learning_rate": 9.652871762179874e-06,
+      "loss": 0.9032,
+      "step": 1112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3392776250839233,
+      "learning_rate": 9.65209432209942e-06,
+      "loss": 0.8636,
+      "step": 1113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7041798830032349,
+      "learning_rate": 9.651316043780779e-06,
+      "loss": 1.1044,
+      "step": 1114
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3258178234100342,
+      "learning_rate": 9.650536927364183e-06,
+      "loss": 0.8894,
+      "step": 1115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6806868314743042,
+      "learning_rate": 9.649756972990021e-06,
+      "loss": 1.0903,
+      "step": 1116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.313660979270935,
+      "learning_rate": 9.648976180798826e-06,
+      "loss": 0.8364,
+      "step": 1117
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6701874136924744,
+      "learning_rate": 9.648194550931287e-06,
+      "loss": 1.0856,
+      "step": 1118
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6704904437065125,
+      "learning_rate": 9.647412083528241e-06,
+      "loss": 1.0854,
+      "step": 1119
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6450378894805908,
+      "learning_rate": 9.646628778730679e-06,
+      "loss": 1.1052,
+      "step": 1120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.628193736076355,
+      "learning_rate": 9.645844636679735e-06,
+      "loss": 1.0851,
+      "step": 1121
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3209068775177002,
+      "learning_rate": 9.645059657516706e-06,
+      "loss": 0.926,
+      "step": 1122
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.332985758781433,
+      "learning_rate": 9.644273841383032e-06,
+      "loss": 0.8469,
+      "step": 1123
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7050449252128601,
+      "learning_rate": 9.643487188420307e-06,
+      "loss": 1.0617,
+      "step": 1124
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7183394432067871,
+      "learning_rate": 9.642699698770268e-06,
+      "loss": 1.0645,
+      "step": 1125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2775408029556274,
+      "learning_rate": 9.641911372574814e-06,
+      "loss": 0.9302,
+      "step": 1126
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6548232436180115,
+      "learning_rate": 9.64112220997599e-06,
+      "loss": 1.0667,
+      "step": 1127
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3561080694198608,
+      "learning_rate": 9.64033221111599e-06,
+      "loss": 1.003,
+      "step": 1128
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3125938177108765,
+      "learning_rate": 9.639541376137159e-06,
+      "loss": 0.9084,
+      "step": 1129
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.305907130241394,
+      "learning_rate": 9.638749705181995e-06,
+      "loss": 0.8739,
+      "step": 1130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2788923978805542,
+      "learning_rate": 9.637957198393145e-06,
+      "loss": 0.8799,
+      "step": 1131
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2828041315078735,
+      "learning_rate": 9.637163855913407e-06,
+      "loss": 0.8431,
+      "step": 1132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2893279790878296,
+      "learning_rate": 9.63636967788573e-06,
+      "loss": 0.934,
+      "step": 1133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2828775644302368,
+      "learning_rate": 9.635574664453212e-06,
+      "loss": 0.8662,
+      "step": 1134
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.271040678024292,
+      "learning_rate": 9.634778815759103e-06,
+      "loss": 0.9014,
+      "step": 1135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3261120319366455,
+      "learning_rate": 9.633982131946803e-06,
+      "loss": 0.8937,
+      "step": 1136
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3481199741363525,
+      "learning_rate": 9.633184613159863e-06,
+      "loss": 0.9171,
+      "step": 1137
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2496123313903809,
+      "learning_rate": 9.632386259541984e-06,
+      "loss": 0.8486,
+      "step": 1138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2919973134994507,
+      "learning_rate": 9.631587071237016e-06,
+      "loss": 0.8688,
+      "step": 1139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.249024748802185,
+      "learning_rate": 9.630787048388965e-06,
+      "loss": 0.8282,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2986634969711304,
+      "learning_rate": 9.629986191141978e-06,
+      "loss": 0.906,
+      "step": 1141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2693885564804077,
+      "learning_rate": 9.62918449964036e-06,
+      "loss": 0.9357,
+      "step": 1142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3233128786087036,
+      "learning_rate": 9.628381974028564e-06,
+      "loss": 0.9068,
+      "step": 1143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2957978248596191,
+      "learning_rate": 9.627578614451194e-06,
+      "loss": 0.8798,
+      "step": 1144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.262910008430481,
+      "learning_rate": 9.626774421053002e-06,
+      "loss": 0.8585,
+      "step": 1145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8963253498077393,
+      "learning_rate": 9.62596939397889e-06,
+      "loss": 1.0841,
+      "step": 1146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2820640802383423,
+      "learning_rate": 9.625163533373917e-06,
+      "loss": 0.8606,
+      "step": 1147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3561351299285889,
+      "learning_rate": 9.624356839383283e-06,
+      "loss": 0.864,
+      "step": 1148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.783988893032074,
+      "learning_rate": 9.623549312152342e-06,
+      "loss": 1.0989,
+      "step": 1149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3198719024658203,
+      "learning_rate": 9.622740951826601e-06,
+      "loss": 0.8437,
+      "step": 1150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3253427743911743,
+      "learning_rate": 9.621931758551713e-06,
+      "loss": 0.8565,
+      "step": 1151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7401647567749023,
+      "learning_rate": 9.621121732473484e-06,
+      "loss": 1.0928,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.393760085105896,
+      "learning_rate": 9.620310873737864e-06,
+      "loss": 0.8834,
+      "step": 1153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2570570707321167,
+      "learning_rate": 9.619499182490964e-06,
+      "loss": 0.776,
+      "step": 1154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3100600242614746,
+      "learning_rate": 9.618686658879035e-06,
+      "loss": 0.9197,
+      "step": 1155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3726849555969238,
+      "learning_rate": 9.617873303048484e-06,
+      "loss": 0.8843,
+      "step": 1156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4967948198318481,
+      "learning_rate": 9.617059115145861e-06,
+      "loss": 0.9129,
+      "step": 1157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3441284894943237,
+      "learning_rate": 9.616244095317876e-06,
+      "loss": 0.889,
+      "step": 1158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9008550047874451,
+      "learning_rate": 9.61542824371138e-06,
+      "loss": 1.0843,
+      "step": 1159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.389017939567566,
+      "learning_rate": 9.61461156047338e-06,
+      "loss": 0.8655,
+      "step": 1160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7045869827270508,
+      "learning_rate": 9.613794045751027e-06,
+      "loss": 1.1054,
+      "step": 1161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.403588056564331,
+      "learning_rate": 9.612975699691626e-06,
+      "loss": 0.8931,
+      "step": 1162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2892720699310303,
+      "learning_rate": 9.612156522442633e-06,
+      "loss": 0.8216,
+      "step": 1163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2796556949615479,
+      "learning_rate": 9.611336514151651e-06,
+      "loss": 0.9038,
+      "step": 1164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.364021897315979,
+      "learning_rate": 9.61051567496643e-06,
+      "loss": 0.934,
+      "step": 1165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2735645771026611,
+      "learning_rate": 9.609694005034875e-06,
+      "loss": 0.8869,
+      "step": 1166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5087066888809204,
+      "learning_rate": 9.60887150450504e-06,
+      "loss": 0.9082,
+      "step": 1167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2585476636886597,
+      "learning_rate": 9.608048173525126e-06,
+      "loss": 0.8655,
+      "step": 1168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.290898084640503,
+      "learning_rate": 9.607224012243486e-06,
+      "loss": 0.8321,
+      "step": 1169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.339229702949524,
+      "learning_rate": 9.60639902080862e-06,
+      "loss": 0.8795,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2251254320144653,
+      "learning_rate": 9.60557319936918e-06,
+      "loss": 1.0852,
+      "step": 1171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3365437984466553,
+      "learning_rate": 9.604746548073968e-06,
+      "loss": 0.8575,
+      "step": 1172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3421076536178589,
+      "learning_rate": 9.603919067071931e-06,
+      "loss": 0.8926,
+      "step": 1173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3561042547225952,
+      "learning_rate": 9.603090756512172e-06,
+      "loss": 0.8997,
+      "step": 1174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3249894380569458,
+      "learning_rate": 9.60226161654394e-06,
+      "loss": 0.8576,
+      "step": 1175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3647409677505493,
+      "learning_rate": 9.601431647316633e-06,
+      "loss": 0.9156,
+      "step": 1176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3923672437667847,
+      "learning_rate": 9.600600848979796e-06,
+      "loss": 0.8591,
+      "step": 1177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3249385356903076,
+      "learning_rate": 9.599769221683131e-06,
+      "loss": 0.9104,
+      "step": 1178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1375173330307007,
+      "learning_rate": 9.598936765576483e-06,
+      "loss": 1.0885,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9271794557571411,
+      "learning_rate": 9.598103480809849e-06,
+      "loss": 1.0929,
+      "step": 1180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3943918943405151,
+      "learning_rate": 9.597269367533375e-06,
+      "loss": 0.8581,
+      "step": 1181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3908696174621582,
+      "learning_rate": 9.596434425897353e-06,
+      "loss": 0.9236,
+      "step": 1182
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.343238115310669,
+      "learning_rate": 9.59559865605223e-06,
+      "loss": 0.8731,
+      "step": 1183
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8228774070739746,
+      "learning_rate": 9.594762058148596e-06,
+      "loss": 1.1017,
+      "step": 1184
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3217071294784546,
+      "learning_rate": 9.593924632337197e-06,
+      "loss": 0.8228,
+      "step": 1185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4429783821105957,
+      "learning_rate": 9.593086378768924e-06,
+      "loss": 1.0998,
+      "step": 1186
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9170881509780884,
+      "learning_rate": 9.592247297594817e-06,
+      "loss": 1.0798,
+      "step": 1187
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4903218746185303,
+      "learning_rate": 9.591407388966066e-06,
+      "loss": 0.8499,
+      "step": 1188
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.491731882095337,
+      "learning_rate": 9.590566653034008e-06,
+      "loss": 0.8544,
+      "step": 1189
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5842654705047607,
+      "learning_rate": 9.589725089950134e-06,
+      "loss": 0.8325,
+      "step": 1190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.177098274230957,
+      "learning_rate": 9.588882699866082e-06,
+      "loss": 1.1407,
+      "step": 1191
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9007859230041504,
+      "learning_rate": 9.588039482933636e-06,
+      "loss": 1.1041,
+      "step": 1192
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4489290714263916,
+      "learning_rate": 9.587195439304732e-06,
+      "loss": 0.9289,
+      "step": 1193
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5989093780517578,
+      "learning_rate": 9.586350569131453e-06,
+      "loss": 0.9646,
+      "step": 1194
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5988696813583374,
+      "learning_rate": 9.585504872566033e-06,
+      "loss": 0.9003,
+      "step": 1195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4282851219177246,
+      "learning_rate": 9.584658349760854e-06,
+      "loss": 0.9318,
+      "step": 1196
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.469433307647705,
+      "learning_rate": 9.583811000868446e-06,
+      "loss": 0.8993,
+      "step": 1197
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2924600839614868,
+      "learning_rate": 9.58296282604149e-06,
+      "loss": 1.0728,
+      "step": 1198
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3249223232269287,
+      "learning_rate": 9.582113825432812e-06,
+      "loss": 0.8956,
+      "step": 1199
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8915609121322632,
+      "learning_rate": 9.581263999195392e-06,
+      "loss": 1.0951,
+      "step": 1200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4750484228134155,
+      "learning_rate": 9.580413347482356e-06,
+      "loss": 0.8687,
+      "step": 1201
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5127379894256592,
+      "learning_rate": 9.579561870446976e-06,
+      "loss": 0.8857,
+      "step": 1202
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1703109741210938,
+      "learning_rate": 9.57870956824268e-06,
+      "loss": 1.1052,
+      "step": 1203
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5426914691925049,
+      "learning_rate": 9.577856441023035e-06,
+      "loss": 0.9401,
+      "step": 1204
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4267759323120117,
+      "learning_rate": 9.577002488941765e-06,
+      "loss": 0.8957,
+      "step": 1205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.342278242111206,
+      "learning_rate": 9.576147712152737e-06,
+      "loss": 0.8369,
+      "step": 1206
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4328118562698364,
+      "learning_rate": 9.575292110809973e-06,
+      "loss": 0.8639,
+      "step": 1207
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4761296510696411,
+      "learning_rate": 9.574435685067637e-06,
+      "loss": 0.8994,
+      "step": 1208
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4109638929367065,
+      "learning_rate": 9.573578435080045e-06,
+      "loss": 0.9045,
+      "step": 1209
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.105068325996399,
+      "learning_rate": 9.57272036100166e-06,
+      "loss": 1.067,
+      "step": 1210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3625997304916382,
+      "learning_rate": 9.571861462987095e-06,
+      "loss": 0.9042,
+      "step": 1211
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6799518465995789,
+      "learning_rate": 9.57100174119111e-06,
+      "loss": 1.0646,
+      "step": 1212
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4597398042678833,
+      "learning_rate": 9.570141195768613e-06,
+      "loss": 0.8727,
+      "step": 1213
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8325707912445068,
+      "learning_rate": 9.569279826874665e-06,
+      "loss": 1.0875,
+      "step": 1214
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.487650752067566,
+      "learning_rate": 9.568417634664467e-06,
+      "loss": 0.8438,
+      "step": 1215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7707642912864685,
+      "learning_rate": 9.567554619293378e-06,
+      "loss": 1.0906,
+      "step": 1216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3136838674545288,
+      "learning_rate": 9.566690780916898e-06,
+      "loss": 0.8914,
+      "step": 1217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3452651500701904,
+      "learning_rate": 9.565826119690677e-06,
+      "loss": 0.8642,
+      "step": 1218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9352661967277527,
+      "learning_rate": 9.564960635770516e-06,
+      "loss": 1.0933,
+      "step": 1219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3448138236999512,
+      "learning_rate": 9.564094329312362e-06,
+      "loss": 0.8639,
+      "step": 1220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7597174048423767,
+      "learning_rate": 9.56322720047231e-06,
+      "loss": 1.1054,
+      "step": 1221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.275694489479065,
+      "learning_rate": 9.562359249406601e-06,
+      "loss": 0.8484,
+      "step": 1222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.304944396018982,
+      "learning_rate": 9.561490476271633e-06,
+      "loss": 0.8743,
+      "step": 1223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2413218021392822,
+      "learning_rate": 9.560620881223941e-06,
+      "loss": 0.9424,
+      "step": 1224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.326399803161621,
+      "learning_rate": 9.559750464420216e-06,
+      "loss": 0.9004,
+      "step": 1225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.311432957649231,
+      "learning_rate": 9.558879226017293e-06,
+      "loss": 0.8403,
+      "step": 1226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4623609781265259,
+      "learning_rate": 9.558007166172157e-06,
+      "loss": 1.1015,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3119314908981323,
+      "learning_rate": 9.557134285041935e-06,
+      "loss": 0.8429,
+      "step": 1228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2865065336227417,
+      "learning_rate": 9.556260582783915e-06,
+      "loss": 0.9047,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6832343935966492,
+      "learning_rate": 9.555386059555522e-06,
+      "loss": 1.0575,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.497455358505249,
+      "learning_rate": 9.554510715514332e-06,
+      "loss": 0.9468,
+      "step": 1231
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4236962795257568,
+      "learning_rate": 9.553634550818067e-06,
+      "loss": 0.9459,
+      "step": 1232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.36245858669281,
+      "learning_rate": 9.5527575656246e-06,
+      "loss": 0.8859,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2488391399383545,
+      "learning_rate": 9.551879760091952e-06,
+      "loss": 0.84,
+      "step": 1234
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.323630690574646,
+      "learning_rate": 9.551001134378292e-06,
+      "loss": 0.8356,
+      "step": 1235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3133095502853394,
+      "learning_rate": 9.550121688641931e-06,
+      "loss": 0.8632,
+      "step": 1236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3347344398498535,
+      "learning_rate": 9.549241423041333e-06,
+      "loss": 0.8664,
+      "step": 1237
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.440498948097229,
+      "learning_rate": 9.548360337735111e-06,
+      "loss": 0.9179,
+      "step": 1238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3240779638290405,
+      "learning_rate": 9.547478432882023e-06,
+      "loss": 0.9126,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2335702180862427,
+      "learning_rate": 9.546595708640975e-06,
+      "loss": 1.0989,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3202534914016724,
+      "learning_rate": 9.54571216517102e-06,
+      "loss": 0.9045,
+      "step": 1241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2859236001968384,
+      "learning_rate": 9.544827802631358e-06,
+      "loss": 0.897,
+      "step": 1242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3415840864181519,
+      "learning_rate": 9.543942621181341e-06,
+      "loss": 0.9248,
+      "step": 1243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3093382120132446,
+      "learning_rate": 9.543056620980462e-06,
+      "loss": 0.8923,
+      "step": 1244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3017247915267944,
+      "learning_rate": 9.54216980218837e-06,
+      "loss": 0.867,
+      "step": 1245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2403043508529663,
+      "learning_rate": 9.54128216496485e-06,
+      "loss": 0.8657,
+      "step": 1246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3069467544555664,
+      "learning_rate": 9.540393709469848e-06,
+      "loss": 0.9049,
+      "step": 1247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3333988189697266,
+      "learning_rate": 9.539504435863448e-06,
+      "loss": 1.0872,
+      "step": 1248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9836461544036865,
+      "learning_rate": 9.538614344305881e-06,
+      "loss": 1.0714,
+      "step": 1249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.388434886932373,
+      "learning_rate": 9.53772343495753e-06,
+      "loss": 0.9019,
+      "step": 1250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1834601163864136,
+      "learning_rate": 9.536831707978926e-06,
+      "loss": 1.1026,
+      "step": 1251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4207568168640137,
+      "learning_rate": 9.53593916353074e-06,
+      "loss": 0.8936,
+      "step": 1252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3612695932388306,
+      "learning_rate": 9.5350458017738e-06,
+      "loss": 0.9099,
+      "step": 1253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3038290739059448,
+      "learning_rate": 9.534151622869075e-06,
+      "loss": 0.84,
+      "step": 1254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3214681148529053,
+      "learning_rate": 9.533256626977681e-06,
+      "loss": 0.9043,
+      "step": 1255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2767406702041626,
+      "learning_rate": 9.532360814260886e-06,
+      "loss": 0.87,
+      "step": 1256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4350197315216064,
+      "learning_rate": 9.5314641848801e-06,
+      "loss": 1.087,
+      "step": 1257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.378888487815857,
+      "learning_rate": 9.530566738996882e-06,
+      "loss": 0.8295,
+      "step": 1258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3527966737747192,
+      "learning_rate": 9.529668476772939e-06,
+      "loss": 0.8543,
+      "step": 1259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3505772352218628,
+      "learning_rate": 9.528769398370123e-06,
+      "loss": 0.9023,
+      "step": 1260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2763257026672363,
+      "learning_rate": 9.527869503950439e-06,
+      "loss": 0.887,
+      "step": 1261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0764909982681274,
+      "learning_rate": 9.52696879367603e-06,
+      "loss": 1.092,
+      "step": 1262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.392112374305725,
+      "learning_rate": 9.526067267709193e-06,
+      "loss": 0.8886,
+      "step": 1263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4401229619979858,
+      "learning_rate": 9.525164926212369e-06,
+      "loss": 0.8871,
+      "step": 1264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3889787197113037,
+      "learning_rate": 9.524261769348147e-06,
+      "loss": 0.8363,
+      "step": 1265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8490387797355652,
+      "learning_rate": 9.52335779727926e-06,
+      "loss": 1.0964,
+      "step": 1266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.261327862739563,
+      "learning_rate": 9.522453010168594e-06,
+      "loss": 0.8538,
+      "step": 1267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3454959392547607,
+      "learning_rate": 9.521547408179174e-06,
+      "loss": 0.9397,
+      "step": 1268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.389575481414795,
+      "learning_rate": 9.52064099147418e-06,
+      "loss": 0.9285,
+      "step": 1269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3975352048873901,
+      "learning_rate": 9.519733760216934e-06,
+      "loss": 0.8692,
+      "step": 1270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8288314938545227,
+      "learning_rate": 9.518825714570902e-06,
+      "loss": 1.0781,
+      "step": 1271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3543071746826172,
+      "learning_rate": 9.517916854699706e-06,
+      "loss": 0.8902,
+      "step": 1272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3716424703598022,
+      "learning_rate": 9.517007180767103e-06,
+      "loss": 0.9225,
+      "step": 1273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2854536771774292,
+      "learning_rate": 9.51609669293701e-06,
+      "loss": 0.8573,
+      "step": 1274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.344300627708435,
+      "learning_rate": 9.515185391373476e-06,
+      "loss": 0.862,
+      "step": 1275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3230177164077759,
+      "learning_rate": 9.514273276240706e-06,
+      "loss": 0.9074,
+      "step": 1276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3718767166137695,
+      "learning_rate": 9.513360347703054e-06,
+      "loss": 1.0236,
+      "step": 1277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3339229822158813,
+      "learning_rate": 9.51244660592501e-06,
+      "loss": 0.8734,
+      "step": 1278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3103916645050049,
+      "learning_rate": 9.511532051071221e-06,
+      "loss": 0.8621,
+      "step": 1279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3403688669204712,
+      "learning_rate": 9.510616683306473e-06,
+      "loss": 0.8429,
+      "step": 1280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.302468180656433,
+      "learning_rate": 9.509700502795706e-06,
+      "loss": 0.8471,
+      "step": 1281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.401252269744873,
+      "learning_rate": 9.508783509703998e-06,
+      "loss": 0.86,
+      "step": 1282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2720144987106323,
+      "learning_rate": 9.507865704196579e-06,
+      "loss": 0.8733,
+      "step": 1283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.261193037033081,
+      "learning_rate": 9.506947086438824e-06,
+      "loss": 0.8666,
+      "step": 1284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3498575687408447,
+      "learning_rate": 9.506027656596253e-06,
+      "loss": 0.8985,
+      "step": 1285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3396393060684204,
+      "learning_rate": 9.505107414834536e-06,
+      "loss": 0.9054,
+      "step": 1286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4082564115524292,
+      "learning_rate": 9.504186361319486e-06,
+      "loss": 0.868,
+      "step": 1287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3160794973373413,
+      "learning_rate": 9.503264496217062e-06,
+      "loss": 0.8387,
+      "step": 1288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0750242471694946,
+      "learning_rate": 9.50234181969337e-06,
+      "loss": 1.09,
+      "step": 1289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.326075553894043,
+      "learning_rate": 9.501418331914664e-06,
+      "loss": 0.9484,
+      "step": 1290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.30379056930542,
+      "learning_rate": 9.500494033047343e-06,
+      "loss": 0.9211,
+      "step": 1291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.358688473701477,
+      "learning_rate": 9.499568923257951e-06,
+      "loss": 0.9141,
+      "step": 1292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.419224500656128,
+      "learning_rate": 9.498643002713179e-06,
+      "loss": 0.8177,
+      "step": 1293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3071002960205078,
+      "learning_rate": 9.497716271579867e-06,
+      "loss": 0.9047,
+      "step": 1294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1320958137512207,
+      "learning_rate": 9.496788730024992e-06,
+      "loss": 1.0952,
+      "step": 1295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6858528852462769,
+      "learning_rate": 9.495860378215689e-06,
+      "loss": 0.871,
+      "step": 1296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.24560546875,
+      "learning_rate": 9.494931216319229e-06,
+      "loss": 0.9111,
+      "step": 1297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2405071258544922,
+      "learning_rate": 9.494001244503037e-06,
+      "loss": 0.8277,
+      "step": 1298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2887706756591797,
+      "learning_rate": 9.493070462934678e-06,
+      "loss": 0.8739,
+      "step": 1299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3576503992080688,
+      "learning_rate": 9.492138871781865e-06,
+      "loss": 0.8569,
+      "step": 1300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3902487754821777,
+      "learning_rate": 9.491206471212458e-06,
+      "loss": 0.9211,
+      "step": 1301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3582743406295776,
+      "learning_rate": 9.490273261394458e-06,
+      "loss": 0.8754,
+      "step": 1302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2712959051132202,
+      "learning_rate": 9.489339242496021e-06,
+      "loss": 0.8865,
+      "step": 1303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2870492935180664,
+      "learning_rate": 9.48840441468544e-06,
+      "loss": 0.922,
+      "step": 1304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0775716304779053,
+      "learning_rate": 9.487468778131157e-06,
+      "loss": 1.0633,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.345636010169983,
+      "learning_rate": 9.486532333001761e-06,
+      "loss": 0.8549,
+      "step": 1306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3357857465744019,
+      "learning_rate": 9.485595079465984e-06,
+      "loss": 0.8855,
+      "step": 1307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6635072827339172,
+      "learning_rate": 9.484657017692707e-06,
+      "loss": 1.0658,
+      "step": 1308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7013596892356873,
+      "learning_rate": 9.483718147850952e-06,
+      "loss": 1.0785,
+      "step": 1309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.328041434288025,
+      "learning_rate": 9.482778470109894e-06,
+      "loss": 0.8624,
+      "step": 1310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6891000866889954,
+      "learning_rate": 9.481837984638846e-06,
+      "loss": 1.0896,
+      "step": 1311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3114944696426392,
+      "learning_rate": 9.480896691607269e-06,
+      "loss": 0.8234,
+      "step": 1312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2167737483978271,
+      "learning_rate": 9.47995459118477e-06,
+      "loss": 0.8609,
+      "step": 1313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3184031248092651,
+      "learning_rate": 9.479011683541103e-06,
+      "loss": 0.8937,
+      "step": 1314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2277346849441528,
+      "learning_rate": 9.478067968846165e-06,
+      "loss": 0.8649,
+      "step": 1315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2838107347488403,
+      "learning_rate": 9.477123447269999e-06,
+      "loss": 0.9101,
+      "step": 1316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.403786301612854,
+      "learning_rate": 9.476178118982793e-06,
+      "loss": 0.8877,
+      "step": 1317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2215336561203003,
+      "learning_rate": 9.475231984154885e-06,
+      "loss": 0.8714,
+      "step": 1318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3202797174453735,
+      "learning_rate": 9.47428504295675e-06,
+      "loss": 0.8974,
+      "step": 1319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3028230667114258,
+      "learning_rate": 9.473337295559015e-06,
+      "loss": 0.8531,
+      "step": 1320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3592076301574707,
+      "learning_rate": 9.472388742132448e-06,
+      "loss": 0.8208,
+      "step": 1321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2410199642181396,
+      "learning_rate": 9.471439382847968e-06,
+      "loss": 1.0778,
+      "step": 1322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3508834838867188,
+      "learning_rate": 9.470489217876632e-06,
+      "loss": 0.8617,
+      "step": 1323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7309561967849731,
+      "learning_rate": 9.469538247389645e-06,
+      "loss": 1.0913,
+      "step": 1324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3641479015350342,
+      "learning_rate": 9.46858647155836e-06,
+      "loss": 0.8367,
+      "step": 1325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.374817967414856,
+      "learning_rate": 9.467633890554275e-06,
+      "loss": 0.9134,
+      "step": 1326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0533117055892944,
+      "learning_rate": 9.466680504549026e-06,
+      "loss": 1.062,
+      "step": 1327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.372441291809082,
+      "learning_rate": 9.465726313714401e-06,
+      "loss": 0.9029,
+      "step": 1328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4159728288650513,
+      "learning_rate": 9.464771318222332e-06,
+      "loss": 0.89,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7638285160064697,
+      "learning_rate": 9.463815518244891e-06,
+      "loss": 1.1067,
+      "step": 1330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3748371601104736,
+      "learning_rate": 9.462858913954305e-06,
+      "loss": 0.8411,
+      "step": 1331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3961635828018188,
+      "learning_rate": 9.461901505522937e-06,
+      "loss": 0.8822,
+      "step": 1332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3622686862945557,
+      "learning_rate": 9.460943293123297e-06,
+      "loss": 0.8902,
+      "step": 1333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4105125665664673,
+      "learning_rate": 9.45998427692804e-06,
+      "loss": 0.8612,
+      "step": 1334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3900761604309082,
+      "learning_rate": 9.459024457109969e-06,
+      "loss": 0.9191,
+      "step": 1335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2813615798950195,
+      "learning_rate": 9.458063833842028e-06,
+      "loss": 0.928,
+      "step": 1336
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0441322326660156,
+      "learning_rate": 9.457102407297307e-06,
+      "loss": 1.0745,
+      "step": 1337
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7541683912277222,
+      "learning_rate": 9.45614017764904e-06,
+      "loss": 1.0658,
+      "step": 1338
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5000640153884888,
+      "learning_rate": 9.455177145070609e-06,
+      "loss": 0.8474,
+      "step": 1339
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9384921789169312,
+      "learning_rate": 9.454213309735535e-06,
+      "loss": 1.0855,
+      "step": 1340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6191002130508423,
+      "learning_rate": 9.453248671817486e-06,
+      "loss": 0.8923,
+      "step": 1341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3648531436920166,
+      "learning_rate": 9.45228323149028e-06,
+      "loss": 0.8825,
+      "step": 1342
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3106964826583862,
+      "learning_rate": 9.451316988927872e-06,
+      "loss": 0.8739,
+      "step": 1343
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0717412233352661,
+      "learning_rate": 9.450349944304365e-06,
+      "loss": 1.0792,
+      "step": 1344
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3446333408355713,
+      "learning_rate": 9.449382097794006e-06,
+      "loss": 0.8623,
+      "step": 1345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7541404962539673,
+      "learning_rate": 9.448413449571188e-06,
+      "loss": 1.0672,
+      "step": 1346
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6973342299461365,
+      "learning_rate": 9.447443999810443e-06,
+      "loss": 1.0748,
+      "step": 1347
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4621435403823853,
+      "learning_rate": 9.446473748686457e-06,
+      "loss": 0.9031,
+      "step": 1348
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3775291442871094,
+      "learning_rate": 9.44550269637405e-06,
+      "loss": 0.8944,
+      "step": 1349
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.359286904335022,
+      "learning_rate": 9.444530843048194e-06,
+      "loss": 0.8932,
+      "step": 1350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2436904907226562,
+      "learning_rate": 9.443558188884001e-06,
+      "loss": 1.1008,
+      "step": 1351
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.027408480644226,
+      "learning_rate": 9.44258473405673e-06,
+      "loss": 1.0783,
+      "step": 1352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3919404745101929,
+      "learning_rate": 9.441610478741782e-06,
+      "loss": 0.8545,
+      "step": 1353
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.333791971206665,
+      "learning_rate": 9.440635423114704e-06,
+      "loss": 0.8127,
+      "step": 1354
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4610910415649414,
+      "learning_rate": 9.439659567351184e-06,
+      "loss": 0.8554,
+      "step": 1355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3542388677597046,
+      "learning_rate": 9.438682911627062e-06,
+      "loss": 1.0756,
+      "step": 1356
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.025325059890747,
+      "learning_rate": 9.437705456118314e-06,
+      "loss": 1.0877,
+      "step": 1357
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3372937440872192,
+      "learning_rate": 9.436727201001062e-06,
+      "loss": 0.8755,
+      "step": 1358
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8144475817680359,
+      "learning_rate": 9.435748146451573e-06,
+      "loss": 1.086,
+      "step": 1359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5457344055175781,
+      "learning_rate": 9.434768292646259e-06,
+      "loss": 0.895,
+      "step": 1360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5571438074111938,
+      "learning_rate": 9.433787639761675e-06,
+      "loss": 0.8685,
+      "step": 1361
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.427215814590454,
+      "learning_rate": 9.432806187974521e-06,
+      "loss": 0.83,
+      "step": 1362
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.344048023223877,
+      "learning_rate": 9.431823937461637e-06,
+      "loss": 0.876,
+      "step": 1363
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4051865339279175,
+      "learning_rate": 9.430840888400014e-06,
+      "loss": 0.8353,
+      "step": 1364
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3948957920074463,
+      "learning_rate": 9.429857040966779e-06,
+      "loss": 0.852,
+      "step": 1365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5121949911117554,
+      "learning_rate": 9.42887239533921e-06,
+      "loss": 0.8398,
+      "step": 1366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3363308906555176,
+      "learning_rate": 9.427886951694722e-06,
+      "loss": 0.8911,
+      "step": 1367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2899209260940552,
+      "learning_rate": 9.426900710210881e-06,
+      "loss": 0.8611,
+      "step": 1368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3150544166564941,
+      "learning_rate": 9.425913671065391e-06,
+      "loss": 0.8993,
+      "step": 1369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5171475410461426,
+      "learning_rate": 9.424925834436102e-06,
+      "loss": 0.8773,
+      "step": 1370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3817026615142822,
+      "learning_rate": 9.423937200501007e-06,
+      "loss": 1.0965,
+      "step": 1371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1129086017608643,
+      "learning_rate": 9.422947769438244e-06,
+      "loss": 1.0674,
+      "step": 1372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4008008241653442,
+      "learning_rate": 9.421957541426092e-06,
+      "loss": 0.8294,
+      "step": 1373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4906309843063354,
+      "learning_rate": 9.420966516642978e-06,
+      "loss": 0.9099,
+      "step": 1374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4762073755264282,
+      "learning_rate": 9.419974695267469e-06,
+      "loss": 0.8935,
+      "step": 1375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4490885734558105,
+      "learning_rate": 9.418982077478275e-06,
+      "loss": 1.1005,
+      "step": 1376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3017760515213013,
+      "learning_rate": 9.417988663454252e-06,
+      "loss": 0.8242,
+      "step": 1377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.259590744972229,
+      "learning_rate": 9.4169944533744e-06,
+      "loss": 0.8405,
+      "step": 1378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3225200176239014,
+      "learning_rate": 9.415999447417858e-06,
+      "loss": 0.9101,
+      "step": 1379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3160207271575928,
+      "learning_rate": 9.415003645763911e-06,
+      "loss": 0.8519,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2719887495040894,
+      "learning_rate": 9.414007048591992e-06,
+      "loss": 0.9327,
+      "step": 1381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2416648864746094,
+      "learning_rate": 9.413009656081669e-06,
+      "loss": 0.8617,
+      "step": 1382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3929064273834229,
+      "learning_rate": 9.41201146841266e-06,
+      "loss": 0.8771,
+      "step": 1383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.349165678024292,
+      "learning_rate": 9.411012485764821e-06,
+      "loss": 0.9503,
+      "step": 1384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.287613868713379,
+      "learning_rate": 9.410012708318155e-06,
+      "loss": 0.8707,
+      "step": 1385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.280615210533142,
+      "learning_rate": 9.409012136252807e-06,
+      "loss": 0.9202,
+      "step": 1386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.402106523513794,
+      "learning_rate": 9.408010769749064e-06,
+      "loss": 0.8775,
+      "step": 1387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1236222982406616,
+      "learning_rate": 9.40700860898736e-06,
+      "loss": 1.1038,
+      "step": 1388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.410272240638733,
+      "learning_rate": 9.406005654148266e-06,
+      "loss": 0.8868,
+      "step": 1389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2695499658584595,
+      "learning_rate": 9.405001905412503e-06,
+      "loss": 0.8493,
+      "step": 1390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.363513469696045,
+      "learning_rate": 9.40399736296093e-06,
+      "loss": 0.8617,
+      "step": 1391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3159571886062622,
+      "learning_rate": 9.402992026974551e-06,
+      "loss": 0.8637,
+      "step": 1392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3171870708465576,
+      "learning_rate": 9.40198589763451e-06,
+      "loss": 0.8369,
+      "step": 1393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3965818881988525,
+      "learning_rate": 9.400978975122103e-06,
+      "loss": 0.9075,
+      "step": 1394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3738800287246704,
+      "learning_rate": 9.399971259618755e-06,
+      "loss": 0.8446,
+      "step": 1395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3126004934310913,
+      "learning_rate": 9.398962751306046e-06,
+      "loss": 0.8538,
+      "step": 1396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0428509712219238,
+      "learning_rate": 9.397953450365692e-06,
+      "loss": 1.0649,
+      "step": 1397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.342939853668213,
+      "learning_rate": 9.396943356979555e-06,
+      "loss": 0.7951,
+      "step": 1398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3067834377288818,
+      "learning_rate": 9.395932471329639e-06,
+      "loss": 0.8868,
+      "step": 1399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.780885636806488,
+      "learning_rate": 9.39492079359809e-06,
+      "loss": 1.0726,
+      "step": 1400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7081215977668762,
+      "learning_rate": 9.393908323967199e-06,
+      "loss": 1.087,
+      "step": 1401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6571691036224365,
+      "learning_rate": 9.392895062619393e-06,
+      "loss": 1.0764,
+      "step": 1402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2850539684295654,
+      "learning_rate": 9.391881009737252e-06,
+      "loss": 0.8701,
+      "step": 1403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2748438119888306,
+      "learning_rate": 9.390866165503491e-06,
+      "loss": 0.8555,
+      "step": 1404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2586814165115356,
+      "learning_rate": 9.38985053010097e-06,
+      "loss": 0.838,
+      "step": 1405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2586095333099365,
+      "learning_rate": 9.388834103712691e-06,
+      "loss": 1.0833,
+      "step": 1406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.306533694267273,
+      "learning_rate": 9.3878168865218e-06,
+      "loss": 0.8738,
+      "step": 1407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3242919445037842,
+      "learning_rate": 9.386798878711584e-06,
+      "loss": 0.8575,
+      "step": 1408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2372325658798218,
+      "learning_rate": 9.385780080465471e-06,
+      "loss": 0.7953,
+      "step": 1409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2386442422866821,
+      "learning_rate": 9.384760491967038e-06,
+      "loss": 0.8481,
+      "step": 1410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2899249792099,
+      "learning_rate": 9.383740113399996e-06,
+      "loss": 0.8816,
+      "step": 1411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3704153299331665,
+      "learning_rate": 9.382718944948201e-06,
+      "loss": 0.8634,
+      "step": 1412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9288102984428406,
+      "learning_rate": 9.381696986795655e-06,
+      "loss": 1.0706,
+      "step": 1413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2339873313903809,
+      "learning_rate": 9.380674239126497e-06,
+      "loss": 0.8645,
+      "step": 1414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2932755947113037,
+      "learning_rate": 9.379650702125015e-06,
+      "loss": 0.8605,
+      "step": 1415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3151921033859253,
+      "learning_rate": 9.378626375975631e-06,
+      "loss": 0.895,
+      "step": 1416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3235859870910645,
+      "learning_rate": 9.377601260862916e-06,
+      "loss": 0.9176,
+      "step": 1417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2498868703842163,
+      "learning_rate": 9.37657535697158e-06,
+      "loss": 0.8436,
+      "step": 1418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.335750937461853,
+      "learning_rate": 9.375548664486475e-06,
+      "loss": 0.8649,
+      "step": 1419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.291631817817688,
+      "learning_rate": 9.374521183592596e-06,
+      "loss": 0.8473,
+      "step": 1420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3702492713928223,
+      "learning_rate": 9.373492914475078e-06,
+      "loss": 0.8704,
+      "step": 1421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3369075059890747,
+      "learning_rate": 9.372463857319203e-06,
+      "loss": 0.872,
+      "step": 1422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.941275417804718,
+      "learning_rate": 9.37143401231039e-06,
+      "loss": 1.0756,
+      "step": 1423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3841630220413208,
+      "learning_rate": 9.370403379634201e-06,
+      "loss": 0.9095,
+      "step": 1424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.314226746559143,
+      "learning_rate": 9.369371959476343e-06,
+      "loss": 0.8104,
+      "step": 1425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3704729080200195,
+      "learning_rate": 9.368339752022662e-06,
+      "loss": 0.8562,
+      "step": 1426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2893364429473877,
+      "learning_rate": 9.367306757459144e-06,
+      "loss": 0.8988,
+      "step": 1427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3372033834457397,
+      "learning_rate": 9.366272975971922e-06,
+      "loss": 0.8578,
+      "step": 1428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3914544582366943,
+      "learning_rate": 9.365238407747265e-06,
+      "loss": 0.895,
+      "step": 1429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9190211296081543,
+      "learning_rate": 9.364203052971591e-06,
+      "loss": 0.8477,
+      "step": 1430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0451191663742065,
+      "learning_rate": 9.363166911831455e-06,
+      "loss": 1.0846,
+      "step": 1431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3833585977554321,
+      "learning_rate": 9.362129984513548e-06,
+      "loss": 0.9178,
+      "step": 1432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6975071430206299,
+      "learning_rate": 9.361092271204717e-06,
+      "loss": 1.0774,
+      "step": 1433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.249782681465149,
+      "learning_rate": 9.360053772091938e-06,
+      "loss": 0.9,
+      "step": 1434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3407652378082275,
+      "learning_rate": 9.359014487362332e-06,
+      "loss": 0.9291,
+      "step": 1435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9788740873336792,
+      "learning_rate": 9.357974417203166e-06,
+      "loss": 1.0705,
+      "step": 1436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2885857820510864,
+      "learning_rate": 9.356933561801845e-06,
+      "loss": 0.8918,
+      "step": 1437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3439234495162964,
+      "learning_rate": 9.355891921345912e-06,
+      "loss": 0.8709,
+      "step": 1438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8314365744590759,
+      "learning_rate": 9.35484949602306e-06,
+      "loss": 1.0967,
+      "step": 1439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3301591873168945,
+      "learning_rate": 9.353806286021115e-06,
+      "loss": 0.9006,
+      "step": 1440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2714580297470093,
+      "learning_rate": 9.352762291528048e-06,
+      "loss": 0.8727,
+      "step": 1441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2289106845855713,
+      "learning_rate": 9.351717512731974e-06,
+      "loss": 0.8888,
+      "step": 1442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2012544870376587,
+      "learning_rate": 9.350671949821143e-06,
+      "loss": 0.8799,
+      "step": 1443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.330269694328308,
+      "learning_rate": 9.349625602983954e-06,
+      "loss": 0.9055,
+      "step": 1444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.401658058166504,
+      "learning_rate": 9.34857847240894e-06,
+      "loss": 0.8662,
+      "step": 1445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.283130407333374,
+      "learning_rate": 9.34753055828478e-06,
+      "loss": 0.8721,
+      "step": 1446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9084926247596741,
+      "learning_rate": 9.346481860800291e-06,
+      "loss": 1.1,
+      "step": 1447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7307529449462891,
+      "learning_rate": 9.345432380144434e-06,
+      "loss": 1.0952,
+      "step": 1448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.675530731678009,
+      "learning_rate": 9.34438211650631e-06,
+      "loss": 1.0828,
+      "step": 1449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4996581077575684,
+      "learning_rate": 9.343331070075158e-06,
+      "loss": 0.9124,
+      "step": 1450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4485065937042236,
+      "learning_rate": 9.342279241040364e-06,
+      "loss": 0.8654,
+      "step": 1451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.351298213005066,
+      "learning_rate": 9.34122662959145e-06,
+      "loss": 0.875,
+      "step": 1452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3224705457687378,
+      "learning_rate": 9.340173235918083e-06,
+      "loss": 0.893,
+      "step": 1453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2619439363479614,
+      "learning_rate": 9.339119060210069e-06,
+      "loss": 0.8306,
+      "step": 1454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2806123495101929,
+      "learning_rate": 9.338064102657353e-06,
+      "loss": 0.8184,
+      "step": 1455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3615152835845947,
+      "learning_rate": 9.337008363450024e-06,
+      "loss": 0.8716,
+      "step": 1456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3716827630996704,
+      "learning_rate": 9.335951842778309e-06,
+      "loss": 0.9599,
+      "step": 1457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5577949285507202,
+      "learning_rate": 9.334894540832578e-06,
+      "loss": 1.1033,
+      "step": 1458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3044816255569458,
+      "learning_rate": 9.33383645780334e-06,
+      "loss": 0.8904,
+      "step": 1459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2791855335235596,
+      "learning_rate": 9.332777593881247e-06,
+      "loss": 0.9327,
+      "step": 1460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7877978086471558,
+      "learning_rate": 9.331717949257094e-06,
+      "loss": 1.0618,
+      "step": 1461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2669956684112549,
+      "learning_rate": 9.330657524121807e-06,
+      "loss": 0.8504,
+      "step": 1462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8390896916389465,
+      "learning_rate": 9.329596318666462e-06,
+      "loss": 1.0973,
+      "step": 1463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3893136978149414,
+      "learning_rate": 9.328534333082271e-06,
+      "loss": 0.8892,
+      "step": 1464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3871034383773804,
+      "learning_rate": 9.327471567560592e-06,
+      "loss": 0.8688,
+      "step": 1465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8883200287818909,
+      "learning_rate": 9.326408022292912e-06,
+      "loss": 1.0639,
+      "step": 1466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2652045488357544,
+      "learning_rate": 9.325343697470874e-06,
+      "loss": 0.7865,
+      "step": 1467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3024524450302124,
+      "learning_rate": 9.324278593286248e-06,
+      "loss": 0.8794,
+      "step": 1468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7621671557426453,
+      "learning_rate": 9.323212709930953e-06,
+      "loss": 1.0722,
+      "step": 1469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3129148483276367,
+      "learning_rate": 9.322146047597045e-06,
+      "loss": 0.8603,
+      "step": 1470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.292138934135437,
+      "learning_rate": 9.32107860647672e-06,
+      "loss": 0.9033,
+      "step": 1471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2761929035186768,
+      "learning_rate": 9.320010386762314e-06,
+      "loss": 0.8307,
+      "step": 1472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3439552783966064,
+      "learning_rate": 9.318941388646306e-06,
+      "loss": 0.8668,
+      "step": 1473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3058017492294312,
+      "learning_rate": 9.317871612321312e-06,
+      "loss": 0.9143,
+      "step": 1474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2533154487609863,
+      "learning_rate": 9.316801057980091e-06,
+      "loss": 0.9127,
+      "step": 1475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.371646761894226,
+      "learning_rate": 9.315729725815542e-06,
+      "loss": 0.8411,
+      "step": 1476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3244438171386719,
+      "learning_rate": 9.3146576160207e-06,
+      "loss": 0.9028,
+      "step": 1477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3033204078674316,
+      "learning_rate": 9.313584728788745e-06,
+      "loss": 0.8277,
+      "step": 1478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9115082621574402,
+      "learning_rate": 9.312511064312995e-06,
+      "loss": 1.0695,
+      "step": 1479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.793706476688385,
+      "learning_rate": 9.31143662278691e-06,
+      "loss": 1.1062,
+      "step": 1480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.266632318496704,
+      "learning_rate": 9.310361404404086e-06,
+      "loss": 0.852,
+      "step": 1481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2862836122512817,
+      "learning_rate": 9.309285409358263e-06,
+      "loss": 0.9172,
+      "step": 1482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9836291074752808,
+      "learning_rate": 9.30820863784332e-06,
+      "loss": 1.0606,
+      "step": 1483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3130632638931274,
+      "learning_rate": 9.307131090053271e-06,
+      "loss": 0.8412,
+      "step": 1484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2469388246536255,
+      "learning_rate": 9.306052766182278e-06,
+      "loss": 0.8295,
+      "step": 1485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7488835453987122,
+      "learning_rate": 9.304973666424638e-06,
+      "loss": 1.0859,
+      "step": 1486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2386302947998047,
+      "learning_rate": 9.303893790974789e-06,
+      "loss": 0.8417,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5215493440628052,
+      "learning_rate": 9.302813140027307e-06,
+      "loss": 0.8438,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8437498807907104,
+      "learning_rate": 9.301731713776912e-06,
+      "loss": 1.0841,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2989917993545532,
+      "learning_rate": 9.300649512418458e-06,
+      "loss": 0.8852,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3477925062179565,
+      "learning_rate": 9.299566536146943e-06,
+      "loss": 0.8953,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3211296796798706,
+      "learning_rate": 9.298482785157504e-06,
+      "loss": 0.8706,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.356980562210083,
+      "learning_rate": 9.297398259645415e-06,
+      "loss": 0.8301,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2872543334960938,
+      "learning_rate": 9.296312959806091e-06,
+      "loss": 0.9,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3247249126434326,
+      "learning_rate": 9.295226885835091e-06,
+      "loss": 0.845,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4972600936889648,
+      "learning_rate": 9.294140037928105e-06,
+      "loss": 0.8876,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8505331873893738,
+      "learning_rate": 9.293052416280967e-06,
+      "loss": 1.0513,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3231993913650513,
+      "learning_rate": 9.291964021089653e-06,
+      "loss": 0.8164,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3240830898284912,
+      "learning_rate": 9.290874852550274e-06,
+      "loss": 0.8528,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.686992347240448,
+      "learning_rate": 9.289784910859082e-06,
+      "loss": 1.1065,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3465070724487305,
+      "learning_rate": 9.288694196212468e-06,
+      "loss": 0.8607,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3532224893569946,
+      "learning_rate": 9.287602708806965e-06,
+      "loss": 0.8869,
+      "step": 1502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8767019510269165,
+      "learning_rate": 9.286510448839237e-06,
+      "loss": 0.8164,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3188328742980957,
+      "learning_rate": 9.2854174165061e-06,
+      "loss": 0.7925,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3470687866210938,
+      "learning_rate": 9.284323612004498e-06,
+      "loss": 0.8531,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.923248291015625,
+      "learning_rate": 9.283229035531518e-06,
+      "loss": 1.083,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8053075671195984,
+      "learning_rate": 9.282133687284389e-06,
+      "loss": 1.097,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3473455905914307,
+      "learning_rate": 9.281037567460474e-06,
+      "loss": 0.8461,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2714478969573975,
+      "learning_rate": 9.27994067625728e-06,
+      "loss": 0.7924,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3690831661224365,
+      "learning_rate": 9.278843013872447e-06,
+      "loss": 0.8847,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2790180444717407,
+      "learning_rate": 9.27774458050376e-06,
+      "loss": 0.8639,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2967411279678345,
+      "learning_rate": 9.276645376349142e-06,
+      "loss": 0.87,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2684439420700073,
+      "learning_rate": 9.27554540160665e-06,
+      "loss": 0.8641,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3855772018432617,
+      "learning_rate": 9.274444656474482e-06,
+      "loss": 1.0523,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3682068586349487,
+      "learning_rate": 9.27334314115098e-06,
+      "loss": 0.8525,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3405416011810303,
+      "learning_rate": 9.27224085583462e-06,
+      "loss": 0.917,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7964107990264893,
+      "learning_rate": 9.271137800724012e-06,
+      "loss": 1.0949,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3188025951385498,
+      "learning_rate": 9.270033976017918e-06,
+      "loss": 0.8881,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3337230682373047,
+      "learning_rate": 9.268929381915224e-06,
+      "loss": 0.8703,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7489079833030701,
+      "learning_rate": 9.267824018614967e-06,
+      "loss": 1.0436,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7896556258201599,
+      "learning_rate": 9.266717886316315e-06,
+      "loss": 1.086,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3721504211425781,
+      "learning_rate": 9.265610985218574e-06,
+      "loss": 0.8777,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3223201036453247,
+      "learning_rate": 9.264503315521193e-06,
+      "loss": 0.8578,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7487199902534485,
+      "learning_rate": 9.26339487742376e-06,
+      "loss": 1.0823,
+      "step": 1524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2638441324234009,
+      "learning_rate": 9.262285671125993e-06,
+      "loss": 0.8473,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3227863311767578,
+      "learning_rate": 9.261175696827762e-06,
+      "loss": 0.8319,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2686564922332764,
+      "learning_rate": 9.260064954729063e-06,
+      "loss": 0.8194,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3773082494735718,
+      "learning_rate": 9.258953445030035e-06,
+      "loss": 0.8216,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2777972221374512,
+      "learning_rate": 9.257841167930958e-06,
+      "loss": 0.8553,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2678182125091553,
+      "learning_rate": 9.256728123632245e-06,
+      "loss": 0.8508,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4213796854019165,
+      "learning_rate": 9.255614312334456e-06,
+      "loss": 0.9045,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4264744520187378,
+      "learning_rate": 9.254499734238276e-06,
+      "loss": 0.9317,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3810417652130127,
+      "learning_rate": 9.253384389544538e-06,
+      "loss": 0.8696,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.283365249633789,
+      "learning_rate": 9.252268278454211e-06,
+      "loss": 0.8294,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3911384344100952,
+      "learning_rate": 9.251151401168402e-06,
+      "loss": 0.8464,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3143086433410645,
+      "learning_rate": 9.250033757888355e-06,
+      "loss": 0.8214,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3559037446975708,
+      "learning_rate": 9.248915348815452e-06,
+      "loss": 0.8778,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3741432428359985,
+      "learning_rate": 9.247796174151215e-06,
+      "loss": 0.8294,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2459479570388794,
+      "learning_rate": 9.246676234097301e-06,
+      "loss": 0.8671,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3199760913848877,
+      "learning_rate": 9.24555552885551e-06,
+      "loss": 0.8468,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2806264162063599,
+      "learning_rate": 9.244434058627772e-06,
+      "loss": 0.8301,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4054595232009888,
+      "learning_rate": 9.243311823616163e-06,
+      "loss": 0.8665,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3234310150146484,
+      "learning_rate": 9.242188824022889e-06,
+      "loss": 0.9036,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2829822301864624,
+      "learning_rate": 9.241065060050302e-06,
+      "loss": 0.8777,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.272683024406433,
+      "learning_rate": 9.239940531900886e-06,
+      "loss": 0.8455,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9675427675247192,
+      "learning_rate": 9.238815239777263e-06,
+      "loss": 1.0672,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2085546255111694,
+      "learning_rate": 9.237689183882194e-06,
+      "loss": 0.8007,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2890681028366089,
+      "learning_rate": 9.23656236441858e-06,
+      "loss": 0.852,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2857118844985962,
+      "learning_rate": 9.235434781589455e-06,
+      "loss": 0.8395,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3169687986373901,
+      "learning_rate": 9.234306435597993e-06,
+      "loss": 0.9012,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2268459796905518,
+      "learning_rate": 9.233177326647504e-06,
+      "loss": 0.8338,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7658856511116028,
+      "learning_rate": 9.23204745494144e-06,
+      "loss": 1.073,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2632033824920654,
+      "learning_rate": 9.230916820683384e-06,
+      "loss": 0.8444,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3398877382278442,
+      "learning_rate": 9.22978542407706e-06,
+      "loss": 0.8561,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2777714729309082,
+      "learning_rate": 9.22865326532633e-06,
+      "loss": 0.8946,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.258187174797058,
+      "learning_rate": 9.227520344635193e-06,
+      "loss": 0.8673,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6677736043930054,
+      "learning_rate": 9.22638666220778e-06,
+      "loss": 1.091,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2871272563934326,
+      "learning_rate": 9.225252218248368e-06,
+      "loss": 0.8841,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.341727614402771,
+      "learning_rate": 9.224117012961366e-06,
+      "loss": 0.8677,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6827267408370972,
+      "learning_rate": 9.222981046551319e-06,
+      "loss": 1.085,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3317402601242065,
+      "learning_rate": 9.221844319222915e-06,
+      "loss": 0.858,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2399228811264038,
+      "learning_rate": 9.220706831180972e-06,
+      "loss": 0.8563,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.286474585533142,
+      "learning_rate": 9.219568582630449e-06,
+      "loss": 0.8683,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.280511736869812,
+      "learning_rate": 9.218429573776445e-06,
+      "loss": 0.8824,
+      "step": 1564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3472031354904175,
+      "learning_rate": 9.217289804824186e-06,
+      "loss": 0.8853,
+      "step": 1565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7007656097412109,
+      "learning_rate": 9.216149275979047e-06,
+      "loss": 1.0927,
+      "step": 1566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3243904113769531,
+      "learning_rate": 9.215007987446532e-06,
+      "loss": 0.8374,
+      "step": 1567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2801414728164673,
+      "learning_rate": 9.213865939432284e-06,
+      "loss": 0.8651,
+      "step": 1568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3581323623657227,
+      "learning_rate": 9.212723132142085e-06,
+      "loss": 0.8803,
+      "step": 1569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.329050898551941,
+      "learning_rate": 9.21157956578185e-06,
+      "loss": 0.893,
+      "step": 1570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4212745428085327,
+      "learning_rate": 9.210435240557631e-06,
+      "loss": 0.8663,
+      "step": 1571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3064486980438232,
+      "learning_rate": 9.209290156675623e-06,
+      "loss": 0.8685,
+      "step": 1572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3286808729171753,
+      "learning_rate": 9.20814431434215e-06,
+      "loss": 0.889,
+      "step": 1573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2743935585021973,
+      "learning_rate": 9.206997713763675e-06,
+      "loss": 0.826,
+      "step": 1574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3213025331497192,
+      "learning_rate": 9.2058503551468e-06,
+      "loss": 0.8921,
+      "step": 1575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2747143507003784,
+      "learning_rate": 9.204702238698264e-06,
+      "loss": 0.8257,
+      "step": 1576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6204150915145874,
+      "learning_rate": 9.203553364624937e-06,
+      "loss": 0.8437,
+      "step": 1577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2037841081619263,
+      "learning_rate": 9.20240373313383e-06,
+      "loss": 0.8508,
+      "step": 1578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2832289934158325,
+      "learning_rate": 9.201253344432088e-06,
+      "loss": 0.8024,
+      "step": 1579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7162715792655945,
+      "learning_rate": 9.200102198726996e-06,
+      "loss": 1.069,
+      "step": 1580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2727771997451782,
+      "learning_rate": 9.198950296225974e-06,
+      "loss": 0.8486,
+      "step": 1581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.326857328414917,
+      "learning_rate": 9.197797637136579e-06,
+      "loss": 0.9258,
+      "step": 1582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3437412977218628,
+      "learning_rate": 9.196644221666498e-06,
+      "loss": 0.8244,
+      "step": 1583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2700417041778564,
+      "learning_rate": 9.195490050023561e-06,
+      "loss": 0.8384,
+      "step": 1584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3129887580871582,
+      "learning_rate": 9.194335122415734e-06,
+      "loss": 0.8515,
+      "step": 1585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3030418157577515,
+      "learning_rate": 9.193179439051118e-06,
+      "loss": 0.8612,
+      "step": 1586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8185754418373108,
+      "learning_rate": 9.192023000137948e-06,
+      "loss": 1.0826,
+      "step": 1587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7030133008956909,
+      "learning_rate": 9.1908658058846e-06,
+      "loss": 1.0822,
+      "step": 1588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4006680250167847,
+      "learning_rate": 9.189707856499581e-06,
+      "loss": 0.8996,
+      "step": 1589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3641372919082642,
+      "learning_rate": 9.188549152191537e-06,
+      "loss": 0.8205,
+      "step": 1590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3378156423568726,
+      "learning_rate": 9.187389693169249e-06,
+      "loss": 0.8519,
+      "step": 1591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.865035891532898,
+      "learning_rate": 9.186229479641634e-06,
+      "loss": 1.0673,
+      "step": 1592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2558718919754028,
+      "learning_rate": 9.185068511817744e-06,
+      "loss": 0.8362,
+      "step": 1593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.462432861328125,
+      "learning_rate": 9.183906789906772e-06,
+      "loss": 0.8782,
+      "step": 1594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4179378747940063,
+      "learning_rate": 9.182744314118038e-06,
+      "loss": 0.9081,
+      "step": 1595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3399766683578491,
+      "learning_rate": 9.181581084661007e-06,
+      "loss": 0.8729,
+      "step": 1596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2431527376174927,
+      "learning_rate": 9.180417101745273e-06,
+      "loss": 0.8265,
+      "step": 1597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3418056964874268,
+      "learning_rate": 9.179252365580573e-06,
+      "loss": 0.8972,
+      "step": 1598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3325402736663818,
+      "learning_rate": 9.178086876376768e-06,
+      "loss": 0.8485,
+      "step": 1599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.982277512550354,
+      "learning_rate": 9.176920634343867e-06,
+      "loss": 1.0835,
+      "step": 1600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3052537441253662,
+      "learning_rate": 9.175753639692009e-06,
+      "loss": 0.8522,
+      "step": 1601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3392248153686523,
+      "learning_rate": 9.174585892631466e-06,
+      "loss": 0.8855,
+      "step": 1602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2902987003326416,
+      "learning_rate": 9.17341739337265e-06,
+      "loss": 0.86,
+      "step": 1603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3656911849975586,
+      "learning_rate": 9.17224814212611e-06,
+      "loss": 0.8857,
+      "step": 1604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9641063213348389,
+      "learning_rate": 9.171078139102525e-06,
+      "loss": 1.0679,
+      "step": 1605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8268311023712158,
+      "learning_rate": 9.169907384512713e-06,
+      "loss": 1.0822,
+      "step": 1606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4208472967147827,
+      "learning_rate": 9.168735878567625e-06,
+      "loss": 0.8528,
+      "step": 1607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7481525540351868,
+      "learning_rate": 9.16756362147835e-06,
+      "loss": 1.0743,
+      "step": 1608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2764308452606201,
+      "learning_rate": 9.166390613456113e-06,
+      "loss": 0.8815,
+      "step": 1609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9395394921302795,
+      "learning_rate": 9.16521685471227e-06,
+      "loss": 1.0816,
+      "step": 1610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2930620908737183,
+      "learning_rate": 9.164042345458317e-06,
+      "loss": 0.9053,
+      "step": 1611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.255264401435852,
+      "learning_rate": 9.162867085905881e-06,
+      "loss": 0.8235,
+      "step": 1612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2359026670455933,
+      "learning_rate": 9.161691076266727e-06,
+      "loss": 0.9084,
+      "step": 1613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7851872444152832,
+      "learning_rate": 9.160514316752756e-06,
+      "loss": 1.0728,
+      "step": 1614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4467363357543945,
+      "learning_rate": 9.159336807576e-06,
+      "loss": 0.8724,
+      "step": 1615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.323252558708191,
+      "learning_rate": 9.158158548948631e-06,
+      "loss": 0.8647,
+      "step": 1616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7344477772712708,
+      "learning_rate": 9.156979541082952e-06,
+      "loss": 1.0728,
+      "step": 1617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3390424251556396,
+      "learning_rate": 9.155799784191404e-06,
+      "loss": 0.8558,
+      "step": 1618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2571468353271484,
+      "learning_rate": 9.154619278486561e-06,
+      "loss": 0.8655,
+      "step": 1619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6900335550308228,
+      "learning_rate": 9.15343802418113e-06,
+      "loss": 1.0761,
+      "step": 1620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6689826846122742,
+      "learning_rate": 9.15225602148796e-06,
+      "loss": 1.0735,
+      "step": 1621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.345324993133545,
+      "learning_rate": 9.151073270620027e-06,
+      "loss": 0.8746,
+      "step": 1622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7471703290939331,
+      "learning_rate": 9.149889771790448e-06,
+      "loss": 1.077,
+      "step": 1623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2268095016479492,
+      "learning_rate": 9.14870552521247e-06,
+      "loss": 0.829,
+      "step": 1624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6505094170570374,
+      "learning_rate": 9.147520531099477e-06,
+      "loss": 1.0707,
+      "step": 1625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6616891026496887,
+      "learning_rate": 9.146334789664985e-06,
+      "loss": 1.0458,
+      "step": 1626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3264750242233276,
+      "learning_rate": 9.145148301122652e-06,
+      "loss": 0.8908,
+      "step": 1627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.296651840209961,
+      "learning_rate": 9.14396106568626e-06,
+      "loss": 0.858,
+      "step": 1628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6865271925926208,
+      "learning_rate": 9.142773083569735e-06,
+      "loss": 1.0703,
+      "step": 1629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2900117635726929,
+      "learning_rate": 9.141584354987134e-06,
+      "loss": 0.9407,
+      "step": 1630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2842940092086792,
+      "learning_rate": 9.140394880152645e-06,
+      "loss": 0.9266,
+      "step": 1631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.287392497062683,
+      "learning_rate": 9.139204659280596e-06,
+      "loss": 0.8409,
+      "step": 1632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.194199562072754,
+      "learning_rate": 9.138013692585446e-06,
+      "loss": 0.8469,
+      "step": 1633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2988991737365723,
+      "learning_rate": 9.136821980281792e-06,
+      "loss": 0.8412,
+      "step": 1634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.337073802947998,
+      "learning_rate": 9.13562952258436e-06,
+      "loss": 0.8707,
+      "step": 1635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8087652921676636,
+      "learning_rate": 9.134436319708012e-06,
+      "loss": 1.0697,
+      "step": 1636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2624655961990356,
+      "learning_rate": 9.13324237186775e-06,
+      "loss": 0.9195,
+      "step": 1637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3031694889068604,
+      "learning_rate": 9.132047679278702e-06,
+      "loss": 0.9378,
+      "step": 1638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3163673877716064,
+      "learning_rate": 9.130852242156135e-06,
+      "loss": 0.8585,
+      "step": 1639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.324325442314148,
+      "learning_rate": 9.129656060715447e-06,
+      "loss": 0.8551,
+      "step": 1640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2503502368927002,
+      "learning_rate": 9.128459135172176e-06,
+      "loss": 0.8757,
+      "step": 1641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4378740787506104,
+      "learning_rate": 9.127261465741987e-06,
+      "loss": 0.8285,
+      "step": 1642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.266660213470459,
+      "learning_rate": 9.126063052640683e-06,
+      "loss": 0.8945,
+      "step": 1643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3243935108184814,
+      "learning_rate": 9.124863896084204e-06,
+      "loss": 0.9047,
+      "step": 1644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3227267265319824,
+      "learning_rate": 9.123663996288614e-06,
+      "loss": 0.901,
+      "step": 1645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7611749768257141,
+      "learning_rate": 9.122463353470118e-06,
+      "loss": 1.0577,
+      "step": 1646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2740439176559448,
+      "learning_rate": 9.121261967845059e-06,
+      "loss": 0.8556,
+      "step": 1647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.711050808429718,
+      "learning_rate": 9.120059839629905e-06,
+      "loss": 1.068,
+      "step": 1648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7022916078567505,
+      "learning_rate": 9.118856969041262e-06,
+      "loss": 1.0886,
+      "step": 1649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2658891677856445,
+      "learning_rate": 9.117653356295868e-06,
+      "loss": 0.8279,
+      "step": 1650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3794409036636353,
+      "learning_rate": 9.1164490016106e-06,
+      "loss": 0.8607,
+      "step": 1651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2983168363571167,
+      "learning_rate": 9.11524390520246e-06,
+      "loss": 0.7788,
+      "step": 1652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8767660856246948,
+      "learning_rate": 9.114038067288594e-06,
+      "loss": 1.0833,
+      "step": 1653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3398394584655762,
+      "learning_rate": 9.112831488086271e-06,
+      "loss": 0.9232,
+      "step": 1654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.295066237449646,
+      "learning_rate": 9.1116241678129e-06,
+      "loss": 0.9123,
+      "step": 1655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6759821176528931,
+      "learning_rate": 9.110416106686025e-06,
+      "loss": 1.0589,
+      "step": 1656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6531561017036438,
+      "learning_rate": 9.109207304923318e-06,
+      "loss": 1.0845,
+      "step": 1657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.656116783618927,
+      "learning_rate": 9.107997762742586e-06,
+      "loss": 1.0739,
+      "step": 1658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3029693365097046,
+      "learning_rate": 9.106787480361771e-06,
+      "loss": 0.8055,
+      "step": 1659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4105992317199707,
+      "learning_rate": 9.105576457998952e-06,
+      "loss": 0.8888,
+      "step": 1660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3367174863815308,
+      "learning_rate": 9.104364695872334e-06,
+      "loss": 0.8376,
+      "step": 1661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3206970691680908,
+      "learning_rate": 9.103152194200258e-06,
+      "loss": 0.9135,
+      "step": 1662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.254521131515503,
+      "learning_rate": 9.101938953201198e-06,
+      "loss": 0.8622,
+      "step": 1663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3419466018676758,
+      "learning_rate": 9.100724973093764e-06,
+      "loss": 0.9306,
+      "step": 1664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2524337768554688,
+      "learning_rate": 9.099510254096698e-06,
+      "loss": 0.8714,
+      "step": 1665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3376166820526123,
+      "learning_rate": 9.09829479642887e-06,
+      "loss": 0.8681,
+      "step": 1666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2777754068374634,
+      "learning_rate": 9.097078600309293e-06,
+      "loss": 0.9208,
+      "step": 1667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2569925785064697,
+      "learning_rate": 9.095861665957104e-06,
+      "loss": 0.9272,
+      "step": 1668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3506686687469482,
+      "learning_rate": 9.094643993591575e-06,
+      "loss": 0.8887,
+      "step": 1669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3260306119918823,
+      "learning_rate": 9.093425583432117e-06,
+      "loss": 0.9084,
+      "step": 1670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2962387800216675,
+      "learning_rate": 9.092206435698265e-06,
+      "loss": 0.8338,
+      "step": 1671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9726826548576355,
+      "learning_rate": 9.090986550609695e-06,
+      "loss": 1.0899,
+      "step": 1672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.854941725730896,
+      "learning_rate": 9.08976592838621e-06,
+      "loss": 1.0917,
+      "step": 1673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3625329732894897,
+      "learning_rate": 9.088544569247748e-06,
+      "loss": 0.8772,
+      "step": 1674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2881193161010742,
+      "learning_rate": 9.08732247341438e-06,
+      "loss": 0.8648,
+      "step": 1675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.382682204246521,
+      "learning_rate": 9.08609964110631e-06,
+      "loss": 0.8933,
+      "step": 1676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2786017656326294,
+      "learning_rate": 9.084876072543874e-06,
+      "loss": 1.0977,
+      "step": 1677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1453545093536377,
+      "learning_rate": 9.08365176794754e-06,
+      "loss": 1.0592,
+      "step": 1678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.291301965713501,
+      "learning_rate": 9.082426727537909e-06,
+      "loss": 0.8335,
+      "step": 1679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3396730422973633,
+      "learning_rate": 9.081200951535719e-06,
+      "loss": 0.8662,
+      "step": 1680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3299226760864258,
+      "learning_rate": 9.079974440161831e-06,
+      "loss": 0.8699,
+      "step": 1681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3570890426635742,
+      "learning_rate": 9.07874719363725e-06,
+      "loss": 0.9356,
+      "step": 1682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2809946537017822,
+      "learning_rate": 9.077519212183101e-06,
+      "loss": 0.8142,
+      "step": 1683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.268712043762207,
+      "learning_rate": 9.076290496020653e-06,
+      "loss": 0.8826,
+      "step": 1684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2749519348144531,
+      "learning_rate": 9.0750610453713e-06,
+      "loss": 0.8638,
+      "step": 1685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3090721368789673,
+      "learning_rate": 9.073830860456574e-06,
+      "loss": 0.9214,
+      "step": 1686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0256130695343018,
+      "learning_rate": 9.07259994149813e-06,
+      "loss": 1.0634,
+      "step": 1687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2402422428131104,
+      "learning_rate": 9.071368288717768e-06,
+      "loss": 0.872,
+      "step": 1688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2540857791900635,
+      "learning_rate": 9.070135902337408e-06,
+      "loss": 0.7857,
+      "step": 1689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.268458366394043,
+      "learning_rate": 9.068902782579111e-06,
+      "loss": 0.8658,
+      "step": 1690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3571828603744507,
+      "learning_rate": 9.067668929665066e-06,
+      "loss": 0.8942,
+      "step": 1691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8214802742004395,
+      "learning_rate": 9.066434343817593e-06,
+      "loss": 1.0871,
+      "step": 1692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7498634457588196,
+      "learning_rate": 9.06519902525915e-06,
+      "loss": 1.0853,
+      "step": 1693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6633102297782898,
+      "learning_rate": 9.06396297421232e-06,
+      "loss": 1.0754,
+      "step": 1694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3980287313461304,
+      "learning_rate": 9.062726190899822e-06,
+      "loss": 0.9034,
+      "step": 1695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.912337064743042,
+      "learning_rate": 9.061488675544505e-06,
+      "loss": 1.0732,
+      "step": 1696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2675312757492065,
+      "learning_rate": 9.060250428369353e-06,
+      "loss": 0.8752,
+      "step": 1697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3098629713058472,
+      "learning_rate": 9.059011449597476e-06,
+      "loss": 0.8134,
+      "step": 1698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8037959933280945,
+      "learning_rate": 9.057771739452125e-06,
+      "loss": 1.0863,
+      "step": 1699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2994356155395508,
+      "learning_rate": 9.056531298156672e-06,
+      "loss": 0.8485,
+      "step": 1700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3698545694351196,
+      "learning_rate": 9.05529012593463e-06,
+      "loss": 0.9436,
+      "step": 1701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7547337412834167,
+      "learning_rate": 9.054048223009638e-06,
+      "loss": 1.0949,
+      "step": 1702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.24457848072052,
+      "learning_rate": 9.052805589605465e-06,
+      "loss": 0.9117,
+      "step": 1703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3910436630249023,
+      "learning_rate": 9.051562225946023e-06,
+      "loss": 0.8525,
+      "step": 1704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.263262152671814,
+      "learning_rate": 9.05031813225534e-06,
+      "loss": 0.9258,
+      "step": 1705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2512948513031006,
+      "learning_rate": 9.049073308757588e-06,
+      "loss": 0.8143,
+      "step": 1706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2592626810073853,
+      "learning_rate": 9.047827755677063e-06,
+      "loss": 0.9062,
+      "step": 1707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8882735371589661,
+      "learning_rate": 9.046581473238196e-06,
+      "loss": 1.07,
+      "step": 1708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3638882637023926,
+      "learning_rate": 9.045334461665551e-06,
+      "loss": 0.8566,
+      "step": 1709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3606510162353516,
+      "learning_rate": 9.044086721183816e-06,
+      "loss": 0.8669,
+      "step": 1710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3252989053726196,
+      "learning_rate": 9.04283825201782e-06,
+      "loss": 0.8453,
+      "step": 1711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2331699132919312,
+      "learning_rate": 9.041589054392514e-06,
+      "loss": 0.8544,
+      "step": 1712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8695377707481384,
+      "learning_rate": 9.04033912853299e-06,
+      "loss": 1.0702,
+      "step": 1713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2237684726715088,
+      "learning_rate": 9.039088474664462e-06,
+      "loss": 0.892,
+      "step": 1714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3017386198043823,
+      "learning_rate": 9.037837093012281e-06,
+      "loss": 0.9188,
+      "step": 1715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.282936692237854,
+      "learning_rate": 9.036584983801926e-06,
+      "loss": 0.8609,
+      "step": 1716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3765010833740234,
+      "learning_rate": 9.035332147259011e-06,
+      "loss": 0.8513,
+      "step": 1717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.774401068687439,
+      "learning_rate": 9.034078583609275e-06,
+      "loss": 1.0743,
+      "step": 1718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2637157440185547,
+      "learning_rate": 9.032824293078595e-06,
+      "loss": 0.8612,
+      "step": 1719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.300920844078064,
+      "learning_rate": 9.031569275892971e-06,
+      "loss": 0.8609,
+      "step": 1720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6664302349090576,
+      "learning_rate": 9.030313532278544e-06,
+      "loss": 1.0722,
+      "step": 1721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2555747032165527,
+      "learning_rate": 9.029057062461577e-06,
+      "loss": 0.9106,
+      "step": 1722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.297909140586853,
+      "learning_rate": 9.02779986666847e-06,
+      "loss": 0.8748,
+      "step": 1723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2842434644699097,
+      "learning_rate": 9.026541945125747e-06,
+      "loss": 0.8549,
+      "step": 1724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3050992488861084,
+      "learning_rate": 9.025283298060067e-06,
+      "loss": 0.8475,
+      "step": 1725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7506802678108215,
+      "learning_rate": 9.024023925698222e-06,
+      "loss": 1.0433,
+      "step": 1726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2919895648956299,
+      "learning_rate": 9.022763828267132e-06,
+      "loss": 0.8715,
+      "step": 1727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3808095455169678,
+      "learning_rate": 9.021503005993846e-06,
+      "loss": 0.9202,
+      "step": 1728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2560158967971802,
+      "learning_rate": 9.020241459105548e-06,
+      "loss": 0.8584,
+      "step": 1729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6649852991104126,
+      "learning_rate": 9.018979187829546e-06,
+      "loss": 1.0751,
+      "step": 1730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6650665402412415,
+      "learning_rate": 9.017716192393287e-06,
+      "loss": 1.0784,
+      "step": 1731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.266905426979065,
+      "learning_rate": 9.016452473024341e-06,
+      "loss": 0.8937,
+      "step": 1732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6476612091064453,
+      "learning_rate": 9.015188029950413e-06,
+      "loss": 1.0743,
+      "step": 1733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6488327383995056,
+      "learning_rate": 9.013922863399335e-06,
+      "loss": 1.0814,
+      "step": 1734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.263551115989685,
+      "learning_rate": 9.012656973599074e-06,
+      "loss": 0.8396,
+      "step": 1735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3219302892684937,
+      "learning_rate": 9.011390360777722e-06,
+      "loss": 0.9082,
+      "step": 1736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2683286666870117,
+      "learning_rate": 9.010123025163502e-06,
+      "loss": 0.8434,
+      "step": 1737
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3021469116210938,
+      "learning_rate": 9.008854966984775e-06,
+      "loss": 0.9072,
+      "step": 1738
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.353756308555603,
+      "learning_rate": 9.007586186470022e-06,
+      "loss": 0.901,
+      "step": 1739
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2891960144042969,
+      "learning_rate": 9.00631668384786e-06,
+      "loss": 0.9321,
+      "step": 1740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.295147180557251,
+      "learning_rate": 9.005046459347031e-06,
+      "loss": 0.8481,
+      "step": 1741
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8654597997665405,
+      "learning_rate": 9.003775513196416e-06,
+      "loss": 1.0821,
+      "step": 1742
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2984306812286377,
+      "learning_rate": 9.002503845625016e-06,
+      "loss": 0.8645,
+      "step": 1743
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3128533363342285,
+      "learning_rate": 9.001231456861971e-06,
+      "loss": 0.8789,
+      "step": 1744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.311476707458496,
+      "learning_rate": 8.999958347136542e-06,
+      "loss": 0.8834,
+      "step": 1745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8002899885177612,
+      "learning_rate": 8.998684516678126e-06,
+      "loss": 1.069,
+      "step": 1746
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7157211899757385,
+      "learning_rate": 8.99740996571625e-06,
+      "loss": 1.0897,
+      "step": 1747
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2763283252716064,
+      "learning_rate": 8.996134694480568e-06,
+      "loss": 0.8016,
+      "step": 1748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6984684467315674,
+      "learning_rate": 8.994858703200864e-06,
+      "loss": 1.0686,
+      "step": 1749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7143226861953735,
+      "learning_rate": 8.993581992107054e-06,
+      "loss": 1.0667,
+      "step": 1750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.196711540222168,
+      "learning_rate": 8.992304561429182e-06,
+      "loss": 0.8219,
+      "step": 1751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4284312725067139,
+      "learning_rate": 8.991026411397421e-06,
+      "loss": 0.8607,
+      "step": 1752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3033111095428467,
+      "learning_rate": 8.989747542242075e-06,
+      "loss": 0.8828,
+      "step": 1753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2685092687606812,
+      "learning_rate": 8.988467954193578e-06,
+      "loss": 0.8335,
+      "step": 1754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7729329466819763,
+      "learning_rate": 8.987187647482492e-06,
+      "loss": 1.0588,
+      "step": 1755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7212794423103333,
+      "learning_rate": 8.985906622339509e-06,
+      "loss": 1.0633,
+      "step": 1756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3342154026031494,
+      "learning_rate": 8.984624878995452e-06,
+      "loss": 0.8978,
+      "step": 1757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3049262762069702,
+      "learning_rate": 8.983342417681271e-06,
+      "loss": 0.8644,
+      "step": 1758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2463408708572388,
+      "learning_rate": 8.982059238628047e-06,
+      "loss": 0.8214,
+      "step": 1759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8703387975692749,
+      "learning_rate": 8.98077534206699e-06,
+      "loss": 1.0638,
+      "step": 1760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.276430368423462,
+      "learning_rate": 8.979490728229438e-06,
+      "loss": 0.8581,
+      "step": 1761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.28687584400177,
+      "learning_rate": 8.978205397346859e-06,
+      "loss": 0.8563,
+      "step": 1762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2638177871704102,
+      "learning_rate": 8.976919349650853e-06,
+      "loss": 0.8125,
+      "step": 1763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.324032187461853,
+      "learning_rate": 8.975632585373143e-06,
+      "loss": 0.8038,
+      "step": 1764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7563796639442444,
+      "learning_rate": 8.97434510474559e-06,
+      "loss": 1.084,
+      "step": 1765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3570705652236938,
+      "learning_rate": 8.973056908000173e-06,
+      "loss": 0.9153,
+      "step": 1766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2855883836746216,
+      "learning_rate": 8.97176799536901e-06,
+      "loss": 0.8377,
+      "step": 1767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2689019441604614,
+      "learning_rate": 8.970478367084342e-06,
+      "loss": 0.9136,
+      "step": 1768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2478238344192505,
+      "learning_rate": 8.969188023378539e-06,
+      "loss": 0.8792,
+      "step": 1769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2481746673583984,
+      "learning_rate": 8.967896964484106e-06,
+      "loss": 0.865,
+      "step": 1770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2619715929031372,
+      "learning_rate": 8.96660519063367e-06,
+      "loss": 0.8659,
+      "step": 1771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3706191778182983,
+      "learning_rate": 8.96531270205999e-06,
+      "loss": 0.8628,
+      "step": 1772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2848198413848877,
+      "learning_rate": 8.964019498995955e-06,
+      "loss": 0.8706,
+      "step": 1773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8487981557846069,
+      "learning_rate": 8.962725581674575e-06,
+      "loss": 1.0736,
+      "step": 1774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.279005527496338,
+      "learning_rate": 8.961430950329004e-06,
+      "loss": 0.8311,
+      "step": 1775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.400479793548584,
+      "learning_rate": 8.960135605192506e-06,
+      "loss": 0.8808,
+      "step": 1776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.245587944984436,
+      "learning_rate": 8.95883954649849e-06,
+      "loss": 0.7917,
+      "step": 1777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7407777309417725,
+      "learning_rate": 8.957542774480483e-06,
+      "loss": 1.0762,
+      "step": 1778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6954470872879028,
+      "learning_rate": 8.956245289372142e-06,
+      "loss": 1.0506,
+      "step": 1779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6447061896324158,
+      "learning_rate": 8.95494709140726e-06,
+      "loss": 1.0767,
+      "step": 1780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6487475037574768,
+      "learning_rate": 8.95364818081975e-06,
+      "loss": 1.0837,
+      "step": 1781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3299518823623657,
+      "learning_rate": 8.952348557843655e-06,
+      "loss": 0.8162,
+      "step": 1782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.275231957435608,
+      "learning_rate": 8.95104822271315e-06,
+      "loss": 0.8564,
+      "step": 1783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2345670461654663,
+      "learning_rate": 8.949747175662535e-06,
+      "loss": 0.8485,
+      "step": 1784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3401345014572144,
+      "learning_rate": 8.94844541692624e-06,
+      "loss": 0.8938,
+      "step": 1785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2904781103134155,
+      "learning_rate": 8.947142946738821e-06,
+      "loss": 0.8019,
+      "step": 1786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.337654948234558,
+      "learning_rate": 8.945839765334964e-06,
+      "loss": 0.9083,
+      "step": 1787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.275899052619934,
+      "learning_rate": 8.944535872949487e-06,
+      "loss": 0.8695,
+      "step": 1788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0646928548812866,
+      "learning_rate": 8.943231269817327e-06,
+      "loss": 1.0812,
+      "step": 1789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2292882204055786,
+      "learning_rate": 8.941925956173556e-06,
+      "loss": 0.8569,
+      "step": 1790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3806543350219727,
+      "learning_rate": 8.94061993225337e-06,
+      "loss": 0.8307,
+      "step": 1791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3697901964187622,
+      "learning_rate": 8.939313198292098e-06,
+      "loss": 0.9061,
+      "step": 1792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7365099787712097,
+      "learning_rate": 8.938005754525194e-06,
+      "loss": 1.0759,
+      "step": 1793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3107945919036865,
+      "learning_rate": 8.936697601188238e-06,
+      "loss": 0.8886,
+      "step": 1794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2718381881713867,
+      "learning_rate": 8.935388738516942e-06,
+      "loss": 0.9233,
+      "step": 1795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2722970247268677,
+      "learning_rate": 8.934079166747142e-06,
+      "loss": 0.8147,
+      "step": 1796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.318584680557251,
+      "learning_rate": 8.932768886114803e-06,
+      "loss": 0.8833,
+      "step": 1797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2533003091812134,
+      "learning_rate": 8.93145789685602e-06,
+      "loss": 0.8933,
+      "step": 1798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2481677532196045,
+      "learning_rate": 8.93014619920701e-06,
+      "loss": 0.8162,
+      "step": 1799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2660287618637085,
+      "learning_rate": 8.928833793404125e-06,
+      "loss": 0.8419,
+      "step": 1800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3671151399612427,
+      "learning_rate": 8.927520679683842e-06,
+      "loss": 0.881,
+      "step": 1801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.003148078918457,
+      "learning_rate": 8.926206858282759e-06,
+      "loss": 1.0333,
+      "step": 1802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7259765267372131,
+      "learning_rate": 8.924892329437611e-06,
+      "loss": 1.0624,
+      "step": 1803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6608260869979858,
+      "learning_rate": 8.923577093385258e-06,
+      "loss": 1.0545,
+      "step": 1804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3326034545898438,
+      "learning_rate": 8.922261150362683e-06,
+      "loss": 0.8769,
+      "step": 1805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3795536756515503,
+      "learning_rate": 8.920944500607e-06,
+      "loss": 0.8899,
+      "step": 1806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2317980527877808,
+      "learning_rate": 8.91962714435545e-06,
+      "loss": 1.0673,
+      "step": 1807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.273640751838684,
+      "learning_rate": 8.918309081845402e-06,
+      "loss": 0.8835,
+      "step": 1808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.237221360206604,
+      "learning_rate": 8.91699031331435e-06,
+      "loss": 0.8943,
+      "step": 1809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.296996831893921,
+      "learning_rate": 8.915670838999917e-06,
+      "loss": 0.87,
+      "step": 1810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7768531441688538,
+      "learning_rate": 8.914350659139852e-06,
+      "loss": 1.0805,
+      "step": 1811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7172191739082336,
+      "learning_rate": 8.913029773972033e-06,
+      "loss": 1.0291,
+      "step": 1812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3254262208938599,
+      "learning_rate": 8.911708183734462e-06,
+      "loss": 0.8612,
+      "step": 1813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4713387489318848,
+      "learning_rate": 8.910385888665271e-06,
+      "loss": 0.7788,
+      "step": 1814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2638458013534546,
+      "learning_rate": 8.909062889002717e-06,
+      "loss": 0.8812,
+      "step": 1815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8435591459274292,
+      "learning_rate": 8.907739184985188e-06,
+      "loss": 1.0719,
+      "step": 1816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2564040422439575,
+      "learning_rate": 8.906414776851192e-06,
+      "loss": 0.8294,
+      "step": 1817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7429515719413757,
+      "learning_rate": 8.905089664839371e-06,
+      "loss": 1.0614,
+      "step": 1818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2689279317855835,
+      "learning_rate": 8.903763849188487e-06,
+      "loss": 0.8622,
+      "step": 1819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7472085952758789,
+      "learning_rate": 8.902437330137433e-06,
+      "loss": 1.0608,
+      "step": 1820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2060976028442383,
+      "learning_rate": 8.901110107925229e-06,
+      "loss": 0.8302,
+      "step": 1821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7289021015167236,
+      "learning_rate": 8.899782182791021e-06,
+      "loss": 1.0933,
+      "step": 1822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.241314172744751,
+      "learning_rate": 8.898453554974082e-06,
+      "loss": 0.9108,
+      "step": 1823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7128763794898987,
+      "learning_rate": 8.89712422471381e-06,
+      "loss": 1.0746,
+      "step": 1824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.291453242301941,
+      "learning_rate": 8.895794192249729e-06,
+      "loss": 0.8367,
+      "step": 1825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3055419921875,
+      "learning_rate": 8.894463457821494e-06,
+      "loss": 0.8139,
+      "step": 1826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7100529670715332,
+      "learning_rate": 8.89313202166888e-06,
+      "loss": 1.0896,
+      "step": 1827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3020634651184082,
+      "learning_rate": 8.891799884031796e-06,
+      "loss": 0.7708,
+      "step": 1828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.654345691204071,
+      "learning_rate": 8.89046704515027e-06,
+      "loss": 1.0715,
+      "step": 1829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2943476438522339,
+      "learning_rate": 8.889133505264462e-06,
+      "loss": 0.8789,
+      "step": 1830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.233794927597046,
+      "learning_rate": 8.887799264614653e-06,
+      "loss": 0.8116,
+      "step": 1831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7006654143333435,
+      "learning_rate": 8.886464323441259e-06,
+      "loss": 1.0719,
+      "step": 1832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3361471891403198,
+      "learning_rate": 8.88512868198481e-06,
+      "loss": 0.8372,
+      "step": 1833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3509457111358643,
+      "learning_rate": 8.883792340485973e-06,
+      "loss": 0.8553,
+      "step": 1834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6707969903945923,
+      "learning_rate": 8.882455299185536e-06,
+      "loss": 1.0726,
+      "step": 1835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2769793272018433,
+      "learning_rate": 8.881117558324414e-06,
+      "loss": 0.8682,
+      "step": 1836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2159932851791382,
+      "learning_rate": 8.879779118143646e-06,
+      "loss": 0.8855,
+      "step": 1837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6350764632225037,
+      "learning_rate": 8.878439978884401e-06,
+      "loss": 1.0674,
+      "step": 1838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3375568389892578,
+      "learning_rate": 8.877100140787972e-06,
+      "loss": 0.8503,
+      "step": 1839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3279846906661987,
+      "learning_rate": 8.875759604095777e-06,
+      "loss": 0.8915,
+      "step": 1840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3122185468673706,
+      "learning_rate": 8.874418369049362e-06,
+      "loss": 0.8318,
+      "step": 1841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3352247476577759,
+      "learning_rate": 8.873076435890396e-06,
+      "loss": 0.8754,
+      "step": 1842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6992528438568115,
+      "learning_rate": 8.871733804860676e-06,
+      "loss": 1.0676,
+      "step": 1843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2804183959960938,
+      "learning_rate": 8.870390476202126e-06,
+      "loss": 0.8513,
+      "step": 1844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3295038938522339,
+      "learning_rate": 8.869046450156792e-06,
+      "loss": 0.8894,
+      "step": 1845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2392350435256958,
+      "learning_rate": 8.867701726966846e-06,
+      "loss": 0.8393,
+      "step": 1846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2620208263397217,
+      "learning_rate": 8.866356306874589e-06,
+      "loss": 0.8437,
+      "step": 1847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6795346736907959,
+      "learning_rate": 8.865010190122445e-06,
+      "loss": 1.0628,
+      "step": 1848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.259092926979065,
+      "learning_rate": 8.863663376952967e-06,
+      "loss": 0.8441,
+      "step": 1849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2960121631622314,
+      "learning_rate": 8.862315867608827e-06,
+      "loss": 0.8983,
+      "step": 1850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7129955291748047,
+      "learning_rate": 8.860967662332827e-06,
+      "loss": 1.0669,
+      "step": 1851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2770737409591675,
+      "learning_rate": 8.859618761367894e-06,
+      "loss": 0.8783,
+      "step": 1852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6917423605918884,
+      "learning_rate": 8.858269164957081e-06,
+      "loss": 1.0895,
+      "step": 1853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6406493186950684,
+      "learning_rate": 8.856918873343563e-06,
+      "loss": 1.0803,
+      "step": 1854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3387750387191772,
+      "learning_rate": 8.855567886770644e-06,
+      "loss": 0.9107,
+      "step": 1855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.306354284286499,
+      "learning_rate": 8.854216205481751e-06,
+      "loss": 0.8529,
+      "step": 1856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2636277675628662,
+      "learning_rate": 8.85286382972044e-06,
+      "loss": 0.888,
+      "step": 1857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.342434048652649,
+      "learning_rate": 8.851510759730382e-06,
+      "loss": 0.8858,
+      "step": 1858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3060681819915771,
+      "learning_rate": 8.850156995755385e-06,
+      "loss": 0.9243,
+      "step": 1859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1298211812973022,
+      "learning_rate": 8.848802538039378e-06,
+      "loss": 1.0493,
+      "step": 1860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2705069780349731,
+      "learning_rate": 8.847447386826411e-06,
+      "loss": 0.8838,
+      "step": 1861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2685141563415527,
+      "learning_rate": 8.846091542360665e-06,
+      "loss": 0.8627,
+      "step": 1862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7633891105651855,
+      "learning_rate": 8.84473500488644e-06,
+      "loss": 1.0563,
+      "step": 1863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.155350923538208,
+      "learning_rate": 8.843377774648166e-06,
+      "loss": 0.8908,
+      "step": 1864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2922497987747192,
+      "learning_rate": 8.842019851890396e-06,
+      "loss": 0.8076,
+      "step": 1865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2701361179351807,
+      "learning_rate": 8.840661236857807e-06,
+      "loss": 0.9024,
+      "step": 1866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.209060788154602,
+      "learning_rate": 8.8393019297952e-06,
+      "loss": 0.8473,
+      "step": 1867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2802071571350098,
+      "learning_rate": 8.837941930947501e-06,
+      "loss": 0.8528,
+      "step": 1868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3580456972122192,
+      "learning_rate": 8.836581240559765e-06,
+      "loss": 0.8606,
+      "step": 1869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3347755670547485,
+      "learning_rate": 8.835219858877167e-06,
+      "loss": 0.7817,
+      "step": 1870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2790679931640625,
+      "learning_rate": 8.833857786145006e-06,
+      "loss": 0.866,
+      "step": 1871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2491192817687988,
+      "learning_rate": 8.832495022608709e-06,
+      "loss": 0.8919,
+      "step": 1872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2935471534729004,
+      "learning_rate": 8.831131568513824e-06,
+      "loss": 0.8218,
+      "step": 1873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.30674147605896,
+      "learning_rate": 8.829767424106028e-06,
+      "loss": 0.8228,
+      "step": 1874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3013381958007812,
+      "learning_rate": 8.828402589631116e-06,
+      "loss": 0.8704,
+      "step": 1875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3105649948120117,
+      "learning_rate": 8.827037065335014e-06,
+      "loss": 0.9284,
+      "step": 1876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3137614727020264,
+      "learning_rate": 8.825670851463769e-06,
+      "loss": 0.9029,
+      "step": 1877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8992375731468201,
+      "learning_rate": 8.824303948263548e-06,
+      "loss": 1.0839,
+      "step": 1878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7392756938934326,
+      "learning_rate": 8.822936355980653e-06,
+      "loss": 1.0473,
+      "step": 1879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6923186779022217,
+      "learning_rate": 8.821568074861498e-06,
+      "loss": 1.0699,
+      "step": 1880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2982527017593384,
+      "learning_rate": 8.820199105152631e-06,
+      "loss": 0.8697,
+      "step": 1881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.289975643157959,
+      "learning_rate": 8.818829447100719e-06,
+      "loss": 0.846,
+      "step": 1882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.288438081741333,
+      "learning_rate": 8.817459100952552e-06,
+      "loss": 0.7792,
+      "step": 1883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2305828332901,
+      "learning_rate": 8.816088066955048e-06,
+      "loss": 0.8556,
+      "step": 1884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3245494365692139,
+      "learning_rate": 8.814716345355249e-06,
+      "loss": 0.8744,
+      "step": 1885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2153843641281128,
+      "learning_rate": 8.813343936400314e-06,
+      "loss": 0.8514,
+      "step": 1886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3815885782241821,
+      "learning_rate": 8.811970840337534e-06,
+      "loss": 0.9304,
+      "step": 1887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2582796812057495,
+      "learning_rate": 8.810597057414316e-06,
+      "loss": 0.7985,
+      "step": 1888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.24312162399292,
+      "learning_rate": 8.809222587878203e-06,
+      "loss": 0.8587,
+      "step": 1889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.185123324394226,
+      "learning_rate": 8.807847431976848e-06,
+      "loss": 0.8461,
+      "step": 1890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3060132265090942,
+      "learning_rate": 8.806471589958035e-06,
+      "loss": 0.8449,
+      "step": 1891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.252765417098999,
+      "learning_rate": 8.805095062069671e-06,
+      "loss": 0.8325,
+      "step": 1892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2611225843429565,
+      "learning_rate": 8.803717848559785e-06,
+      "loss": 0.8643,
+      "step": 1893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2899560928344727,
+      "learning_rate": 8.80233994967653e-06,
+      "loss": 0.8645,
+      "step": 1894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2736467123031616,
+      "learning_rate": 8.800961365668187e-06,
+      "loss": 0.824,
+      "step": 1895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3861703872680664,
+      "learning_rate": 8.799582096783149e-06,
+      "loss": 0.8814,
+      "step": 1896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2920624017715454,
+      "learning_rate": 8.798202143269945e-06,
+      "loss": 0.8175,
+      "step": 1897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5603231191635132,
+      "learning_rate": 8.796821505377222e-06,
+      "loss": 1.0891,
+      "step": 1898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2529518604278564,
+      "learning_rate": 8.795440183353747e-06,
+      "loss": 1.0609,
+      "step": 1899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3532220125198364,
+      "learning_rate": 8.794058177448418e-06,
+      "loss": 0.8698,
+      "step": 1900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.292570948600769,
+      "learning_rate": 8.792675487910249e-06,
+      "loss": 0.837,
+      "step": 1901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2581837177276611,
+      "learning_rate": 8.791292114988382e-06,
+      "loss": 0.8303,
+      "step": 1902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3472720384597778,
+      "learning_rate": 8.789908058932077e-06,
+      "loss": 1.0778,
+      "step": 1903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1985386610031128,
+      "learning_rate": 8.788523319990725e-06,
+      "loss": 1.0673,
+      "step": 1904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9283960461616516,
+      "learning_rate": 8.78713789841383e-06,
+      "loss": 1.0363,
+      "step": 1905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4137272834777832,
+      "learning_rate": 8.785751794451028e-06,
+      "loss": 0.8237,
+      "step": 1906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4374568462371826,
+      "learning_rate": 8.784365008352075e-06,
+      "loss": 0.8278,
+      "step": 1907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.437732458114624,
+      "learning_rate": 8.782977540366846e-06,
+      "loss": 0.8863,
+      "step": 1908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3391906023025513,
+      "learning_rate": 8.781589390745344e-06,
+      "loss": 0.856,
+      "step": 1909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3066058158874512,
+      "learning_rate": 8.780200559737692e-06,
+      "loss": 0.8305,
+      "step": 1910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2867823839187622,
+      "learning_rate": 8.77881104759414e-06,
+      "loss": 0.847,
+      "step": 1911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3275237083435059,
+      "learning_rate": 8.777420854565054e-06,
+      "loss": 0.8271,
+      "step": 1912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3925204277038574,
+      "learning_rate": 8.776029980900924e-06,
+      "loss": 0.8387,
+      "step": 1913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.359645128250122,
+      "learning_rate": 8.77463842685237e-06,
+      "loss": 0.8567,
+      "step": 1914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3467267751693726,
+      "learning_rate": 8.773246192670128e-06,
+      "loss": 0.8938,
+      "step": 1915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3122711181640625,
+      "learning_rate": 8.771853278605053e-06,
+      "loss": 0.8056,
+      "step": 1916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.334834337234497,
+      "learning_rate": 8.770459684908135e-06,
+      "loss": 0.7712,
+      "step": 1917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.483920097351074,
+      "learning_rate": 8.769065411830473e-06,
+      "loss": 1.1135,
+      "step": 1918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.369441270828247,
+      "learning_rate": 8.767670459623298e-06,
+      "loss": 0.8324,
+      "step": 1919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.305923342704773,
+      "learning_rate": 8.766274828537957e-06,
+      "loss": 0.8272,
+      "step": 1920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1235196590423584,
+      "learning_rate": 8.764878518825923e-06,
+      "loss": 1.0911,
+      "step": 1921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2653162479400635,
+      "learning_rate": 8.76348153073879e-06,
+      "loss": 0.8883,
+      "step": 1922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3397270441055298,
+      "learning_rate": 8.762083864528274e-06,
+      "loss": 0.8313,
+      "step": 1923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3801798820495605,
+      "learning_rate": 8.760685520446216e-06,
+      "loss": 0.95,
+      "step": 1924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.294053077697754,
+      "learning_rate": 8.759286498744574e-06,
+      "loss": 0.853,
+      "step": 1925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2820075750350952,
+      "learning_rate": 8.757886799675433e-06,
+      "loss": 1.0787,
+      "step": 1926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2635024785995483,
+      "learning_rate": 8.756486423490995e-06,
+      "loss": 0.8525,
+      "step": 1927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9956100583076477,
+      "learning_rate": 8.75508537044359e-06,
+      "loss": 1.0824,
+      "step": 1928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.325456976890564,
+      "learning_rate": 8.753683640785667e-06,
+      "loss": 0.8756,
+      "step": 1929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.323514461517334,
+      "learning_rate": 8.752281234769795e-06,
+      "loss": 0.8778,
+      "step": 1930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.444161057472229,
+      "learning_rate": 8.750878152648667e-06,
+      "loss": 0.8403,
+      "step": 1931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.427542805671692,
+      "learning_rate": 8.749474394675097e-06,
+      "loss": 0.8271,
+      "step": 1932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4247180223464966,
+      "learning_rate": 8.748069961102024e-06,
+      "loss": 0.9206,
+      "step": 1933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6704890727996826,
+      "learning_rate": 8.746664852182503e-06,
+      "loss": 1.0802,
+      "step": 1934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1519042253494263,
+      "learning_rate": 8.745259068169715e-06,
+      "loss": 1.0694,
+      "step": 1935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7072116136550903,
+      "learning_rate": 8.743852609316962e-06,
+      "loss": 1.0783,
+      "step": 1936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7863869667053223,
+      "learning_rate": 8.742445475877666e-06,
+      "loss": 0.8719,
+      "step": 1937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8336175680160522,
+      "learning_rate": 8.741037668105371e-06,
+      "loss": 0.9107,
+      "step": 1938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6812655925750732,
+      "learning_rate": 8.739629186253744e-06,
+      "loss": 0.8581,
+      "step": 1939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5878610610961914,
+      "learning_rate": 8.738220030576573e-06,
+      "loss": 0.9182,
+      "step": 1940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3409312963485718,
+      "learning_rate": 8.736810201327766e-06,
+      "loss": 0.8023,
+      "step": 1941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2863434553146362,
+      "learning_rate": 8.735399698761353e-06,
+      "loss": 0.8478,
+      "step": 1942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4747267961502075,
+      "learning_rate": 8.733988523131487e-06,
+      "loss": 0.8583,
+      "step": 1943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4846693277359009,
+      "learning_rate": 8.732576674692438e-06,
+      "loss": 0.8464,
+      "step": 1944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.580715298652649,
+      "learning_rate": 8.731164153698604e-06,
+      "loss": 0.9065,
+      "step": 1945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5150569677352905,
+      "learning_rate": 8.729750960404498e-06,
+      "loss": 0.8703,
+      "step": 1946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3399661779403687,
+      "learning_rate": 8.728337095064756e-06,
+      "loss": 0.8746,
+      "step": 1947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3303695917129517,
+      "learning_rate": 8.726922557934136e-06,
+      "loss": 0.8673,
+      "step": 1948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.370926022529602,
+      "learning_rate": 8.725507349267518e-06,
+      "loss": 0.8411,
+      "step": 1949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3513365983963013,
+      "learning_rate": 8.724091469319898e-06,
+      "loss": 0.8651,
+      "step": 1950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0079431533813477,
+      "learning_rate": 8.722674918346399e-06,
+      "loss": 1.0911,
+      "step": 1951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3024940490722656,
+      "learning_rate": 8.721257696602264e-06,
+      "loss": 1.0952,
+      "step": 1952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5159296989440918,
+      "learning_rate": 8.719839804342852e-06,
+      "loss": 0.8502,
+      "step": 1953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.394816517829895,
+      "learning_rate": 8.718421241823648e-06,
+      "loss": 0.8386,
+      "step": 1954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4089313745498657,
+      "learning_rate": 8.717002009300253e-06,
+      "loss": 0.8343,
+      "step": 1955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3045310974121094,
+      "learning_rate": 8.715582107028396e-06,
+      "loss": 0.863,
+      "step": 1956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.340588092803955,
+      "learning_rate": 8.714161535263921e-06,
+      "loss": 0.8769,
+      "step": 1957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.275490403175354,
+      "learning_rate": 8.712740294262791e-06,
+      "loss": 0.8145,
+      "step": 1958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3657301664352417,
+      "learning_rate": 8.711318384281096e-06,
+      "loss": 0.8635,
+      "step": 1959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9504239559173584,
+      "learning_rate": 8.709895805575042e-06,
+      "loss": 1.0951,
+      "step": 1960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.27976393699646,
+      "learning_rate": 8.708472558400956e-06,
+      "loss": 0.8469,
+      "step": 1961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2499003410339355,
+      "learning_rate": 8.707048643015285e-06,
+      "loss": 0.7946,
+      "step": 1962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3144197463989258,
+      "learning_rate": 8.705624059674598e-06,
+      "loss": 0.8854,
+      "step": 1963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3021539449691772,
+      "learning_rate": 8.704198808635587e-06,
+      "loss": 0.8388,
+      "step": 1964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0866750478744507,
+      "learning_rate": 8.702772890155055e-06,
+      "loss": 1.0726,
+      "step": 1965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7604942321777344,
+      "learning_rate": 8.701346304489937e-06,
+      "loss": 1.0666,
+      "step": 1966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7544482946395874,
+      "learning_rate": 8.699919051897278e-06,
+      "loss": 1.0896,
+      "step": 1967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3449636697769165,
+      "learning_rate": 8.698491132634252e-06,
+      "loss": 0.8708,
+      "step": 1968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3008641004562378,
+      "learning_rate": 8.697062546958145e-06,
+      "loss": 0.8612,
+      "step": 1969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.368179440498352,
+      "learning_rate": 8.69563329512637e-06,
+      "loss": 0.8548,
+      "step": 1970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4446871280670166,
+      "learning_rate": 8.694203377396453e-06,
+      "loss": 1.08,
+      "step": 1971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.410772442817688,
+      "learning_rate": 8.692772794026048e-06,
+      "loss": 0.8378,
+      "step": 1972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3266545534133911,
+      "learning_rate": 8.691341545272923e-06,
+      "loss": 0.822,
+      "step": 1973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3193920850753784,
+      "learning_rate": 8.689909631394967e-06,
+      "loss": 0.8431,
+      "step": 1974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3573565483093262,
+      "learning_rate": 8.68847705265019e-06,
+      "loss": 0.8295,
+      "step": 1975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3348208665847778,
+      "learning_rate": 8.687043809296722e-06,
+      "loss": 0.8958,
+      "step": 1976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2751635313034058,
+      "learning_rate": 8.685609901592813e-06,
+      "loss": 0.8517,
+      "step": 1977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2541040182113647,
+      "learning_rate": 8.684175329796829e-06,
+      "loss": 0.9014,
+      "step": 1978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9109680652618408,
+      "learning_rate": 8.682740094167262e-06,
+      "loss": 1.0569,
+      "step": 1979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3059806823730469,
+      "learning_rate": 8.681304194962716e-06,
+      "loss": 0.8413,
+      "step": 1980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6964817047119141,
+      "learning_rate": 8.679867632441923e-06,
+      "loss": 1.0662,
+      "step": 1981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.676534116268158,
+      "learning_rate": 8.678430406863728e-06,
+      "loss": 1.0645,
+      "step": 1982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4674651622772217,
+      "learning_rate": 8.676992518487097e-06,
+      "loss": 0.8835,
+      "step": 1983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3493527173995972,
+      "learning_rate": 8.675553967571118e-06,
+      "loss": 0.8938,
+      "step": 1984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2724926471710205,
+      "learning_rate": 8.674114754374996e-06,
+      "loss": 0.9002,
+      "step": 1985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2761167287826538,
+      "learning_rate": 8.672674879158054e-06,
+      "loss": 0.9305,
+      "step": 1986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3087843656539917,
+      "learning_rate": 8.671234342179737e-06,
+      "loss": 0.8347,
+      "step": 1987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.362865924835205,
+      "learning_rate": 8.66979314369961e-06,
+      "loss": 0.8741,
+      "step": 1988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3547240495681763,
+      "learning_rate": 8.668351283977353e-06,
+      "loss": 0.8483,
+      "step": 1989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3044947385787964,
+      "learning_rate": 8.66690876327277e-06,
+      "loss": 0.8736,
+      "step": 1990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.245009183883667,
+      "learning_rate": 8.665465581845777e-06,
+      "loss": 0.878,
+      "step": 1991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2732857465744019,
+      "learning_rate": 8.664021739956422e-06,
+      "loss": 1.0487,
+      "step": 1992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0328216552734375,
+      "learning_rate": 8.662577237864857e-06,
+      "loss": 1.088,
+      "step": 1993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3441593647003174,
+      "learning_rate": 8.661132075831363e-06,
+      "loss": 0.7984,
+      "step": 1994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3433098793029785,
+      "learning_rate": 8.659686254116335e-06,
+      "loss": 0.8481,
+      "step": 1995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.347385048866272,
+      "learning_rate": 8.65823977298029e-06,
+      "loss": 0.8511,
+      "step": 1996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4919662475585938,
+      "learning_rate": 8.656792632683862e-06,
+      "loss": 1.0828,
+      "step": 1997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2986359596252441,
+      "learning_rate": 8.655344833487803e-06,
+      "loss": 0.8248,
+      "step": 1998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2179113626480103,
+      "learning_rate": 8.653896375652986e-06,
+      "loss": 1.0845,
+      "step": 1999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3354798555374146,
+      "learning_rate": 8.652447259440402e-06,
+      "loss": 0.8091,
+      "step": 2000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3192139863967896,
+      "learning_rate": 8.650997485111159e-06,
+      "loss": 0.8523,
+      "step": 2001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9370669722557068,
+      "learning_rate": 8.649547052926485e-06,
+      "loss": 1.0651,
+      "step": 2002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.260258436203003,
+      "learning_rate": 8.648095963147726e-06,
+      "loss": 0.8438,
+      "step": 2003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4653054475784302,
+      "learning_rate": 8.646644216036348e-06,
+      "loss": 0.8781,
+      "step": 2004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9015095829963684,
+      "learning_rate": 8.645191811853934e-06,
+      "loss": 1.0681,
+      "step": 2005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7417618036270142,
+      "learning_rate": 8.643738750862186e-06,
+      "loss": 1.0676,
+      "step": 2006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2881509065628052,
+      "learning_rate": 8.64228503332292e-06,
+      "loss": 0.8646,
+      "step": 2007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8354318737983704,
+      "learning_rate": 8.64083065949808e-06,
+      "loss": 1.0373,
+      "step": 2008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.389532446861267,
+      "learning_rate": 8.63937562964972e-06,
+      "loss": 0.8297,
+      "step": 2009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3537027835845947,
+      "learning_rate": 8.637919944040015e-06,
+      "loss": 0.8246,
+      "step": 2010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2696675062179565,
+      "learning_rate": 8.636463602931255e-06,
+      "loss": 0.7947,
+      "step": 2011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.307424545288086,
+      "learning_rate": 8.635006606585856e-06,
+      "loss": 0.9145,
+      "step": 2012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.258249044418335,
+      "learning_rate": 8.633548955266344e-06,
+      "loss": 0.8675,
+      "step": 2013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1610230207443237,
+      "learning_rate": 8.632090649235368e-06,
+      "loss": 1.0406,
+      "step": 2014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2825210094451904,
+      "learning_rate": 8.63063168875569e-06,
+      "loss": 0.8938,
+      "step": 2015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.363271951675415,
+      "learning_rate": 8.629172074090196e-06,
+      "loss": 0.8101,
+      "step": 2016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3558621406555176,
+      "learning_rate": 8.627711805501884e-06,
+      "loss": 0.8401,
+      "step": 2017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6679145097732544,
+      "learning_rate": 8.626250883253874e-06,
+      "loss": 1.0564,
+      "step": 2018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.278344750404358,
+      "learning_rate": 8.624789307609404e-06,
+      "loss": 0.8244,
+      "step": 2019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7056128978729248,
+      "learning_rate": 8.623327078831826e-06,
+      "loss": 1.0609,
+      "step": 2020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.292506217956543,
+      "learning_rate": 8.621864197184616e-06,
+      "loss": 0.836,
+      "step": 2021
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2514880895614624,
+      "learning_rate": 8.620400662931356e-06,
+      "loss": 0.8502,
+      "step": 2022
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.292537808418274,
+      "learning_rate": 8.618936476335759e-06,
+      "loss": 0.8276,
+      "step": 2023
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2109473943710327,
+      "learning_rate": 8.61747163766165e-06,
+      "loss": 0.8564,
+      "step": 2024
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2065536975860596,
+      "learning_rate": 8.616006147172967e-06,
+      "loss": 0.8082,
+      "step": 2025
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2800276279449463,
+      "learning_rate": 8.614540005133773e-06,
+      "loss": 0.8817,
+      "step": 2026
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2281490564346313,
+      "learning_rate": 8.613073211808242e-06,
+      "loss": 0.8433,
+      "step": 2027
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2509658336639404,
+      "learning_rate": 8.611605767460671e-06,
+      "loss": 0.8371,
+      "step": 2028
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2478877305984497,
+      "learning_rate": 8.610137672355472e-06,
+      "loss": 0.8461,
+      "step": 2029
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.316353678703308,
+      "learning_rate": 8.60866892675717e-06,
+      "loss": 0.8785,
+      "step": 2030
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.281317114830017,
+      "learning_rate": 8.607199530930418e-06,
+      "loss": 0.8435,
+      "step": 2031
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3595694303512573,
+      "learning_rate": 8.605729485139971e-06,
+      "loss": 0.8063,
+      "step": 2032
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2809197902679443,
+      "learning_rate": 8.604258789650714e-06,
+      "loss": 0.8558,
+      "step": 2033
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3660931587219238,
+      "learning_rate": 8.602787444727645e-06,
+      "loss": 0.8218,
+      "step": 2034
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3095648288726807,
+      "learning_rate": 8.601315450635876e-06,
+      "loss": 0.8638,
+      "step": 2035
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3191345930099487,
+      "learning_rate": 8.599842807640642e-06,
+      "loss": 0.8573,
+      "step": 2036
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.315412163734436,
+      "learning_rate": 8.598369516007288e-06,
+      "loss": 0.8367,
+      "step": 2037
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8077951073646545,
+      "learning_rate": 8.596895576001278e-06,
+      "loss": 1.063,
+      "step": 2038
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3003404140472412,
+      "learning_rate": 8.5954209878882e-06,
+      "loss": 0.9184,
+      "step": 2039
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.261550784111023,
+      "learning_rate": 8.593945751933747e-06,
+      "loss": 0.8626,
+      "step": 2040
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7069729566574097,
+      "learning_rate": 8.592469868403737e-06,
+      "loss": 1.0671,
+      "step": 2041
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.241598129272461,
+      "learning_rate": 8.590993337564102e-06,
+      "loss": 0.8718,
+      "step": 2042
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6864539384841919,
+      "learning_rate": 8.589516159680893e-06,
+      "loss": 1.0653,
+      "step": 2043
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3502180576324463,
+      "learning_rate": 8.58803833502027e-06,
+      "loss": 0.7965,
+      "step": 2044
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.355806589126587,
+      "learning_rate": 8.586559863848521e-06,
+      "loss": 0.7983,
+      "step": 2045
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2749450206756592,
+      "learning_rate": 8.585080746432042e-06,
+      "loss": 0.8788,
+      "step": 2046
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2048343420028687,
+      "learning_rate": 8.583600983037347e-06,
+      "loss": 0.8491,
+      "step": 2047
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2967233657836914,
+      "learning_rate": 8.582120573931068e-06,
+      "loss": 0.8595,
+      "step": 2048
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.263377070426941,
+      "learning_rate": 8.580639519379955e-06,
+      "loss": 0.861,
+      "step": 2049
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2960383892059326,
+      "learning_rate": 8.579157819650867e-06,
+      "loss": 0.84,
+      "step": 2050
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2637882232666016,
+      "learning_rate": 8.577675475010789e-06,
+      "loss": 0.9008,
+      "step": 2051
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1961098909378052,
+      "learning_rate": 8.576192485726816e-06,
+      "loss": 0.7678,
+      "step": 2052
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3340755701065063,
+      "learning_rate": 8.574708852066159e-06,
+      "loss": 0.8091,
+      "step": 2053
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.323927640914917,
+      "learning_rate": 8.573224574296148e-06,
+      "loss": 0.8534,
+      "step": 2054
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3466746807098389,
+      "learning_rate": 8.571739652684227e-06,
+      "loss": 0.8545,
+      "step": 2055
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9067081809043884,
+      "learning_rate": 8.570254087497959e-06,
+      "loss": 1.0564,
+      "step": 2056
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2833993434906006,
+      "learning_rate": 8.56876787900502e-06,
+      "loss": 0.8545,
+      "step": 2057
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2997359037399292,
+      "learning_rate": 8.5672810274732e-06,
+      "loss": 0.9147,
+      "step": 2058
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3560856580734253,
+      "learning_rate": 8.56579353317041e-06,
+      "loss": 0.8131,
+      "step": 2059
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3102887868881226,
+      "learning_rate": 8.564305396364674e-06,
+      "loss": 0.8446,
+      "step": 2060
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.352183222770691,
+      "learning_rate": 8.562816617324132e-06,
+      "loss": 0.8226,
+      "step": 2061
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2321275472640991,
+      "learning_rate": 8.561327196317039e-06,
+      "loss": 0.8737,
+      "step": 2062
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.281092882156372,
+      "learning_rate": 8.55983713361177e-06,
+      "loss": 0.8519,
+      "step": 2063
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2192951440811157,
+      "learning_rate": 8.558346429476807e-06,
+      "loss": 0.8002,
+      "step": 2064
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8812317848205566,
+      "learning_rate": 8.556855084180755e-06,
+      "loss": 1.0598,
+      "step": 2065
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2907503843307495,
+      "learning_rate": 8.555363097992334e-06,
+      "loss": 0.8453,
+      "step": 2066
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6925197839736938,
+      "learning_rate": 8.553870471180374e-06,
+      "loss": 1.0592,
+      "step": 2067
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6498041749000549,
+      "learning_rate": 8.552377204013828e-06,
+      "loss": 1.0756,
+      "step": 2068
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3045166730880737,
+      "learning_rate": 8.550883296761757e-06,
+      "loss": 0.7676,
+      "step": 2069
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3451416492462158,
+      "learning_rate": 8.549388749693344e-06,
+      "loss": 0.8562,
+      "step": 2070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3188577890396118,
+      "learning_rate": 8.547893563077882e-06,
+      "loss": 0.839,
+      "step": 2071
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3451582193374634,
+      "learning_rate": 8.546397737184782e-06,
+      "loss": 0.7962,
+      "step": 2072
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.292868971824646,
+      "learning_rate": 8.54490127228357e-06,
+      "loss": 0.8254,
+      "step": 2073
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3429176807403564,
+      "learning_rate": 8.543404168643887e-06,
+      "loss": 0.8187,
+      "step": 2074
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2207163572311401,
+      "learning_rate": 8.541906426535487e-06,
+      "loss": 1.0813,
+      "step": 2075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3381308317184448,
+      "learning_rate": 8.540408046228242e-06,
+      "loss": 0.8126,
+      "step": 2076
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3356101512908936,
+      "learning_rate": 8.53890902799214e-06,
+      "loss": 0.8804,
+      "step": 2077
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2490941286087036,
+      "learning_rate": 8.537409372097278e-06,
+      "loss": 0.8106,
+      "step": 2078
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7178091406822205,
+      "learning_rate": 8.535909078813872e-06,
+      "loss": 1.063,
+      "step": 2079
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7316675186157227,
+      "learning_rate": 8.534408148412257e-06,
+      "loss": 1.0711,
+      "step": 2080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6954258680343628,
+      "learning_rate": 8.532906581162872e-06,
+      "loss": 1.0609,
+      "step": 2081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3789459466934204,
+      "learning_rate": 8.531404377336283e-06,
+      "loss": 0.8598,
+      "step": 2082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.292380690574646,
+      "learning_rate": 8.52990153720316e-06,
+      "loss": 0.8886,
+      "step": 2083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.365106225013733,
+      "learning_rate": 8.528398061034295e-06,
+      "loss": 0.8464,
+      "step": 2084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2891864776611328,
+      "learning_rate": 8.526893949100592e-06,
+      "loss": 0.8058,
+      "step": 2085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2300176620483398,
+      "learning_rate": 8.525389201673069e-06,
+      "loss": 0.832,
+      "step": 2086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2913756370544434,
+      "learning_rate": 8.52388381902286e-06,
+      "loss": 0.8299,
+      "step": 2087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2732232809066772,
+      "learning_rate": 8.522377801421211e-06,
+      "loss": 0.8271,
+      "step": 2088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1113044023513794,
+      "learning_rate": 8.520871149139485e-06,
+      "loss": 1.0576,
+      "step": 2089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.369467854499817,
+      "learning_rate": 8.51936386244916e-06,
+      "loss": 0.8226,
+      "step": 2090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3965648412704468,
+      "learning_rate": 8.517855941621822e-06,
+      "loss": 0.8503,
+      "step": 2091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4219321012496948,
+      "learning_rate": 8.516347386929179e-06,
+      "loss": 0.8153,
+      "step": 2092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7998663783073425,
+      "learning_rate": 8.514838198643049e-06,
+      "loss": 1.0559,
+      "step": 2093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2231522798538208,
+      "learning_rate": 8.513328377035368e-06,
+      "loss": 0.8134,
+      "step": 2094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.301601529121399,
+      "learning_rate": 8.51181792237818e-06,
+      "loss": 0.8434,
+      "step": 2095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7171523571014404,
+      "learning_rate": 8.51030683494365e-06,
+      "loss": 1.0535,
+      "step": 2096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2459185123443604,
+      "learning_rate": 8.508795115004049e-06,
+      "loss": 0.838,
+      "step": 2097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7157946825027466,
+      "learning_rate": 8.50728276283177e-06,
+      "loss": 1.0678,
+      "step": 2098
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2852526903152466,
+      "learning_rate": 8.505769778699314e-06,
+      "loss": 0.8967,
+      "step": 2099
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3011096715927124,
+      "learning_rate": 8.504256162879302e-06,
+      "loss": 0.8509,
+      "step": 2100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7333881855010986,
+      "learning_rate": 8.502741915644459e-06,
+      "loss": 1.073,
+      "step": 2101
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2817645072937012,
+      "learning_rate": 8.501227037267635e-06,
+      "loss": 0.8447,
+      "step": 2102
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.309291958808899,
+      "learning_rate": 8.499711528021785e-06,
+      "loss": 0.883,
+      "step": 2103
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3102920055389404,
+      "learning_rate": 8.498195388179986e-06,
+      "loss": 0.856,
+      "step": 2104
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3376411199569702,
+      "learning_rate": 8.496678618015419e-06,
+      "loss": 0.9061,
+      "step": 2105
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4023765325546265,
+      "learning_rate": 8.495161217801386e-06,
+      "loss": 0.8863,
+      "step": 2106
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1947153806686401,
+      "learning_rate": 8.4936431878113e-06,
+      "loss": 0.8132,
+      "step": 2107
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2363944053649902,
+      "learning_rate": 8.492124528318686e-06,
+      "loss": 0.8426,
+      "step": 2108
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2886425256729126,
+      "learning_rate": 8.490605239597182e-06,
+      "loss": 0.842,
+      "step": 2109
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3204377889633179,
+      "learning_rate": 8.489085321920547e-06,
+      "loss": 0.8303,
+      "step": 2110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.224958062171936,
+      "learning_rate": 8.487564775562643e-06,
+      "loss": 0.8027,
+      "step": 2111
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2945879697799683,
+      "learning_rate": 8.486043600797453e-06,
+      "loss": 0.8253,
+      "step": 2112
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.213448405265808,
+      "learning_rate": 8.484521797899065e-06,
+      "loss": 0.8026,
+      "step": 2113
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4012451171875,
+      "learning_rate": 8.482999367141692e-06,
+      "loss": 0.9002,
+      "step": 2114
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8171733021736145,
+      "learning_rate": 8.481476308799645e-06,
+      "loss": 1.0646,
+      "step": 2115
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2786657810211182,
+      "learning_rate": 8.479952623147366e-06,
+      "loss": 0.8674,
+      "step": 2116
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7058357000350952,
+      "learning_rate": 8.478428310459393e-06,
+      "loss": 1.0403,
+      "step": 2117
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.668547511100769,
+      "learning_rate": 8.476903371010387e-06,
+      "loss": 1.0558,
+      "step": 2118
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3617972135543823,
+      "learning_rate": 8.47537780507512e-06,
+      "loss": 0.8848,
+      "step": 2119
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.282021164894104,
+      "learning_rate": 8.473851612928477e-06,
+      "loss": 0.8844,
+      "step": 2120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.232177495956421,
+      "learning_rate": 8.472324794845453e-06,
+      "loss": 0.7971,
+      "step": 2121
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8160005211830139,
+      "learning_rate": 8.470797351101159e-06,
+      "loss": 1.0555,
+      "step": 2122
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3841344118118286,
+      "learning_rate": 8.469269281970818e-06,
+      "loss": 0.848,
+      "step": 2123
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.231188416481018,
+      "learning_rate": 8.467740587729765e-06,
+      "loss": 0.8155,
+      "step": 2124
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.269723653793335,
+      "learning_rate": 8.466211268653447e-06,
+      "loss": 0.85,
+      "step": 2125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2742301225662231,
+      "learning_rate": 8.464681325017425e-06,
+      "loss": 0.8987,
+      "step": 2126
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2293083667755127,
+      "learning_rate": 8.463150757097373e-06,
+      "loss": 0.8457,
+      "step": 2127
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7923334836959839,
+      "learning_rate": 8.461619565169077e-06,
+      "loss": 1.0535,
+      "step": 2128
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7963456511497498,
+      "learning_rate": 8.460087749508435e-06,
+      "loss": 1.0575,
+      "step": 2129
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2880580425262451,
+      "learning_rate": 8.458555310391453e-06,
+      "loss": 0.8463,
+      "step": 2130
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3567508459091187,
+      "learning_rate": 8.45702224809426e-06,
+      "loss": 0.8582,
+      "step": 2131
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3330193758010864,
+      "learning_rate": 8.455488562893087e-06,
+      "loss": 0.8307,
+      "step": 2132
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.290679693222046,
+      "learning_rate": 8.453954255064283e-06,
+      "loss": 0.8264,
+      "step": 2133
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1895647048950195,
+      "learning_rate": 8.452419324884307e-06,
+      "loss": 0.8654,
+      "step": 2134
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2426303625106812,
+      "learning_rate": 8.450883772629728e-06,
+      "loss": 0.875,
+      "step": 2135
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2320258617401123,
+      "learning_rate": 8.449347598577236e-06,
+      "loss": 0.7865,
+      "step": 2136
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1199358701705933,
+      "learning_rate": 8.44781080300362e-06,
+      "loss": 1.0428,
+      "step": 2137
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.306275725364685,
+      "learning_rate": 8.446273386185792e-06,
+      "loss": 0.8404,
+      "step": 2138
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3679805994033813,
+      "learning_rate": 8.44473534840077e-06,
+      "loss": 0.8481,
+      "step": 2139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3810871839523315,
+      "learning_rate": 8.443196689925684e-06,
+      "loss": 0.8902,
+      "step": 2140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3239946365356445,
+      "learning_rate": 8.44165741103778e-06,
+      "loss": 0.91,
+      "step": 2141
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3441720008850098,
+      "learning_rate": 8.440117512014414e-06,
+      "loss": 0.8835,
+      "step": 2142
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8158602714538574,
+      "learning_rate": 8.43857699313305e-06,
+      "loss": 1.0444,
+      "step": 2143
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.328433632850647,
+      "learning_rate": 8.437035854671267e-06,
+      "loss": 0.8697,
+      "step": 2144
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2639670372009277,
+      "learning_rate": 8.435494096906755e-06,
+      "loss": 0.8596,
+      "step": 2145
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7439870834350586,
+      "learning_rate": 8.433951720117319e-06,
+      "loss": 1.0679,
+      "step": 2146
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6821069121360779,
+      "learning_rate": 8.43240872458087e-06,
+      "loss": 1.0363,
+      "step": 2147
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6345691680908203,
+      "learning_rate": 8.430865110575432e-06,
+      "loss": 1.0393,
+      "step": 2148
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6294697523117065,
+      "learning_rate": 8.429320878379142e-06,
+      "loss": 1.0518,
+      "step": 2149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4293125867843628,
+      "learning_rate": 8.427776028270248e-06,
+      "loss": 0.9045,
+      "step": 2150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7274340391159058,
+      "learning_rate": 8.42623056052711e-06,
+      "loss": 1.0749,
+      "step": 2151
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3431564569473267,
+      "learning_rate": 8.424684475428196e-06,
+      "loss": 0.8138,
+      "step": 2152
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3383474349975586,
+      "learning_rate": 8.423137773252088e-06,
+      "loss": 0.8178,
+      "step": 2153
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.307834506034851,
+      "learning_rate": 8.421590454277482e-06,
+      "loss": 0.8326,
+      "step": 2154
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.220627784729004,
+      "learning_rate": 8.420042518783178e-06,
+      "loss": 0.8212,
+      "step": 2155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2885347604751587,
+      "learning_rate": 8.418493967048092e-06,
+      "loss": 0.8266,
+      "step": 2156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2887239456176758,
+      "learning_rate": 8.41694479935125e-06,
+      "loss": 0.8616,
+      "step": 2157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.801588773727417,
+      "learning_rate": 8.415395015971791e-06,
+      "loss": 1.0622,
+      "step": 2158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.24701988697052,
+      "learning_rate": 8.413844617188957e-06,
+      "loss": 0.8401,
+      "step": 2159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4075746536254883,
+      "learning_rate": 8.412293603282115e-06,
+      "loss": 0.8307,
+      "step": 2160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3132729530334473,
+      "learning_rate": 8.41074197453073e-06,
+      "loss": 0.8064,
+      "step": 2161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6990694999694824,
+      "learning_rate": 8.40918973121438e-06,
+      "loss": 1.0683,
+      "step": 2162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3122400045394897,
+      "learning_rate": 8.40763687361276e-06,
+      "loss": 0.8234,
+      "step": 2163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3203343152999878,
+      "learning_rate": 8.406083402005673e-06,
+      "loss": 0.8147,
+      "step": 2164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7041956782341003,
+      "learning_rate": 8.404529316673028e-06,
+      "loss": 1.0552,
+      "step": 2165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2229079008102417,
+      "learning_rate": 8.402974617894853e-06,
+      "loss": 0.7962,
+      "step": 2166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3331260681152344,
+      "learning_rate": 8.401419305951273e-06,
+      "loss": 0.8155,
+      "step": 2167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2503989934921265,
+      "learning_rate": 8.39986338112254e-06,
+      "loss": 0.8511,
+      "step": 2168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6891682744026184,
+      "learning_rate": 8.398306843689005e-06,
+      "loss": 1.0721,
+      "step": 2169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3050427436828613,
+      "learning_rate": 8.396749693931133e-06,
+      "loss": 0.8801,
+      "step": 2170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.238756537437439,
+      "learning_rate": 8.3951919321295e-06,
+      "loss": 0.8299,
+      "step": 2171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6410030722618103,
+      "learning_rate": 8.393633558564792e-06,
+      "loss": 1.0471,
+      "step": 2172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6486477255821228,
+      "learning_rate": 8.392074573517804e-06,
+      "loss": 1.0397,
+      "step": 2173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.275058627128601,
+      "learning_rate": 8.39051497726944e-06,
+      "loss": 0.8498,
+      "step": 2174
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2463288307189941,
+      "learning_rate": 8.38895477010072e-06,
+      "loss": 0.8181,
+      "step": 2175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6398757100105286,
+      "learning_rate": 8.387393952292767e-06,
+      "loss": 1.0598,
+      "step": 2176
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2357101440429688,
+      "learning_rate": 8.38583252412682e-06,
+      "loss": 0.7948,
+      "step": 2177
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3581864833831787,
+      "learning_rate": 8.384270485884221e-06,
+      "loss": 0.8705,
+      "step": 2178
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2885698080062866,
+      "learning_rate": 8.382707837846429e-06,
+      "loss": 0.8561,
+      "step": 2179
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.223678708076477,
+      "learning_rate": 8.38114458029501e-06,
+      "loss": 0.7991,
+      "step": 2180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2290151119232178,
+      "learning_rate": 8.37958071351164e-06,
+      "loss": 0.8557,
+      "step": 2181
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7594748139381409,
+      "learning_rate": 8.378016237778102e-06,
+      "loss": 1.0449,
+      "step": 2182
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2904711961746216,
+      "learning_rate": 8.376451153376292e-06,
+      "loss": 0.8554,
+      "step": 2183
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2991963624954224,
+      "learning_rate": 8.374885460588218e-06,
+      "loss": 0.8253,
+      "step": 2184
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6935288906097412,
+      "learning_rate": 8.373319159695989e-06,
+      "loss": 1.0512,
+      "step": 2185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6824633479118347,
+      "learning_rate": 8.371752250981836e-06,
+      "loss": 1.0664,
+      "step": 2186
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.651799201965332,
+      "learning_rate": 8.370184734728089e-06,
+      "loss": 1.0615,
+      "step": 2187
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3565657138824463,
+      "learning_rate": 8.368616611217188e-06,
+      "loss": 0.832,
+      "step": 2188
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.239324688911438,
+      "learning_rate": 8.36704788073169e-06,
+      "loss": 0.8432,
+      "step": 2189
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2958683967590332,
+      "learning_rate": 8.365478543554259e-06,
+      "loss": 0.868,
+      "step": 2190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2655532360076904,
+      "learning_rate": 8.36390859996766e-06,
+      "loss": 0.8251,
+      "step": 2191
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2392398118972778,
+      "learning_rate": 8.362338050254777e-06,
+      "loss": 0.8466,
+      "step": 2192
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3292738199234009,
+      "learning_rate": 8.360766894698598e-06,
+      "loss": 0.8061,
+      "step": 2193
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3308783769607544,
+      "learning_rate": 8.359195133582223e-06,
+      "loss": 0.8834,
+      "step": 2194
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.529285192489624,
+      "learning_rate": 8.35762276718886e-06,
+      "loss": 0.8686,
+      "step": 2195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2944306135177612,
+      "learning_rate": 8.356049795801826e-06,
+      "loss": 0.8558,
+      "step": 2196
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9774466156959534,
+      "learning_rate": 8.354476219704546e-06,
+      "loss": 1.0652,
+      "step": 2197
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3497653007507324,
+      "learning_rate": 8.352902039180559e-06,
+      "loss": 0.8357,
+      "step": 2198
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.284282922744751,
+      "learning_rate": 8.351327254513504e-06,
+      "loss": 0.8199,
+      "step": 2199
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2400554418563843,
+      "learning_rate": 8.349751865987134e-06,
+      "loss": 0.8325,
+      "step": 2200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3974705934524536,
+      "learning_rate": 8.348175873885314e-06,
+      "loss": 0.8411,
+      "step": 2201
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.273522973060608,
+      "learning_rate": 8.346599278492012e-06,
+      "loss": 0.8219,
+      "step": 2202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3020727634429932,
+      "learning_rate": 8.345022080091308e-06,
+      "loss": 0.8083,
+      "step": 2203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8737990260124207,
+      "learning_rate": 8.343444278967388e-06,
+      "loss": 1.0674,
+      "step": 2204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3403209447860718,
+      "learning_rate": 8.341865875404551e-06,
+      "loss": 0.8711,
+      "step": 2205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7410245537757874,
+      "learning_rate": 8.3402868696872e-06,
+      "loss": 1.0385,
+      "step": 2206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3131184577941895,
+      "learning_rate": 8.338707262099847e-06,
+      "loss": 0.8177,
+      "step": 2207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2571330070495605,
+      "learning_rate": 8.337127052927117e-06,
+      "loss": 0.8068,
+      "step": 2208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8811030983924866,
+      "learning_rate": 8.335546242453738e-06,
+      "loss": 1.0557,
+      "step": 2209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2309426069259644,
+      "learning_rate": 8.33396483096455e-06,
+      "loss": 0.8024,
+      "step": 2210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2587406635284424,
+      "learning_rate": 8.332382818744499e-06,
+      "loss": 0.8758,
+      "step": 2211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3209506273269653,
+      "learning_rate": 8.33080020607864e-06,
+      "loss": 0.8191,
+      "step": 2212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2637529373168945,
+      "learning_rate": 8.329216993252135e-06,
+      "loss": 0.8252,
+      "step": 2213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7439998388290405,
+      "learning_rate": 8.327633180550258e-06,
+      "loss": 1.0665,
+      "step": 2214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2387902736663818,
+      "learning_rate": 8.326048768258388e-06,
+      "loss": 0.7969,
+      "step": 2215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.315648078918457,
+      "learning_rate": 8.32446375666201e-06,
+      "loss": 0.8267,
+      "step": 2216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2925833463668823,
+      "learning_rate": 8.322878146046723e-06,
+      "loss": 0.8608,
+      "step": 2217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2636542320251465,
+      "learning_rate": 8.321291936698227e-06,
+      "loss": 0.8486,
+      "step": 2218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4524047374725342,
+      "learning_rate": 8.319705128902336e-06,
+      "loss": 0.8604,
+      "step": 2219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3300044536590576,
+      "learning_rate": 8.318117722944967e-06,
+      "loss": 0.8424,
+      "step": 2220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7916128039360046,
+      "learning_rate": 8.31652971911215e-06,
+      "loss": 1.0378,
+      "step": 2221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3194878101348877,
+      "learning_rate": 8.314941117690017e-06,
+      "loss": 0.8761,
+      "step": 2222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.238194465637207,
+      "learning_rate": 8.31335191896481e-06,
+      "loss": 0.8221,
+      "step": 2223
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3378403186798096,
+      "learning_rate": 8.311762123222882e-06,
+      "loss": 0.9219,
+      "step": 2224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.753247082233429,
+      "learning_rate": 8.310171730750687e-06,
+      "loss": 1.0557,
+      "step": 2225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4257842302322388,
+      "learning_rate": 8.308580741834793e-06,
+      "loss": 0.8502,
+      "step": 2226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4900792837142944,
+      "learning_rate": 8.306989156761872e-06,
+      "loss": 0.8556,
+      "step": 2227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.257476568222046,
+      "learning_rate": 8.3053969758187e-06,
+      "loss": 0.8607,
+      "step": 2228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.353279709815979,
+      "learning_rate": 8.30380419929217e-06,
+      "loss": 0.8391,
+      "step": 2229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.367388129234314,
+      "learning_rate": 8.302210827469272e-06,
+      "loss": 0.8285,
+      "step": 2230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7864716649055481,
+      "learning_rate": 8.300616860637111e-06,
+      "loss": 1.0433,
+      "step": 2231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.182981252670288,
+      "learning_rate": 8.299022299082897e-06,
+      "loss": 0.8443,
+      "step": 2232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6303523182868958,
+      "learning_rate": 8.297427143093943e-06,
+      "loss": 1.0568,
+      "step": 2233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6367732286453247,
+      "learning_rate": 8.295831392957674e-06,
+      "loss": 1.0692,
+      "step": 2234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.675452470779419,
+      "learning_rate": 8.294235048961618e-06,
+      "loss": 1.0428,
+      "step": 2235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3675307035446167,
+      "learning_rate": 8.292638111393418e-06,
+      "loss": 0.8744,
+      "step": 2236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.329197883605957,
+      "learning_rate": 8.291040580540814e-06,
+      "loss": 0.8087,
+      "step": 2237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2379142045974731,
+      "learning_rate": 8.289442456691658e-06,
+      "loss": 0.8654,
+      "step": 2238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3058524131774902,
+      "learning_rate": 8.287843740133908e-06,
+      "loss": 0.883,
+      "step": 2239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3559993505477905,
+      "learning_rate": 8.28624443115563e-06,
+      "loss": 0.8756,
+      "step": 2240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2826493978500366,
+      "learning_rate": 8.284644530044996e-06,
+      "loss": 0.8656,
+      "step": 2241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3119560480117798,
+      "learning_rate": 8.283044037090282e-06,
+      "loss": 0.8269,
+      "step": 2242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2626519203186035,
+      "learning_rate": 8.281442952579876e-06,
+      "loss": 0.8284,
+      "step": 2243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9827457070350647,
+      "learning_rate": 8.279841276802267e-06,
+      "loss": 1.0801,
+      "step": 2244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3297693729400635,
+      "learning_rate": 8.278239010046055e-06,
+      "loss": 0.8286,
+      "step": 2245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2829421758651733,
+      "learning_rate": 8.276636152599944e-06,
+      "loss": 0.8834,
+      "step": 2246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3850988149642944,
+      "learning_rate": 8.275032704752744e-06,
+      "loss": 0.8646,
+      "step": 2247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3403866291046143,
+      "learning_rate": 8.273428666793375e-06,
+      "loss": 0.9103,
+      "step": 2248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.79488605260849,
+      "learning_rate": 8.271824039010856e-06,
+      "loss": 1.05,
+      "step": 2249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2922897338867188,
+      "learning_rate": 8.270218821694324e-06,
+      "loss": 0.8709,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.642350971698761,
+      "learning_rate": 8.268613015133009e-06,
+      "loss": 1.0384,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4987260103225708,
+      "learning_rate": 8.267006619616258e-06,
+      "loss": 0.8898,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6983321309089661,
+      "learning_rate": 8.265399635433515e-06,
+      "loss": 1.0838,
+      "step": 2253
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.311261534690857,
+      "learning_rate": 8.26379206287434e-06,
+      "loss": 0.8382,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.383606195449829,
+      "learning_rate": 8.26218390222839e-06,
+      "loss": 0.94,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2880290746688843,
+      "learning_rate": 8.26057515378543e-06,
+      "loss": 0.8532,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2318313121795654,
+      "learning_rate": 8.258965817835338e-06,
+      "loss": 0.7925,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.293599247932434,
+      "learning_rate": 8.257355894668088e-06,
+      "loss": 0.7994,
+      "step": 2258
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2868762016296387,
+      "learning_rate": 8.255745384573766e-06,
+      "loss": 0.8602,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3142626285552979,
+      "learning_rate": 8.254134287842562e-06,
+      "loss": 0.8186,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.306653380393982,
+      "learning_rate": 8.25252260476477e-06,
+      "loss": 0.887,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.223941445350647,
+      "learning_rate": 8.250910335630795e-06,
+      "loss": 0.7836,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8135960698127747,
+      "learning_rate": 8.24929748073114e-06,
+      "loss": 1.0438,
+      "step": 2263
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7143467664718628,
+      "learning_rate": 8.247684040356419e-06,
+      "loss": 1.0604,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6474387645721436,
+      "learning_rate": 8.246070014797349e-06,
+      "loss": 1.0571,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7069191336631775,
+      "learning_rate": 8.244455404344756e-06,
+      "loss": 1.0796,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.421578049659729,
+      "learning_rate": 8.242840209289564e-06,
+      "loss": 0.8832,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7588387131690979,
+      "learning_rate": 8.241224429922813e-06,
+      "loss": 1.0551,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6704986095428467,
+      "learning_rate": 8.239608066535637e-06,
+      "loss": 1.0828,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3382351398468018,
+      "learning_rate": 8.237991119419286e-06,
+      "loss": 0.8868,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7444342374801636,
+      "learning_rate": 8.236373588865107e-06,
+      "loss": 1.0527,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7449299693107605,
+      "learning_rate": 8.234755475164553e-06,
+      "loss": 1.0535,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3799817562103271,
+      "learning_rate": 8.233136778609188e-06,
+      "loss": 0.8604,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2726234197616577,
+      "learning_rate": 8.231517499490673e-06,
+      "loss": 0.8764,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3013056516647339,
+      "learning_rate": 8.22989763810078e-06,
+      "loss": 0.8808,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2826999425888062,
+      "learning_rate": 8.228277194731385e-06,
+      "loss": 0.8839,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7117277383804321,
+      "learning_rate": 8.226656169674467e-06,
+      "loss": 0.8161,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0700881481170654,
+      "learning_rate": 8.22503456322211e-06,
+      "loss": 1.0483,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2542366981506348,
+      "learning_rate": 8.223412375666504e-06,
+      "loss": 0.8473,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2828962802886963,
+      "learning_rate": 8.221789607299943e-06,
+      "loss": 0.8543,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7101644277572632,
+      "learning_rate": 8.220166258414826e-06,
+      "loss": 1.0605,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2218552827835083,
+      "learning_rate": 8.218542329303654e-06,
+      "loss": 0.8175,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3214446306228638,
+      "learning_rate": 8.216917820259038e-06,
+      "loss": 0.8562,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.350572109222412,
+      "learning_rate": 8.215292731573687e-06,
+      "loss": 0.8406,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2979488372802734,
+      "learning_rate": 8.213667063540423e-06,
+      "loss": 0.8968,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.280500054359436,
+      "learning_rate": 8.21204081645216e-06,
+      "loss": 0.8328,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.234437346458435,
+      "learning_rate": 8.21041399060193e-06,
+      "loss": 0.8021,
+      "step": 2287
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3395960330963135,
+      "learning_rate": 8.208786586282862e-06,
+      "loss": 0.8433,
+      "step": 2288
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2233288288116455,
+      "learning_rate": 8.207158603788187e-06,
+      "loss": 0.7966,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1613348722457886,
+      "learning_rate": 8.205530043411243e-06,
+      "loss": 1.0448,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3112993240356445,
+      "learning_rate": 8.203900905445477e-06,
+      "loss": 0.881,
+      "step": 2291
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.355731725692749,
+      "learning_rate": 8.20227119018443e-06,
+      "loss": 0.8323,
+      "step": 2292
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2913293838500977,
+      "learning_rate": 8.200640897921755e-06,
+      "loss": 0.8517,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3255492448806763,
+      "learning_rate": 8.199010028951206e-06,
+      "loss": 0.824,
+      "step": 2294
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2856396436691284,
+      "learning_rate": 8.197378583566644e-06,
+      "loss": 0.7743,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.364687204360962,
+      "learning_rate": 8.195746562062027e-06,
+      "loss": 0.8523,
+      "step": 2296
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.98493492603302,
+      "learning_rate": 8.194113964731423e-06,
+      "loss": 1.0776,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.428605318069458,
+      "learning_rate": 8.192480791869002e-06,
+      "loss": 0.8739,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.725376307964325,
+      "learning_rate": 8.190847043769037e-06,
+      "loss": 1.0622,
+      "step": 2299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.381938099861145,
+      "learning_rate": 8.189212720725904e-06,
+      "loss": 0.815,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4764595031738281,
+      "learning_rate": 8.187577823034086e-06,
+      "loss": 0.7926,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3050110340118408,
+      "learning_rate": 8.185942350988165e-06,
+      "loss": 0.8716,
+      "step": 2302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2068325281143188,
+      "learning_rate": 8.184306304882832e-06,
+      "loss": 0.7979,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3532794713974,
+      "learning_rate": 8.182669685012872e-06,
+      "loss": 0.8432,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2489885091781616,
+      "learning_rate": 8.181032491673186e-06,
+      "loss": 0.7801,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3200814723968506,
+      "learning_rate": 8.179394725158769e-06,
+      "loss": 0.8782,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2832105159759521,
+      "learning_rate": 8.177756385764723e-06,
+      "loss": 0.8191,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2181081771850586,
+      "learning_rate": 8.176117473786252e-06,
+      "loss": 0.77,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2567659616470337,
+      "learning_rate": 8.174477989518663e-06,
+      "loss": 0.8563,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3770227432250977,
+      "learning_rate": 8.172837933257368e-06,
+      "loss": 0.8609,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3625712394714355,
+      "learning_rate": 8.171197305297878e-06,
+      "loss": 1.0615,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3532320261001587,
+      "learning_rate": 8.169556105935812e-06,
+      "loss": 0.8754,
+      "step": 2312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2988698482513428,
+      "learning_rate": 8.16791433546689e-06,
+      "loss": 0.8129,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2933070659637451,
+      "learning_rate": 8.166271994186935e-06,
+      "loss": 0.7913,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6808357238769531,
+      "learning_rate": 8.164629082391871e-06,
+      "loss": 1.0452,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2803293466567993,
+      "learning_rate": 8.162985600377726e-06,
+      "loss": 0.8772,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3249911069869995,
+      "learning_rate": 8.161341548440631e-06,
+      "loss": 0.8914,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8592790961265564,
+      "learning_rate": 8.15969692687682e-06,
+      "loss": 1.0617,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3951386213302612,
+      "learning_rate": 8.158051735982633e-06,
+      "loss": 0.8988,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2435624599456787,
+      "learning_rate": 8.156405976054504e-06,
+      "loss": 0.8381,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.774296224117279,
+      "learning_rate": 8.154759647388974e-06,
+      "loss": 1.0723,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3755284547805786,
+      "learning_rate": 8.153112750282692e-06,
+      "loss": 0.8573,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3396536111831665,
+      "learning_rate": 8.1514652850324e-06,
+      "loss": 0.8799,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3425244092941284,
+      "learning_rate": 8.149817251934948e-06,
+      "loss": 0.8991,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2665830850601196,
+      "learning_rate": 8.148168651287287e-06,
+      "loss": 0.8075,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3224929571151733,
+      "learning_rate": 8.146519483386473e-06,
+      "loss": 0.8385,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.294895052909851,
+      "learning_rate": 8.144869748529655e-06,
+      "loss": 0.8302,
+      "step": 2327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2835818529129028,
+      "learning_rate": 8.143219447014096e-06,
+      "loss": 0.7986,
+      "step": 2328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3077954053878784,
+      "learning_rate": 8.141568579137157e-06,
+      "loss": 0.9108,
+      "step": 2329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9773204922676086,
+      "learning_rate": 8.139917145196295e-06,
+      "loss": 1.0584,
+      "step": 2330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8817781805992126,
+      "learning_rate": 8.138265145489077e-06,
+      "loss": 1.0441,
+      "step": 2331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.285596489906311,
+      "learning_rate": 8.136612580313167e-06,
+      "loss": 0.8269,
+      "step": 2332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6903011798858643,
+      "learning_rate": 8.134959449966335e-06,
+      "loss": 1.0603,
+      "step": 2333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.273403286933899,
+      "learning_rate": 8.133305754746447e-06,
+      "loss": 0.8362,
+      "step": 2334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3642973899841309,
+      "learning_rate": 8.131651494951478e-06,
+      "loss": 0.7996,
+      "step": 2335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2098721265792847,
+      "learning_rate": 8.129996670879499e-06,
+      "loss": 0.8424,
+      "step": 2336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2771353721618652,
+      "learning_rate": 8.128341282828684e-06,
+      "loss": 0.8592,
+      "step": 2337
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.235375165939331,
+      "learning_rate": 8.126685331097312e-06,
+      "loss": 0.806,
+      "step": 2338
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2775158882141113,
+      "learning_rate": 8.125028815983758e-06,
+      "loss": 0.8384,
+      "step": 2339
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4018257856369019,
+      "learning_rate": 8.123371737786503e-06,
+      "loss": 0.8392,
+      "step": 2340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2893955707550049,
+      "learning_rate": 8.121714096804128e-06,
+      "loss": 0.8824,
+      "step": 2341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.279536485671997,
+      "learning_rate": 8.120055893335313e-06,
+      "loss": 0.8372,
+      "step": 2342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3208245038986206,
+      "learning_rate": 8.118397127678845e-06,
+      "loss": 0.8346,
+      "step": 2343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.30107581615448,
+      "learning_rate": 8.116737800133605e-06,
+      "loss": 0.8415,
+      "step": 2344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5558140277862549,
+      "learning_rate": 8.115077910998583e-06,
+      "loss": 1.056,
+      "step": 2345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3060343265533447,
+      "learning_rate": 8.113417460572865e-06,
+      "loss": 0.8618,
+      "step": 2346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9976842403411865,
+      "learning_rate": 8.111756449155638e-06,
+      "loss": 1.0544,
+      "step": 2347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.303991436958313,
+      "learning_rate": 8.110094877046193e-06,
+      "loss": 0.8633,
+      "step": 2348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3717190027236938,
+      "learning_rate": 8.10843274454392e-06,
+      "loss": 0.9116,
+      "step": 2349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3188121318817139,
+      "learning_rate": 8.10677005194831e-06,
+      "loss": 0.9095,
+      "step": 2350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.342751145362854,
+      "learning_rate": 8.105106799558957e-06,
+      "loss": 0.8701,
+      "step": 2351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3564246892929077,
+      "learning_rate": 8.103442987675554e-06,
+      "loss": 0.8316,
+      "step": 2352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3090510368347168,
+      "learning_rate": 8.101778616597894e-06,
+      "loss": 0.8952,
+      "step": 2353
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3092594146728516,
+      "learning_rate": 8.100113686625872e-06,
+      "loss": 0.8759,
+      "step": 2354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5692179203033447,
+      "learning_rate": 8.098448198059485e-06,
+      "loss": 0.8221,
+      "step": 2355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6256866455078125,
+      "learning_rate": 8.09678215119883e-06,
+      "loss": 1.0691,
+      "step": 2356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4786216020584106,
+      "learning_rate": 8.0951155463441e-06,
+      "loss": 1.0638,
+      "step": 2357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3777687549591064,
+      "learning_rate": 8.093448383795596e-06,
+      "loss": 0.8344,
+      "step": 2358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7922725081443787,
+      "learning_rate": 8.091780663853715e-06,
+      "loss": 1.0701,
+      "step": 2359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3967421054840088,
+      "learning_rate": 8.090112386818954e-06,
+      "loss": 0.8564,
+      "step": 2360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.320028305053711,
+      "learning_rate": 8.088443552991912e-06,
+      "loss": 0.8248,
+      "step": 2361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3821243047714233,
+      "learning_rate": 8.086774162673288e-06,
+      "loss": 0.9057,
+      "step": 2362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3206249475479126,
+      "learning_rate": 8.085104216163885e-06,
+      "loss": 1.0457,
+      "step": 2363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.305389642715454,
+      "learning_rate": 8.083433713764597e-06,
+      "loss": 0.8515,
+      "step": 2364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1963725090026855,
+      "learning_rate": 8.081762655776429e-06,
+      "loss": 1.0417,
+      "step": 2365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6679192781448364,
+      "learning_rate": 8.080091042500474e-06,
+      "loss": 0.8654,
+      "step": 2366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9292422533035278,
+      "learning_rate": 8.07841887423794e-06,
+      "loss": 1.0538,
+      "step": 2367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.778257429599762,
+      "learning_rate": 8.076746151290118e-06,
+      "loss": 1.0569,
+      "step": 2368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5355569124221802,
+      "learning_rate": 8.075072873958415e-06,
+      "loss": 0.8678,
+      "step": 2369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7751597762107849,
+      "learning_rate": 8.073399042544329e-06,
+      "loss": 1.06,
+      "step": 2370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2980690002441406,
+      "learning_rate": 8.071724657349458e-06,
+      "loss": 0.8481,
+      "step": 2371
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8481086492538452,
+      "learning_rate": 8.070049718675502e-06,
+      "loss": 1.0485,
+      "step": 2372
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3212261199951172,
+      "learning_rate": 8.068374226824259e-06,
+      "loss": 0.8553,
+      "step": 2373
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2332789897918701,
+      "learning_rate": 8.066698182097628e-06,
+      "loss": 0.798,
+      "step": 2374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2814580202102661,
+      "learning_rate": 8.06502158479761e-06,
+      "loss": 0.8589,
+      "step": 2375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0122061967849731,
+      "learning_rate": 8.063344435226299e-06,
+      "loss": 1.0662,
+      "step": 2376
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.196740746498108,
+      "learning_rate": 8.061666733685897e-06,
+      "loss": 0.7765,
+      "step": 2377
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2837917804718018,
+      "learning_rate": 8.059988480478695e-06,
+      "loss": 0.8142,
+      "step": 2378
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.311537265777588,
+      "learning_rate": 8.058309675907094e-06,
+      "loss": 0.846,
+      "step": 2379
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.299961805343628,
+      "learning_rate": 8.056630320273588e-06,
+      "loss": 0.8143,
+      "step": 2380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2565068006515503,
+      "learning_rate": 8.054950413880772e-06,
+      "loss": 0.8161,
+      "step": 2381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.27354896068573,
+      "learning_rate": 8.05326995703134e-06,
+      "loss": 0.8965,
+      "step": 2382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3007864952087402,
+      "learning_rate": 8.051588950028086e-06,
+      "loss": 0.7964,
+      "step": 2383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2189915180206299,
+      "learning_rate": 8.049907393173902e-06,
+      "loss": 0.8364,
+      "step": 2384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2646139860153198,
+      "learning_rate": 8.048225286771777e-06,
+      "loss": 0.8281,
+      "step": 2385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3162851333618164,
+      "learning_rate": 8.046542631124806e-06,
+      "loss": 0.8628,
+      "step": 2386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.205352783203125,
+      "learning_rate": 8.044859426536176e-06,
+      "loss": 0.8032,
+      "step": 2387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.316967248916626,
+      "learning_rate": 8.043175673309176e-06,
+      "loss": 0.8486,
+      "step": 2388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7502685785293579,
+      "learning_rate": 8.041491371747192e-06,
+      "loss": 1.0478,
+      "step": 2389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.680019736289978,
+      "learning_rate": 8.039806522153713e-06,
+      "loss": 1.0468,
+      "step": 2390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4303315877914429,
+      "learning_rate": 8.038121124832322e-06,
+      "loss": 0.8634,
+      "step": 2391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.342356562614441,
+      "learning_rate": 8.036435180086701e-06,
+      "loss": 0.9024,
+      "step": 2392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3191437721252441,
+      "learning_rate": 8.034748688220633e-06,
+      "loss": 0.8514,
+      "step": 2393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7458620071411133,
+      "learning_rate": 8.033061649538003e-06,
+      "loss": 1.0537,
+      "step": 2394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3250089883804321,
+      "learning_rate": 8.031374064342782e-06,
+      "loss": 0.8549,
+      "step": 2395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.311827540397644,
+      "learning_rate": 8.029685932939057e-06,
+      "loss": 0.8622,
+      "step": 2396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2432093620300293,
+      "learning_rate": 8.027997255630996e-06,
+      "loss": 0.8098,
+      "step": 2397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2596198320388794,
+      "learning_rate": 8.026308032722882e-06,
+      "loss": 0.781,
+      "step": 2398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7353680729866028,
+      "learning_rate": 8.024618264519079e-06,
+      "loss": 1.0605,
+      "step": 2399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3314157724380493,
+      "learning_rate": 8.022927951324065e-06,
+      "loss": 0.8687,
+      "step": 2400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7200467586517334,
+      "learning_rate": 8.021237093442403e-06,
+      "loss": 1.0586,
+      "step": 2401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3268356323242188,
+      "learning_rate": 8.019545691178764e-06,
+      "loss": 0.8863,
+      "step": 2402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.30378258228302,
+      "learning_rate": 8.017853744837915e-06,
+      "loss": 0.8739,
+      "step": 2403
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.49485445022583,
+      "learning_rate": 8.01616125472472e-06,
+      "loss": 0.8691,
+      "step": 2404
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2088567018508911,
+      "learning_rate": 8.014468221144136e-06,
+      "loss": 0.8354,
+      "step": 2405
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3744698762893677,
+      "learning_rate": 8.012774644401224e-06,
+      "loss": 0.8017,
+      "step": 2406
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5287193059921265,
+      "learning_rate": 8.011080524801145e-06,
+      "loss": 0.8696,
+      "step": 2407
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9511661529541016,
+      "learning_rate": 8.009385862649149e-06,
+      "loss": 1.0413,
+      "step": 2408
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.767482340335846,
+      "learning_rate": 8.007690658250593e-06,
+      "loss": 1.0534,
+      "step": 2409
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2829058170318604,
+      "learning_rate": 8.005994911910925e-06,
+      "loss": 0.8898,
+      "step": 2410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7969670295715332,
+      "learning_rate": 8.004298623935694e-06,
+      "loss": 1.0154,
+      "step": 2411
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3663973808288574,
+      "learning_rate": 8.002601794630547e-06,
+      "loss": 0.8787,
+      "step": 2412
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3126904964447021,
+      "learning_rate": 8.000904424301225e-06,
+      "loss": 0.8281,
+      "step": 2413
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.296817660331726,
+      "learning_rate": 7.999206513253573e-06,
+      "loss": 0.9001,
+      "step": 2414
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2749918699264526,
+      "learning_rate": 7.997508061793525e-06,
+      "loss": 0.8243,
+      "step": 2415
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3949772119522095,
+      "learning_rate": 7.995809070227116e-06,
+      "loss": 0.8923,
+      "step": 2416
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3495936393737793,
+      "learning_rate": 7.994109538860484e-06,
+      "loss": 0.8639,
+      "step": 2417
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3644553422927856,
+      "learning_rate": 7.992409467999855e-06,
+      "loss": 0.7971,
+      "step": 2418
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.327644944190979,
+      "learning_rate": 7.990708857951561e-06,
+      "loss": 0.8886,
+      "step": 2419
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.233739972114563,
+      "learning_rate": 7.98900770902202e-06,
+      "loss": 1.0697,
+      "step": 2420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9672421216964722,
+      "learning_rate": 7.987306021517761e-06,
+      "loss": 1.0395,
+      "step": 2421
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2891470193862915,
+      "learning_rate": 7.985603795745397e-06,
+      "loss": 0.8911,
+      "step": 2422
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3081644773483276,
+      "learning_rate": 7.983901032011648e-06,
+      "loss": 0.8867,
+      "step": 2423
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2711498737335205,
+      "learning_rate": 7.982197730623322e-06,
+      "loss": 0.8579,
+      "step": 2424
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.30775785446167,
+      "learning_rate": 7.980493891887333e-06,
+      "loss": 0.8354,
+      "step": 2425
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3012700080871582,
+      "learning_rate": 7.978789516110684e-06,
+      "loss": 0.8094,
+      "step": 2426
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2689217329025269,
+      "learning_rate": 7.977084603600482e-06,
+      "loss": 0.8392,
+      "step": 2427
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2579792737960815,
+      "learning_rate": 7.975379154663923e-06,
+      "loss": 0.8459,
+      "step": 2428
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3215928077697754,
+      "learning_rate": 7.973673169608308e-06,
+      "loss": 0.8792,
+      "step": 2429
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5018366575241089,
+      "learning_rate": 7.971966648741023e-06,
+      "loss": 0.8111,
+      "step": 2430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2950294017791748,
+      "learning_rate": 7.970259592369564e-06,
+      "loss": 0.8157,
+      "step": 2431
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2643022537231445,
+      "learning_rate": 7.968552000801516e-06,
+      "loss": 0.7773,
+      "step": 2432
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.150285482406616,
+      "learning_rate": 7.96684387434456e-06,
+      "loss": 1.0699,
+      "step": 2433
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8598589897155762,
+      "learning_rate": 7.965135213306474e-06,
+      "loss": 1.0871,
+      "step": 2434
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4148564338684082,
+      "learning_rate": 7.963426017995137e-06,
+      "loss": 0.8906,
+      "step": 2435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4011112451553345,
+      "learning_rate": 7.961716288718518e-06,
+      "loss": 0.8413,
+      "step": 2436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.924346923828125,
+      "learning_rate": 7.960006025784687e-06,
+      "loss": 1.0649,
+      "step": 2437
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4519782066345215,
+      "learning_rate": 7.958295229501804e-06,
+      "loss": 0.8611,
+      "step": 2438
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3417631387710571,
+      "learning_rate": 7.956583900178133e-06,
+      "loss": 0.8528,
+      "step": 2439
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9775850772857666,
+      "learning_rate": 7.954872038122028e-06,
+      "loss": 1.0545,
+      "step": 2440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4304217100143433,
+      "learning_rate": 7.95315964364194e-06,
+      "loss": 0.8459,
+      "step": 2441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4154038429260254,
+      "learning_rate": 7.95144671704642e-06,
+      "loss": 0.8199,
+      "step": 2442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3670209646224976,
+      "learning_rate": 7.94973325864411e-06,
+      "loss": 0.8666,
+      "step": 2443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3826408386230469,
+      "learning_rate": 7.94801926874375e-06,
+      "loss": 0.8486,
+      "step": 2444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2132552862167358,
+      "learning_rate": 7.946304747654176e-06,
+      "loss": 0.8562,
+      "step": 2445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3096131086349487,
+      "learning_rate": 7.94458969568432e-06,
+      "loss": 0.8737,
+      "step": 2446
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2261825799942017,
+      "learning_rate": 7.942874113143206e-06,
+      "loss": 1.0574,
+      "step": 2447
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9805470108985901,
+      "learning_rate": 7.941158000339956e-06,
+      "loss": 1.0485,
+      "step": 2448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7299032211303711,
+      "learning_rate": 7.939441357583794e-06,
+      "loss": 1.0642,
+      "step": 2449
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5213171243667603,
+      "learning_rate": 7.937724185184027e-06,
+      "loss": 0.8441,
+      "step": 2450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3726316690444946,
+      "learning_rate": 7.93600648345007e-06,
+      "loss": 0.8567,
+      "step": 2451
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2997819185256958,
+      "learning_rate": 7.93428825269142e-06,
+      "loss": 0.8908,
+      "step": 2452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.32731294631958,
+      "learning_rate": 7.932569493217681e-06,
+      "loss": 1.0517,
+      "step": 2453
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.223536729812622,
+      "learning_rate": 7.93085020533855e-06,
+      "loss": 1.0607,
+      "step": 2454
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0071001052856445,
+      "learning_rate": 7.929130389363811e-06,
+      "loss": 1.073,
+      "step": 2455
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3793420791625977,
+      "learning_rate": 7.927410045603355e-06,
+      "loss": 0.8143,
+      "step": 2456
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5402896404266357,
+      "learning_rate": 7.925689174367158e-06,
+      "loss": 0.8981,
+      "step": 2457
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.433461308479309,
+      "learning_rate": 7.923967775965298e-06,
+      "loss": 0.8478,
+      "step": 2458
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4200477600097656,
+      "learning_rate": 7.922245850707945e-06,
+      "loss": 0.8534,
+      "step": 2459
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4835337400436401,
+      "learning_rate": 7.920523398905365e-06,
+      "loss": 1.0406,
+      "step": 2460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3136411905288696,
+      "learning_rate": 7.918800420867916e-06,
+      "loss": 1.0752,
+      "step": 2461
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3444551229476929,
+      "learning_rate": 7.917076916906056e-06,
+      "loss": 0.7836,
+      "step": 2462
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3789544105529785,
+      "learning_rate": 7.915352887330332e-06,
+      "loss": 0.865,
+      "step": 2463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.470974326133728,
+      "learning_rate": 7.913628332451389e-06,
+      "loss": 0.8736,
+      "step": 2464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.16569185256958,
+      "learning_rate": 7.911903252579966e-06,
+      "loss": 1.0624,
+      "step": 2465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4029771089553833,
+      "learning_rate": 7.9101776480269e-06,
+      "loss": 0.8616,
+      "step": 2466
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2402920722961426,
+      "learning_rate": 7.908451519103115e-06,
+      "loss": 0.8456,
+      "step": 2467
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2809373140335083,
+      "learning_rate": 7.906724866119634e-06,
+      "loss": 0.8614,
+      "step": 2468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3424105644226074,
+      "learning_rate": 7.90499768938758e-06,
+      "loss": 0.8499,
+      "step": 2469
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3007092475891113,
+      "learning_rate": 7.903269989218156e-06,
+      "loss": 0.8435,
+      "step": 2470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3375861644744873,
+      "learning_rate": 7.901541765922674e-06,
+      "loss": 0.8825,
+      "step": 2471
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.268302321434021,
+      "learning_rate": 7.899813019812532e-06,
+      "loss": 0.8193,
+      "step": 2472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2723933458328247,
+      "learning_rate": 7.898083751199223e-06,
+      "loss": 0.8238,
+      "step": 2473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.275877833366394,
+      "learning_rate": 7.896353960394339e-06,
+      "loss": 0.9374,
+      "step": 2474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0687806606292725,
+      "learning_rate": 7.89462364770956e-06,
+      "loss": 1.0244,
+      "step": 2475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2721234560012817,
+      "learning_rate": 7.892892813456663e-06,
+      "loss": 0.8948,
+      "step": 2476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2734066247940063,
+      "learning_rate": 7.891161457947518e-06,
+      "loss": 0.8975,
+      "step": 2477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3277411460876465,
+      "learning_rate": 7.88942958149409e-06,
+      "loss": 0.8765,
+      "step": 2478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2908236980438232,
+      "learning_rate": 7.887697184408438e-06,
+      "loss": 0.7944,
+      "step": 2479
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4502190351486206,
+      "learning_rate": 7.885964267002715e-06,
+      "loss": 0.8624,
+      "step": 2480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3119754791259766,
+      "learning_rate": 7.884230829589165e-06,
+      "loss": 0.8363,
+      "step": 2481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7373291254043579,
+      "learning_rate": 7.882496872480128e-06,
+      "loss": 1.0334,
+      "step": 2482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2061450481414795,
+      "learning_rate": 7.880762395988037e-06,
+      "loss": 0.8663,
+      "step": 2483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6549899578094482,
+      "learning_rate": 7.879027400425421e-06,
+      "loss": 1.0499,
+      "step": 2484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.255236268043518,
+      "learning_rate": 7.877291886104897e-06,
+      "loss": 0.8195,
+      "step": 2485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.65566086769104,
+      "learning_rate": 7.875555853339184e-06,
+      "loss": 1.0371,
+      "step": 2486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3055273294448853,
+      "learning_rate": 7.873819302441082e-06,
+      "loss": 0.8103,
+      "step": 2487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.332232117652893,
+      "learning_rate": 7.872082233723498e-06,
+      "loss": 0.8442,
+      "step": 2488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2872645854949951,
+      "learning_rate": 7.870344647499422e-06,
+      "loss": 0.8232,
+      "step": 2489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6854805946350098,
+      "learning_rate": 7.868606544081945e-06,
+      "loss": 1.0557,
+      "step": 2490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5983079671859741,
+      "learning_rate": 7.866867923784243e-06,
+      "loss": 0.8854,
+      "step": 2491
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.297674298286438,
+      "learning_rate": 7.865128786919594e-06,
+      "loss": 0.829,
+      "step": 2492
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3193473815917969,
+      "learning_rate": 7.86338913380136e-06,
+      "loss": 0.8132,
+      "step": 2493
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.299903392791748,
+      "learning_rate": 7.861648964743004e-06,
+      "loss": 0.8547,
+      "step": 2494
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2772645950317383,
+      "learning_rate": 7.859908280058077e-06,
+      "loss": 0.8607,
+      "step": 2495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2404985427856445,
+      "learning_rate": 7.858167080060224e-06,
+      "loss": 0.8676,
+      "step": 2496
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2106467485427856,
+      "learning_rate": 7.856425365063187e-06,
+      "loss": 0.8448,
+      "step": 2497
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.243106484413147,
+      "learning_rate": 7.854683135380791e-06,
+      "loss": 0.8128,
+      "step": 2498
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2317794561386108,
+      "learning_rate": 7.852940391326967e-06,
+      "loss": 0.8109,
+      "step": 2499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2543375492095947,
+      "learning_rate": 7.851197133215725e-06,
+      "loss": 0.8252,
+      "step": 2500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7806945443153381,
+      "learning_rate": 7.849453361361179e-06,
+      "loss": 1.0687,
+      "step": 2501
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2608129978179932,
+      "learning_rate": 7.847709076077527e-06,
+      "loss": 0.8167,
+      "step": 2502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2217901945114136,
+      "learning_rate": 7.845964277679066e-06,
+      "loss": 0.8039,
+      "step": 2503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2519893646240234,
+      "learning_rate": 7.844218966480181e-06,
+      "loss": 0.8345,
+      "step": 2504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2750500440597534,
+      "learning_rate": 7.842473142795353e-06,
+      "loss": 0.7984,
+      "step": 2505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.175591230392456,
+      "learning_rate": 7.84072680693915e-06,
+      "loss": 0.8018,
+      "step": 2506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.269715666770935,
+      "learning_rate": 7.838979959226239e-06,
+      "loss": 0.8177,
+      "step": 2507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3822591304779053,
+      "learning_rate": 7.837232599971376e-06,
+      "loss": 0.8733,
+      "step": 2508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2983927726745605,
+      "learning_rate": 7.835484729489407e-06,
+      "loss": 0.8564,
+      "step": 2509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8333718776702881,
+      "learning_rate": 7.833736348095274e-06,
+      "loss": 1.0354,
+      "step": 2510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2349369525909424,
+      "learning_rate": 7.831987456104008e-06,
+      "loss": 0.7904,
+      "step": 2511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6882080435752869,
+      "learning_rate": 7.830238053830735e-06,
+      "loss": 1.0571,
+      "step": 2512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6499983072280884,
+      "learning_rate": 7.828488141590671e-06,
+      "loss": 1.0694,
+      "step": 2513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2733501195907593,
+      "learning_rate": 7.826737719699122e-06,
+      "loss": 0.7933,
+      "step": 2514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2781227827072144,
+      "learning_rate": 7.824986788471489e-06,
+      "loss": 0.8215,
+      "step": 2515
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2844797372817993,
+      "learning_rate": 7.823235348223267e-06,
+      "loss": 0.7958,
+      "step": 2516
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7984095215797424,
+      "learning_rate": 7.821483399270033e-06,
+      "loss": 1.0626,
+      "step": 2517
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2803195714950562,
+      "learning_rate": 7.819730941927467e-06,
+      "loss": 0.8047,
+      "step": 2518
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2378897666931152,
+      "learning_rate": 7.817977976511334e-06,
+      "loss": 0.7874,
+      "step": 2519
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2440996170043945,
+      "learning_rate": 7.816224503337495e-06,
+      "loss": 0.8195,
+      "step": 2520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2215795516967773,
+      "learning_rate": 7.814470522721898e-06,
+      "loss": 0.8248,
+      "step": 2521
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.330870270729065,
+      "learning_rate": 7.812716034980581e-06,
+      "loss": 0.8667,
+      "step": 2522
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.352516531944275,
+      "learning_rate": 7.81096104042968e-06,
+      "loss": 0.837,
+      "step": 2523
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3091157674789429,
+      "learning_rate": 7.809205539385419e-06,
+      "loss": 0.8685,
+      "step": 2524
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2769447565078735,
+      "learning_rate": 7.807449532164112e-06,
+      "loss": 0.8667,
+      "step": 2525
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0442728996276855,
+      "learning_rate": 7.805693019082168e-06,
+      "loss": 1.0405,
+      "step": 2526
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8505499362945557,
+      "learning_rate": 7.80393600045608e-06,
+      "loss": 1.0425,
+      "step": 2527
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6560333371162415,
+      "learning_rate": 7.80217847660244e-06,
+      "loss": 1.0473,
+      "step": 2528
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4268993139266968,
+      "learning_rate": 7.800420447837926e-06,
+      "loss": 0.824,
+      "step": 2529
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.396239995956421,
+      "learning_rate": 7.798661914479308e-06,
+      "loss": 0.8662,
+      "step": 2530
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2881433963775635,
+      "learning_rate": 7.79690287684345e-06,
+      "loss": 0.8663,
+      "step": 2531
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.359786033630371,
+      "learning_rate": 7.795143335247306e-06,
+      "loss": 0.8157,
+      "step": 2532
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2889009714126587,
+      "learning_rate": 7.793383290007914e-06,
+      "loss": 0.8955,
+      "step": 2533
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2896366119384766,
+      "learning_rate": 7.79162274144241e-06,
+      "loss": 0.8519,
+      "step": 2534
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2965589761734009,
+      "learning_rate": 7.78986168986802e-06,
+      "loss": 0.8149,
+      "step": 2535
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3436747789382935,
+      "learning_rate": 7.788100135602058e-06,
+      "loss": 0.862,
+      "step": 2536
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3191670179367065,
+      "learning_rate": 7.78633807896193e-06,
+      "loss": 0.911,
+      "step": 2537
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3801847696304321,
+      "learning_rate": 7.784575520265132e-06,
+      "loss": 0.8307,
+      "step": 2538
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7564704418182373,
+      "learning_rate": 7.782812459829254e-06,
+      "loss": 1.0317,
+      "step": 2539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2932320833206177,
+      "learning_rate": 7.781048897971969e-06,
+      "loss": 0.8259,
+      "step": 2540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3421549797058105,
+      "learning_rate": 7.779284835011047e-06,
+      "loss": 0.8534,
+      "step": 2541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3511643409729004,
+      "learning_rate": 7.777520271264345e-06,
+      "loss": 0.835,
+      "step": 2542
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3365064859390259,
+      "learning_rate": 7.77575520704981e-06,
+      "loss": 0.7747,
+      "step": 2543
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3266938924789429,
+      "learning_rate": 7.773989642685485e-06,
+      "loss": 0.8215,
+      "step": 2544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9853667616844177,
+      "learning_rate": 7.772223578489492e-06,
+      "loss": 1.0508,
+      "step": 2545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.22501540184021,
+      "learning_rate": 7.770457014780054e-06,
+      "loss": 0.7729,
+      "step": 2546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3209078311920166,
+      "learning_rate": 7.76868995187548e-06,
+      "loss": 0.8444,
+      "step": 2547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.798454761505127,
+      "learning_rate": 7.766922390094163e-06,
+      "loss": 1.051,
+      "step": 2548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3354198932647705,
+      "learning_rate": 7.765154329754596e-06,
+      "loss": 0.8437,
+      "step": 2549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2597715854644775,
+      "learning_rate": 7.763385771175358e-06,
+      "loss": 0.8539,
+      "step": 2550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.257220983505249,
+      "learning_rate": 7.761616714675114e-06,
+      "loss": 0.7746,
+      "step": 2551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8412327170372009,
+      "learning_rate": 7.759847160572622e-06,
+      "loss": 1.0371,
+      "step": 2552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2304428815841675,
+      "learning_rate": 7.75807710918673e-06,
+      "loss": 0.8889,
+      "step": 2553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.062561511993408,
+      "learning_rate": 7.756306560836375e-06,
+      "loss": 0.8524,
+      "step": 2554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.668779194355011,
+      "learning_rate": 7.754535515840584e-06,
+      "loss": 1.0525,
+      "step": 2555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.342820644378662,
+      "learning_rate": 7.752763974518474e-06,
+      "loss": 0.8418,
+      "step": 2556
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.708640456199646,
+      "learning_rate": 7.750991937189245e-06,
+      "loss": 1.0542,
+      "step": 2557
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3221486806869507,
+      "learning_rate": 7.749219404172198e-06,
+      "loss": 0.8212,
+      "step": 2558
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2574487924575806,
+      "learning_rate": 7.747446375786715e-06,
+      "loss": 0.8749,
+      "step": 2559
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2693716287612915,
+      "learning_rate": 7.745672852352268e-06,
+      "loss": 0.8299,
+      "step": 2560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.286001443862915,
+      "learning_rate": 7.743898834188421e-06,
+      "loss": 0.9351,
+      "step": 2561
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1930619478225708,
+      "learning_rate": 7.742124321614826e-06,
+      "loss": 0.7895,
+      "step": 2562
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2663344144821167,
+      "learning_rate": 7.74034931495122e-06,
+      "loss": 0.829,
+      "step": 2563
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7658432722091675,
+      "learning_rate": 7.738573814517438e-06,
+      "loss": 1.0691,
+      "step": 2564
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6873274445533752,
+      "learning_rate": 7.736797820633397e-06,
+      "loss": 1.0478,
+      "step": 2565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6273059844970703,
+      "learning_rate": 7.735021333619102e-06,
+      "loss": 1.0499,
+      "step": 2566
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2987314462661743,
+      "learning_rate": 7.733244353794655e-06,
+      "loss": 0.8369,
+      "step": 2567
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7454431056976318,
+      "learning_rate": 7.731466881480235e-06,
+      "loss": 1.0619,
+      "step": 2568
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6981526613235474,
+      "learning_rate": 7.72968891699612e-06,
+      "loss": 1.0567,
+      "step": 2569
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.334380865097046,
+      "learning_rate": 7.727910460662673e-06,
+      "loss": 0.756,
+      "step": 2570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2751524448394775,
+      "learning_rate": 7.726131512800343e-06,
+      "loss": 0.7987,
+      "step": 2571
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7176579236984253,
+      "learning_rate": 7.72435207372967e-06,
+      "loss": 1.0362,
+      "step": 2572
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3107401132583618,
+      "learning_rate": 7.722572143771288e-06,
+      "loss": 0.8896,
+      "step": 2573
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2261414527893066,
+      "learning_rate": 7.720791723245907e-06,
+      "loss": 0.8648,
+      "step": 2574
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3013050556182861,
+      "learning_rate": 7.719010812474334e-06,
+      "loss": 0.8849,
+      "step": 2575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2206703424453735,
+      "learning_rate": 7.717229411777463e-06,
+      "loss": 0.7808,
+      "step": 2576
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.390152931213379,
+      "learning_rate": 7.715447521476279e-06,
+      "loss": 0.8936,
+      "step": 2577
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2874579429626465,
+      "learning_rate": 7.713665141891847e-06,
+      "loss": 0.8084,
+      "step": 2578
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.287598967552185,
+      "learning_rate": 7.711882273345329e-06,
+      "loss": 0.807,
+      "step": 2579
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6010318994522095,
+      "learning_rate": 7.710098916157968e-06,
+      "loss": 0.8405,
+      "step": 2580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9325595498085022,
+      "learning_rate": 7.708315070651104e-06,
+      "loss": 1.0592,
+      "step": 2581
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8337389826774597,
+      "learning_rate": 7.706530737146154e-06,
+      "loss": 1.0494,
+      "step": 2582
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3103801012039185,
+      "learning_rate": 7.704745915964632e-06,
+      "loss": 0.8268,
+      "step": 2583
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2624452114105225,
+      "learning_rate": 7.70296060742813e-06,
+      "loss": 0.8601,
+      "step": 2584
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8445315957069397,
+      "learning_rate": 7.701174811858343e-06,
+      "loss": 1.0553,
+      "step": 2585
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9497881531715393,
+      "learning_rate": 7.699388529577038e-06,
+      "loss": 1.0482,
+      "step": 2586
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3802552223205566,
+      "learning_rate": 7.697601760906082e-06,
+      "loss": 0.8339,
+      "step": 2587
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.294763207435608,
+      "learning_rate": 7.69581450616742e-06,
+      "loss": 0.8891,
+      "step": 2588
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2968716621398926,
+      "learning_rate": 7.694026765683089e-06,
+      "loss": 0.8375,
+      "step": 2589
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2589976787567139,
+      "learning_rate": 7.692238539775211e-06,
+      "loss": 0.8124,
+      "step": 2590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2676384449005127,
+      "learning_rate": 7.690449828766007e-06,
+      "loss": 0.8904,
+      "step": 2591
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9646385312080383,
+      "learning_rate": 7.688660632977764e-06,
+      "loss": 1.0453,
+      "step": 2592
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2234407663345337,
+      "learning_rate": 7.686870952732878e-06,
+      "loss": 0.8019,
+      "step": 2593
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3883187770843506,
+      "learning_rate": 7.685080788353819e-06,
+      "loss": 0.8477,
+      "step": 2594
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2558010816574097,
+      "learning_rate": 7.683290140163147e-06,
+      "loss": 0.8159,
+      "step": 2595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6898724436759949,
+      "learning_rate": 7.681499008483509e-06,
+      "loss": 1.0479,
+      "step": 2596
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.246668815612793,
+      "learning_rate": 7.679707393637645e-06,
+      "loss": 0.8086,
+      "step": 2597
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6829752326011658,
+      "learning_rate": 7.677915295948374e-06,
+      "loss": 1.0562,
+      "step": 2598
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2916669845581055,
+      "learning_rate": 7.676122715738605e-06,
+      "loss": 0.8593,
+      "step": 2599
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6418684124946594,
+      "learning_rate": 7.674329653331338e-06,
+      "loss": 1.0428,
+      "step": 2600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5934301614761353,
+      "learning_rate": 7.672536109049652e-06,
+      "loss": 0.8704,
+      "step": 2601
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2990731000900269,
+      "learning_rate": 7.670742083216719e-06,
+      "loss": 0.7937,
+      "step": 2602
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7142493724822998,
+      "learning_rate": 7.668947576155796e-06,
+      "loss": 1.026,
+      "step": 2603
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2786521911621094,
+      "learning_rate": 7.667152588190225e-06,
+      "loss": 0.7861,
+      "step": 2604
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2171430587768555,
+      "learning_rate": 7.665357119643438e-06,
+      "loss": 0.8377,
+      "step": 2605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7262842059135437,
+      "learning_rate": 7.663561170838948e-06,
+      "loss": 1.0689,
+      "step": 2606
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.262280821800232,
+      "learning_rate": 7.661764742100363e-06,
+      "loss": 0.8656,
+      "step": 2607
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2808278799057007,
+      "learning_rate": 7.659967833751372e-06,
+      "loss": 0.8391,
+      "step": 2608
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2966384887695312,
+      "learning_rate": 7.658170446115747e-06,
+      "loss": 0.7663,
+      "step": 2609
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7389931678771973,
+      "learning_rate": 7.656372579517356e-06,
+      "loss": 1.0545,
+      "step": 2610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.664376437664032,
+      "learning_rate": 7.654574234280142e-06,
+      "loss": 1.054,
+      "step": 2611
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2464044094085693,
+      "learning_rate": 7.652775410728145e-06,
+      "loss": 0.8029,
+      "step": 2612
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3509541749954224,
+      "learning_rate": 7.650976109185484e-06,
+      "loss": 0.8309,
+      "step": 2613
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2998528480529785,
+      "learning_rate": 7.649176329976368e-06,
+      "loss": 0.8557,
+      "step": 2614
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2693666219711304,
+      "learning_rate": 7.647376073425088e-06,
+      "loss": 0.8861,
+      "step": 2615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2937744855880737,
+      "learning_rate": 7.645575339856026e-06,
+      "loss": 0.8817,
+      "step": 2616
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1433987617492676,
+      "learning_rate": 7.643774129593644e-06,
+      "loss": 1.0535,
+      "step": 2617
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2218538522720337,
+      "learning_rate": 7.641972442962498e-06,
+      "loss": 0.8252,
+      "step": 2618
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2936040163040161,
+      "learning_rate": 7.64017028028722e-06,
+      "loss": 0.8699,
+      "step": 2619
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6564664840698242,
+      "learning_rate": 7.638367641892538e-06,
+      "loss": 1.0567,
+      "step": 2620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2961831092834473,
+      "learning_rate": 7.636564528103256e-06,
+      "loss": 0.8328,
+      "step": 2621
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3119202852249146,
+      "learning_rate": 7.63476093924427e-06,
+      "loss": 0.8145,
+      "step": 2622
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8658624291419983,
+      "learning_rate": 7.632956875640563e-06,
+      "loss": 1.0599,
+      "step": 2623
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.235541582107544,
+      "learning_rate": 7.631152337617197e-06,
+      "loss": 0.7905,
+      "step": 2624
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3507096767425537,
+      "learning_rate": 7.629347325499321e-06,
+      "loss": 0.8759,
+      "step": 2625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2283000946044922,
+      "learning_rate": 7.6275418396121745e-06,
+      "loss": 0.8147,
+      "step": 2626
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7090712785720825,
+      "learning_rate": 7.62573588028108e-06,
+      "loss": 1.0591,
+      "step": 2627
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2846288681030273,
+      "learning_rate": 7.623929447831442e-06,
+      "loss": 0.8042,
+      "step": 2628
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3286707401275635,
+      "learning_rate": 7.622122542588755e-06,
+      "loss": 0.8255,
+      "step": 2629
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7196866273880005,
+      "learning_rate": 7.6203151648785935e-06,
+      "loss": 1.0363,
+      "step": 2630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3640694618225098,
+      "learning_rate": 7.618507315026622e-06,
+      "loss": 0.8559,
+      "step": 2631
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3674308061599731,
+      "learning_rate": 7.61669899335859e-06,
+      "loss": 0.848,
+      "step": 2632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.685600221157074,
+      "learning_rate": 7.614890200200326e-06,
+      "loss": 1.0212,
+      "step": 2633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2772520780563354,
+      "learning_rate": 7.6130809358777514e-06,
+      "loss": 0.8162,
+      "step": 2634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2635962963104248,
+      "learning_rate": 7.611271200716867e-06,
+      "loss": 0.8258,
+      "step": 2635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6850273609161377,
+      "learning_rate": 7.609460995043758e-06,
+      "loss": 1.0398,
+      "step": 2636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.224825143814087,
+      "learning_rate": 7.607650319184601e-06,
+      "loss": 0.8445,
+      "step": 2637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3007365465164185,
+      "learning_rate": 7.60583917346565e-06,
+      "loss": 0.907,
+      "step": 2638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6555388569831848,
+      "learning_rate": 7.604027558213247e-06,
+      "loss": 1.0663,
+      "step": 2639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2755221128463745,
+      "learning_rate": 7.602215473753819e-06,
+      "loss": 0.8683,
+      "step": 2640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3264508247375488,
+      "learning_rate": 7.600402920413876e-06,
+      "loss": 0.8856,
+      "step": 2641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.465968132019043,
+      "learning_rate": 7.598589898520011e-06,
+      "loss": 0.8708,
+      "step": 2642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2313069105148315,
+      "learning_rate": 7.596776408398908e-06,
+      "loss": 0.8149,
+      "step": 2643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2203158140182495,
+      "learning_rate": 7.594962450377328e-06,
+      "loss": 0.8092,
+      "step": 2644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3059965372085571,
+      "learning_rate": 7.59314802478212e-06,
+      "loss": 0.8707,
+      "step": 2645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.198370099067688,
+      "learning_rate": 7.591333131940214e-06,
+      "loss": 0.8013,
+      "step": 2646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2312397956848145,
+      "learning_rate": 7.589517772178632e-06,
+      "loss": 0.8221,
+      "step": 2647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7094905376434326,
+      "learning_rate": 7.5877019458244705e-06,
+      "loss": 1.0505,
+      "step": 2648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6825560331344604,
+      "learning_rate": 7.585885653204916e-06,
+      "loss": 1.0255,
+      "step": 2649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.269484281539917,
+      "learning_rate": 7.5840688946472365e-06,
+      "loss": 0.8557,
+      "step": 2650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6474770307540894,
+      "learning_rate": 7.582251670478787e-06,
+      "loss": 1.0284,
+      "step": 2651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3621948957443237,
+      "learning_rate": 7.580433981027e-06,
+      "loss": 0.8379,
+      "step": 2652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3511321544647217,
+      "learning_rate": 7.578615826619402e-06,
+      "loss": 0.9096,
+      "step": 2653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.735480546951294,
+      "learning_rate": 7.576797207583592e-06,
+      "loss": 1.0349,
+      "step": 2654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6882060766220093,
+      "learning_rate": 7.574978124247261e-06,
+      "loss": 1.0666,
+      "step": 2655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6312455534934998,
+      "learning_rate": 7.57315857693818e-06,
+      "loss": 1.0644,
+      "step": 2656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6713507771492004,
+      "learning_rate": 7.5713385659842056e-06,
+      "loss": 1.0485,
+      "step": 2657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3109878301620483,
+      "learning_rate": 7.569518091713273e-06,
+      "loss": 0.8448,
+      "step": 2658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3903013467788696,
+      "learning_rate": 7.567697154453411e-06,
+      "loss": 0.8198,
+      "step": 2659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3332114219665527,
+      "learning_rate": 7.5658757545327174e-06,
+      "loss": 0.8073,
+      "step": 2660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3037081956863403,
+      "learning_rate": 7.564053892279388e-06,
+      "loss": 0.8995,
+      "step": 2661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2978622913360596,
+      "learning_rate": 7.562231568021692e-06,
+      "loss": 0.8329,
+      "step": 2662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3116317987442017,
+      "learning_rate": 7.560408782087987e-06,
+      "loss": 0.8504,
+      "step": 2663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8212358951568604,
+      "learning_rate": 7.5585855348067084e-06,
+      "loss": 1.0236,
+      "step": 2664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2870222330093384,
+      "learning_rate": 7.556761826506382e-06,
+      "loss": 0.8033,
+      "step": 2665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.322788119316101,
+      "learning_rate": 7.554937657515611e-06,
+      "loss": 0.8539,
+      "step": 2666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7093355059623718,
+      "learning_rate": 7.553113028163086e-06,
+      "loss": 1.0595,
+      "step": 2667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.237036108970642,
+      "learning_rate": 7.551287938777574e-06,
+      "loss": 0.8085,
+      "step": 2668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2830079793930054,
+      "learning_rate": 7.549462389687932e-06,
+      "loss": 0.8335,
+      "step": 2669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3387119770050049,
+      "learning_rate": 7.547636381223095e-06,
+      "loss": 0.8191,
+      "step": 2670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.322961449623108,
+      "learning_rate": 7.545809913712084e-06,
+      "loss": 0.8596,
+      "step": 2671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.245439887046814,
+      "learning_rate": 7.543982987484001e-06,
+      "loss": 0.8082,
+      "step": 2672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2903600931167603,
+      "learning_rate": 7.54215560286803e-06,
+      "loss": 0.8536,
+      "step": 2673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2654783725738525,
+      "learning_rate": 7.54032776019344e-06,
+      "loss": 0.8236,
+      "step": 2674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2684671878814697,
+      "learning_rate": 7.5384994597895784e-06,
+      "loss": 0.8031,
+      "step": 2675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3509429693222046,
+      "learning_rate": 7.536670701985882e-06,
+      "loss": 0.9005,
+      "step": 2676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8219190835952759,
+      "learning_rate": 7.534841487111861e-06,
+      "loss": 1.0406,
+      "step": 2677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2607899904251099,
+      "learning_rate": 7.533011815497115e-06,
+      "loss": 0.8502,
+      "step": 2678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3500337600708008,
+      "learning_rate": 7.531181687471325e-06,
+      "loss": 0.8687,
+      "step": 2679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3462703227996826,
+      "learning_rate": 7.5293511033642534e-06,
+      "loss": 0.8404,
+      "step": 2680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6980384588241577,
+      "learning_rate": 7.527520063505738e-06,
+      "loss": 1.0654,
+      "step": 2681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6791280508041382,
+      "learning_rate": 7.52568856822571e-06,
+      "loss": 1.0226,
+      "step": 2682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6357405781745911,
+      "learning_rate": 7.523856617854179e-06,
+      "loss": 1.0491,
+      "step": 2683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6664731502532959,
+      "learning_rate": 7.522024212721233e-06,
+      "loss": 1.0424,
+      "step": 2684
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6464449763298035,
+      "learning_rate": 7.520191353157042e-06,
+      "loss": 1.0373,
+      "step": 2685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3402069807052612,
+      "learning_rate": 7.518358039491863e-06,
+      "loss": 0.8622,
+      "step": 2686
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2244513034820557,
+      "learning_rate": 7.516524272056032e-06,
+      "loss": 0.7667,
+      "step": 2687
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3732932806015015,
+      "learning_rate": 7.514690051179965e-06,
+      "loss": 0.8652,
+      "step": 2688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.251632571220398,
+      "learning_rate": 7.512855377194162e-06,
+      "loss": 0.8279,
+      "step": 2689
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.240238904953003,
+      "learning_rate": 7.511020250429202e-06,
+      "loss": 0.8217,
+      "step": 2690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.252609133720398,
+      "learning_rate": 7.509184671215751e-06,
+      "loss": 0.8483,
+      "step": 2691
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1815812587738037,
+      "learning_rate": 7.507348639884553e-06,
+      "loss": 0.7572,
+      "step": 2692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2822682857513428,
+      "learning_rate": 7.50551215676643e-06,
+      "loss": 0.8733,
+      "step": 2693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8870078325271606,
+      "learning_rate": 7.50367522219229e-06,
+      "loss": 1.0597,
+      "step": 2694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.296649694442749,
+      "learning_rate": 7.501837836493124e-06,
+      "loss": 0.8857,
+      "step": 2695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3060263395309448,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.821,
+      "step": 2696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7347646355628967,
+      "learning_rate": 7.498161713044069e-06,
+      "loss": 1.0441,
+      "step": 2697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2044130563735962,
+      "learning_rate": 7.496322975956562e-06,
+      "loss": 0.8822,
+      "step": 2698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7162137627601624,
+      "learning_rate": 7.494483789068791e-06,
+      "loss": 1.0463,
+      "step": 2699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3431298732757568,
+      "learning_rate": 7.492644152712155e-06,
+      "loss": 0.7766,
+      "step": 2700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2916085720062256,
+      "learning_rate": 7.490804067218124e-06,
+      "loss": 0.8361,
+      "step": 2701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2451971769332886,
+      "learning_rate": 7.488963532918255e-06,
+      "loss": 0.8682,
+      "step": 2702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2836402654647827,
+      "learning_rate": 7.487122550144187e-06,
+      "loss": 0.8509,
+      "step": 2703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3267080783843994,
+      "learning_rate": 7.4852811192276375e-06,
+      "loss": 0.8728,
+      "step": 2704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2258824110031128,
+      "learning_rate": 7.483439240500403e-06,
+      "loss": 0.8394,
+      "step": 2705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.326572299003601,
+      "learning_rate": 7.481596914294363e-06,
+      "loss": 0.8237,
+      "step": 2706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.291773796081543,
+      "learning_rate": 7.479754140941477e-06,
+      "loss": 0.8369,
+      "step": 2707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2916988134384155,
+      "learning_rate": 7.477910920773789e-06,
+      "loss": 0.863,
+      "step": 2708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3066625595092773,
+      "learning_rate": 7.476067254123415e-06,
+      "loss": 0.8569,
+      "step": 2709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3168648481369019,
+      "learning_rate": 7.474223141322559e-06,
+      "loss": 0.8168,
+      "step": 2710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8290144801139832,
+      "learning_rate": 7.472378582703501e-06,
+      "loss": 1.0375,
+      "step": 2711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2481895685195923,
+      "learning_rate": 7.470533578598605e-06,
+      "loss": 0.8797,
+      "step": 2712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2099411487579346,
+      "learning_rate": 7.468688129340314e-06,
+      "loss": 0.8421,
+      "step": 2713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.21030592918396,
+      "learning_rate": 7.466842235261147e-06,
+      "loss": 0.8101,
+      "step": 2714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3270996809005737,
+      "learning_rate": 7.464995896693709e-06,
+      "loss": 0.8962,
+      "step": 2715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2403401136398315,
+      "learning_rate": 7.4631491139706815e-06,
+      "loss": 0.7612,
+      "step": 2716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2276872396469116,
+      "learning_rate": 7.46130188742483e-06,
+      "loss": 0.7896,
+      "step": 2717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3421711921691895,
+      "learning_rate": 7.459454217388993e-06,
+      "loss": 0.8541,
+      "step": 2718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7647614479064941,
+      "learning_rate": 7.457606104196095e-06,
+      "loss": 1.0554,
+      "step": 2719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2093449831008911,
+      "learning_rate": 7.455757548179141e-06,
+      "loss": 0.7632,
+      "step": 2720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2943353652954102,
+      "learning_rate": 7.453908549671211e-06,
+      "loss": 0.8792,
+      "step": 2721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.222739815711975,
+      "learning_rate": 7.452059109005468e-06,
+      "loss": 0.8318,
+      "step": 2722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.280491590499878,
+      "learning_rate": 7.4502092265151515e-06,
+      "loss": 0.8806,
+      "step": 2723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.368239164352417,
+      "learning_rate": 7.448358902533583e-06,
+      "loss": 0.8091,
+      "step": 2724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.283385157585144,
+      "learning_rate": 7.4465081373941684e-06,
+      "loss": 0.8502,
+      "step": 2725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2828912734985352,
+      "learning_rate": 7.4446569314303815e-06,
+      "loss": 0.8144,
+      "step": 2726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.252037525177002,
+      "learning_rate": 7.442805284975785e-06,
+      "loss": 0.8483,
+      "step": 2727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3137975931167603,
+      "learning_rate": 7.440953198364019e-06,
+      "loss": 0.8061,
+      "step": 2728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3000609874725342,
+      "learning_rate": 7.4391006719288005e-06,
+      "loss": 0.8536,
+      "step": 2729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3107235431671143,
+      "learning_rate": 7.4372477060039274e-06,
+      "loss": 0.8466,
+      "step": 2730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.305612325668335,
+      "learning_rate": 7.435394300923275e-06,
+      "loss": 0.8399,
+      "step": 2731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2579092979431152,
+      "learning_rate": 7.433540457020803e-06,
+      "loss": 0.897,
+      "step": 2732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.289204716682434,
+      "learning_rate": 7.431686174630544e-06,
+      "loss": 0.8296,
+      "step": 2733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2825934886932373,
+      "learning_rate": 7.429831454086614e-06,
+      "loss": 0.812,
+      "step": 2734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.423568844795227,
+      "learning_rate": 7.427976295723203e-06,
+      "loss": 0.8419,
+      "step": 2735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7363172173500061,
+      "learning_rate": 7.426120699874584e-06,
+      "loss": 1.0316,
+      "step": 2736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2591129541397095,
+      "learning_rate": 7.424264666875111e-06,
+      "loss": 0.8653,
+      "step": 2737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.303109049797058,
+      "learning_rate": 7.422408197059209e-06,
+      "loss": 0.8071,
+      "step": 2738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3309978246688843,
+      "learning_rate": 7.420551290761388e-06,
+      "loss": 0.7806,
+      "step": 2739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6466371417045593,
+      "learning_rate": 7.418693948316236e-06,
+      "loss": 1.0442,
+      "step": 2740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2396917343139648,
+      "learning_rate": 7.416836170058417e-06,
+      "loss": 0.875,
+      "step": 2741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2882047891616821,
+      "learning_rate": 7.414977956322676e-06,
+      "loss": 0.859,
+      "step": 2742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2481006383895874,
+      "learning_rate": 7.4131193074438335e-06,
+      "loss": 0.8165,
+      "step": 2743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.250991702079773,
+      "learning_rate": 7.411260223756793e-06,
+      "loss": 0.8406,
+      "step": 2744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2338812351226807,
+      "learning_rate": 7.409400705596532e-06,
+      "loss": 0.8488,
+      "step": 2745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.267181158065796,
+      "learning_rate": 7.4075407532981104e-06,
+      "loss": 0.8571,
+      "step": 2746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2468432188034058,
+      "learning_rate": 7.405680367196661e-06,
+      "loss": 0.8345,
+      "step": 2747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2891825437545776,
+      "learning_rate": 7.403819547627399e-06,
+      "loss": 0.8235,
+      "step": 2748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7178151607513428,
+      "learning_rate": 7.401958294925615e-06,
+      "loss": 1.0446,
+      "step": 2749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2175098657608032,
+      "learning_rate": 7.400096609426684e-06,
+      "loss": 0.8238,
+      "step": 2750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6611570715904236,
+      "learning_rate": 7.3982344914660464e-06,
+      "loss": 1.0258,
+      "step": 2751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.313321590423584,
+      "learning_rate": 7.396371941379233e-06,
+      "loss": 0.89,
+      "step": 2752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3327058553695679,
+      "learning_rate": 7.394508959501847e-06,
+      "loss": 0.8732,
+      "step": 2753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2531752586364746,
+      "learning_rate": 7.392645546169571e-06,
+      "loss": 0.8114,
+      "step": 2754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3157471418380737,
+      "learning_rate": 7.390781701718161e-06,
+      "loss": 0.8578,
+      "step": 2755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2650245428085327,
+      "learning_rate": 7.388917426483458e-06,
+      "loss": 0.8347,
+      "step": 2756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2358821630477905,
+      "learning_rate": 7.387052720801373e-06,
+      "loss": 0.8667,
+      "step": 2757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9056904315948486,
+      "learning_rate": 7.385187585007902e-06,
+      "loss": 1.0544,
+      "step": 2758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2540171146392822,
+      "learning_rate": 7.383322019439112e-06,
+      "loss": 0.8452,
+      "step": 2759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.251724362373352,
+      "learning_rate": 7.381456024431149e-06,
+      "loss": 0.788,
+      "step": 2760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6260056495666504,
+      "learning_rate": 7.379589600320242e-06,
+      "loss": 1.016,
+      "step": 2761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2491825819015503,
+      "learning_rate": 7.377722747442689e-06,
+      "loss": 0.8503,
+      "step": 2762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3449183702468872,
+      "learning_rate": 7.375855466134871e-06,
+      "loss": 0.8393,
+      "step": 2763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3008158206939697,
+      "learning_rate": 7.373987756733243e-06,
+      "loss": 0.8225,
+      "step": 2764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3429607152938843,
+      "learning_rate": 7.372119619574339e-06,
+      "loss": 0.8934,
+      "step": 2765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2604460716247559,
+      "learning_rate": 7.370251054994772e-06,
+      "loss": 0.8016,
+      "step": 2766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2944107055664062,
+      "learning_rate": 7.368382063331225e-06,
+      "loss": 0.8622,
+      "step": 2767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0412492752075195,
+      "learning_rate": 7.366512644920465e-06,
+      "loss": 1.0578,
+      "step": 2768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2474303245544434,
+      "learning_rate": 7.364642800099333e-06,
+      "loss": 0.8273,
+      "step": 2769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2665696144104004,
+      "learning_rate": 7.362772529204748e-06,
+      "loss": 0.737,
+      "step": 2770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2254283428192139,
+      "learning_rate": 7.360901832573704e-06,
+      "loss": 0.847,
+      "step": 2771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2413480281829834,
+      "learning_rate": 7.3590307105432725e-06,
+      "loss": 0.8446,
+      "step": 2772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3049520254135132,
+      "learning_rate": 7.357159163450603e-06,
+      "loss": 0.8661,
+      "step": 2773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8844630122184753,
+      "learning_rate": 7.35528719163292e-06,
+      "loss": 1.0377,
+      "step": 2774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.295513391494751,
+      "learning_rate": 7.353414795427524e-06,
+      "loss": 0.8714,
+      "step": 2775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3031649589538574,
+      "learning_rate": 7.3515419751717945e-06,
+      "loss": 0.8697,
+      "step": 2776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3365111351013184,
+      "learning_rate": 7.349668731203183e-06,
+      "loss": 0.8431,
+      "step": 2777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2947700023651123,
+      "learning_rate": 7.347795063859226e-06,
+      "loss": 0.8591,
+      "step": 2778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2917511463165283,
+      "learning_rate": 7.345920973477524e-06,
+      "loss": 0.8408,
+      "step": 2779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2240729331970215,
+      "learning_rate": 7.344046460395764e-06,
+      "loss": 0.779,
+      "step": 2780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.251871109008789,
+      "learning_rate": 7.342171524951704e-06,
+      "loss": 0.8725,
+      "step": 2781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7196139097213745,
+      "learning_rate": 7.340296167483178e-06,
+      "loss": 1.0729,
+      "step": 2782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.301884412765503,
+      "learning_rate": 7.338420388328102e-06,
+      "loss": 0.7694,
+      "step": 2783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3317946195602417,
+      "learning_rate": 7.336544187824459e-06,
+      "loss": 0.8035,
+      "step": 2784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6397833824157715,
+      "learning_rate": 7.334667566310316e-06,
+      "loss": 1.0646,
+      "step": 2785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.653681755065918,
+      "learning_rate": 7.332790524123809e-06,
+      "loss": 1.0176,
+      "step": 2786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3156651258468628,
+      "learning_rate": 7.330913061603158e-06,
+      "loss": 0.8504,
+      "step": 2787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6595556139945984,
+      "learning_rate": 7.329035179086646e-06,
+      "loss": 1.0577,
+      "step": 2788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6535671949386597,
+      "learning_rate": 7.3271568769126456e-06,
+      "loss": 1.049,
+      "step": 2789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2775517702102661,
+      "learning_rate": 7.325278155419598e-06,
+      "loss": 0.8359,
+      "step": 2790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3147900104522705,
+      "learning_rate": 7.32339901494602e-06,
+      "loss": 0.892,
+      "step": 2791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6936156153678894,
+      "learning_rate": 7.321519455830502e-06,
+      "loss": 1.0267,
+      "step": 2792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2216532230377197,
+      "learning_rate": 7.319639478411719e-06,
+      "loss": 0.8084,
+      "step": 2793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7468457818031311,
+      "learning_rate": 7.317759083028408e-06,
+      "loss": 1.0556,
+      "step": 2794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2221174240112305,
+      "learning_rate": 7.315878270019395e-06,
+      "loss": 0.791,
+      "step": 2795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4055688381195068,
+      "learning_rate": 7.313997039723568e-06,
+      "loss": 0.8502,
+      "step": 2796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6537927985191345,
+      "learning_rate": 7.312115392479904e-06,
+      "loss": 1.0446,
+      "step": 2797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.277134895324707,
+      "learning_rate": 7.310233328627441e-06,
+      "loss": 0.8612,
+      "step": 2798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2538710832595825,
+      "learning_rate": 7.3083508485053015e-06,
+      "loss": 0.804,
+      "step": 2799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2278568744659424,
+      "learning_rate": 7.306467952452681e-06,
+      "loss": 0.8538,
+      "step": 2800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.288352608680725,
+      "learning_rate": 7.304584640808849e-06,
+      "loss": 0.8348,
+      "step": 2801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2816497087478638,
+      "learning_rate": 7.30270091391315e-06,
+      "loss": 0.8407,
+      "step": 2802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7010305523872375,
+      "learning_rate": 7.300816772105004e-06,
+      "loss": 1.0348,
+      "step": 2803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3167706727981567,
+      "learning_rate": 7.298932215723904e-06,
+      "loss": 0.8054,
+      "step": 2804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2728781700134277,
+      "learning_rate": 7.297047245109421e-06,
+      "loss": 0.8612,
+      "step": 2805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3781365156173706,
+      "learning_rate": 7.295161860601198e-06,
+      "loss": 0.8328,
+      "step": 2806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1988152265548706,
+      "learning_rate": 7.293276062538952e-06,
+      "loss": 0.822,
+      "step": 2807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3148807287216187,
+      "learning_rate": 7.291389851262476e-06,
+      "loss": 0.8553,
+      "step": 2808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.238926649093628,
+      "learning_rate": 7.289503227111641e-06,
+      "loss": 0.7961,
+      "step": 2809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2738085985183716,
+      "learning_rate": 7.287616190426383e-06,
+      "loss": 0.8254,
+      "step": 2810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.259247064590454,
+      "learning_rate": 7.285728741546721e-06,
+      "loss": 0.8477,
+      "step": 2811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2521620988845825,
+      "learning_rate": 7.283840880812745e-06,
+      "loss": 0.8175,
+      "step": 2812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2805122137069702,
+      "learning_rate": 7.2819526085646185e-06,
+      "loss": 0.8559,
+      "step": 2813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2745094299316406,
+      "learning_rate": 7.2800639251425805e-06,
+      "loss": 0.8582,
+      "step": 2814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2234736680984497,
+      "learning_rate": 7.2781748308869456e-06,
+      "loss": 0.8483,
+      "step": 2815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.373611569404602,
+      "learning_rate": 7.2762853261380965e-06,
+      "loss": 0.9046,
+      "step": 2816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.730300784111023,
+      "learning_rate": 7.274395411236496e-06,
+      "loss": 1.0443,
+      "step": 2817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2971150875091553,
+      "learning_rate": 7.272505086522681e-06,
+      "loss": 0.7738,
+      "step": 2818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2584205865859985,
+      "learning_rate": 7.270614352337255e-06,
+      "loss": 0.8478,
+      "step": 2819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2412110567092896,
+      "learning_rate": 7.268723209020905e-06,
+      "loss": 0.8449,
+      "step": 2820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2507939338684082,
+      "learning_rate": 7.266831656914382e-06,
+      "loss": 0.8273,
+      "step": 2821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2334891557693481,
+      "learning_rate": 7.264939696358521e-06,
+      "loss": 0.7899,
+      "step": 2822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.267279028892517,
+      "learning_rate": 7.2630473276942214e-06,
+      "loss": 0.8337,
+      "step": 2823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2745047807693481,
+      "learning_rate": 7.261154551262459e-06,
+      "loss": 0.8213,
+      "step": 2824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2685693502426147,
+      "learning_rate": 7.259261367404287e-06,
+      "loss": 0.8593,
+      "step": 2825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2485283613204956,
+      "learning_rate": 7.257367776460829e-06,
+      "loss": 0.8397,
+      "step": 2826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.274105429649353,
+      "learning_rate": 7.25547377877328e-06,
+      "loss": 0.7968,
+      "step": 2827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3032187223434448,
+      "learning_rate": 7.2535793746829095e-06,
+      "loss": 0.8155,
+      "step": 2828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8225850462913513,
+      "learning_rate": 7.251684564531063e-06,
+      "loss": 1.0453,
+      "step": 2829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2425453662872314,
+      "learning_rate": 7.249789348659157e-06,
+      "loss": 0.813,
+      "step": 2830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.281583547592163,
+      "learning_rate": 7.247893727408681e-06,
+      "loss": 0.8411,
+      "step": 2831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2589370012283325,
+      "learning_rate": 7.245997701121197e-06,
+      "loss": 0.8993,
+      "step": 2832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3330706357955933,
+      "learning_rate": 7.2441012701383395e-06,
+      "loss": 0.8878,
+      "step": 2833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2288769483566284,
+      "learning_rate": 7.2422044348018205e-06,
+      "loss": 0.8187,
+      "step": 2834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2827693223953247,
+      "learning_rate": 7.24030719545342e-06,
+      "loss": 0.8444,
+      "step": 2835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2558414936065674,
+      "learning_rate": 7.23840955243499e-06,
+      "loss": 0.821,
+      "step": 2836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2846511602401733,
+      "learning_rate": 7.236511506088461e-06,
+      "loss": 0.9047,
+      "step": 2837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2345991134643555,
+      "learning_rate": 7.234613056755831e-06,
+      "loss": 0.8664,
+      "step": 2838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1772981882095337,
+      "learning_rate": 7.232714204779173e-06,
+      "loss": 0.7997,
+      "step": 2839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.248042106628418,
+      "learning_rate": 7.230814950500631e-06,
+      "loss": 0.8082,
+      "step": 2840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.268566370010376,
+      "learning_rate": 7.228915294262423e-06,
+      "loss": 0.8201,
+      "step": 2841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.221276044845581,
+      "learning_rate": 7.2270152364068395e-06,
+      "loss": 0.8664,
+      "step": 2842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.305989384651184,
+      "learning_rate": 7.2251147772762405e-06,
+      "loss": 0.8651,
+      "step": 2843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4160962104797363,
+      "learning_rate": 7.223213917213061e-06,
+      "loss": 0.8693,
+      "step": 2844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3727253675460815,
+      "learning_rate": 7.22131265655981e-06,
+      "loss": 0.7976,
+      "step": 2845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.169970154762268,
+      "learning_rate": 7.219410995659065e-06,
+      "loss": 0.7835,
+      "step": 2846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.271941900253296,
+      "learning_rate": 7.217508934853477e-06,
+      "loss": 0.9052,
+      "step": 2847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2848469018936157,
+      "learning_rate": 7.215606474485768e-06,
+      "loss": 0.8724,
+      "step": 2848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2244559526443481,
+      "learning_rate": 7.213703614898734e-06,
+      "loss": 0.8052,
+      "step": 2849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3597817420959473,
+      "learning_rate": 7.211800356435244e-06,
+      "loss": 0.8734,
+      "step": 2850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2028509378433228,
+      "learning_rate": 7.209896699438232e-06,
+      "loss": 0.7892,
+      "step": 2851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2633638381958008,
+      "learning_rate": 7.207992644250712e-06,
+      "loss": 0.7758,
+      "step": 2852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2669669389724731,
+      "learning_rate": 7.206088191215767e-06,
+      "loss": 0.7999,
+      "step": 2853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2609277963638306,
+      "learning_rate": 7.2041833406765495e-06,
+      "loss": 0.8451,
+      "step": 2854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1574777364730835,
+      "learning_rate": 7.202278092976286e-06,
+      "loss": 1.0647,
+      "step": 2855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7804716229438782,
+      "learning_rate": 7.2003724484582725e-06,
+      "loss": 1.0479,
+      "step": 2856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3288475275039673,
+      "learning_rate": 7.19846640746588e-06,
+      "loss": 0.8187,
+      "step": 2857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2296926975250244,
+      "learning_rate": 7.196559970342547e-06,
+      "loss": 0.8064,
+      "step": 2858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2920323610305786,
+      "learning_rate": 7.1946531374317865e-06,
+      "loss": 0.7922,
+      "step": 2859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3316175937652588,
+      "learning_rate": 7.19274590907718e-06,
+      "loss": 0.842,
+      "step": 2860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.346356987953186,
+      "learning_rate": 7.1908382856223826e-06,
+      "loss": 0.8923,
+      "step": 2861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.252722144126892,
+      "learning_rate": 7.188930267411118e-06,
+      "loss": 0.8156,
+      "step": 2862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2179621458053589,
+      "learning_rate": 7.1870218547871865e-06,
+      "loss": 0.8404,
+      "step": 2863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.260043978691101,
+      "learning_rate": 7.185113048094451e-06,
+      "loss": 0.8335,
+      "step": 2864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2975237369537354,
+      "learning_rate": 7.183203847676853e-06,
+      "loss": 0.8664,
+      "step": 2865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2779159545898438,
+      "learning_rate": 7.1812942538784015e-06,
+      "loss": 0.8144,
+      "step": 2866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.553035020828247,
+      "learning_rate": 7.179384267043177e-06,
+      "loss": 1.0759,
+      "step": 2867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.240421175956726,
+      "learning_rate": 7.177473887515329e-06,
+      "loss": 0.8337,
+      "step": 2868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2476160526275635,
+      "learning_rate": 7.175563115639082e-06,
+      "loss": 0.8273,
+      "step": 2869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3889448642730713,
+      "learning_rate": 7.173651951758727e-06,
+      "loss": 0.7943,
+      "step": 2870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.378594160079956,
+      "learning_rate": 7.171740396218628e-06,
+      "loss": 0.8586,
+      "step": 2871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1363191604614258,
+      "learning_rate": 7.1698284493632184e-06,
+      "loss": 1.0551,
+      "step": 2872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3890388011932373,
+      "learning_rate": 7.167916111537003e-06,
+      "loss": 0.8274,
+      "step": 2873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7164141535758972,
+      "learning_rate": 7.1660033830845565e-06,
+      "loss": 1.0448,
+      "step": 2874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7187297940254211,
+      "learning_rate": 7.164090264350525e-06,
+      "loss": 1.0559,
+      "step": 2875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3597805500030518,
+      "learning_rate": 7.162176755679623e-06,
+      "loss": 0.8101,
+      "step": 2876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2824201583862305,
+      "learning_rate": 7.160262857416635e-06,
+      "loss": 0.8333,
+      "step": 2877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.273154377937317,
+      "learning_rate": 7.158348569906419e-06,
+      "loss": 0.8206,
+      "step": 2878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.344057559967041,
+      "learning_rate": 7.156433893493901e-06,
+      "loss": 0.8553,
+      "step": 2879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3124703168869019,
+      "learning_rate": 7.154518828524077e-06,
+      "loss": 0.8335,
+      "step": 2880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4367212057113647,
+      "learning_rate": 7.152603375342013e-06,
+      "loss": 0.8888,
+      "step": 2881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3517839908599854,
+      "learning_rate": 7.150687534292844e-06,
+      "loss": 0.9008,
+      "step": 2882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2619558572769165,
+      "learning_rate": 7.148771305721779e-06,
+      "loss": 0.8119,
+      "step": 2883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2891145944595337,
+      "learning_rate": 7.146854689974091e-06,
+      "loss": 1.0426,
+      "step": 2884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.362887978553772,
+      "learning_rate": 7.144937687395126e-06,
+      "loss": 0.8256,
+      "step": 2885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8599793910980225,
+      "learning_rate": 7.143020298330301e-06,
+      "loss": 1.0563,
+      "step": 2886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7270867228507996,
+      "learning_rate": 7.141102523125101e-06,
+      "loss": 1.059,
+      "step": 2887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.397987723350525,
+      "learning_rate": 7.139184362125078e-06,
+      "loss": 0.8235,
+      "step": 2888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8787017464637756,
+      "learning_rate": 7.137265815675858e-06,
+      "loss": 1.0568,
+      "step": 2889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3366947174072266,
+      "learning_rate": 7.1353468841231335e-06,
+      "loss": 0.8411,
+      "step": 2890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2684530019760132,
+      "learning_rate": 7.133427567812669e-06,
+      "loss": 0.8848,
+      "step": 2891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0781230926513672,
+      "learning_rate": 7.131507867090297e-06,
+      "loss": 1.0656,
+      "step": 2892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9289823174476624,
+      "learning_rate": 7.129587782301917e-06,
+      "loss": 1.0421,
+      "step": 2893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3171056509017944,
+      "learning_rate": 7.1276673137934995e-06,
+      "loss": 0.8623,
+      "step": 2894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7552684545516968,
+      "learning_rate": 7.125746461911086e-06,
+      "loss": 1.0365,
+      "step": 2895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4269651174545288,
+      "learning_rate": 7.123825227000786e-06,
+      "loss": 0.8214,
+      "step": 2896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3370500802993774,
+      "learning_rate": 7.121903609408776e-06,
+      "loss": 0.8492,
+      "step": 2897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3077919483184814,
+      "learning_rate": 7.119981609481301e-06,
+      "loss": 0.8369,
+      "step": 2898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1545467376708984,
+      "learning_rate": 7.118059227564682e-06,
+      "loss": 1.0316,
+      "step": 2899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3532863855361938,
+      "learning_rate": 7.1161364640052995e-06,
+      "loss": 0.8284,
+      "step": 2900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2683351039886475,
+      "learning_rate": 7.114213319149607e-06,
+      "loss": 0.7722,
+      "step": 2901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3425896167755127,
+      "learning_rate": 7.112289793344129e-06,
+      "loss": 0.8684,
+      "step": 2902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8276830315589905,
+      "learning_rate": 7.110365886935455e-06,
+      "loss": 1.0613,
+      "step": 2903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2106422185897827,
+      "learning_rate": 7.108441600270244e-06,
+      "loss": 0.7814,
+      "step": 2904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.294831395149231,
+      "learning_rate": 7.106516933695224e-06,
+      "loss": 0.8258,
+      "step": 2905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6950622797012329,
+      "learning_rate": 7.104591887557191e-06,
+      "loss": 1.0157,
+      "step": 2906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2780781984329224,
+      "learning_rate": 7.1026664622030095e-06,
+      "loss": 0.8343,
+      "step": 2907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7388509511947632,
+      "learning_rate": 7.100740657979615e-06,
+      "loss": 1.0186,
+      "step": 2908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7474695444107056,
+      "learning_rate": 7.0988144752340045e-06,
+      "loss": 1.0507,
+      "step": 2909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3144596815109253,
+      "learning_rate": 7.09688791431325e-06,
+      "loss": 0.8408,
+      "step": 2910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3160717487335205,
+      "learning_rate": 7.09496097556449e-06,
+      "loss": 0.8265,
+      "step": 2911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2735190391540527,
+      "learning_rate": 7.0930336593349294e-06,
+      "loss": 0.8757,
+      "step": 2912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.274863362312317,
+      "learning_rate": 7.091105965971839e-06,
+      "loss": 0.818,
+      "step": 2913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.290814757347107,
+      "learning_rate": 7.089177895822565e-06,
+      "loss": 0.8852,
+      "step": 2914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.271418571472168,
+      "learning_rate": 7.087249449234514e-06,
+      "loss": 0.8431,
+      "step": 2915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9959194660186768,
+      "learning_rate": 7.0853206265551655e-06,
+      "loss": 1.0358,
+      "step": 2916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.403885841369629,
+      "learning_rate": 7.083391428132062e-06,
+      "loss": 0.8944,
+      "step": 2917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3407607078552246,
+      "learning_rate": 7.08146185431282e-06,
+      "loss": 0.8903,
+      "step": 2918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3266568183898926,
+      "learning_rate": 7.079531905445116e-06,
+      "loss": 0.8585,
+      "step": 2919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3050228357315063,
+      "learning_rate": 7.0776015818767015e-06,
+      "loss": 0.8444,
+      "step": 2920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1946403980255127,
+      "learning_rate": 7.07567088395539e-06,
+      "loss": 0.8547,
+      "step": 2921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2401626110076904,
+      "learning_rate": 7.073739812029063e-06,
+      "loss": 0.8707,
+      "step": 2922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.38413667678833,
+      "learning_rate": 7.071808366445675e-06,
+      "loss": 0.8892,
+      "step": 2923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9676514863967896,
+      "learning_rate": 7.0698765475532425e-06,
+      "loss": 1.0721,
+      "step": 2924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2946784496307373,
+      "learning_rate": 7.067944355699848e-06,
+      "loss": 0.8411,
+      "step": 2925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6598960757255554,
+      "learning_rate": 7.066011791233647e-06,
+      "loss": 1.0344,
+      "step": 2926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3551067113876343,
+      "learning_rate": 7.064078854502856e-06,
+      "loss": 0.821,
+      "step": 2927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.283556342124939,
+      "learning_rate": 7.062145545855763e-06,
+      "loss": 0.8235,
+      "step": 2928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.337345838546753,
+      "learning_rate": 7.060211865640723e-06,
+      "loss": 0.839,
+      "step": 2929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9805741906166077,
+      "learning_rate": 7.058277814206152e-06,
+      "loss": 1.0215,
+      "step": 2930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2460635900497437,
+      "learning_rate": 7.056343391900542e-06,
+      "loss": 0.8531,
+      "step": 2931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6701040267944336,
+      "learning_rate": 7.054408599072444e-06,
+      "loss": 1.0498,
+      "step": 2932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2070411443710327,
+      "learning_rate": 7.05247343607048e-06,
+      "loss": 0.8335,
+      "step": 2933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3234004974365234,
+      "learning_rate": 7.050537903243338e-06,
+      "loss": 0.7907,
+      "step": 2934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.323266863822937,
+      "learning_rate": 7.048602000939772e-06,
+      "loss": 0.858,
+      "step": 2935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.303736686706543,
+      "learning_rate": 7.046665729508601e-06,
+      "loss": 0.8657,
+      "step": 2936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2250422239303589,
+      "learning_rate": 7.044729089298718e-06,
+      "loss": 0.8543,
+      "step": 2937
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2912378311157227,
+      "learning_rate": 7.042792080659068e-06,
+      "loss": 0.8455,
+      "step": 2938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9270570278167725,
+      "learning_rate": 7.040854703938678e-06,
+      "loss": 1.043,
+      "step": 2939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.357468605041504,
+      "learning_rate": 7.038916959486632e-06,
+      "loss": 0.796,
+      "step": 2940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3452321290969849,
+      "learning_rate": 7.036978847652083e-06,
+      "loss": 0.8934,
+      "step": 2941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.285821557044983,
+      "learning_rate": 7.0350403687842495e-06,
+      "loss": 0.8273,
+      "step": 2942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2708114385604858,
+      "learning_rate": 7.0331015232324175e-06,
+      "loss": 0.8342,
+      "step": 2943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2529218196868896,
+      "learning_rate": 7.0311623113459375e-06,
+      "loss": 0.8336,
+      "step": 2944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2447495460510254,
+      "learning_rate": 7.029222733474227e-06,
+      "loss": 0.8595,
+      "step": 2945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8295711874961853,
+      "learning_rate": 7.027282789966766e-06,
+      "loss": 1.04,
+      "step": 2946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3592381477355957,
+      "learning_rate": 7.0253424811731095e-06,
+      "loss": 0.8562,
+      "step": 2947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2488857507705688,
+      "learning_rate": 7.023401807442868e-06,
+      "loss": 0.7978,
+      "step": 2948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.402176022529602,
+      "learning_rate": 7.021460769125723e-06,
+      "loss": 0.8527,
+      "step": 2949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.300858736038208,
+      "learning_rate": 7.01951936657142e-06,
+      "loss": 0.8085,
+      "step": 2950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.207065463066101,
+      "learning_rate": 7.0175776001297735e-06,
+      "loss": 0.8234,
+      "step": 2951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2630831003189087,
+      "learning_rate": 7.015635470150658e-06,
+      "loss": 0.8372,
+      "step": 2952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2569385766983032,
+      "learning_rate": 7.013692976984018e-06,
+      "loss": 0.7932,
+      "step": 2953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2650424242019653,
+      "learning_rate": 7.0117501209798586e-06,
+      "loss": 0.811,
+      "step": 2954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.249558448791504,
+      "learning_rate": 7.0098069024882595e-06,
+      "loss": 0.8231,
+      "step": 2955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3177170753479004,
+      "learning_rate": 7.007863321859356e-06,
+      "loss": 0.8094,
+      "step": 2956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3059971332550049,
+      "learning_rate": 7.005919379443353e-06,
+      "loss": 0.8043,
+      "step": 2957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0638045072555542,
+      "learning_rate": 7.003975075590518e-06,
+      "loss": 1.0475,
+      "step": 2958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2583049535751343,
+      "learning_rate": 7.0020304106511905e-06,
+      "loss": 0.8117,
+      "step": 2959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6596599817276001,
+      "learning_rate": 7.0000853849757655e-06,
+      "loss": 1.0446,
+      "step": 2960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3383564949035645,
+      "learning_rate": 6.9981399989147105e-06,
+      "loss": 0.8569,
+      "step": 2961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2627750635147095,
+      "learning_rate": 6.996194252818552e-06,
+      "loss": 0.8053,
+      "step": 2962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9620859026908875,
+      "learning_rate": 6.99424814703789e-06,
+      "loss": 1.0522,
+      "step": 2963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.177931308746338,
+      "learning_rate": 6.9923016819233765e-06,
+      "loss": 0.8526,
+      "step": 2964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2496998310089111,
+      "learning_rate": 6.990354857825742e-06,
+      "loss": 0.8128,
+      "step": 2965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7417113780975342,
+      "learning_rate": 6.988407675095772e-06,
+      "loss": 1.0228,
+      "step": 2966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2464687824249268,
+      "learning_rate": 6.986460134084321e-06,
+      "loss": 0.8275,
+      "step": 2967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.25774347782135,
+      "learning_rate": 6.984512235142307e-06,
+      "loss": 0.8703,
+      "step": 2968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2951887845993042,
+      "learning_rate": 6.982563978620711e-06,
+      "loss": 0.8421,
+      "step": 2969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2629098892211914,
+      "learning_rate": 6.980615364870581e-06,
+      "loss": 0.8266,
+      "step": 2970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3346569538116455,
+      "learning_rate": 6.978666394243028e-06,
+      "loss": 0.7692,
+      "step": 2971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9033862948417664,
+      "learning_rate": 6.976717067089228e-06,
+      "loss": 1.0332,
+      "step": 2972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5153002738952637,
+      "learning_rate": 6.974767383760419e-06,
+      "loss": 0.8544,
+      "step": 2973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.657181441783905,
+      "learning_rate": 6.972817344607907e-06,
+      "loss": 1.0493,
+      "step": 2974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6665467023849487,
+      "learning_rate": 6.970866949983059e-06,
+      "loss": 1.051,
+      "step": 2975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.669784665107727,
+      "learning_rate": 6.968916200237308e-06,
+      "loss": 1.0305,
+      "step": 2976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4082361459732056,
+      "learning_rate": 6.966965095722148e-06,
+      "loss": 0.7895,
+      "step": 2977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4106746912002563,
+      "learning_rate": 6.965013636789141e-06,
+      "loss": 0.8617,
+      "step": 2978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.33734929561615,
+      "learning_rate": 6.9630618237899095e-06,
+      "loss": 0.8227,
+      "step": 2979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3657679557800293,
+      "learning_rate": 6.961109657076144e-06,
+      "loss": 0.9079,
+      "step": 2980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2157481908798218,
+      "learning_rate": 6.959157136999593e-06,
+      "loss": 0.7853,
+      "step": 2981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3641479015350342,
+      "learning_rate": 6.957204263912071e-06,
+      "loss": 0.8301,
+      "step": 2982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.287866473197937,
+      "learning_rate": 6.955251038165459e-06,
+      "loss": 0.8253,
+      "step": 2983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2531886100769043,
+      "learning_rate": 6.953297460111699e-06,
+      "loss": 0.822,
+      "step": 2984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0112714767456055,
+      "learning_rate": 6.951343530102794e-06,
+      "loss": 1.0662,
+      "step": 2985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2130917310714722,
+      "learning_rate": 6.9493892484908165e-06,
+      "loss": 0.8048,
+      "step": 2986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7068005800247192,
+      "learning_rate": 6.947434615627897e-06,
+      "loss": 1.0362,
+      "step": 2987
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3282498121261597,
+      "learning_rate": 6.9454796318662345e-06,
+      "loss": 0.8822,
+      "step": 2988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6487564444541931,
+      "learning_rate": 6.943524297558082e-06,
+      "loss": 1.0549,
+      "step": 2989
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3548694849014282,
+      "learning_rate": 6.941568613055765e-06,
+      "loss": 0.8372,
+      "step": 2990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7665804028511047,
+      "learning_rate": 6.93961257871167e-06,
+      "loss": 1.0263,
+      "step": 2991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3742420673370361,
+      "learning_rate": 6.937656194878245e-06,
+      "loss": 0.8808,
+      "step": 2992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.279531478881836,
+      "learning_rate": 6.935699461907999e-06,
+      "loss": 0.8219,
+      "step": 2993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.275760531425476,
+      "learning_rate": 6.933742380153507e-06,
+      "loss": 0.8626,
+      "step": 2994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2896785736083984,
+      "learning_rate": 6.931784949967407e-06,
+      "loss": 0.8655,
+      "step": 2995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3236750364303589,
+      "learning_rate": 6.929827171702399e-06,
+      "loss": 0.9035,
+      "step": 2996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.301511526107788,
+      "learning_rate": 6.927869045711244e-06,
+      "loss": 0.8007,
+      "step": 2997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2749481201171875,
+      "learning_rate": 6.9259105723467675e-06,
+      "loss": 0.8008,
+      "step": 2998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3246076107025146,
+      "learning_rate": 6.923951751961859e-06,
+      "loss": 0.8378,
+      "step": 2999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2497378587722778,
+      "learning_rate": 6.921992584909467e-06,
+      "loss": 0.8763,
+      "step": 3000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2708700895309448,
+      "learning_rate": 6.920033071542604e-06,
+      "loss": 0.8871,
+      "step": 3001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.235849142074585,
+      "learning_rate": 6.9180732122143465e-06,
+      "loss": 0.7944,
+      "step": 3002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.267148733139038,
+      "learning_rate": 6.916113007277832e-06,
+      "loss": 0.8337,
+      "step": 3003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2565304040908813,
+      "learning_rate": 6.91415245708626e-06,
+      "loss": 0.8805,
+      "step": 3004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3273481130599976,
+      "learning_rate": 6.912191561992891e-06,
+      "loss": 0.8174,
+      "step": 3005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2829430103302002,
+      "learning_rate": 6.91023032235105e-06,
+      "loss": 0.8405,
+      "step": 3006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8448411226272583,
+      "learning_rate": 6.908268738514125e-06,
+      "loss": 1.0477,
+      "step": 3007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2474268674850464,
+      "learning_rate": 6.906306810835561e-06,
+      "loss": 0.8523,
+      "step": 3008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2695434093475342,
+      "learning_rate": 6.904344539668872e-06,
+      "loss": 0.8387,
+      "step": 3009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4005435705184937,
+      "learning_rate": 6.9023819253676264e-06,
+      "loss": 0.8845,
+      "step": 3010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7055203914642334,
+      "learning_rate": 6.90041896828546e-06,
+      "loss": 1.0486,
+      "step": 3011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6531649231910706,
+      "learning_rate": 6.8984556687760675e-06,
+      "loss": 1.0548,
+      "step": 3012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2599622011184692,
+      "learning_rate": 6.896492027193209e-06,
+      "loss": 0.893,
+      "step": 3013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2699707746505737,
+      "learning_rate": 6.894528043890699e-06,
+      "loss": 0.8608,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6576747298240662,
+      "learning_rate": 6.892563719222422e-06,
+      "loss": 1.0537,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6495311260223389,
+      "learning_rate": 6.8905990535423175e-06,
+      "loss": 1.0581,
+      "step": 3016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1897445917129517,
+      "learning_rate": 6.888634047204391e-06,
+      "loss": 0.815,
+      "step": 3017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6119108200073242,
+      "learning_rate": 6.886668700562706e-06,
+      "loss": 1.027,
+      "step": 3018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2855043411254883,
+      "learning_rate": 6.8847030139713885e-06,
+      "loss": 0.8332,
+      "step": 3019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3190646171569824,
+      "learning_rate": 6.882736987784628e-06,
+      "loss": 0.8565,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2614656686782837,
+      "learning_rate": 6.880770622356672e-06,
+      "loss": 0.7552,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2529594898223877,
+      "learning_rate": 6.878803918041828e-06,
+      "loss": 0.856,
+      "step": 3022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3334892988204956,
+      "learning_rate": 6.876836875194471e-06,
+      "loss": 0.8463,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8504834175109863,
+      "learning_rate": 6.874869494169031e-06,
+      "loss": 1.0697,
+      "step": 3024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2780265808105469,
+      "learning_rate": 6.87290177532e-06,
+      "loss": 0.8789,
+      "step": 3025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1945598125457764,
+      "learning_rate": 6.8709337190019335e-06,
+      "loss": 0.8059,
+      "step": 3026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2995221614837646,
+      "learning_rate": 6.8689653255694426e-06,
+      "loss": 0.8177,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2728736400604248,
+      "learning_rate": 6.866996595377206e-06,
+      "loss": 0.8466,
+      "step": 3028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.252349853515625,
+      "learning_rate": 6.865027528779958e-06,
+      "loss": 0.8519,
+      "step": 3029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2936235666275024,
+      "learning_rate": 6.863058126132496e-06,
+      "loss": 0.8178,
+      "step": 3030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2490698099136353,
+      "learning_rate": 6.861088387789676e-06,
+      "loss": 0.8567,
+      "step": 3031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8257250189781189,
+      "learning_rate": 6.8591183141064156e-06,
+      "loss": 1.0437,
+      "step": 3032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.270664930343628,
+      "learning_rate": 6.8571479054376945e-06,
+      "loss": 0.7866,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2891241312026978,
+      "learning_rate": 6.855177162138549e-06,
+      "loss": 0.8676,
+      "step": 3034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2825944423675537,
+      "learning_rate": 6.853206084564077e-06,
+      "loss": 0.8187,
+      "step": 3035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3212093114852905,
+      "learning_rate": 6.85123467306944e-06,
+      "loss": 0.8424,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6813771724700928,
+      "learning_rate": 6.849262928009857e-06,
+      "loss": 1.0303,
+      "step": 3037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2969274520874023,
+      "learning_rate": 6.847290849740604e-06,
+      "loss": 0.87,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3099604845046997,
+      "learning_rate": 6.845318438617023e-06,
+      "loss": 0.8791,
+      "step": 3039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6458196043968201,
+      "learning_rate": 6.843345694994512e-06,
+      "loss": 1.0506,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2473407983779907,
+      "learning_rate": 6.841372619228531e-06,
+      "loss": 0.8192,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2987524271011353,
+      "learning_rate": 6.839399211674598e-06,
+      "loss": 0.8198,
+      "step": 3042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.27738618850708,
+      "learning_rate": 6.837425472688291e-06,
+      "loss": 0.8633,
+      "step": 3043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2495741844177246,
+      "learning_rate": 6.835451402625251e-06,
+      "loss": 0.8804,
+      "step": 3044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.277329444885254,
+      "learning_rate": 6.833477001841172e-06,
+      "loss": 0.8453,
+      "step": 3045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6944372653961182,
+      "learning_rate": 6.831502270691816e-06,
+      "loss": 1.032,
+      "step": 3046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2545708417892456,
+      "learning_rate": 6.8295272095329975e-06,
+      "loss": 0.8723,
+      "step": 3047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.288841962814331,
+      "learning_rate": 6.827551818720594e-06,
+      "loss": 0.8742,
+      "step": 3048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2532033920288086,
+      "learning_rate": 6.82557609861054e-06,
+      "loss": 0.8469,
+      "step": 3049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.244672179222107,
+      "learning_rate": 6.823600049558834e-06,
+      "loss": 0.894,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2336561679840088,
+      "learning_rate": 6.821623671921529e-06,
+      "loss": 0.7992,
+      "step": 3051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2848756313323975,
+      "learning_rate": 6.819646966054737e-06,
+      "loss": 0.9027,
+      "step": 3052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2421746253967285,
+      "learning_rate": 6.8176699323146335e-06,
+      "loss": 0.8983,
+      "step": 3053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2299535274505615,
+      "learning_rate": 6.815692571057451e-06,
+      "loss": 0.7725,
+      "step": 3054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2635186910629272,
+      "learning_rate": 6.813714882639477e-06,
+      "loss": 0.8908,
+      "step": 3055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2301254272460938,
+      "learning_rate": 6.811736867417065e-06,
+      "loss": 0.8196,
+      "step": 3056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2781635522842407,
+      "learning_rate": 6.809758525746623e-06,
+      "loss": 0.827,
+      "step": 3057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2145894765853882,
+      "learning_rate": 6.8077798579846186e-06,
+      "loss": 0.8446,
+      "step": 3058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7328004240989685,
+      "learning_rate": 6.8058008644875774e-06,
+      "loss": 1.0443,
+      "step": 3059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.298415184020996,
+      "learning_rate": 6.803821545612086e-06,
+      "loss": 0.8245,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.265503168106079,
+      "learning_rate": 6.8018419017147875e-06,
+      "loss": 0.858,
+      "step": 3061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.305755853652954,
+      "learning_rate": 6.799861933152385e-06,
+      "loss": 0.8615,
+      "step": 3062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.24689781665802,
+      "learning_rate": 6.7978816402816385e-06,
+      "loss": 0.8727,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2133574485778809,
+      "learning_rate": 6.795901023459367e-06,
+      "loss": 0.8483,
+      "step": 3064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6986088752746582,
+      "learning_rate": 6.793920083042449e-06,
+      "loss": 1.0441,
+      "step": 3065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.285955548286438,
+      "learning_rate": 6.791938819387821e-06,
+      "loss": 0.8046,
+      "step": 3066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2928717136383057,
+      "learning_rate": 6.789957232852475e-06,
+      "loss": 0.8543,
+      "step": 3067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6150689125061035,
+      "learning_rate": 6.787975323793465e-06,
+      "loss": 1.0479,
+      "step": 3068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2755489349365234,
+      "learning_rate": 6.785993092567903e-06,
+      "loss": 0.8511,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2439240217208862,
+      "learning_rate": 6.784010539532956e-06,
+      "loss": 0.8067,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6730513572692871,
+      "learning_rate": 6.782027665045849e-06,
+      "loss": 1.047,
+      "step": 3071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2281417846679688,
+      "learning_rate": 6.780044469463869e-06,
+      "loss": 0.8687,
+      "step": 3072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2863178253173828,
+      "learning_rate": 6.778060953144358e-06,
+      "loss": 0.8452,
+      "step": 3073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6649810075759888,
+      "learning_rate": 6.776077116444713e-06,
+      "loss": 1.0586,
+      "step": 3074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3012428283691406,
+      "learning_rate": 6.7740929597223985e-06,
+      "loss": 0.8796,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2729579210281372,
+      "learning_rate": 6.772108483334921e-06,
+      "loss": 0.838,
+      "step": 3076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.707554817199707,
+      "learning_rate": 6.770123687639862e-06,
+      "loss": 1.0292,
+      "step": 3077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2899081707000732,
+      "learning_rate": 6.768138572994849e-06,
+      "loss": 0.8406,
+      "step": 3078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2623339891433716,
+      "learning_rate": 6.766153139757569e-06,
+      "loss": 0.8245,
+      "step": 3079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2251836061477661,
+      "learning_rate": 6.764167388285767e-06,
+      "loss": 0.8572,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.192222237586975,
+      "learning_rate": 6.762181318937248e-06,
+      "loss": 0.7974,
+      "step": 3081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2493019104003906,
+      "learning_rate": 6.7601949320698725e-06,
+      "loss": 0.8982,
+      "step": 3082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2603614330291748,
+      "learning_rate": 6.758208228041558e-06,
+      "loss": 0.8564,
+      "step": 3083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2969081401824951,
+      "learning_rate": 6.7562212072102736e-06,
+      "loss": 0.819,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2966957092285156,
+      "learning_rate": 6.754233869934057e-06,
+      "loss": 0.8254,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2109451293945312,
+      "learning_rate": 6.752246216570995e-06,
+      "loss": 0.8096,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7776377201080322,
+      "learning_rate": 6.7502582474792345e-06,
+      "loss": 1.0278,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2620656490325928,
+      "learning_rate": 6.7482699630169735e-06,
+      "loss": 0.8203,
+      "step": 3088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2580486536026,
+      "learning_rate": 6.7462813635424775e-06,
+      "loss": 0.7668,
+      "step": 3089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2978641986846924,
+      "learning_rate": 6.744292449414056e-06,
+      "loss": 0.8422,
+      "step": 3090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.247689962387085,
+      "learning_rate": 6.7423032209900875e-06,
+      "loss": 0.8137,
+      "step": 3091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2331955432891846,
+      "learning_rate": 6.740313678628997e-06,
+      "loss": 0.838,
+      "step": 3092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7803306579589844,
+      "learning_rate": 6.7383238226892745e-06,
+      "loss": 1.0181,
+      "step": 3093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2314506769180298,
+      "learning_rate": 6.7363336535294585e-06,
+      "loss": 0.8324,
+      "step": 3094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2736390829086304,
+      "learning_rate": 6.734343171508151e-06,
+      "loss": 0.8627,
+      "step": 3095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2103761434555054,
+      "learning_rate": 6.732352376984003e-06,
+      "loss": 0.8061,
+      "step": 3096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2637054920196533,
+      "learning_rate": 6.730361270315733e-06,
+      "loss": 0.7977,
+      "step": 3097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3057150840759277,
+      "learning_rate": 6.728369851862103e-06,
+      "loss": 0.8102,
+      "step": 3098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.659233033657074,
+      "learning_rate": 6.726378121981939e-06,
+      "loss": 1.0602,
+      "step": 3099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3233031034469604,
+      "learning_rate": 6.72438608103412e-06,
+      "loss": 0.8178,
+      "step": 3100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.261354684829712,
+      "learning_rate": 6.722393729377584e-06,
+      "loss": 0.868,
+      "step": 3101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6557329297065735,
+      "learning_rate": 6.720401067371321e-06,
+      "loss": 1.0675,
+      "step": 3102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3086514472961426,
+      "learning_rate": 6.718408095374381e-06,
+      "loss": 0.8458,
+      "step": 3103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.627705991268158,
+      "learning_rate": 6.716414813745866e-06,
+      "loss": 1.0498,
+      "step": 3104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2687721252441406,
+      "learning_rate": 6.714421222844938e-06,
+      "loss": 0.7854,
+      "step": 3105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6687136292457581,
+      "learning_rate": 6.712427323030811e-06,
+      "loss": 1.0183,
+      "step": 3106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3406189680099487,
+      "learning_rate": 6.7104331146627565e-06,
+      "loss": 0.8054,
+      "step": 3107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3041709661483765,
+      "learning_rate": 6.708438598100099e-06,
+      "loss": 0.8262,
+      "step": 3108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.293502688407898,
+      "learning_rate": 6.706443773702225e-06,
+      "loss": 0.8208,
+      "step": 3109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6351795792579651,
+      "learning_rate": 6.704448641828568e-06,
+      "loss": 1.0616,
+      "step": 3110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.195331335067749,
+      "learning_rate": 6.702453202838624e-06,
+      "loss": 0.8114,
+      "step": 3111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2420653104782104,
+      "learning_rate": 6.70045745709194e-06,
+      "loss": 0.8231,
+      "step": 3112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6804537773132324,
+      "learning_rate": 6.69846140494812e-06,
+      "loss": 1.0245,
+      "step": 3113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6558142900466919,
+      "learning_rate": 6.696465046766824e-06,
+      "loss": 1.0452,
+      "step": 3114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2849830389022827,
+      "learning_rate": 6.6944683829077626e-06,
+      "loss": 0.8318,
+      "step": 3115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6530535221099854,
+      "learning_rate": 6.692471413730709e-06,
+      "loss": 1.0335,
+      "step": 3116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2607474327087402,
+      "learning_rate": 6.690474139595485e-06,
+      "loss": 0.813,
+      "step": 3117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.254752516746521,
+      "learning_rate": 6.688476560861971e-06,
+      "loss": 0.8604,
+      "step": 3118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2460746765136719,
+      "learning_rate": 6.6864786778901e-06,
+      "loss": 0.7894,
+      "step": 3119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2149513959884644,
+      "learning_rate": 6.684480491039861e-06,
+      "loss": 0.8314,
+      "step": 3120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2363662719726562,
+      "learning_rate": 6.682482000671296e-06,
+      "loss": 0.8301,
+      "step": 3121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2776429653167725,
+      "learning_rate": 6.680483207144508e-06,
+      "loss": 0.8122,
+      "step": 3122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8452882766723633,
+      "learning_rate": 6.678484110819645e-06,
+      "loss": 1.0619,
+      "step": 3123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.241146445274353,
+      "learning_rate": 6.6764847120569155e-06,
+      "loss": 0.8175,
+      "step": 3124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.26650071144104,
+      "learning_rate": 6.674485011216582e-06,
+      "loss": 0.8368,
+      "step": 3125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2637354135513306,
+      "learning_rate": 6.6724850086589636e-06,
+      "loss": 0.8065,
+      "step": 3126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2061195373535156,
+      "learning_rate": 6.670484704744425e-06,
+      "loss": 0.8112,
+      "step": 3127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2289897203445435,
+      "learning_rate": 6.668484099833396e-06,
+      "loss": 0.858,
+      "step": 3128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7567646503448486,
+      "learning_rate": 6.666483194286354e-06,
+      "loss": 1.0535,
+      "step": 3129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2587306499481201,
+      "learning_rate": 6.664481988463833e-06,
+      "loss": 0.884,
+      "step": 3130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.303517460823059,
+      "learning_rate": 6.66248048272642e-06,
+      "loss": 0.8429,
+      "step": 3131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6557027101516724,
+      "learning_rate": 6.660478677434757e-06,
+      "loss": 1.0419,
+      "step": 3132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3165332078933716,
+      "learning_rate": 6.658476572949539e-06,
+      "loss": 0.8248,
+      "step": 3133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2094236612319946,
+      "learning_rate": 6.656474169631517e-06,
+      "loss": 0.7967,
+      "step": 3134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3002657890319824,
+      "learning_rate": 6.654471467841492e-06,
+      "loss": 0.8267,
+      "step": 3135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2969719171524048,
+      "learning_rate": 6.652468467940322e-06,
+      "loss": 0.8211,
+      "step": 3136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7849142551422119,
+      "learning_rate": 6.650465170288918e-06,
+      "loss": 1.033,
+      "step": 3137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2320160865783691,
+      "learning_rate": 6.648461575248245e-06,
+      "loss": 0.8157,
+      "step": 3138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2416926622390747,
+      "learning_rate": 6.646457683179319e-06,
+      "loss": 0.8477,
+      "step": 3139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1707853078842163,
+      "learning_rate": 6.644453494443215e-06,
+      "loss": 0.7547,
+      "step": 3140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2964086532592773,
+      "learning_rate": 6.642449009401054e-06,
+      "loss": 0.8235,
+      "step": 3141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6744616627693176,
+      "learning_rate": 6.640444228414018e-06,
+      "loss": 1.0448,
+      "step": 3142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.242992639541626,
+      "learning_rate": 6.638439151843336e-06,
+      "loss": 0.8259,
+      "step": 3143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2610151767730713,
+      "learning_rate": 6.636433780050293e-06,
+      "loss": 0.8261,
+      "step": 3144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6336978077888489,
+      "learning_rate": 6.6344281133962295e-06,
+      "loss": 1.0536,
+      "step": 3145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2433085441589355,
+      "learning_rate": 6.6324221522425344e-06,
+      "loss": 0.8611,
+      "step": 3146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6295402646064758,
+      "learning_rate": 6.630415896950655e-06,
+      "loss": 1.0522,
+      "step": 3147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6258330941200256,
+      "learning_rate": 6.628409347882086e-06,
+      "loss": 1.0306,
+      "step": 3148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6134485006332397,
+      "learning_rate": 6.626402505398377e-06,
+      "loss": 1.0204,
+      "step": 3149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2565559148788452,
+      "learning_rate": 6.624395369861133e-06,
+      "loss": 0.8166,
+      "step": 3150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3011916875839233,
+      "learning_rate": 6.622387941632012e-06,
+      "loss": 0.7893,
+      "step": 3151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.264585256576538,
+      "learning_rate": 6.620380221072717e-06,
+      "loss": 0.7756,
+      "step": 3152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1795819997787476,
+      "learning_rate": 6.618372208545014e-06,
+      "loss": 0.822,
+      "step": 3153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2272905111312866,
+      "learning_rate": 6.6163639044107155e-06,
+      "loss": 0.806,
+      "step": 3154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.276229977607727,
+      "learning_rate": 6.6143553090316885e-06,
+      "loss": 0.8469,
+      "step": 3155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.337058186531067,
+      "learning_rate": 6.612346422769851e-06,
+      "loss": 0.8274,
+      "step": 3156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3403674364089966,
+      "learning_rate": 6.610337245987175e-06,
+      "loss": 0.8741,
+      "step": 3157
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8189079165458679,
+      "learning_rate": 6.6083277790456855e-06,
+      "loss": 1.0381,
+      "step": 3158
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7351385354995728,
+      "learning_rate": 6.606318022307458e-06,
+      "loss": 1.0342,
+      "step": 3159
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2631311416625977,
+      "learning_rate": 6.604307976134619e-06,
+      "loss": 0.7476,
+      "step": 3160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.359329104423523,
+      "learning_rate": 6.602297640889348e-06,
+      "loss": 0.8432,
+      "step": 3161
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2724566459655762,
+      "learning_rate": 6.600287016933881e-06,
+      "loss": 0.8217,
+      "step": 3162
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.274717926979065,
+      "learning_rate": 6.598276104630503e-06,
+      "loss": 0.8523,
+      "step": 3163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2160208225250244,
+      "learning_rate": 6.596264904341547e-06,
+      "loss": 0.7877,
+      "step": 3164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3158622980117798,
+      "learning_rate": 6.5942534164294035e-06,
+      "loss": 0.8641,
+      "step": 3165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2003742456436157,
+      "learning_rate": 6.592241641256511e-06,
+      "loss": 0.8471,
+      "step": 3166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.280346155166626,
+      "learning_rate": 6.5902295791853645e-06,
+      "loss": 0.861,
+      "step": 3167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3209179639816284,
+      "learning_rate": 6.588217230578504e-06,
+      "loss": 0.8149,
+      "step": 3168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2978181838989258,
+      "learning_rate": 6.586204595798526e-06,
+      "loss": 0.8741,
+      "step": 3169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.258528709411621,
+      "learning_rate": 6.58419167520808e-06,
+      "loss": 0.7893,
+      "step": 3170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3073625564575195,
+      "learning_rate": 6.582178469169862e-06,
+      "loss": 0.8144,
+      "step": 3171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2251795530319214,
+      "learning_rate": 6.5801649780466215e-06,
+      "loss": 0.8595,
+      "step": 3172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2118141651153564,
+      "learning_rate": 6.57815120220116e-06,
+      "loss": 0.7855,
+      "step": 3173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3518544435501099,
+      "learning_rate": 6.576137141996328e-06,
+      "loss": 0.8001,
+      "step": 3174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1155824661254883,
+      "learning_rate": 6.574122797795035e-06,
+      "loss": 1.051,
+      "step": 3175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3218315839767456,
+      "learning_rate": 6.572108169960229e-06,
+      "loss": 0.8148,
+      "step": 3176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7282642126083374,
+      "learning_rate": 6.570093258854921e-06,
+      "loss": 1.0207,
+      "step": 3177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6798320412635803,
+      "learning_rate": 6.568078064842166e-06,
+      "loss": 1.028,
+      "step": 3178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3373188972473145,
+      "learning_rate": 6.566062588285072e-06,
+      "loss": 0.808,
+      "step": 3179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9185531735420227,
+      "learning_rate": 6.5640468295467976e-06,
+      "loss": 1.0597,
+      "step": 3180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8302538990974426,
+      "learning_rate": 6.562030788990552e-06,
+      "loss": 1.0435,
+      "step": 3181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4344767332077026,
+      "learning_rate": 6.560014466979598e-06,
+      "loss": 0.8193,
+      "step": 3182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2978579998016357,
+      "learning_rate": 6.557997863877244e-06,
+      "loss": 0.8159,
+      "step": 3183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2701858282089233,
+      "learning_rate": 6.555980980046857e-06,
+      "loss": 0.7989,
+      "step": 3184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2465697526931763,
+      "learning_rate": 6.553963815851842e-06,
+      "loss": 0.8455,
+      "step": 3185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3177146911621094,
+      "learning_rate": 6.551946371655667e-06,
+      "loss": 0.8482,
+      "step": 3186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2919880151748657,
+      "learning_rate": 6.549928647821844e-06,
+      "loss": 0.7897,
+      "step": 3187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5507673025131226,
+      "learning_rate": 6.5479106447139375e-06,
+      "loss": 1.0605,
+      "step": 3188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2121453285217285,
+      "learning_rate": 6.545892362695561e-06,
+      "loss": 1.0443,
+      "step": 3189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7860432267189026,
+      "learning_rate": 6.5438738021303785e-06,
+      "loss": 1.0688,
+      "step": 3190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.749329149723053,
+      "learning_rate": 6.541854963382106e-06,
+      "loss": 1.0473,
+      "step": 3191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.488519310951233,
+      "learning_rate": 6.539835846814507e-06,
+      "loss": 0.8111,
+      "step": 3192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4143102169036865,
+      "learning_rate": 6.537816452791397e-06,
+      "loss": 1.0333,
+      "step": 3193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2828987836837769,
+      "learning_rate": 6.53579678167664e-06,
+      "loss": 0.8146,
+      "step": 3194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2145992517471313,
+      "learning_rate": 6.53377683383415e-06,
+      "loss": 1.0379,
+      "step": 3195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3270175457000732,
+      "learning_rate": 6.531756609627895e-06,
+      "loss": 0.8395,
+      "step": 3196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3027976751327515,
+      "learning_rate": 6.529736109421885e-06,
+      "loss": 0.835,
+      "step": 3197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.341961145401001,
+      "learning_rate": 6.527715333580186e-06,
+      "loss": 0.8495,
+      "step": 3198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0916224718093872,
+      "learning_rate": 6.525694282466912e-06,
+      "loss": 1.0218,
+      "step": 3199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2825192213058472,
+      "learning_rate": 6.523672956446228e-06,
+      "loss": 0.8592,
+      "step": 3200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.277367353439331,
+      "learning_rate": 6.521651355882343e-06,
+      "loss": 0.8496,
+      "step": 3201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3120925426483154,
+      "learning_rate": 6.5196294811395215e-06,
+      "loss": 0.8156,
+      "step": 3202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.274811863899231,
+      "learning_rate": 6.517607332582075e-06,
+      "loss": 0.8098,
+      "step": 3203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8613308668136597,
+      "learning_rate": 6.515584910574366e-06,
+      "loss": 1.0126,
+      "step": 3204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7323770523071289,
+      "learning_rate": 6.513562215480802e-06,
+      "loss": 1.0539,
+      "step": 3205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2946196794509888,
+      "learning_rate": 6.511539247665846e-06,
+      "loss": 0.8445,
+      "step": 3206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2473515272140503,
+      "learning_rate": 6.5095160074940026e-06,
+      "loss": 0.7858,
+      "step": 3207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.282252550125122,
+      "learning_rate": 6.507492495329833e-06,
+      "loss": 0.8153,
+      "step": 3208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2701749801635742,
+      "learning_rate": 6.505468711537943e-06,
+      "loss": 0.8117,
+      "step": 3209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.313422679901123,
+      "learning_rate": 6.503444656482987e-06,
+      "loss": 0.8595,
+      "step": 3210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2713239192962646,
+      "learning_rate": 6.501420330529671e-06,
+      "loss": 0.8616,
+      "step": 3211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.262614130973816,
+      "learning_rate": 6.49939573404275e-06,
+      "loss": 0.872,
+      "step": 3212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1425724029541016,
+      "learning_rate": 6.497370867387023e-06,
+      "loss": 1.0157,
+      "step": 3213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3309834003448486,
+      "learning_rate": 6.49534573092734e-06,
+      "loss": 0.8961,
+      "step": 3214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2151076793670654,
+      "learning_rate": 6.493320325028604e-06,
+      "loss": 0.777,
+      "step": 3215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7135636806488037,
+      "learning_rate": 6.491294650055764e-06,
+      "loss": 1.0457,
+      "step": 3216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3649344444274902,
+      "learning_rate": 6.489268706373812e-06,
+      "loss": 0.8195,
+      "step": 3217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.303420066833496,
+      "learning_rate": 6.4872424943477965e-06,
+      "loss": 0.8233,
+      "step": 3218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3154162168502808,
+      "learning_rate": 6.485216014342808e-06,
+      "loss": 0.8609,
+      "step": 3219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3549294471740723,
+      "learning_rate": 6.4831892667239886e-06,
+      "loss": 0.8586,
+      "step": 3220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2814302444458008,
+      "learning_rate": 6.48116225185653e-06,
+      "loss": 0.8065,
+      "step": 3221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.275475263595581,
+      "learning_rate": 6.479134970105667e-06,
+      "loss": 0.7734,
+      "step": 3222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9518083333969116,
+      "learning_rate": 6.477107421836691e-06,
+      "loss": 1.0275,
+      "step": 3223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3400224447250366,
+      "learning_rate": 6.47507960741493e-06,
+      "loss": 0.8429,
+      "step": 3224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3572417497634888,
+      "learning_rate": 6.4730515272057705e-06,
+      "loss": 0.826,
+      "step": 3225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.292293906211853,
+      "learning_rate": 6.471023181574639e-06,
+      "loss": 0.8407,
+      "step": 3226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7303999066352844,
+      "learning_rate": 6.468994570887015e-06,
+      "loss": 1.0518,
+      "step": 3227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2772717475891113,
+      "learning_rate": 6.466965695508424e-06,
+      "loss": 0.8214,
+      "step": 3228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2168406248092651,
+      "learning_rate": 6.4649365558044385e-06,
+      "loss": 0.8209,
+      "step": 3229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2613664865493774,
+      "learning_rate": 6.462907152140679e-06,
+      "loss": 0.7905,
+      "step": 3230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2750037908554077,
+      "learning_rate": 6.4608774848828145e-06,
+      "loss": 0.8654,
+      "step": 3231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2399652004241943,
+      "learning_rate": 6.458847554396562e-06,
+      "loss": 0.8768,
+      "step": 3232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3028558492660522,
+      "learning_rate": 6.456817361047682e-06,
+      "loss": 0.8502,
+      "step": 3233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2540022134780884,
+      "learning_rate": 6.4547869052019876e-06,
+      "loss": 0.8674,
+      "step": 3234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3687978982925415,
+      "learning_rate": 6.452756187225336e-06,
+      "loss": 0.8541,
+      "step": 3235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2515044212341309,
+      "learning_rate": 6.450725207483632e-06,
+      "loss": 0.8598,
+      "step": 3236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8262393474578857,
+      "learning_rate": 6.448693966342828e-06,
+      "loss": 1.0298,
+      "step": 3237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7239108085632324,
+      "learning_rate": 6.446662464168923e-06,
+      "loss": 1.0369,
+      "step": 3238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3136482238769531,
+      "learning_rate": 6.444630701327965e-06,
+      "loss": 0.8516,
+      "step": 3239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3855243921279907,
+      "learning_rate": 6.442598678186045e-06,
+      "loss": 0.8144,
+      "step": 3240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2961817979812622,
+      "learning_rate": 6.440566395109307e-06,
+      "loss": 0.8017,
+      "step": 3241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2637332677841187,
+      "learning_rate": 6.438533852463933e-06,
+      "loss": 0.8258,
+      "step": 3242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3706679344177246,
+      "learning_rate": 6.436501050616163e-06,
+      "loss": 0.8359,
+      "step": 3243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2620594501495361,
+      "learning_rate": 6.434467989932272e-06,
+      "loss": 0.7913,
+      "step": 3244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.259866714477539,
+      "learning_rate": 6.432434670778594e-06,
+      "loss": 0.8329,
+      "step": 3245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.055430293083191,
+      "learning_rate": 6.4304010935214935e-06,
+      "loss": 1.028,
+      "step": 3246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3077974319458008,
+      "learning_rate": 6.428367258527399e-06,
+      "loss": 0.7805,
+      "step": 3247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.293461799621582,
+      "learning_rate": 6.4263331661627735e-06,
+      "loss": 0.8535,
+      "step": 3248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.28246009349823,
+      "learning_rate": 6.424298816794132e-06,
+      "loss": 0.8519,
+      "step": 3249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2343201637268066,
+      "learning_rate": 6.422264210788032e-06,
+      "loss": 0.8156,
+      "step": 3250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2578487396240234,
+      "learning_rate": 6.420229348511081e-06,
+      "loss": 0.8222,
+      "step": 3251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3055429458618164,
+      "learning_rate": 6.418194230329931e-06,
+      "loss": 0.8581,
+      "step": 3252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4003429412841797,
+      "learning_rate": 6.416158856611278e-06,
+      "loss": 0.8685,
+      "step": 3253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2658947706222534,
+      "learning_rate": 6.414123227721867e-06,
+      "loss": 0.8328,
+      "step": 3254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2361339330673218,
+      "learning_rate": 6.412087344028489e-06,
+      "loss": 0.8374,
+      "step": 3255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.302671194076538,
+      "learning_rate": 6.410051205897979e-06,
+      "loss": 0.8406,
+      "step": 3256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9995531439781189,
+      "learning_rate": 6.408014813697219e-06,
+      "loss": 1.0614,
+      "step": 3257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2326327562332153,
+      "learning_rate": 6.4059781677931365e-06,
+      "loss": 0.8339,
+      "step": 3258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.70982825756073,
+      "learning_rate": 6.403941268552705e-06,
+      "loss": 1.0398,
+      "step": 3259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6371957659721375,
+      "learning_rate": 6.401904116342945e-06,
+      "loss": 1.0173,
+      "step": 3260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2887425422668457,
+      "learning_rate": 6.399866711530917e-06,
+      "loss": 0.8234,
+      "step": 3261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3546842336654663,
+      "learning_rate": 6.397829054483735e-06,
+      "loss": 0.8689,
+      "step": 3262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3018255233764648,
+      "learning_rate": 6.3957911455685524e-06,
+      "loss": 0.8323,
+      "step": 3263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.346253752708435,
+      "learning_rate": 6.393752985152572e-06,
+      "loss": 0.8788,
+      "step": 3264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2266935110092163,
+      "learning_rate": 6.391714573603037e-06,
+      "loss": 0.8885,
+      "step": 3265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2121329307556152,
+      "learning_rate": 6.389675911287241e-06,
+      "loss": 1.033,
+      "step": 3266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3355660438537598,
+      "learning_rate": 6.387636998572519e-06,
+      "loss": 0.8372,
+      "step": 3267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3059245347976685,
+      "learning_rate": 6.385597835826256e-06,
+      "loss": 0.8212,
+      "step": 3268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7782707810401917,
+      "learning_rate": 6.383558423415875e-06,
+      "loss": 1.0178,
+      "step": 3269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.731459379196167,
+      "learning_rate": 6.38151876170885e-06,
+      "loss": 1.0488,
+      "step": 3270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2691112756729126,
+      "learning_rate": 6.379478851072697e-06,
+      "loss": 0.8439,
+      "step": 3271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3250726461410522,
+      "learning_rate": 6.377438691874977e-06,
+      "loss": 0.8673,
+      "step": 3272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.206027626991272,
+      "learning_rate": 6.3753982844832985e-06,
+      "loss": 0.7919,
+      "step": 3273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2214137315750122,
+      "learning_rate": 6.373357629265309e-06,
+      "loss": 0.7384,
+      "step": 3274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2840287685394287,
+      "learning_rate": 6.371316726588707e-06,
+      "loss": 0.8651,
+      "step": 3275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0361956357955933,
+      "learning_rate": 6.369275576821232e-06,
+      "loss": 1.0422,
+      "step": 3276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9552755355834961,
+      "learning_rate": 6.3672341803306695e-06,
+      "loss": 1.0391,
+      "step": 3277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4348386526107788,
+      "learning_rate": 6.3651925374848476e-06,
+      "loss": 0.8117,
+      "step": 3278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3573235273361206,
+      "learning_rate": 6.363150648651639e-06,
+      "loss": 0.8544,
+      "step": 3279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.725721001625061,
+      "learning_rate": 6.361108514198966e-06,
+      "loss": 1.0322,
+      "step": 3280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3802398443222046,
+      "learning_rate": 6.359066134494787e-06,
+      "loss": 0.8587,
+      "step": 3281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.240106225013733,
+      "learning_rate": 6.357023509907108e-06,
+      "loss": 0.8491,
+      "step": 3282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.279549241065979,
+      "learning_rate": 6.354980640803983e-06,
+      "loss": 0.8415,
+      "step": 3283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2869853973388672,
+      "learning_rate": 6.352937527553503e-06,
+      "loss": 0.8635,
+      "step": 3284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9277810454368591,
+      "learning_rate": 6.350894170523809e-06,
+      "loss": 1.0287,
+      "step": 3285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2370388507843018,
+      "learning_rate": 6.348850570083084e-06,
+      "loss": 0.8669,
+      "step": 3286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7572387456893921,
+      "learning_rate": 6.34680672659955e-06,
+      "loss": 1.0254,
+      "step": 3287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6298990249633789,
+      "learning_rate": 6.344762640441484e-06,
+      "loss": 1.0771,
+      "step": 3288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3028311729431152,
+      "learning_rate": 6.342718311977194e-06,
+      "loss": 0.8023,
+      "step": 3289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.374403953552246,
+      "learning_rate": 6.340673741575041e-06,
+      "loss": 0.798,
+      "step": 3290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7530023455619812,
+      "learning_rate": 6.338628929603423e-06,
+      "loss": 1.0314,
+      "step": 3291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3521498441696167,
+      "learning_rate": 6.336583876430788e-06,
+      "loss": 0.9351,
+      "step": 3292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2383849620819092,
+      "learning_rate": 6.334538582425624e-06,
+      "loss": 0.8232,
+      "step": 3293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7673928141593933,
+      "learning_rate": 6.332493047956461e-06,
+      "loss": 1.0273,
+      "step": 3294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2155545949935913,
+      "learning_rate": 6.330447273391873e-06,
+      "loss": 0.7871,
+      "step": 3295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6698043346405029,
+      "learning_rate": 6.328401259100481e-06,
+      "loss": 1.0541,
+      "step": 3296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3710155487060547,
+      "learning_rate": 6.326355005450944e-06,
+      "loss": 0.8553,
+      "step": 3297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3401696681976318,
+      "learning_rate": 6.324308512811968e-06,
+      "loss": 0.8202,
+      "step": 3298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6374285221099854,
+      "learning_rate": 6.322261781552298e-06,
+      "loss": 1.0377,
+      "step": 3299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2922574281692505,
+      "learning_rate": 6.320214812040728e-06,
+      "loss": 0.8583,
+      "step": 3300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7105167508125305,
+      "learning_rate": 6.318167604646089e-06,
+      "loss": 1.0539,
+      "step": 3301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2151544094085693,
+      "learning_rate": 6.316120159737259e-06,
+      "loss": 0.8291,
+      "step": 3302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6753464341163635,
+      "learning_rate": 6.314072477683155e-06,
+      "loss": 1.0527,
+      "step": 3303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2705239057540894,
+      "learning_rate": 6.3120245588527405e-06,
+      "loss": 0.8721,
+      "step": 3304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6237471103668213,
+      "learning_rate": 6.30997640361502e-06,
+      "loss": 1.024,
+      "step": 3305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.638044536113739,
+      "learning_rate": 6.30792801233904e-06,
+      "loss": 1.0313,
+      "step": 3306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2580679655075073,
+      "learning_rate": 6.305879385393889e-06,
+      "loss": 0.8264,
+      "step": 3307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2828845977783203,
+      "learning_rate": 6.303830523148702e-06,
+      "loss": 0.8297,
+      "step": 3308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.220296859741211,
+      "learning_rate": 6.301781425972653e-06,
+      "loss": 0.8684,
+      "step": 3309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2945301532745361,
+      "learning_rate": 6.299732094234955e-06,
+      "loss": 0.8351,
+      "step": 3310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2228449583053589,
+      "learning_rate": 6.29768252830487e-06,
+      "loss": 0.8596,
+      "step": 3311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2164865732192993,
+      "learning_rate": 6.2956327285517e-06,
+      "loss": 0.893,
+      "step": 3312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2768715620040894,
+      "learning_rate": 6.293582695344789e-06,
+      "loss": 0.8529,
+      "step": 3313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9054147601127625,
+      "learning_rate": 6.2915324290535185e-06,
+      "loss": 1.0173,
+      "step": 3314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2745821475982666,
+      "learning_rate": 6.289481930047319e-06,
+      "loss": 0.8248,
+      "step": 3315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.698155403137207,
+      "learning_rate": 6.28743119869566e-06,
+      "loss": 1.0418,
+      "step": 3316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2952635288238525,
+      "learning_rate": 6.285380235368052e-06,
+      "loss": 0.8183,
+      "step": 3317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2972882986068726,
+      "learning_rate": 6.283329040434048e-06,
+      "loss": 0.8047,
+      "step": 3318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2711400985717773,
+      "learning_rate": 6.281277614263244e-06,
+      "loss": 0.8254,
+      "step": 3319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.280392050743103,
+      "learning_rate": 6.279225957225274e-06,
+      "loss": 0.8016,
+      "step": 3320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9261526465415955,
+      "learning_rate": 6.27717406968982e-06,
+      "loss": 1.0498,
+      "step": 3321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.237675666809082,
+      "learning_rate": 6.275121952026597e-06,
+      "loss": 0.8064,
+      "step": 3322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3045090436935425,
+      "learning_rate": 6.273069604605369e-06,
+      "loss": 0.8779,
+      "step": 3323
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2638356685638428,
+      "learning_rate": 6.271017027795937e-06,
+      "loss": 0.807,
+      "step": 3324
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4670872688293457,
+      "learning_rate": 6.268964221968147e-06,
+      "loss": 0.8119,
+      "step": 3325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2759391069412231,
+      "learning_rate": 6.266911187491882e-06,
+      "loss": 0.814,
+      "step": 3326
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6743872165679932,
+      "learning_rate": 6.264857924737068e-06,
+      "loss": 1.0352,
+      "step": 3327
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2184100151062012,
+      "learning_rate": 6.2628044340736734e-06,
+      "loss": 0.8214,
+      "step": 3328
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2630012035369873,
+      "learning_rate": 6.260750715871707e-06,
+      "loss": 0.8961,
+      "step": 3329
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2461662292480469,
+      "learning_rate": 6.2586967705012206e-06,
+      "loss": 0.8343,
+      "step": 3330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2042288780212402,
+      "learning_rate": 6.256642598332299e-06,
+      "loss": 0.8056,
+      "step": 3331
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.246346116065979,
+      "learning_rate": 6.254588199735076e-06,
+      "loss": 0.8322,
+      "step": 3332
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3242446184158325,
+      "learning_rate": 6.252533575079725e-06,
+      "loss": 0.8347,
+      "step": 3333
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.248073697090149,
+      "learning_rate": 6.250478724736457e-06,
+      "loss": 0.7696,
+      "step": 3334
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2598519325256348,
+      "learning_rate": 6.2484236490755264e-06,
+      "loss": 0.7854,
+      "step": 3335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2239220142364502,
+      "learning_rate": 6.246368348467228e-06,
+      "loss": 0.8489,
+      "step": 3336
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2310059070587158,
+      "learning_rate": 6.2443128232818935e-06,
+      "loss": 0.8311,
+      "step": 3337
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6652606725692749,
+      "learning_rate": 6.242257073889902e-06,
+      "loss": 1.0277,
+      "step": 3338
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2415058612823486,
+      "learning_rate": 6.240201100661667e-06,
+      "loss": 0.8439,
+      "step": 3339
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2532440423965454,
+      "learning_rate": 6.238144903967643e-06,
+      "loss": 0.7599,
+      "step": 3340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.237271785736084,
+      "learning_rate": 6.236088484178327e-06,
+      "loss": 0.8099,
+      "step": 3341
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.299513816833496,
+      "learning_rate": 6.234031841664257e-06,
+      "loss": 0.8159,
+      "step": 3342
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2541218996047974,
+      "learning_rate": 6.231974976796007e-06,
+      "loss": 0.8809,
+      "step": 3343
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6467993855476379,
+      "learning_rate": 6.229917889944195e-06,
+      "loss": 1.0532,
+      "step": 3344
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6350182294845581,
+      "learning_rate": 6.227860581479474e-06,
+      "loss": 1.0293,
+      "step": 3345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2508807182312012,
+      "learning_rate": 6.225803051772547e-06,
+      "loss": 0.776,
+      "step": 3346
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2610642910003662,
+      "learning_rate": 6.223745301194145e-06,
+      "loss": 0.7793,
+      "step": 3347
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2435837984085083,
+      "learning_rate": 6.2216873301150445e-06,
+      "loss": 0.8285,
+      "step": 3348
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3304839134216309,
+      "learning_rate": 6.219629138906063e-06,
+      "loss": 0.8576,
+      "step": 3349
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2788594961166382,
+      "learning_rate": 6.217570727938056e-06,
+      "loss": 0.8557,
+      "step": 3350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2632651329040527,
+      "learning_rate": 6.215512097581916e-06,
+      "loss": 0.8278,
+      "step": 3351
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8726982474327087,
+      "learning_rate": 6.21345324820858e-06,
+      "loss": 1.0505,
+      "step": 3352
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2793279886245728,
+      "learning_rate": 6.211394180189021e-06,
+      "loss": 0.8137,
+      "step": 3353
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7049834132194519,
+      "learning_rate": 6.209334893894254e-06,
+      "loss": 1.0391,
+      "step": 3354
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2657948732376099,
+      "learning_rate": 6.207275389695329e-06,
+      "loss": 0.884,
+      "step": 3355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2407876253128052,
+      "learning_rate": 6.205215667963339e-06,
+      "loss": 0.8239,
+      "step": 3356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2431128025054932,
+      "learning_rate": 6.203155729069417e-06,
+      "loss": 0.849,
+      "step": 3357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.299599051475525,
+      "learning_rate": 6.201095573384732e-06,
+      "loss": 0.8159,
+      "step": 3358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8625061511993408,
+      "learning_rate": 6.199035201280492e-06,
+      "loss": 1.025,
+      "step": 3359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2722247838974,
+      "learning_rate": 6.196974613127949e-06,
+      "loss": 0.8383,
+      "step": 3360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3350574970245361,
+      "learning_rate": 6.194913809298386e-06,
+      "loss": 0.8409,
+      "step": 3361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2902600765228271,
+      "learning_rate": 6.192852790163133e-06,
+      "loss": 0.8251,
+      "step": 3362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2694236040115356,
+      "learning_rate": 6.190791556093553e-06,
+      "loss": 0.9105,
+      "step": 3363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6449236273765564,
+      "learning_rate": 6.188730107461049e-06,
+      "loss": 1.0426,
+      "step": 3364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.245007872581482,
+      "learning_rate": 6.186668444637065e-06,
+      "loss": 0.7958,
+      "step": 3365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2676438093185425,
+      "learning_rate": 6.184606567993081e-06,
+      "loss": 0.8065,
+      "step": 3366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.338681697845459,
+      "learning_rate": 6.182544477900618e-06,
+      "loss": 0.898,
+      "step": 3367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2887310981750488,
+      "learning_rate": 6.180482174731232e-06,
+      "loss": 0.8497,
+      "step": 3368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1835836172103882,
+      "learning_rate": 6.1784196588565205e-06,
+      "loss": 0.8507,
+      "step": 3369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7019469738006592,
+      "learning_rate": 6.1763569306481175e-06,
+      "loss": 1.0113,
+      "step": 3370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2296017408370972,
+      "learning_rate": 6.174293990477698e-06,
+      "loss": 0.8275,
+      "step": 3371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.239554762840271,
+      "learning_rate": 6.172230838716968e-06,
+      "loss": 0.8411,
+      "step": 3372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6358869075775146,
+      "learning_rate": 6.170167475737684e-06,
+      "loss": 1.0387,
+      "step": 3373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2915802001953125,
+      "learning_rate": 6.168103901911628e-06,
+      "loss": 0.8041,
+      "step": 3374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6289359927177429,
+      "learning_rate": 6.1660401176106265e-06,
+      "loss": 1.0469,
+      "step": 3375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.289228081703186,
+      "learning_rate": 6.163976123206542e-06,
+      "loss": 0.7942,
+      "step": 3376
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.182784914970398,
+      "learning_rate": 6.1619119190712795e-06,
+      "loss": 0.8017,
+      "step": 3377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2834264039993286,
+      "learning_rate": 6.159847505576772e-06,
+      "loss": 0.7845,
+      "step": 3378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2510274648666382,
+      "learning_rate": 6.157782883095001e-06,
+      "loss": 0.8532,
+      "step": 3379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2282649278640747,
+      "learning_rate": 6.155718051997977e-06,
+      "loss": 0.8097,
+      "step": 3380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6856256127357483,
+      "learning_rate": 6.153653012657756e-06,
+      "loss": 1.0506,
+      "step": 3381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.313794493675232,
+      "learning_rate": 6.151587765446424e-06,
+      "loss": 0.8165,
+      "step": 3382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.262157917022705,
+      "learning_rate": 6.149522310736111e-06,
+      "loss": 0.822,
+      "step": 3383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.370390772819519,
+      "learning_rate": 6.147456648898975e-06,
+      "loss": 0.8168,
+      "step": 3384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.626015841960907,
+      "learning_rate": 6.1453907803072244e-06,
+      "loss": 1.0309,
+      "step": 3385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2574381828308105,
+      "learning_rate": 6.143324705333094e-06,
+      "loss": 0.8293,
+      "step": 3386
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3757644891738892,
+      "learning_rate": 6.141258424348862e-06,
+      "loss": 0.8912,
+      "step": 3387
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.274771809577942,
+      "learning_rate": 6.139191937726839e-06,
+      "loss": 0.8239,
+      "step": 3388
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.25979483127594,
+      "learning_rate": 6.137125245839379e-06,
+      "loss": 0.7958,
+      "step": 3389
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6623035669326782,
+      "learning_rate": 6.135058349058864e-06,
+      "loss": 1.0549,
+      "step": 3390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6842257380485535,
+      "learning_rate": 6.132991247757724e-06,
+      "loss": 1.0318,
+      "step": 3391
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1313778162002563,
+      "learning_rate": 6.130923942308413e-06,
+      "loss": 0.7832,
+      "step": 3392
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.225713849067688,
+      "learning_rate": 6.128856433083436e-06,
+      "loss": 0.8039,
+      "step": 3393
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6431071162223816,
+      "learning_rate": 6.126788720455321e-06,
+      "loss": 1.0186,
+      "step": 3394
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2593657970428467,
+      "learning_rate": 6.124720804796644e-06,
+      "loss": 0.828,
+      "step": 3395
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2510569095611572,
+      "learning_rate": 6.122652686480008e-06,
+      "loss": 0.8252,
+      "step": 3396
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1774184703826904,
+      "learning_rate": 6.1205843658780626e-06,
+      "loss": 0.7936,
+      "step": 3397
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2393025159835815,
+      "learning_rate": 6.118515843363484e-06,
+      "loss": 0.828,
+      "step": 3398
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7250487804412842,
+      "learning_rate": 6.116447119308992e-06,
+      "loss": 1.0208,
+      "step": 3399
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6878812313079834,
+      "learning_rate": 6.114378194087335e-06,
+      "loss": 0.9969,
+      "step": 3400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2241910696029663,
+      "learning_rate": 6.1123090680713085e-06,
+      "loss": 0.8309,
+      "step": 3401
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6278878450393677,
+      "learning_rate": 6.110239741633735e-06,
+      "loss": 1.0497,
+      "step": 3402
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.311607003211975,
+      "learning_rate": 6.108170215147479e-06,
+      "loss": 0.8437,
+      "step": 3403
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2686485052108765,
+      "learning_rate": 6.1061004889854355e-06,
+      "loss": 0.8213,
+      "step": 3404
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.297542691230774,
+      "learning_rate": 6.104030563520538e-06,
+      "loss": 0.8972,
+      "step": 3405
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.211349368095398,
+      "learning_rate": 6.101960439125759e-06,
+      "loss": 0.8203,
+      "step": 3406
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.201511025428772,
+      "learning_rate": 6.099890116174103e-06,
+      "loss": 0.8348,
+      "step": 3407
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3451647758483887,
+      "learning_rate": 6.097819595038612e-06,
+      "loss": 0.7807,
+      "step": 3408
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8670550584793091,
+      "learning_rate": 6.0957488760923586e-06,
+      "loss": 1.0485,
+      "step": 3409
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2922741174697876,
+      "learning_rate": 6.0936779597084626e-06,
+      "loss": 0.7985,
+      "step": 3410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.345735788345337,
+      "learning_rate": 6.091606846260069e-06,
+      "loss": 0.7894,
+      "step": 3411
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6897585988044739,
+      "learning_rate": 6.089535536120361e-06,
+      "loss": 1.0212,
+      "step": 3412
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2690491676330566,
+      "learning_rate": 6.087464029662556e-06,
+      "loss": 0.878,
+      "step": 3413
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2979819774627686,
+      "learning_rate": 6.085392327259915e-06,
+      "loss": 0.797,
+      "step": 3414
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.327212929725647,
+      "learning_rate": 6.083320429285722e-06,
+      "loss": 0.8555,
+      "step": 3415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.261387825012207,
+      "learning_rate": 6.081248336113305e-06,
+      "loss": 0.81,
+      "step": 3416
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7267676591873169,
+      "learning_rate": 6.079176048116022e-06,
+      "loss": 1.0222,
+      "step": 3417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.235216736793518,
+      "learning_rate": 6.077103565667271e-06,
+      "loss": 0.7996,
+      "step": 3418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.326816439628601,
+      "learning_rate": 6.075030889140483e-06,
+      "loss": 0.8239,
+      "step": 3419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.233298659324646,
+      "learning_rate": 6.0729580189091206e-06,
+      "loss": 0.7832,
+      "step": 3420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6489242911338806,
+      "learning_rate": 6.070884955346685e-06,
+      "loss": 1.0447,
+      "step": 3421
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.325913667678833,
+      "learning_rate": 6.0688116988267144e-06,
+      "loss": 0.8,
+      "step": 3422
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2566243410110474,
+      "learning_rate": 6.066738249722774e-06,
+      "loss": 0.7832,
+      "step": 3423
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2833586931228638,
+      "learning_rate": 6.064664608408474e-06,
+      "loss": 0.8485,
+      "step": 3424
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2769712209701538,
+      "learning_rate": 6.062590775257447e-06,
+      "loss": 0.8364,
+      "step": 3425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2017892599105835,
+      "learning_rate": 6.060516750643373e-06,
+      "loss": 0.8129,
+      "step": 3426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6959932446479797,
+      "learning_rate": 6.0584425349399565e-06,
+      "loss": 1.0411,
+      "step": 3427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.264050006866455,
+      "learning_rate": 6.056368128520943e-06,
+      "loss": 0.8803,
+      "step": 3428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6405385732650757,
+      "learning_rate": 6.054293531760106e-06,
+      "loss": 1.0398,
+      "step": 3429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2913730144500732,
+      "learning_rate": 6.052218745031262e-06,
+      "loss": 0.8416,
+      "step": 3430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2403308153152466,
+      "learning_rate": 6.050143768708252e-06,
+      "loss": 0.8049,
+      "step": 3431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7079811096191406,
+      "learning_rate": 6.048068603164961e-06,
+      "loss": 1.0403,
+      "step": 3432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6791108250617981,
+      "learning_rate": 6.045993248775296e-06,
+      "loss": 1.0208,
+      "step": 3433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2861366271972656,
+      "learning_rate": 6.043917705913212e-06,
+      "loss": 0.8383,
+      "step": 3434
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2804481983184814,
+      "learning_rate": 6.0418419749526845e-06,
+      "loss": 0.8669,
+      "step": 3435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.210191249847412,
+      "learning_rate": 6.039766056267736e-06,
+      "loss": 0.7634,
+      "step": 3436
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2265523672103882,
+      "learning_rate": 6.037689950232411e-06,
+      "loss": 0.7717,
+      "step": 3437
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1880512237548828,
+      "learning_rate": 6.035613657220794e-06,
+      "loss": 0.7262,
+      "step": 3438
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7371399998664856,
+      "learning_rate": 6.033537177607005e-06,
+      "loss": 1.028,
+      "step": 3439
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2001680135726929,
+      "learning_rate": 6.031460511765191e-06,
+      "loss": 0.8021,
+      "step": 3440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2718313932418823,
+      "learning_rate": 6.029383660069539e-06,
+      "loss": 0.8182,
+      "step": 3441
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2984561920166016,
+      "learning_rate": 6.027306622894265e-06,
+      "loss": 0.812,
+      "step": 3442
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2172712087631226,
+      "learning_rate": 6.025229400613621e-06,
+      "loss": 0.768,
+      "step": 3443
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.297870397567749,
+      "learning_rate": 6.0231519936018915e-06,
+      "loss": 0.7859,
+      "step": 3444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.706224262714386,
+      "learning_rate": 6.021074402233393e-06,
+      "loss": 1.0563,
+      "step": 3445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2265329360961914,
+      "learning_rate": 6.0189966268824785e-06,
+      "loss": 0.8646,
+      "step": 3446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6605444550514221,
+      "learning_rate": 6.016918667923533e-06,
+      "loss": 1.0333,
+      "step": 3447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6332717537879944,
+      "learning_rate": 6.014840525730971e-06,
+      "loss": 1.0186,
+      "step": 3448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.289260745048523,
+      "learning_rate": 6.012762200679243e-06,
+      "loss": 0.8123,
+      "step": 3449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3150802850723267,
+      "learning_rate": 6.010683693142835e-06,
+      "loss": 0.8364,
+      "step": 3450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2828418016433716,
+      "learning_rate": 6.008605003496261e-06,
+      "loss": 0.7915,
+      "step": 3451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.324878215789795,
+      "learning_rate": 6.006526132114071e-06,
+      "loss": 0.8134,
+      "step": 3452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2234101295471191,
+      "learning_rate": 6.004447079370845e-06,
+      "loss": 0.7509,
+      "step": 3453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1999905109405518,
+      "learning_rate": 6.002367845641199e-06,
+      "loss": 0.8151,
+      "step": 3454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8470258116722107,
+      "learning_rate": 6.00028843129978e-06,
+      "loss": 1.0306,
+      "step": 3455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.329847812652588,
+      "learning_rate": 5.998208836721266e-06,
+      "loss": 0.8438,
+      "step": 3456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6849929690361023,
+      "learning_rate": 5.996129062280371e-06,
+      "loss": 1.0287,
+      "step": 3457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2245920896530151,
+      "learning_rate": 5.994049108351838e-06,
+      "loss": 0.729,
+      "step": 3458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2419102191925049,
+      "learning_rate": 5.9919689753104445e-06,
+      "loss": 0.8139,
+      "step": 3459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2344619035720825,
+      "learning_rate": 5.989888663530999e-06,
+      "loss": 0.7756,
+      "step": 3460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.27533757686615,
+      "learning_rate": 5.987808173388343e-06,
+      "loss": 0.7932,
+      "step": 3461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9135928750038147,
+      "learning_rate": 5.985727505257349e-06,
+      "loss": 1.0393,
+      "step": 3462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3131287097930908,
+      "learning_rate": 5.983646659512925e-06,
+      "loss": 0.854,
+      "step": 3463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2364221811294556,
+      "learning_rate": 5.981565636530005e-06,
+      "loss": 0.7977,
+      "step": 3464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3262851238250732,
+      "learning_rate": 5.9794844366835615e-06,
+      "loss": 0.8248,
+      "step": 3465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2505775690078735,
+      "learning_rate": 5.977403060348593e-06,
+      "loss": 0.7858,
+      "step": 3466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.256067156791687,
+      "learning_rate": 5.9753215079001355e-06,
+      "loss": 0.8534,
+      "step": 3467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.230394721031189,
+      "learning_rate": 5.973239779713251e-06,
+      "loss": 0.8301,
+      "step": 3468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2120758295059204,
+      "learning_rate": 5.9711578761630375e-06,
+      "loss": 0.7758,
+      "step": 3469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2692527770996094,
+      "learning_rate": 5.969075797624622e-06,
+      "loss": 0.8348,
+      "step": 3470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7182345986366272,
+      "learning_rate": 5.966993544473166e-06,
+      "loss": 1.0365,
+      "step": 3471
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2705059051513672,
+      "learning_rate": 5.964911117083858e-06,
+      "loss": 0.7904,
+      "step": 3472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.297895073890686,
+      "learning_rate": 5.962828515831922e-06,
+      "loss": 0.8241,
+      "step": 3473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3023625612258911,
+      "learning_rate": 5.960745741092611e-06,
+      "loss": 0.7879,
+      "step": 3474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2613043785095215,
+      "learning_rate": 5.958662793241212e-06,
+      "loss": 0.7907,
+      "step": 3475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2167541980743408,
+      "learning_rate": 5.95657967265304e-06,
+      "loss": 0.8734,
+      "step": 3476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6605958938598633,
+      "learning_rate": 5.9544963797034405e-06,
+      "loss": 1.0257,
+      "step": 3477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2540984153747559,
+      "learning_rate": 5.952412914767795e-06,
+      "loss": 0.8702,
+      "step": 3478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.649009644985199,
+      "learning_rate": 5.950329278221509e-06,
+      "loss": 1.0352,
+      "step": 3479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2783029079437256,
+      "learning_rate": 5.948245470440027e-06,
+      "loss": 0.8183,
+      "step": 3480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6219145655632019,
+      "learning_rate": 5.946161491798818e-06,
+      "loss": 1.0473,
+      "step": 3481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6179258227348328,
+      "learning_rate": 5.944077342673384e-06,
+      "loss": 1.0374,
+      "step": 3482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1825714111328125,
+      "learning_rate": 5.941993023439258e-06,
+      "loss": 0.8274,
+      "step": 3483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6010535955429077,
+      "learning_rate": 5.939908534472004e-06,
+      "loss": 1.0094,
+      "step": 3484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2205548286437988,
+      "learning_rate": 5.937823876147215e-06,
+      "loss": 0.8475,
+      "step": 3485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6122511625289917,
+      "learning_rate": 5.935739048840514e-06,
+      "loss": 1.0274,
+      "step": 3486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2171636819839478,
+      "learning_rate": 5.933654052927559e-06,
+      "loss": 0.8357,
+      "step": 3487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6111154556274414,
+      "learning_rate": 5.931568888784034e-06,
+      "loss": 1.0353,
+      "step": 3488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6106334328651428,
+      "learning_rate": 5.929483556785653e-06,
+      "loss": 1.0282,
+      "step": 3489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2349324226379395,
+      "learning_rate": 5.9273980573081635e-06,
+      "loss": 0.8362,
+      "step": 3490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2420103549957275,
+      "learning_rate": 5.9253123907273405e-06,
+      "loss": 0.8227,
+      "step": 3491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6319237947463989,
+      "learning_rate": 5.923226557418992e-06,
+      "loss": 1.0414,
+      "step": 3492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2158030271530151,
+      "learning_rate": 5.921140557758951e-06,
+      "loss": 0.8246,
+      "step": 3493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3009158372879028,
+      "learning_rate": 5.919054392123087e-06,
+      "loss": 0.8196,
+      "step": 3494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.288640022277832,
+      "learning_rate": 5.9169680608872925e-06,
+      "loss": 0.8546,
+      "step": 3495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2299060821533203,
+      "learning_rate": 5.914881564427497e-06,
+      "loss": 0.8204,
+      "step": 3496
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2855113744735718,
+      "learning_rate": 5.912794903119654e-06,
+      "loss": 0.8209,
+      "step": 3497
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2374975681304932,
+      "learning_rate": 5.910708077339749e-06,
+      "loss": 0.7915,
+      "step": 3498
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7180941700935364,
+      "learning_rate": 5.908621087463795e-06,
+      "loss": 1.0423,
+      "step": 3499
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.259125828742981,
+      "learning_rate": 5.9065339338678414e-06,
+      "loss": 0.8485,
+      "step": 3500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6518394947052002,
+      "learning_rate": 5.904446616927959e-06,
+      "loss": 1.0395,
+      "step": 3501
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2071963548660278,
+      "learning_rate": 5.902359137020251e-06,
+      "loss": 0.8176,
+      "step": 3502
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6384567618370056,
+      "learning_rate": 5.900271494520851e-06,
+      "loss": 1.0263,
+      "step": 3503
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2392420768737793,
+      "learning_rate": 5.898183689805922e-06,
+      "loss": 0.8005,
+      "step": 3504
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4012023210525513,
+      "learning_rate": 5.8960957232516535e-06,
+      "loss": 0.8052,
+      "step": 3505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2040573358535767,
+      "learning_rate": 5.894007595234267e-06,
+      "loss": 0.79,
+      "step": 3506
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1919574737548828,
+      "learning_rate": 5.891919306130012e-06,
+      "loss": 0.8182,
+      "step": 3507
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2422351837158203,
+      "learning_rate": 5.8898308563151674e-06,
+      "loss": 0.836,
+      "step": 3508
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3121265172958374,
+      "learning_rate": 5.88774224616604e-06,
+      "loss": 0.8284,
+      "step": 3509
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3355929851531982,
+      "learning_rate": 5.8856534760589675e-06,
+      "loss": 0.8367,
+      "step": 3510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9418318271636963,
+      "learning_rate": 5.883564546370314e-06,
+      "loss": 1.0166,
+      "step": 3511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.269180417060852,
+      "learning_rate": 5.881475457476474e-06,
+      "loss": 0.8036,
+      "step": 3512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2896543741226196,
+      "learning_rate": 5.879386209753872e-06,
+      "loss": 0.8785,
+      "step": 3513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1709553003311157,
+      "learning_rate": 5.877296803578955e-06,
+      "loss": 0.8294,
+      "step": 3514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2603181600570679,
+      "learning_rate": 5.875207239328208e-06,
+      "loss": 0.811,
+      "step": 3515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2351670265197754,
+      "learning_rate": 5.8731175173781355e-06,
+      "loss": 0.8372,
+      "step": 3516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.697597086429596,
+      "learning_rate": 5.871027638105277e-06,
+      "loss": 1.0452,
+      "step": 3517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6885037422180176,
+      "learning_rate": 5.868937601886194e-06,
+      "loss": 1.0122,
+      "step": 3518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6236408352851868,
+      "learning_rate": 5.866847409097485e-06,
+      "loss": 1.0249,
+      "step": 3519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2843046188354492,
+      "learning_rate": 5.864757060115768e-06,
+      "loss": 0.8313,
+      "step": 3520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2686392068862915,
+      "learning_rate": 5.862666555317695e-06,
+      "loss": 0.8649,
+      "step": 3521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7442857623100281,
+      "learning_rate": 5.860575895079941e-06,
+      "loss": 1.0443,
+      "step": 3522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2519147396087646,
+      "learning_rate": 5.858485079779214e-06,
+      "loss": 0.7985,
+      "step": 3523
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2486671209335327,
+      "learning_rate": 5.856394109792247e-06,
+      "loss": 0.7402,
+      "step": 3524
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2237039804458618,
+      "learning_rate": 5.854302985495803e-06,
+      "loss": 0.8337,
+      "step": 3525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2819271087646484,
+      "learning_rate": 5.8522117072666695e-06,
+      "loss": 0.8286,
+      "step": 3526
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2221733331680298,
+      "learning_rate": 5.850120275481665e-06,
+      "loss": 0.7697,
+      "step": 3527
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2879935503005981,
+      "learning_rate": 5.848028690517634e-06,
+      "loss": 0.8405,
+      "step": 3528
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2703369855880737,
+      "learning_rate": 5.84593695275145e-06,
+      "loss": 0.8377,
+      "step": 3529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2577824592590332,
+      "learning_rate": 5.843845062560012e-06,
+      "loss": 0.8305,
+      "step": 3530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3211898803710938,
+      "learning_rate": 5.841753020320247e-06,
+      "loss": 0.834,
+      "step": 3531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7852920293807983,
+      "learning_rate": 5.83966082640911e-06,
+      "loss": 1.0449,
+      "step": 3532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.29731023311615,
+      "learning_rate": 5.837568481203586e-06,
+      "loss": 0.8016,
+      "step": 3533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7348714470863342,
+      "learning_rate": 5.83547598508068e-06,
+      "loss": 1.0242,
+      "step": 3534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6532610058784485,
+      "learning_rate": 5.833383338417431e-06,
+      "loss": 1.0045,
+      "step": 3535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.233376145362854,
+      "learning_rate": 5.831290541590904e-06,
+      "loss": 0.8665,
+      "step": 3536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.257711410522461,
+      "learning_rate": 5.829197594978189e-06,
+      "loss": 0.7599,
+      "step": 3537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3294997215270996,
+      "learning_rate": 5.8271044989564015e-06,
+      "loss": 0.8212,
+      "step": 3538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7767366766929626,
+      "learning_rate": 5.825011253902688e-06,
+      "loss": 1.0386,
+      "step": 3539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.747369110584259,
+      "learning_rate": 5.822917860194222e-06,
+      "loss": 1.0262,
+      "step": 3540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.264349341392517,
+      "learning_rate": 5.8208243182082004e-06,
+      "loss": 0.8583,
+      "step": 3541
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.229507327079773,
+      "learning_rate": 5.818730628321846e-06,
+      "loss": 0.8204,
+      "step": 3542
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2498204708099365,
+      "learning_rate": 5.816636790912413e-06,
+      "loss": 0.8059,
+      "step": 3543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6525168418884277,
+      "learning_rate": 5.814542806357181e-06,
+      "loss": 1.0294,
+      "step": 3544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1874887943267822,
+      "learning_rate": 5.812448675033453e-06,
+      "loss": 0.7748,
+      "step": 3545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6980419158935547,
+      "learning_rate": 5.8103543973185606e-06,
+      "loss": 1.0328,
+      "step": 3546
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6527331471443176,
+      "learning_rate": 5.80825997358986e-06,
+      "loss": 1.0552,
+      "step": 3547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2971206903457642,
+      "learning_rate": 5.806165404224737e-06,
+      "loss": 0.8723,
+      "step": 3548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2633788585662842,
+      "learning_rate": 5.804070689600603e-06,
+      "loss": 0.8495,
+      "step": 3549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.288820743560791,
+      "learning_rate": 5.8019758300948926e-06,
+      "loss": 0.8691,
+      "step": 3550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.765614926815033,
+      "learning_rate": 5.799880826085067e-06,
+      "loss": 1.0403,
+      "step": 3551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1962076425552368,
+      "learning_rate": 5.797785677948619e-06,
+      "loss": 0.7388,
+      "step": 3552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7014071345329285,
+      "learning_rate": 5.795690386063057e-06,
+      "loss": 1.0288,
+      "step": 3553
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2138935327529907,
+      "learning_rate": 5.7935949508059286e-06,
+      "loss": 0.8402,
+      "step": 3554
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3126105070114136,
+      "learning_rate": 5.791499372554794e-06,
+      "loss": 0.7997,
+      "step": 3555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2821882963180542,
+      "learning_rate": 5.7894036516872484e-06,
+      "loss": 0.8319,
+      "step": 3556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2410485744476318,
+      "learning_rate": 5.787307788580909e-06,
+      "loss": 0.8866,
+      "step": 3557
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2217411994934082,
+      "learning_rate": 5.785211783613421e-06,
+      "loss": 0.8608,
+      "step": 3558
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2296158075332642,
+      "learning_rate": 5.783115637162446e-06,
+      "loss": 0.8709,
+      "step": 3559
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2877849340438843,
+      "learning_rate": 5.781019349605688e-06,
+      "loss": 0.8038,
+      "step": 3560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1851149797439575,
+      "learning_rate": 5.778922921320861e-06,
+      "loss": 0.8133,
+      "step": 3561
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2251300811767578,
+      "learning_rate": 5.776826352685712e-06,
+      "loss": 0.8327,
+      "step": 3562
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2545593976974487,
+      "learning_rate": 5.7747296440780095e-06,
+      "loss": 0.823,
+      "step": 3563
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2170064449310303,
+      "learning_rate": 5.7726327958755535e-06,
+      "loss": 0.8011,
+      "step": 3564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2891184091567993,
+      "learning_rate": 5.7705358084561615e-06,
+      "loss": 0.8374,
+      "step": 3565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2579395771026611,
+      "learning_rate": 5.76843868219768e-06,
+      "loss": 0.8794,
+      "step": 3566
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2458479404449463,
+      "learning_rate": 5.766341417477979e-06,
+      "loss": 0.8044,
+      "step": 3567
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2657225131988525,
+      "learning_rate": 5.764244014674955e-06,
+      "loss": 0.791,
+      "step": 3568
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1639002561569214,
+      "learning_rate": 5.762146474166531e-06,
+      "loss": 0.821,
+      "step": 3569
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2115799188613892,
+      "learning_rate": 5.760048796330651e-06,
+      "loss": 0.821,
+      "step": 3570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.874882161617279,
+      "learning_rate": 5.757950981545283e-06,
+      "loss": 1.0461,
+      "step": 3571
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2547259330749512,
+      "learning_rate": 5.755853030188426e-06,
+      "loss": 0.7663,
+      "step": 3572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6305436491966248,
+      "learning_rate": 5.753754942638097e-06,
+      "loss": 1.0436,
+      "step": 3573
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6401702165603638,
+      "learning_rate": 5.751656719272343e-06,
+      "loss": 1.0214,
+      "step": 3574
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2869157791137695,
+      "learning_rate": 5.749558360469227e-06,
+      "loss": 0.8756,
+      "step": 3575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3040639162063599,
+      "learning_rate": 5.747459866606848e-06,
+      "loss": 0.8744,
+      "step": 3576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7446020245552063,
+      "learning_rate": 5.74536123806332e-06,
+      "loss": 1.0366,
+      "step": 3577
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.236698031425476,
+      "learning_rate": 5.743262475216789e-06,
+      "loss": 0.9103,
+      "step": 3578
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1982958316802979,
+      "learning_rate": 5.741163578445412e-06,
+      "loss": 0.8033,
+      "step": 3579
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2317880392074585,
+      "learning_rate": 5.739064548127389e-06,
+      "loss": 0.8172,
+      "step": 3580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2607005834579468,
+      "learning_rate": 5.736965384640927e-06,
+      "loss": 0.8175,
+      "step": 3581
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.219626545906067,
+      "learning_rate": 5.734866088364268e-06,
+      "loss": 0.8854,
+      "step": 3582
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.235374927520752,
+      "learning_rate": 5.732766659675671e-06,
+      "loss": 0.7873,
+      "step": 3583
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2334736585617065,
+      "learning_rate": 5.7306670989534244e-06,
+      "loss": 0.8681,
+      "step": 3584
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1722739934921265,
+      "learning_rate": 5.728567406575837e-06,
+      "loss": 0.8167,
+      "step": 3585
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.244261384010315,
+      "learning_rate": 5.726467582921241e-06,
+      "loss": 0.8244,
+      "step": 3586
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.778134822845459,
+      "learning_rate": 5.724367628367995e-06,
+      "loss": 1.0266,
+      "step": 3587
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3254276514053345,
+      "learning_rate": 5.7222675432944786e-06,
+      "loss": 0.8552,
+      "step": 3588
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3024260997772217,
+      "learning_rate": 5.720167328079096e-06,
+      "loss": 0.8225,
+      "step": 3589
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2845851182937622,
+      "learning_rate": 5.718066983100274e-06,
+      "loss": 0.7672,
+      "step": 3590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4071017503738403,
+      "learning_rate": 5.7159665087364655e-06,
+      "loss": 0.7587,
+      "step": 3591
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6770083904266357,
+      "learning_rate": 5.7138659053661435e-06,
+      "loss": 1.0355,
+      "step": 3592
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2659573554992676,
+      "learning_rate": 5.711765173367806e-06,
+      "loss": 0.8456,
+      "step": 3593
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2069636583328247,
+      "learning_rate": 5.709664313119972e-06,
+      "loss": 0.8328,
+      "step": 3594
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1837430000305176,
+      "learning_rate": 5.707563325001188e-06,
+      "loss": 0.7443,
+      "step": 3595
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2409985065460205,
+      "learning_rate": 5.705462209390018e-06,
+      "loss": 0.8284,
+      "step": 3596
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.318311095237732,
+      "learning_rate": 5.703360966665053e-06,
+      "loss": 0.8634,
+      "step": 3597
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3050683736801147,
+      "learning_rate": 5.701259597204906e-06,
+      "loss": 0.8288,
+      "step": 3598
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2556500434875488,
+      "learning_rate": 5.699158101388214e-06,
+      "loss": 0.7941,
+      "step": 3599
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6599319577217102,
+      "learning_rate": 5.6970564795936314e-06,
+      "loss": 1.0206,
+      "step": 3600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.247117519378662,
+      "learning_rate": 5.694954732199842e-06,
+      "loss": 0.8072,
+      "step": 3601
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6226639151573181,
+      "learning_rate": 5.692852859585548e-06,
+      "loss": 1.0444,
+      "step": 3602
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2146644592285156,
+      "learning_rate": 5.690750862129477e-06,
+      "loss": 0.7971,
+      "step": 3603
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.621841311454773,
+      "learning_rate": 5.688648740210377e-06,
+      "loss": 1.033,
+      "step": 3604
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2343666553497314,
+      "learning_rate": 5.686546494207019e-06,
+      "loss": 0.8343,
+      "step": 3605
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2563185691833496,
+      "learning_rate": 5.684444124498197e-06,
+      "loss": 0.8729,
+      "step": 3606
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6410859823226929,
+      "learning_rate": 5.682341631462725e-06,
+      "loss": 1.0323,
+      "step": 3607
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.31291663646698,
+      "learning_rate": 5.680239015479443e-06,
+      "loss": 0.8311,
+      "step": 3608
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1924583911895752,
+      "learning_rate": 5.678136276927212e-06,
+      "loss": 0.7948,
+      "step": 3609
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2347559928894043,
+      "learning_rate": 5.67603341618491e-06,
+      "loss": 0.7951,
+      "step": 3610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6439375281333923,
+      "learning_rate": 5.673930433631445e-06,
+      "loss": 1.018,
+      "step": 3611
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6606663465499878,
+      "learning_rate": 5.6718273296457415e-06,
+      "loss": 1.0285,
+      "step": 3612
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3166265487670898,
+      "learning_rate": 5.669724104606749e-06,
+      "loss": 0.8503,
+      "step": 3613
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6097078919410706,
+      "learning_rate": 5.667620758893435e-06,
+      "loss": 1.0423,
+      "step": 3614
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2725721597671509,
+      "learning_rate": 5.6655172928847935e-06,
+      "loss": 0.8128,
+      "step": 3615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6340359449386597,
+      "learning_rate": 5.663413706959835e-06,
+      "loss": 1.0244,
+      "step": 3616
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6443120241165161,
+      "learning_rate": 5.6613100014975985e-06,
+      "loss": 1.0406,
+      "step": 3617
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.237058162689209,
+      "learning_rate": 5.659206176877137e-06,
+      "loss": 0.8204,
+      "step": 3618
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2507939338684082,
+      "learning_rate": 5.65710223347753e-06,
+      "loss": 0.822,
+      "step": 3619
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6208112835884094,
+      "learning_rate": 5.654998171677875e-06,
+      "loss": 1.0397,
+      "step": 3620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2317724227905273,
+      "learning_rate": 5.652893991857294e-06,
+      "loss": 0.8077,
+      "step": 3621
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6199010014533997,
+      "learning_rate": 5.65078969439493e-06,
+      "loss": 1.0267,
+      "step": 3622
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6230006814002991,
+      "learning_rate": 5.648685279669944e-06,
+      "loss": 1.0356,
+      "step": 3623
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.260152816772461,
+      "learning_rate": 5.646580748061522e-06,
+      "loss": 0.8536,
+      "step": 3624
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3621753454208374,
+      "learning_rate": 5.644476099948868e-06,
+      "loss": 0.7895,
+      "step": 3625
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2357532978057861,
+      "learning_rate": 5.642371335711209e-06,
+      "loss": 0.8136,
+      "step": 3626
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2278600931167603,
+      "learning_rate": 5.640266455727791e-06,
+      "loss": 0.8412,
+      "step": 3627
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2222795486450195,
+      "learning_rate": 5.6381614603778835e-06,
+      "loss": 0.7981,
+      "step": 3628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6724081039428711,
+      "learning_rate": 5.636056350040774e-06,
+      "loss": 1.0337,
+      "step": 3629
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.220899224281311,
+      "learning_rate": 5.633951125095774e-06,
+      "loss": 0.8088,
+      "step": 3630
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.224700689315796,
+      "learning_rate": 5.6318457859222135e-06,
+      "loss": 0.8241,
+      "step": 3631
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2473098039627075,
+      "learning_rate": 5.62974033289944e-06,
+      "loss": 0.8524,
+      "step": 3632
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1748511791229248,
+      "learning_rate": 5.627634766406828e-06,
+      "loss": 0.8172,
+      "step": 3633
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.653656542301178,
+      "learning_rate": 5.625529086823769e-06,
+      "loss": 1.0401,
+      "step": 3634
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2892436981201172,
+      "learning_rate": 5.623423294529674e-06,
+      "loss": 0.9045,
+      "step": 3635
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3364299535751343,
+      "learning_rate": 5.621317389903977e-06,
+      "loss": 0.8326,
+      "step": 3636
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3062598705291748,
+      "learning_rate": 5.619211373326128e-06,
+      "loss": 0.8786,
+      "step": 3637
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6582712531089783,
+      "learning_rate": 5.617105245175604e-06,
+      "loss": 1.0297,
+      "step": 3638
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3249176740646362,
+      "learning_rate": 5.614999005831894e-06,
+      "loss": 0.796,
+      "step": 3639
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6600614190101624,
+      "learning_rate": 5.612892655674512e-06,
+      "loss": 1.0474,
+      "step": 3640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6455150842666626,
+      "learning_rate": 5.610786195082991e-06,
+      "loss": 1.0023,
+      "step": 3641
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2815417051315308,
+      "learning_rate": 5.608679624436887e-06,
+      "loss": 0.8071,
+      "step": 3642
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6448112726211548,
+      "learning_rate": 5.606572944115768e-06,
+      "loss": 1.0514,
+      "step": 3643
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2829891443252563,
+      "learning_rate": 5.604466154499229e-06,
+      "loss": 0.8207,
+      "step": 3644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2258154153823853,
+      "learning_rate": 5.602359255966881e-06,
+      "loss": 0.8167,
+      "step": 3645
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2290526628494263,
+      "learning_rate": 5.600252248898358e-06,
+      "loss": 0.837,
+      "step": 3646
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6728348731994629,
+      "learning_rate": 5.598145133673308e-06,
+      "loss": 1.0289,
+      "step": 3647
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6895084381103516,
+      "learning_rate": 5.596037910671405e-06,
+      "loss": 1.0336,
+      "step": 3648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2612271308898926,
+      "learning_rate": 5.593930580272337e-06,
+      "loss": 0.8204,
+      "step": 3649
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.501835584640503,
+      "learning_rate": 5.591823142855815e-06,
+      "loss": 0.8509,
+      "step": 3650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6214595437049866,
+      "learning_rate": 5.589715598801566e-06,
+      "loss": 1.0166,
+      "step": 3651
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2956233024597168,
+      "learning_rate": 5.5876079484893395e-06,
+      "loss": 0.8654,
+      "step": 3652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2359542846679688,
+      "learning_rate": 5.585500192298903e-06,
+      "loss": 0.7944,
+      "step": 3653
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2007431983947754,
+      "learning_rate": 5.583392330610042e-06,
+      "loss": 0.8052,
+      "step": 3654
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1995011568069458,
+      "learning_rate": 5.581284363802564e-06,
+      "loss": 0.7682,
+      "step": 3655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.26311194896698,
+      "learning_rate": 5.579176292256288e-06,
+      "loss": 0.812,
+      "step": 3656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.265154480934143,
+      "learning_rate": 5.577068116351063e-06,
+      "loss": 0.8217,
+      "step": 3657
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1908860206604004,
+      "learning_rate": 5.574959836466748e-06,
+      "loss": 0.8106,
+      "step": 3658
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8370785713195801,
+      "learning_rate": 5.572851452983226e-06,
+      "loss": 1.0288,
+      "step": 3659
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7683365941047668,
+      "learning_rate": 5.570742966280392e-06,
+      "loss": 1.0178,
+      "step": 3660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.32356858253479,
+      "learning_rate": 5.568634376738167e-06,
+      "loss": 0.8405,
+      "step": 3661
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.193405032157898,
+      "learning_rate": 5.566525684736487e-06,
+      "loss": 0.8251,
+      "step": 3662
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2772588729858398,
+      "learning_rate": 5.5644168906553076e-06,
+      "loss": 0.7808,
+      "step": 3663
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3004111051559448,
+      "learning_rate": 5.5623079948746006e-06,
+      "loss": 0.8173,
+      "step": 3664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0003604888916016,
+      "learning_rate": 5.560198997774358e-06,
+      "loss": 1.0291,
+      "step": 3665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8521328568458557,
+      "learning_rate": 5.55808989973459e-06,
+      "loss": 1.0453,
+      "step": 3666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3993885517120361,
+      "learning_rate": 5.555980701135326e-06,
+      "loss": 0.7185,
+      "step": 3667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2273666858673096,
+      "learning_rate": 5.553871402356611e-06,
+      "loss": 0.7533,
+      "step": 3668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8929486870765686,
+      "learning_rate": 5.551762003778508e-06,
+      "loss": 1.0219,
+      "step": 3669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3706204891204834,
+      "learning_rate": 5.549652505781101e-06,
+      "loss": 0.8264,
+      "step": 3670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2929034233093262,
+      "learning_rate": 5.547542908744491e-06,
+      "loss": 0.8656,
+      "step": 3671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2844951152801514,
+      "learning_rate": 5.545433213048793e-06,
+      "loss": 0.8428,
+      "step": 3672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2988768815994263,
+      "learning_rate": 5.5433234190741445e-06,
+      "loss": 0.7688,
+      "step": 3673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2648602724075317,
+      "learning_rate": 5.541213527200699e-06,
+      "loss": 0.8066,
+      "step": 3674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3700276613235474,
+      "learning_rate": 5.539103537808629e-06,
+      "loss": 0.8278,
+      "step": 3675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2720286846160889,
+      "learning_rate": 5.536993451278121e-06,
+      "loss": 0.7819,
+      "step": 3676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9234135746955872,
+      "learning_rate": 5.534883267989381e-06,
+      "loss": 1.0488,
+      "step": 3677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.243483543395996,
+      "learning_rate": 5.5327729883226345e-06,
+      "loss": 0.8191,
+      "step": 3678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6694689989089966,
+      "learning_rate": 5.530662612658122e-06,
+      "loss": 1.0413,
+      "step": 3679
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2914083003997803,
+      "learning_rate": 5.5285521413761e-06,
+      "loss": 0.8411,
+      "step": 3680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.281036138534546,
+      "learning_rate": 5.526441574856845e-06,
+      "loss": 0.8082,
+      "step": 3681
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3099288940429688,
+      "learning_rate": 5.524330913480652e-06,
+      "loss": 0.8389,
+      "step": 3682
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3002359867095947,
+      "learning_rate": 5.522220157627829e-06,
+      "loss": 0.8008,
+      "step": 3683
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3056786060333252,
+      "learning_rate": 5.520109307678702e-06,
+      "loss": 0.874,
+      "step": 3684
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3572965860366821,
+      "learning_rate": 5.517998364013615e-06,
+      "loss": 0.8748,
+      "step": 3685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.261283278465271,
+      "learning_rate": 5.5158873270129286e-06,
+      "loss": 0.8275,
+      "step": 3686
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2971839904785156,
+      "learning_rate": 5.513776197057023e-06,
+      "loss": 0.8686,
+      "step": 3687
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3075042963027954,
+      "learning_rate": 5.511664974526288e-06,
+      "loss": 0.8063,
+      "step": 3688
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.113528847694397,
+      "learning_rate": 5.509553659801137e-06,
+      "loss": 1.0291,
+      "step": 3689
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2869353294372559,
+      "learning_rate": 5.507442253261997e-06,
+      "loss": 0.8072,
+      "step": 3690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3126065731048584,
+      "learning_rate": 5.505330755289315e-06,
+      "loss": 0.83,
+      "step": 3691
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3140103816986084,
+      "learning_rate": 5.5032191662635445e-06,
+      "loss": 0.8272,
+      "step": 3692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3165065050125122,
+      "learning_rate": 5.501107486565168e-06,
+      "loss": 0.8161,
+      "step": 3693
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6459331512451172,
+      "learning_rate": 5.498995716574679e-06,
+      "loss": 1.0175,
+      "step": 3694
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2206908464431763,
+      "learning_rate": 5.496883856672584e-06,
+      "loss": 0.7941,
+      "step": 3695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.264884352684021,
+      "learning_rate": 5.4947719072394115e-06,
+      "loss": 0.8242,
+      "step": 3696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2281174659729004,
+      "learning_rate": 5.4926598686557e-06,
+      "loss": 0.8526,
+      "step": 3697
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3030858039855957,
+      "learning_rate": 5.490547741302012e-06,
+      "loss": 0.7843,
+      "step": 3698
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2528736591339111,
+      "learning_rate": 5.488435525558917e-06,
+      "loss": 0.8215,
+      "step": 3699
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.381070613861084,
+      "learning_rate": 5.486323221807009e-06,
+      "loss": 0.8298,
+      "step": 3700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2076992988586426,
+      "learning_rate": 5.484210830426888e-06,
+      "loss": 0.8243,
+      "step": 3701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2101858854293823,
+      "learning_rate": 5.482098351799181e-06,
+      "loss": 0.8244,
+      "step": 3702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2857677936553955,
+      "learning_rate": 5.4799857863045235e-06,
+      "loss": 0.8705,
+      "step": 3703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2770832777023315,
+      "learning_rate": 5.477873134323568e-06,
+      "loss": 0.8809,
+      "step": 3704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2400920391082764,
+      "learning_rate": 5.475760396236982e-06,
+      "loss": 0.8292,
+      "step": 3705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2571582794189453,
+      "learning_rate": 5.473647572425451e-06,
+      "loss": 0.8877,
+      "step": 3706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2583686113357544,
+      "learning_rate": 5.471534663269674e-06,
+      "loss": 0.785,
+      "step": 3707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3131728172302246,
+      "learning_rate": 5.469421669150367e-06,
+      "loss": 0.8541,
+      "step": 3708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2473318576812744,
+      "learning_rate": 5.467308590448257e-06,
+      "loss": 0.8727,
+      "step": 3709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.266913890838623,
+      "learning_rate": 5.465195427544092e-06,
+      "loss": 0.8298,
+      "step": 3710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2895218133926392,
+      "learning_rate": 5.4630821808186315e-06,
+      "loss": 0.8163,
+      "step": 3711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.260604977607727,
+      "learning_rate": 5.460968850652653e-06,
+      "loss": 0.8373,
+      "step": 3712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8249561786651611,
+      "learning_rate": 5.458855437426942e-06,
+      "loss": 1.0264,
+      "step": 3713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2629518508911133,
+      "learning_rate": 5.45674194152231e-06,
+      "loss": 0.8306,
+      "step": 3714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6710664629936218,
+      "learning_rate": 5.4546283633195744e-06,
+      "loss": 1.0379,
+      "step": 3715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6715431809425354,
+      "learning_rate": 5.4525147031995715e-06,
+      "loss": 1.0391,
+      "step": 3716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2822651863098145,
+      "learning_rate": 5.4504009615431505e-06,
+      "loss": 0.7843,
+      "step": 3717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3228788375854492,
+      "learning_rate": 5.448287138731179e-06,
+      "loss": 0.8583,
+      "step": 3718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7353352904319763,
+      "learning_rate": 5.446173235144532e-06,
+      "loss": 1.0224,
+      "step": 3719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.253600835800171,
+      "learning_rate": 5.4440592511641065e-06,
+      "loss": 0.8567,
+      "step": 3720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.256108283996582,
+      "learning_rate": 5.441945187170809e-06,
+      "loss": 0.8878,
+      "step": 3721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.207448959350586,
+      "learning_rate": 5.4398310435455655e-06,
+      "loss": 0.7742,
+      "step": 3722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2476434707641602,
+      "learning_rate": 5.437716820669311e-06,
+      "loss": 0.8662,
+      "step": 3723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2930818796157837,
+      "learning_rate": 5.4356025189229975e-06,
+      "loss": 0.7847,
+      "step": 3724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2214897871017456,
+      "learning_rate": 5.43348813868759e-06,
+      "loss": 0.8322,
+      "step": 3725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2186142206192017,
+      "learning_rate": 5.431373680344069e-06,
+      "loss": 0.8272,
+      "step": 3726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1909555196762085,
+      "learning_rate": 5.429259144273428e-06,
+      "loss": 0.792,
+      "step": 3727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9049088954925537,
+      "learning_rate": 5.427144530856676e-06,
+      "loss": 1.0265,
+      "step": 3728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7593244314193726,
+      "learning_rate": 5.4250298404748335e-06,
+      "loss": 1.0192,
+      "step": 3729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3197728395462036,
+      "learning_rate": 5.4229150735089365e-06,
+      "loss": 0.8552,
+      "step": 3730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.373611330986023,
+      "learning_rate": 5.420800230340036e-06,
+      "loss": 0.8186,
+      "step": 3731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6483821868896484,
+      "learning_rate": 5.4186853113491945e-06,
+      "loss": 1.0219,
+      "step": 3732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.673839271068573,
+      "learning_rate": 5.416570316917488e-06,
+      "loss": 1.0179,
+      "step": 3733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3298759460449219,
+      "learning_rate": 5.414455247426008e-06,
+      "loss": 0.8784,
+      "step": 3734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3680555820465088,
+      "learning_rate": 5.412340103255858e-06,
+      "loss": 0.8321,
+      "step": 3735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.692715048789978,
+      "learning_rate": 5.410224884788157e-06,
+      "loss": 1.0343,
+      "step": 3736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2255454063415527,
+      "learning_rate": 5.408109592404033e-06,
+      "loss": 0.7974,
+      "step": 3737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2554419040679932,
+      "learning_rate": 5.405994226484632e-06,
+      "loss": 0.7988,
+      "step": 3738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.274903416633606,
+      "learning_rate": 5.403878787411113e-06,
+      "loss": 0.8392,
+      "step": 3739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3477505445480347,
+      "learning_rate": 5.401763275564644e-06,
+      "loss": 0.8381,
+      "step": 3740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3236993551254272,
+      "learning_rate": 5.3996476913264095e-06,
+      "loss": 0.9227,
+      "step": 3741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2448089122772217,
+      "learning_rate": 5.3975320350776075e-06,
+      "loss": 0.7941,
+      "step": 3742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2012310028076172,
+      "learning_rate": 5.395416307199447e-06,
+      "loss": 0.8645,
+      "step": 3743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6615229845046997,
+      "learning_rate": 5.393300508073149e-06,
+      "loss": 1.025,
+      "step": 3744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2650922536849976,
+      "learning_rate": 5.391184638079951e-06,
+      "loss": 0.7814,
+      "step": 3745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2357031106948853,
+      "learning_rate": 5.389068697601102e-06,
+      "loss": 0.8966,
+      "step": 3746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2525899410247803,
+      "learning_rate": 5.3869526870178625e-06,
+      "loss": 0.8508,
+      "step": 3747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6687286496162415,
+      "learning_rate": 5.384836606711503e-06,
+      "loss": 1.0215,
+      "step": 3748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2171417474746704,
+      "learning_rate": 5.382720457063314e-06,
+      "loss": 0.8225,
+      "step": 3749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2321946620941162,
+      "learning_rate": 5.380604238454592e-06,
+      "loss": 0.8263,
+      "step": 3750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.295219898223877,
+      "learning_rate": 5.378487951266649e-06,
+      "loss": 0.8503,
+      "step": 3751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2108001708984375,
+      "learning_rate": 5.376371595880807e-06,
+      "loss": 0.751,
+      "step": 3752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.282712697982788,
+      "learning_rate": 5.374255172678403e-06,
+      "loss": 0.8818,
+      "step": 3753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6875323057174683,
+      "learning_rate": 5.372138682040784e-06,
+      "loss": 1.0123,
+      "step": 3754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2823617458343506,
+      "learning_rate": 5.3700221243493135e-06,
+      "loss": 0.812,
+      "step": 3755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6630176305770874,
+      "learning_rate": 5.367905499985359e-06,
+      "loss": 1.0047,
+      "step": 3756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2903192043304443,
+      "learning_rate": 5.3657888093303055e-06,
+      "loss": 0.7442,
+      "step": 3757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2819364070892334,
+      "learning_rate": 5.363672052765552e-06,
+      "loss": 0.8429,
+      "step": 3758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2727398872375488,
+      "learning_rate": 5.361555230672506e-06,
+      "loss": 0.8996,
+      "step": 3759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2559232711791992,
+      "learning_rate": 5.3594383434325845e-06,
+      "loss": 0.7903,
+      "step": 3760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2470085620880127,
+      "learning_rate": 5.357321391427221e-06,
+      "loss": 0.8545,
+      "step": 3761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2827861309051514,
+      "learning_rate": 5.355204375037859e-06,
+      "loss": 0.802,
+      "step": 3762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2919270992279053,
+      "learning_rate": 5.353087294645954e-06,
+      "loss": 0.8602,
+      "step": 3763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3060885667800903,
+      "learning_rate": 5.35097015063297e-06,
+      "loss": 0.8347,
+      "step": 3764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2044572830200195,
+      "learning_rate": 5.348852943380387e-06,
+      "loss": 0.8157,
+      "step": 3765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7870275974273682,
+      "learning_rate": 5.346735673269693e-06,
+      "loss": 1.0008,
+      "step": 3766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.233031153678894,
+      "learning_rate": 5.344618340682389e-06,
+      "loss": 0.7419,
+      "step": 3767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2234835624694824,
+      "learning_rate": 5.342500945999988e-06,
+      "loss": 0.8521,
+      "step": 3768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6878315210342407,
+      "learning_rate": 5.340383489604011e-06,
+      "loss": 1.0245,
+      "step": 3769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3682324886322021,
+      "learning_rate": 5.3382659718759935e-06,
+      "loss": 0.8454,
+      "step": 3770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6425873041152954,
+      "learning_rate": 5.336148393197479e-06,
+      "loss": 1.045,
+      "step": 3771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3056650161743164,
+      "learning_rate": 5.334030753950027e-06,
+      "loss": 0.8384,
+      "step": 3772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2537692785263062,
+      "learning_rate": 5.331913054515201e-06,
+      "loss": 0.7555,
+      "step": 3773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2701579332351685,
+      "learning_rate": 5.329795295274582e-06,
+      "loss": 0.9062,
+      "step": 3774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.261278748512268,
+      "learning_rate": 5.327677476609755e-06,
+      "loss": 0.8344,
+      "step": 3775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.303965449333191,
+      "learning_rate": 5.3255595989023235e-06,
+      "loss": 0.8339,
+      "step": 3776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.271486759185791,
+      "learning_rate": 5.3234416625338945e-06,
+      "loss": 0.8336,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7664963006973267,
+      "learning_rate": 5.32132366788609e-06,
+      "loss": 1.0206,
+      "step": 3778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.212839126586914,
+      "learning_rate": 5.3192056153405415e-06,
+      "loss": 0.8284,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.706427812576294,
+      "learning_rate": 5.317087505278889e-06,
+      "loss": 1.0383,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6075137257575989,
+      "learning_rate": 5.314969338082786e-06,
+      "loss": 1.0391,
+      "step": 3781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.242242455482483,
+      "learning_rate": 5.312851114133895e-06,
+      "loss": 0.8498,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6618263721466064,
+      "learning_rate": 5.310732833813885e-06,
+      "loss": 1.0319,
+      "step": 3783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2650259733200073,
+      "learning_rate": 5.308614497504444e-06,
+      "loss": 0.8188,
+      "step": 3784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2485965490341187,
+      "learning_rate": 5.30649610558726e-06,
+      "loss": 0.8202,
+      "step": 3785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.252366065979004,
+      "learning_rate": 5.304377658444038e-06,
+      "loss": 0.8145,
+      "step": 3786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2161319255828857,
+      "learning_rate": 5.3022591564564895e-06,
+      "loss": 0.8509,
+      "step": 3787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7632567882537842,
+      "learning_rate": 5.300140600006339e-06,
+      "loss": 1.0093,
+      "step": 3788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.278090476989746,
+      "learning_rate": 5.298021989475317e-06,
+      "loss": 0.8232,
+      "step": 3789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.302208662033081,
+      "learning_rate": 5.2959033252451666e-06,
+      "loss": 0.8176,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3507975339889526,
+      "learning_rate": 5.293784607697638e-06,
+      "loss": 0.8235,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6717748045921326,
+      "learning_rate": 5.291665837214496e-06,
+      "loss": 1.0299,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.290799617767334,
+      "learning_rate": 5.289547014177507e-06,
+      "loss": 0.8377,
+      "step": 3793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1942156553268433,
+      "learning_rate": 5.287428138968455e-06,
+      "loss": 0.8061,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2652878761291504,
+      "learning_rate": 5.2853092119691265e-06,
+      "loss": 0.7612,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2828965187072754,
+      "learning_rate": 5.283190233561325e-06,
+      "loss": 0.8444,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7088223695755005,
+      "learning_rate": 5.281071204126855e-06,
+      "loss": 1.026,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.310884714126587,
+      "learning_rate": 5.278952124047535e-06,
+      "loss": 0.8728,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.287872076034546,
+      "learning_rate": 5.276832993705193e-06,
+      "loss": 0.8404,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6395774483680725,
+      "learning_rate": 5.274713813481666e-06,
+      "loss": 1.0252,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2399721145629883,
+      "learning_rate": 5.272594583758795e-06,
+      "loss": 0.8078,
+      "step": 3801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.195039987564087,
+      "learning_rate": 5.270475304918434e-06,
+      "loss": 0.7856,
+      "step": 3802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6476578712463379,
+      "learning_rate": 5.268355977342448e-06,
+      "loss": 1.0263,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6331608891487122,
+      "learning_rate": 5.266236601412706e-06,
+      "loss": 1.0208,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2592217922210693,
+      "learning_rate": 5.264117177511092e-06,
+      "loss": 0.8491,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.288292407989502,
+      "learning_rate": 5.261997706019491e-06,
+      "loss": 0.8199,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.25655996799469,
+      "learning_rate": 5.2598781873198e-06,
+      "loss": 0.8718,
+      "step": 3807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2150869369506836,
+      "learning_rate": 5.257758621793926e-06,
+      "loss": 0.7796,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2136235237121582,
+      "learning_rate": 5.255639009823785e-06,
+      "loss": 0.7643,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2962703704833984,
+      "learning_rate": 5.253519351791295e-06,
+      "loss": 0.8061,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1923867464065552,
+      "learning_rate": 5.251399648078391e-06,
+      "loss": 0.8492,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2929627895355225,
+      "learning_rate": 5.24927989906701e-06,
+      "loss": 0.8188,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3183610439300537,
+      "learning_rate": 5.2471601051391005e-06,
+      "loss": 0.869,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1831284761428833,
+      "learning_rate": 5.245040266676617e-06,
+      "loss": 0.8102,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2525405883789062,
+      "learning_rate": 5.2429203840615205e-06,
+      "loss": 0.831,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9717052578926086,
+      "learning_rate": 5.240800457675787e-06,
+      "loss": 1.0257,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2606761455535889,
+      "learning_rate": 5.238680487901393e-06,
+      "loss": 0.7945,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2375012636184692,
+      "learning_rate": 5.236560475120326e-06,
+      "loss": 0.7676,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2857567071914673,
+      "learning_rate": 5.23444041971458e-06,
+      "loss": 0.8184,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7188566327095032,
+      "learning_rate": 5.232320322066159e-06,
+      "loss": 1.0323,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.190695881843567,
+      "learning_rate": 5.230200182557073e-06,
+      "loss": 0.8395,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7579325437545776,
+      "learning_rate": 5.228080001569338e-06,
+      "loss": 1.0197,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2737318277359009,
+      "learning_rate": 5.225959779484979e-06,
+      "loss": 0.8066,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2612446546554565,
+      "learning_rate": 5.2238395166860286e-06,
+      "loss": 0.8358,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3024219274520874,
+      "learning_rate": 5.22171921355453e-06,
+      "loss": 0.8837,
+      "step": 3825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7385522127151489,
+      "learning_rate": 5.219598870472526e-06,
+      "loss": 1.0486,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1999852657318115,
+      "learning_rate": 5.217478487822073e-06,
+      "loss": 0.7485,
+      "step": 3827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2073436975479126,
+      "learning_rate": 5.215358065985232e-06,
+      "loss": 0.8323,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7088455557823181,
+      "learning_rate": 5.213237605344072e-06,
+      "loss": 1.0185,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6694773435592651,
+      "learning_rate": 5.211117106280667e-06,
+      "loss": 1.0188,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2977824211120605,
+      "learning_rate": 5.2089965691770995e-06,
+      "loss": 0.8574,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2999382019042969,
+      "learning_rate": 5.20687599441546e-06,
+      "loss": 0.8395,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2322931289672852,
+      "learning_rate": 5.204755382377845e-06,
+      "loss": 0.8341,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2505282163619995,
+      "learning_rate": 5.2026347334463555e-06,
+      "loss": 0.8142,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2214823961257935,
+      "learning_rate": 5.200514048003102e-06,
+      "loss": 0.887,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2464874982833862,
+      "learning_rate": 5.1983933264302e-06,
+      "loss": 0.8515,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2474582195281982,
+      "learning_rate": 5.1962725691097725e-06,
+      "loss": 0.8253,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4248026609420776,
+      "learning_rate": 5.1941517764239465e-06,
+      "loss": 0.7774,
+      "step": 3838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3391191959381104,
+      "learning_rate": 5.192030948754859e-06,
+      "loss": 0.8666,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2575328350067139,
+      "learning_rate": 5.189910086484653e-06,
+      "loss": 0.7869,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.186660885810852,
+      "learning_rate": 5.187789189995474e-06,
+      "loss": 1.0143,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2443050146102905,
+      "learning_rate": 5.185668259669477e-06,
+      "loss": 0.7737,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7942463159561157,
+      "learning_rate": 5.18354729588882e-06,
+      "loss": 1.0273,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.262618899345398,
+      "learning_rate": 5.1814262990356735e-06,
+      "loss": 0.7957,
+      "step": 3844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.753009021282196,
+      "learning_rate": 5.179305269492205e-06,
+      "loss": 1.0426,
+      "step": 3845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3127014636993408,
+      "learning_rate": 5.177184207640597e-06,
+      "loss": 0.85,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3251585960388184,
+      "learning_rate": 5.175063113863027e-06,
+      "loss": 0.8293,
+      "step": 3847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.244962215423584,
+      "learning_rate": 5.172941988541691e-06,
+      "loss": 0.8186,
+      "step": 3848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2452973127365112,
+      "learning_rate": 5.1708208320587806e-06,
+      "loss": 0.8314,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.211859941482544,
+      "learning_rate": 5.168699644796499e-06,
+      "loss": 0.8182,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.223374605178833,
+      "learning_rate": 5.166578427137048e-06,
+      "loss": 0.7812,
+      "step": 3851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3419219255447388,
+      "learning_rate": 5.164457179462645e-06,
+      "loss": 0.868,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.26482093334198,
+      "learning_rate": 5.162335902155506e-06,
+      "loss": 0.8407,
+      "step": 3853
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2369862794876099,
+      "learning_rate": 5.160214595597851e-06,
+      "loss": 0.8317,
+      "step": 3854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2562150955200195,
+      "learning_rate": 5.158093260171908e-06,
+      "loss": 0.8308,
+      "step": 3855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2474015951156616,
+      "learning_rate": 5.155971896259914e-06,
+      "loss": 0.8706,
+      "step": 3856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2250806093215942,
+      "learning_rate": 5.153850504244102e-06,
+      "loss": 0.7354,
+      "step": 3857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1269642114639282,
+      "learning_rate": 5.1517290845067216e-06,
+      "loss": 1.0487,
+      "step": 3858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2645751237869263,
+      "learning_rate": 5.149607637430014e-06,
+      "loss": 0.8594,
+      "step": 3859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7277780175209045,
+      "learning_rate": 5.147486163396238e-06,
+      "loss": 1.0141,
+      "step": 3860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2691224813461304,
+      "learning_rate": 5.145364662787649e-06,
+      "loss": 0.8302,
+      "step": 3861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.246394157409668,
+      "learning_rate": 5.14324313598651e-06,
+      "loss": 0.7415,
+      "step": 3862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.285098910331726,
+      "learning_rate": 5.141121583375088e-06,
+      "loss": 0.8612,
+      "step": 3863
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2099387645721436,
+      "learning_rate": 5.139000005335658e-06,
+      "loss": 0.8491,
+      "step": 3864
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.250780463218689,
+      "learning_rate": 5.136878402250494e-06,
+      "loss": 0.8247,
+      "step": 3865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9379558563232422,
+      "learning_rate": 5.1347567745018785e-06,
+      "loss": 1.0257,
+      "step": 3866
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3153250217437744,
+      "learning_rate": 5.132635122472096e-06,
+      "loss": 0.8715,
+      "step": 3867
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2414952516555786,
+      "learning_rate": 5.1305134465434395e-06,
+      "loss": 0.7761,
+      "step": 3868
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3137282133102417,
+      "learning_rate": 5.1283917470982005e-06,
+      "loss": 0.8487,
+      "step": 3869
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3460074663162231,
+      "learning_rate": 5.126270024518681e-06,
+      "loss": 0.8044,
+      "step": 3870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2966498136520386,
+      "learning_rate": 5.124148279187177e-06,
+      "loss": 0.8378,
+      "step": 3871
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2201626300811768,
+      "learning_rate": 5.122026511486003e-06,
+      "loss": 0.8622,
+      "step": 3872
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2490122318267822,
+      "learning_rate": 5.1199047217974675e-06,
+      "loss": 0.7681,
+      "step": 3873
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2518165111541748,
+      "learning_rate": 5.117782910503884e-06,
+      "loss": 0.8111,
+      "step": 3874
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.704135000705719,
+      "learning_rate": 5.115661077987571e-06,
+      "loss": 1.024,
+      "step": 3875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6668209433555603,
+      "learning_rate": 5.1135392246308534e-06,
+      "loss": 1.016,
+      "step": 3876
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.29795503616333,
+      "learning_rate": 5.111417350816058e-06,
+      "loss": 0.8275,
+      "step": 3877
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3047642707824707,
+      "learning_rate": 5.109295456925513e-06,
+      "loss": 0.8416,
+      "step": 3878
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2719776630401611,
+      "learning_rate": 5.107173543341551e-06,
+      "loss": 0.7945,
+      "step": 3879
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6863534450531006,
+      "learning_rate": 5.1050516104465135e-06,
+      "loss": 1.018,
+      "step": 3880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2827975749969482,
+      "learning_rate": 5.1029296586227375e-06,
+      "loss": 0.8818,
+      "step": 3881
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.336637020111084,
+      "learning_rate": 5.100807688252568e-06,
+      "loss": 0.835,
+      "step": 3882
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2432619333267212,
+      "learning_rate": 5.0986856997183524e-06,
+      "loss": 0.8277,
+      "step": 3883
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6674555540084839,
+      "learning_rate": 5.096563693402442e-06,
+      "loss": 1.0212,
+      "step": 3884
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.309697151184082,
+      "learning_rate": 5.094441669687192e-06,
+      "loss": 0.843,
+      "step": 3885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7184653878211975,
+      "learning_rate": 5.092319628954957e-06,
+      "loss": 1.0125,
+      "step": 3886
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6718325614929199,
+      "learning_rate": 5.090197571588098e-06,
+      "loss": 1.05,
+      "step": 3887
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3096809387207031,
+      "learning_rate": 5.0880754979689786e-06,
+      "loss": 0.7875,
+      "step": 3888
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.276070475578308,
+      "learning_rate": 5.085953408479967e-06,
+      "loss": 0.8269,
+      "step": 3889
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2861641645431519,
+      "learning_rate": 5.083831303503427e-06,
+      "loss": 0.7895,
+      "step": 3890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7302389144897461,
+      "learning_rate": 5.081709183421733e-06,
+      "loss": 1.0477,
+      "step": 3891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2402410507202148,
+      "learning_rate": 5.079587048617262e-06,
+      "loss": 0.8179,
+      "step": 3892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6916017532348633,
+      "learning_rate": 5.0774648994723885e-06,
+      "loss": 1.0118,
+      "step": 3893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2004839181900024,
+      "learning_rate": 5.075342736369492e-06,
+      "loss": 0.7447,
+      "step": 3894
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3265608549118042,
+      "learning_rate": 5.073220559690954e-06,
+      "loss": 0.8324,
+      "step": 3895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2637403011322021,
+      "learning_rate": 5.071098369819162e-06,
+      "loss": 0.8576,
+      "step": 3896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2595163583755493,
+      "learning_rate": 5.0689761671365025e-06,
+      "loss": 0.7938,
+      "step": 3897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2153124809265137,
+      "learning_rate": 5.066853952025362e-06,
+      "loss": 0.822,
+      "step": 3898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2276597023010254,
+      "learning_rate": 5.064731724868134e-06,
+      "loss": 0.8041,
+      "step": 3899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7426479458808899,
+      "learning_rate": 5.062609486047213e-06,
+      "loss": 1.0355,
+      "step": 3900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2354519367218018,
+      "learning_rate": 5.0604872359449945e-06,
+      "loss": 0.8228,
+      "step": 3901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2659392356872559,
+      "learning_rate": 5.058364974943875e-06,
+      "loss": 0.7888,
+      "step": 3902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.232408046722412,
+      "learning_rate": 5.056242703426255e-06,
+      "loss": 0.7703,
+      "step": 3903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.631510853767395,
+      "learning_rate": 5.054120421774536e-06,
+      "loss": 1.0111,
+      "step": 3904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3114515542984009,
+      "learning_rate": 5.051998130371125e-06,
+      "loss": 0.8086,
+      "step": 3905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2582569122314453,
+      "learning_rate": 5.049875829598422e-06,
+      "loss": 0.8049,
+      "step": 3906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6057882308959961,
+      "learning_rate": 5.047753519838837e-06,
+      "loss": 1.0288,
+      "step": 3907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3032078742980957,
+      "learning_rate": 5.0456312014747785e-06,
+      "loss": 0.8545,
+      "step": 3908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2230795621871948,
+      "learning_rate": 5.043508874888658e-06,
+      "loss": 0.8485,
+      "step": 3909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6458452939987183,
+      "learning_rate": 5.041386540462883e-06,
+      "loss": 1.044,
+      "step": 3910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2835687398910522,
+      "learning_rate": 5.039264198579869e-06,
+      "loss": 0.8032,
+      "step": 3911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2058207988739014,
+      "learning_rate": 5.0371418496220315e-06,
+      "loss": 0.889,
+      "step": 3912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.293122410774231,
+      "learning_rate": 5.035019493971784e-06,
+      "loss": 0.7684,
+      "step": 3913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2664483785629272,
+      "learning_rate": 5.0328971320115475e-06,
+      "loss": 0.8599,
+      "step": 3914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2280542850494385,
+      "learning_rate": 5.030774764123735e-06,
+      "loss": 0.81,
+      "step": 3915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3286634683609009,
+      "learning_rate": 5.028652390690768e-06,
+      "loss": 0.7894,
+      "step": 3916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2603504657745361,
+      "learning_rate": 5.026530012095067e-06,
+      "loss": 0.7572,
+      "step": 3917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2505900859832764,
+      "learning_rate": 5.024407628719054e-06,
+      "loss": 0.7508,
+      "step": 3918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7133882641792297,
+      "learning_rate": 5.0222852409451465e-06,
+      "loss": 0.9848,
+      "step": 3919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3045777082443237,
+      "learning_rate": 5.020162849155772e-06,
+      "loss": 0.8974,
+      "step": 3920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6486445069313049,
+      "learning_rate": 5.01804045373335e-06,
+      "loss": 1.0358,
+      "step": 3921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3440806865692139,
+      "learning_rate": 5.015918055060309e-06,
+      "loss": 0.8443,
+      "step": 3922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3106399774551392,
+      "learning_rate": 5.01379565351907e-06,
+      "loss": 0.8393,
+      "step": 3923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2795708179473877,
+      "learning_rate": 5.0116732494920606e-06,
+      "loss": 0.8662,
+      "step": 3924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2471171617507935,
+      "learning_rate": 5.009550843361703e-06,
+      "loss": 0.8377,
+      "step": 3925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7045625448226929,
+      "learning_rate": 5.0074284355104284e-06,
+      "loss": 1.0422,
+      "step": 3926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3142902851104736,
+      "learning_rate": 5.0053060263206585e-06,
+      "loss": 0.8551,
+      "step": 3927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6557985544204712,
+      "learning_rate": 5.003183616174823e-06,
+      "loss": 1.0536,
+      "step": 3928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2418928146362305,
+      "learning_rate": 5.0010612054553455e-06,
+      "loss": 0.7995,
+      "step": 3929
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2758595943450928,
+      "learning_rate": 4.998938794544655e-06,
+      "loss": 0.8561,
+      "step": 3930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6267393827438354,
+      "learning_rate": 4.996816383825179e-06,
+      "loss": 1.0164,
+      "step": 3931
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2511866092681885,
+      "learning_rate": 4.994693973679342e-06,
+      "loss": 0.7667,
+      "step": 3932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3359918594360352,
+      "learning_rate": 4.992571564489574e-06,
+      "loss": 0.8301,
+      "step": 3933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2231931686401367,
+      "learning_rate": 4.990449156638297e-06,
+      "loss": 0.8974,
+      "step": 3934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.246911883354187,
+      "learning_rate": 4.988326750507941e-06,
+      "loss": 0.8527,
+      "step": 3935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6555006504058838,
+      "learning_rate": 4.986204346480931e-06,
+      "loss": 1.0022,
+      "step": 3936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6774218082427979,
+      "learning_rate": 4.984081944939693e-06,
+      "loss": 1.0378,
+      "step": 3937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3821957111358643,
+      "learning_rate": 4.98195954626665e-06,
+      "loss": 0.8136,
+      "step": 3938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6563639044761658,
+      "learning_rate": 4.97983715084423e-06,
+      "loss": 1.0175,
+      "step": 3939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2559748888015747,
+      "learning_rate": 4.977714759054854e-06,
+      "loss": 0.8251,
+      "step": 3940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2527391910552979,
+      "learning_rate": 4.97559237128095e-06,
+      "loss": 0.8214,
+      "step": 3941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2660707235336304,
+      "learning_rate": 4.973469987904934e-06,
+      "loss": 0.7658,
+      "step": 3942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2362128496170044,
+      "learning_rate": 4.9713476093092335e-06,
+      "loss": 0.8152,
+      "step": 3943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7217087745666504,
+      "learning_rate": 4.969225235876266e-06,
+      "loss": 1.0234,
+      "step": 3944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2139458656311035,
+      "learning_rate": 4.967102867988455e-06,
+      "loss": 0.8383,
+      "step": 3945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3034833669662476,
+      "learning_rate": 4.964980506028216e-06,
+      "loss": 0.8303,
+      "step": 3946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.261366844177246,
+      "learning_rate": 4.96285815037797e-06,
+      "loss": 0.7713,
+      "step": 3947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6352205872535706,
+      "learning_rate": 4.960735801420132e-06,
+      "loss": 1.034,
+      "step": 3948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2313514947891235,
+      "learning_rate": 4.958613459537118e-06,
+      "loss": 0.7856,
+      "step": 3949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2842782735824585,
+      "learning_rate": 4.956491125111345e-06,
+      "loss": 0.8573,
+      "step": 3950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2781721353530884,
+      "learning_rate": 4.954368798525223e-06,
+      "loss": 0.8342,
+      "step": 3951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2525923252105713,
+      "learning_rate": 4.952246480161164e-06,
+      "loss": 0.7997,
+      "step": 3952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6284239888191223,
+      "learning_rate": 4.950124170401578e-06,
+      "loss": 1.0031,
+      "step": 3953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.264249563217163,
+      "learning_rate": 4.948001869628877e-06,
+      "loss": 0.8583,
+      "step": 3954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2631093263626099,
+      "learning_rate": 4.945879578225465e-06,
+      "loss": 0.8449,
+      "step": 3955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2942379713058472,
+      "learning_rate": 4.943757296573746e-06,
+      "loss": 0.863,
+      "step": 3956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.337600827217102,
+      "learning_rate": 4.941635025056126e-06,
+      "loss": 0.8209,
+      "step": 3957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2606542110443115,
+      "learning_rate": 4.939512764055007e-06,
+      "loss": 0.79,
+      "step": 3958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1900253295898438,
+      "learning_rate": 4.9373905139527885e-06,
+      "loss": 0.836,
+      "step": 3959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.190750241279602,
+      "learning_rate": 4.935268275131867e-06,
+      "loss": 0.8227,
+      "step": 3960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.664259135723114,
+      "learning_rate": 4.9331460479746386e-06,
+      "loss": 1.0311,
+      "step": 3961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2356820106506348,
+      "learning_rate": 4.931023832863499e-06,
+      "loss": 0.8502,
+      "step": 3962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6372910737991333,
+      "learning_rate": 4.928901630180839e-06,
+      "loss": 1.0164,
+      "step": 3963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6115574240684509,
+      "learning_rate": 4.926779440309047e-06,
+      "loss": 1.0302,
+      "step": 3964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2459810972213745,
+      "learning_rate": 4.924657263630509e-06,
+      "loss": 0.8347,
+      "step": 3965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2635105848312378,
+      "learning_rate": 4.922535100527613e-06,
+      "loss": 0.8098,
+      "step": 3966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6476534605026245,
+      "learning_rate": 4.92041295138274e-06,
+      "loss": 0.8246,
+      "step": 3967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3200496435165405,
+      "learning_rate": 4.918290816578267e-06,
+      "loss": 0.8179,
+      "step": 3968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6840243935585022,
+      "learning_rate": 4.9161686964965735e-06,
+      "loss": 1.0277,
+      "step": 3969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6771604418754578,
+      "learning_rate": 4.914046591520035e-06,
+      "loss": 1.0234,
+      "step": 3970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6642364859580994,
+      "learning_rate": 4.911924502031022e-06,
+      "loss": 1.0398,
+      "step": 3971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6114084720611572,
+      "learning_rate": 4.9098024284119036e-06,
+      "loss": 1.0123,
+      "step": 3972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2821824550628662,
+      "learning_rate": 4.907680371045043e-06,
+      "loss": 0.7818,
+      "step": 3973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3496105670928955,
+      "learning_rate": 4.90555833031281e-06,
+      "loss": 0.8334,
+      "step": 3974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.442533016204834,
+      "learning_rate": 4.9034363065975594e-06,
+      "loss": 0.8805,
+      "step": 3975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2528313398361206,
+      "learning_rate": 4.901314300281648e-06,
+      "loss": 0.8717,
+      "step": 3976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2274911403656006,
+      "learning_rate": 4.899192311747433e-06,
+      "loss": 0.8436,
+      "step": 3977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7330421805381775,
+      "learning_rate": 4.897070341377264e-06,
+      "loss": 0.9965,
+      "step": 3978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.268293023109436,
+      "learning_rate": 4.894948389553489e-06,
+      "loss": 0.7997,
+      "step": 3979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.239206314086914,
+      "learning_rate": 4.89282645665845e-06,
+      "loss": 0.7973,
+      "step": 3980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6780872941017151,
+      "learning_rate": 4.890704543074488e-06,
+      "loss": 1.0238,
+      "step": 3981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2848055362701416,
+      "learning_rate": 4.8885826491839435e-06,
+      "loss": 0.8174,
+      "step": 3982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3128927946090698,
+      "learning_rate": 4.886460775369147e-06,
+      "loss": 0.837,
+      "step": 3983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6265841126441956,
+      "learning_rate": 4.88433892201243e-06,
+      "loss": 1.0326,
+      "step": 3984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2760382890701294,
+      "learning_rate": 4.8822170894961174e-06,
+      "loss": 0.835,
+      "step": 3985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3358070850372314,
+      "learning_rate": 4.880095278202535e-06,
+      "loss": 0.7728,
+      "step": 3986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2822771072387695,
+      "learning_rate": 4.877973488513999e-06,
+      "loss": 0.8566,
+      "step": 3987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3188183307647705,
+      "learning_rate": 4.875851720812823e-06,
+      "loss": 0.8432,
+      "step": 3988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2537851333618164,
+      "learning_rate": 4.873729975481322e-06,
+      "loss": 0.7549,
+      "step": 3989
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1712286472320557,
+      "learning_rate": 4.871608252901801e-06,
+      "loss": 0.8371,
+      "step": 3990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7518023252487183,
+      "learning_rate": 4.869486553456563e-06,
+      "loss": 1.0192,
+      "step": 3991
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3126792907714844,
+      "learning_rate": 4.8673648775279045e-06,
+      "loss": 0.8269,
+      "step": 3992
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2963229417800903,
+      "learning_rate": 4.865243225498122e-06,
+      "loss": 0.8435,
+      "step": 3993
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2165201902389526,
+      "learning_rate": 4.863121597749508e-06,
+      "loss": 0.7628,
+      "step": 3994
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.213618278503418,
+      "learning_rate": 4.860999994664344e-06,
+      "loss": 0.8049,
+      "step": 3995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3353263139724731,
+      "learning_rate": 4.858878416624913e-06,
+      "loss": 0.8545,
+      "step": 3996
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.636618971824646,
+      "learning_rate": 4.856756864013491e-06,
+      "loss": 1.0131,
+      "step": 3997
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.278692603111267,
+      "learning_rate": 4.8546353372123535e-06,
+      "loss": 0.8151,
+      "step": 3998
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.258847713470459,
+      "learning_rate": 4.852513836603765e-06,
+      "loss": 0.8018,
+      "step": 3999
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2146085500717163,
+      "learning_rate": 4.8503923625699865e-06,
+      "loss": 0.8587,
+      "step": 4000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2128013372421265,
+      "learning_rate": 4.848270915493281e-06,
+      "loss": 0.8001,
+      "step": 4001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2456330060958862,
+      "learning_rate": 4.846149495755899e-06,
+      "loss": 0.7815,
+      "step": 4002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6351097226142883,
+      "learning_rate": 4.84402810374009e-06,
+      "loss": 1.0331,
+      "step": 4003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2126713991165161,
+      "learning_rate": 4.841906739828093e-06,
+      "loss": 0.8387,
+      "step": 4004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1858428716659546,
+      "learning_rate": 4.839785404402151e-06,
+      "loss": 0.8694,
+      "step": 4005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2928552627563477,
+      "learning_rate": 4.837664097844497e-06,
+      "loss": 0.7576,
+      "step": 4006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.265821099281311,
+      "learning_rate": 4.8355428205373565e-06,
+      "loss": 0.7589,
+      "step": 4007
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2754158973693848,
+      "learning_rate": 4.833421572862952e-06,
+      "loss": 0.8416,
+      "step": 4008
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3298914432525635,
+      "learning_rate": 4.831300355203503e-06,
+      "loss": 0.8595,
+      "step": 4009
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2009533643722534,
+      "learning_rate": 4.829179167941221e-06,
+      "loss": 0.8364,
+      "step": 4010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.667730450630188,
+      "learning_rate": 4.827058011458312e-06,
+      "loss": 1.017,
+      "step": 4011
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6461138725280762,
+      "learning_rate": 4.824936886136974e-06,
+      "loss": 1.0218,
+      "step": 4012
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2658262252807617,
+      "learning_rate": 4.822815792359406e-06,
+      "loss": 0.8365,
+      "step": 4013
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2261275053024292,
+      "learning_rate": 4.820694730507797e-06,
+      "loss": 0.8272,
+      "step": 4014
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2583518028259277,
+      "learning_rate": 4.81857370096433e-06,
+      "loss": 0.7768,
+      "step": 4015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7067124843597412,
+      "learning_rate": 4.816452704111181e-06,
+      "loss": 1.0299,
+      "step": 4016
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3145217895507812,
+      "learning_rate": 4.814331740330524e-06,
+      "loss": 0.9003,
+      "step": 4017
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2603211402893066,
+      "learning_rate": 4.812210810004528e-06,
+      "loss": 0.8793,
+      "step": 4018
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.264506220817566,
+      "learning_rate": 4.810089913515349e-06,
+      "loss": 0.8178,
+      "step": 4019
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6707747578620911,
+      "learning_rate": 4.8079690512451415e-06,
+      "loss": 1.0145,
+      "step": 4020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2556633949279785,
+      "learning_rate": 4.805848223576054e-06,
+      "loss": 0.7925,
+      "step": 4021
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3703832626342773,
+      "learning_rate": 4.803727430890231e-06,
+      "loss": 0.8273,
+      "step": 4022
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.305376410484314,
+      "learning_rate": 4.801606673569801e-06,
+      "loss": 0.8145,
+      "step": 4023
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.255009651184082,
+      "learning_rate": 4.799485951996899e-06,
+      "loss": 0.8371,
+      "step": 4024
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2016725540161133,
+      "learning_rate": 4.797365266553645e-06,
+      "loss": 0.8091,
+      "step": 4025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2526220083236694,
+      "learning_rate": 4.7952446176221565e-06,
+      "loss": 0.824,
+      "step": 4026
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2239036560058594,
+      "learning_rate": 4.79312400558454e-06,
+      "loss": 0.7989,
+      "step": 4027
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2752691507339478,
+      "learning_rate": 4.791003430822901e-06,
+      "loss": 0.8499,
+      "step": 4028
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3336336612701416,
+      "learning_rate": 4.788882893719334e-06,
+      "loss": 0.8288,
+      "step": 4029
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2467061281204224,
+      "learning_rate": 4.786762394655931e-06,
+      "loss": 0.895,
+      "step": 4030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8313925266265869,
+      "learning_rate": 4.784641934014769e-06,
+      "loss": 1.0238,
+      "step": 4031
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.250105619430542,
+      "learning_rate": 4.782521512177928e-06,
+      "loss": 0.7678,
+      "step": 4032
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2456114292144775,
+      "learning_rate": 4.780401129527475e-06,
+      "loss": 0.7914,
+      "step": 4033
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7738509178161621,
+      "learning_rate": 4.7782807864454724e-06,
+      "loss": 1.0204,
+      "step": 4034
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3499075174331665,
+      "learning_rate": 4.7761604833139706e-06,
+      "loss": 0.8749,
+      "step": 4035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.340709924697876,
+      "learning_rate": 4.7740402205150224e-06,
+      "loss": 0.7723,
+      "step": 4036
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2498292922973633,
+      "learning_rate": 4.7719199984306645e-06,
+      "loss": 0.8214,
+      "step": 4037
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7981082201004028,
+      "learning_rate": 4.76979981744293e-06,
+      "loss": 1.0325,
+      "step": 4038
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.262525200843811,
+      "learning_rate": 4.767679677933841e-06,
+      "loss": 0.8026,
+      "step": 4039
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2979129552841187,
+      "learning_rate": 4.765559580285421e-06,
+      "loss": 0.8858,
+      "step": 4040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6412421464920044,
+      "learning_rate": 4.763439524879675e-06,
+      "loss": 1.0374,
+      "step": 4041
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6281381249427795,
+      "learning_rate": 4.761319512098609e-06,
+      "loss": 1.0009,
+      "step": 4042
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.630214273929596,
+      "learning_rate": 4.759199542324213e-06,
+      "loss": 1.0235,
+      "step": 4043
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2473485469818115,
+      "learning_rate": 4.75707961593848e-06,
+      "loss": 0.7827,
+      "step": 4044
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3200479745864868,
+      "learning_rate": 4.754959733323385e-06,
+      "loss": 0.8728,
+      "step": 4045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3062947988510132,
+      "learning_rate": 4.752839894860902e-06,
+      "loss": 0.8355,
+      "step": 4046
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2857447862625122,
+      "learning_rate": 4.7507201009329904e-06,
+      "loss": 0.8177,
+      "step": 4047
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2969623804092407,
+      "learning_rate": 4.748600351921611e-06,
+      "loss": 0.7659,
+      "step": 4048
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7485570311546326,
+      "learning_rate": 4.746480648208706e-06,
+      "loss": 1.0182,
+      "step": 4049
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3253164291381836,
+      "learning_rate": 4.744360990176218e-06,
+      "loss": 0.8124,
+      "step": 4050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4344923496246338,
+      "learning_rate": 4.742241378206074e-06,
+      "loss": 0.7922,
+      "step": 4051
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3386931419372559,
+      "learning_rate": 4.740121812680201e-06,
+      "loss": 0.8276,
+      "step": 4052
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.314497709274292,
+      "learning_rate": 4.738002293980511e-06,
+      "loss": 0.837,
+      "step": 4053
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6934188008308411,
+      "learning_rate": 4.73588282248891e-06,
+      "loss": 1.0246,
+      "step": 4054
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2540688514709473,
+      "learning_rate": 4.7337633985872936e-06,
+      "loss": 0.8267,
+      "step": 4055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6577440500259399,
+      "learning_rate": 4.731644022657554e-06,
+      "loss": 1.0195,
+      "step": 4056
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6334788799285889,
+      "learning_rate": 4.729524695081567e-06,
+      "loss": 1.0417,
+      "step": 4057
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.246588110923767,
+      "learning_rate": 4.727405416241208e-06,
+      "loss": 0.775,
+      "step": 4058
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2979398965835571,
+      "learning_rate": 4.725286186518337e-06,
+      "loss": 0.8408,
+      "step": 4059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2328006029129028,
+      "learning_rate": 4.723167006294808e-06,
+      "loss": 0.8406,
+      "step": 4060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6601748466491699,
+      "learning_rate": 4.721047875952466e-06,
+      "loss": 1.0306,
+      "step": 4061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6875689625740051,
+      "learning_rate": 4.718928795873146e-06,
+      "loss": 1.0406,
+      "step": 4062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6380907297134399,
+      "learning_rate": 4.716809766438676e-06,
+      "loss": 1.0234,
+      "step": 4063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5879508256912231,
+      "learning_rate": 4.714690788030875e-06,
+      "loss": 1.0119,
+      "step": 4064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2226868867874146,
+      "learning_rate": 4.712571861031548e-06,
+      "loss": 0.7817,
+      "step": 4065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2207916975021362,
+      "learning_rate": 4.7104529858224945e-06,
+      "loss": 0.8225,
+      "step": 4066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3526673316955566,
+      "learning_rate": 4.7083341627855065e-06,
+      "loss": 0.8198,
+      "step": 4067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3233298063278198,
+      "learning_rate": 4.7062153923023635e-06,
+      "loss": 0.8321,
+      "step": 4068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.230621099472046,
+      "learning_rate": 4.704096674754835e-06,
+      "loss": 0.8256,
+      "step": 4069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.747631311416626,
+      "learning_rate": 4.701978010524684e-06,
+      "loss": 1.03,
+      "step": 4070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.274298906326294,
+      "learning_rate": 4.699859399993662e-06,
+      "loss": 0.8408,
+      "step": 4071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2389085292816162,
+      "learning_rate": 4.697740843543512e-06,
+      "loss": 0.8428,
+      "step": 4072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2596062421798706,
+      "learning_rate": 4.695622341555964e-06,
+      "loss": 0.7479,
+      "step": 4073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.324824333190918,
+      "learning_rate": 4.693503894412741e-06,
+      "loss": 0.8017,
+      "step": 4074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3075828552246094,
+      "learning_rate": 4.691385502495557e-06,
+      "loss": 0.8495,
+      "step": 4075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3851916790008545,
+      "learning_rate": 4.689267166186116e-06,
+      "loss": 0.8594,
+      "step": 4076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2874876260757446,
+      "learning_rate": 4.687148885866107e-06,
+      "loss": 0.8231,
+      "step": 4077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7168103456497192,
+      "learning_rate": 4.685030661917215e-06,
+      "loss": 1.0297,
+      "step": 4078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2841092348098755,
+      "learning_rate": 4.682912494721112e-06,
+      "loss": 0.8322,
+      "step": 4079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3533743619918823,
+      "learning_rate": 4.68079438465946e-06,
+      "loss": 0.8303,
+      "step": 4080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2942906618118286,
+      "learning_rate": 4.6786763321139114e-06,
+      "loss": 0.8799,
+      "step": 4081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3279603719711304,
+      "learning_rate": 4.676558337466106e-06,
+      "loss": 0.9034,
+      "step": 4082
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6608389616012573,
+      "learning_rate": 4.674440401097678e-06,
+      "loss": 1.0059,
+      "step": 4083
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.247409462928772,
+      "learning_rate": 4.6723225233902465e-06,
+      "loss": 0.8592,
+      "step": 4084
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2751444578170776,
+      "learning_rate": 4.67020470472542e-06,
+      "loss": 0.8756,
+      "step": 4085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7334144115447998,
+      "learning_rate": 4.6680869454848e-06,
+      "loss": 0.8012,
+      "step": 4086
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2515569925308228,
+      "learning_rate": 4.6659692460499744e-06,
+      "loss": 0.8083,
+      "step": 4087
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2001549005508423,
+      "learning_rate": 4.663851606802522e-06,
+      "loss": 0.7477,
+      "step": 4088
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3407909870147705,
+      "learning_rate": 4.661734028124009e-06,
+      "loss": 0.8058,
+      "step": 4089
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2022114992141724,
+      "learning_rate": 4.659616510395991e-06,
+      "loss": 0.8076,
+      "step": 4090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.242878794670105,
+      "learning_rate": 4.657499054000014e-06,
+      "loss": 0.8171,
+      "step": 4091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6958788633346558,
+      "learning_rate": 4.655381659317613e-06,
+      "loss": 1.034,
+      "step": 4092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.665681779384613,
+      "learning_rate": 4.653264326730309e-06,
+      "loss": 1.0222,
+      "step": 4093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3156424760818481,
+      "learning_rate": 4.651147056619615e-06,
+      "loss": 0.8268,
+      "step": 4094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2565807104110718,
+      "learning_rate": 4.649029849367031e-06,
+      "loss": 0.7712,
+      "step": 4095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.268163800239563,
+      "learning_rate": 4.646912705354049e-06,
+      "loss": 0.851,
+      "step": 4096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2865875959396362,
+      "learning_rate": 4.6447956249621415e-06,
+      "loss": 0.8388,
+      "step": 4097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2872635126113892,
+      "learning_rate": 4.6426786085727806e-06,
+      "loss": 0.8135,
+      "step": 4098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6567814946174622,
+      "learning_rate": 4.640561656567417e-06,
+      "loss": 1.0198,
+      "step": 4099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6808801293373108,
+      "learning_rate": 4.638444769327497e-06,
+      "loss": 1.0234,
+      "step": 4100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2290375232696533,
+      "learning_rate": 4.636327947234448e-06,
+      "loss": 0.7989,
+      "step": 4101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6122003793716431,
+      "learning_rate": 4.634211190669695e-06,
+      "loss": 1.039,
+      "step": 4102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2432372570037842,
+      "learning_rate": 4.6320945000146435e-06,
+      "loss": 0.8125,
+      "step": 4103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3007535934448242,
+      "learning_rate": 4.62997787565069e-06,
+      "loss": 0.8481,
+      "step": 4104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6353008151054382,
+      "learning_rate": 4.627861317959216e-06,
+      "loss": 1.0199,
+      "step": 4105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2472736835479736,
+      "learning_rate": 4.625744827321599e-06,
+      "loss": 0.8064,
+      "step": 4106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2849743366241455,
+      "learning_rate": 4.623628404119194e-06,
+      "loss": 0.817,
+      "step": 4107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.224970817565918,
+      "learning_rate": 4.621512048733354e-06,
+      "loss": 0.7371,
+      "step": 4108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1936060190200806,
+      "learning_rate": 4.619395761545409e-06,
+      "loss": 0.8331,
+      "step": 4109
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2428494691848755,
+      "learning_rate": 4.6172795429366875e-06,
+      "loss": 0.8125,
+      "step": 4110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2548854351043701,
+      "learning_rate": 4.615163393288498e-06,
+      "loss": 0.795,
+      "step": 4111
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7453475594520569,
+      "learning_rate": 4.61304731298214e-06,
+      "loss": 1.0206,
+      "step": 4112
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2145445346832275,
+      "learning_rate": 4.610931302398898e-06,
+      "loss": 0.7791,
+      "step": 4113
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3141717910766602,
+      "learning_rate": 4.60881536192005e-06,
+      "loss": 0.8699,
+      "step": 4114
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3250631093978882,
+      "learning_rate": 4.606699491926852e-06,
+      "loss": 0.8354,
+      "step": 4115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2395200729370117,
+      "learning_rate": 4.604583692800556e-06,
+      "loss": 0.7996,
+      "step": 4116
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.273869276046753,
+      "learning_rate": 4.602467964922393e-06,
+      "loss": 0.7863,
+      "step": 4117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3043162822723389,
+      "learning_rate": 4.600352308673591e-06,
+      "loss": 0.7822,
+      "step": 4118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6911168098449707,
+      "learning_rate": 4.598236724435357e-06,
+      "loss": 1.0268,
+      "step": 4119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.236036777496338,
+      "learning_rate": 4.596121212588889e-06,
+      "loss": 0.7874,
+      "step": 4120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2665499448776245,
+      "learning_rate": 4.594005773515368e-06,
+      "loss": 0.8257,
+      "step": 4121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.240630030632019,
+      "learning_rate": 4.591890407595968e-06,
+      "loss": 0.8203,
+      "step": 4122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.665501594543457,
+      "learning_rate": 4.589775115211845e-06,
+      "loss": 1.0331,
+      "step": 4123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2527031898498535,
+      "learning_rate": 4.587659896744144e-06,
+      "loss": 0.7996,
+      "step": 4124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.17771315574646,
+      "learning_rate": 4.585544752573993e-06,
+      "loss": 0.773,
+      "step": 4125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.231908917427063,
+      "learning_rate": 4.583429683082513e-06,
+      "loss": 0.7873,
+      "step": 4126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1684069633483887,
+      "learning_rate": 4.581314688650807e-06,
+      "loss": 0.7821,
+      "step": 4127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6296792030334473,
+      "learning_rate": 4.579199769659965e-06,
+      "loss": 1.0425,
+      "step": 4128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6168434023857117,
+      "learning_rate": 4.5770849264910635e-06,
+      "loss": 1.021,
+      "step": 4129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3719794750213623,
+      "learning_rate": 4.574970159525168e-06,
+      "loss": 0.8576,
+      "step": 4130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6362971067428589,
+      "learning_rate": 4.572855469143326e-06,
+      "loss": 1.0418,
+      "step": 4131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2730367183685303,
+      "learning_rate": 4.570740855726573e-06,
+      "loss": 0.7367,
+      "step": 4132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5944656133651733,
+      "learning_rate": 4.568626319655933e-06,
+      "loss": 0.8468,
+      "step": 4133
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2471975088119507,
+      "learning_rate": 4.566511861312413e-06,
+      "loss": 0.835,
+      "step": 4134
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3009517192840576,
+      "learning_rate": 4.564397481077005e-06,
+      "loss": 0.8638,
+      "step": 4135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1697158813476562,
+      "learning_rate": 4.56228317933069e-06,
+      "loss": 0.8241,
+      "step": 4136
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2570840120315552,
+      "learning_rate": 4.560168956454435e-06,
+      "loss": 0.8566,
+      "step": 4137
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2708511352539062,
+      "learning_rate": 4.558054812829192e-06,
+      "loss": 0.7964,
+      "step": 4138
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2615147829055786,
+      "learning_rate": 4.555940748835895e-06,
+      "loss": 0.7727,
+      "step": 4139
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.242127537727356,
+      "learning_rate": 4.553826764855468e-06,
+      "loss": 0.8263,
+      "step": 4140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3534854650497437,
+      "learning_rate": 4.551712861268824e-06,
+      "loss": 0.8238,
+      "step": 4141
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3158169984817505,
+      "learning_rate": 4.549599038456851e-06,
+      "loss": 0.7879,
+      "step": 4142
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2765120267868042,
+      "learning_rate": 4.54748529680043e-06,
+      "loss": 0.7908,
+      "step": 4143
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2437669038772583,
+      "learning_rate": 4.5453716366804255e-06,
+      "loss": 0.8257,
+      "step": 4144
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2086760997772217,
+      "learning_rate": 4.543258058477691e-06,
+      "loss": 0.8564,
+      "step": 4145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7290740013122559,
+      "learning_rate": 4.541144562573059e-06,
+      "loss": 1.0502,
+      "step": 4146
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2246907949447632,
+      "learning_rate": 4.539031149347351e-06,
+      "loss": 0.8413,
+      "step": 4147
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6625051498413086,
+      "learning_rate": 4.536917819181369e-06,
+      "loss": 1.0143,
+      "step": 4148
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2948781251907349,
+      "learning_rate": 4.534804572455909e-06,
+      "loss": 0.846,
+      "step": 4149
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2132526636123657,
+      "learning_rate": 4.532691409551745e-06,
+      "loss": 0.8725,
+      "step": 4150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6251712441444397,
+      "learning_rate": 4.530578330849635e-06,
+      "loss": 1.0284,
+      "step": 4151
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6300190687179565,
+      "learning_rate": 4.528465336730326e-06,
+      "loss": 1.0357,
+      "step": 4152
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2757209539413452,
+      "learning_rate": 4.52635242757455e-06,
+      "loss": 0.8096,
+      "step": 4153
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1922545433044434,
+      "learning_rate": 4.5242396037630196e-06,
+      "loss": 0.82,
+      "step": 4154
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6313367486000061,
+      "learning_rate": 4.522126865676434e-06,
+      "loss": 1.0148,
+      "step": 4155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2711358070373535,
+      "learning_rate": 4.520014213695477e-06,
+      "loss": 0.7528,
+      "step": 4156
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6364678740501404,
+      "learning_rate": 4.51790164820082e-06,
+      "loss": 1.0349,
+      "step": 4157
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2262489795684814,
+      "learning_rate": 4.5157891695731135e-06,
+      "loss": 0.8362,
+      "step": 4158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.246058464050293,
+      "learning_rate": 4.513676778192995e-06,
+      "loss": 0.777,
+      "step": 4159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6682671308517456,
+      "learning_rate": 4.511564474441084e-06,
+      "loss": 1.0104,
+      "step": 4160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2875406742095947,
+      "learning_rate": 4.50945225869799e-06,
+      "loss": 0.8883,
+      "step": 4161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.205047607421875,
+      "learning_rate": 4.507340131344302e-06,
+      "loss": 0.8425,
+      "step": 4162
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2510112524032593,
+      "learning_rate": 4.505228092760591e-06,
+      "loss": 0.8365,
+      "step": 4163
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2295557260513306,
+      "learning_rate": 4.503116143327416e-06,
+      "loss": 0.8065,
+      "step": 4164
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7097113728523254,
+      "learning_rate": 4.5010042834253225e-06,
+      "loss": 1.0031,
+      "step": 4165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3187248706817627,
+      "learning_rate": 4.498892513434834e-06,
+      "loss": 0.8461,
+      "step": 4166
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6361199617385864,
+      "learning_rate": 4.496780833736456e-06,
+      "loss": 1.0074,
+      "step": 4167
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2550185918807983,
+      "learning_rate": 4.494669244710688e-06,
+      "loss": 0.7766,
+      "step": 4168
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2877531051635742,
+      "learning_rate": 4.4925577467380035e-06,
+      "loss": 0.7898,
+      "step": 4169
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.251213550567627,
+      "learning_rate": 4.490446340198865e-06,
+      "loss": 0.8186,
+      "step": 4170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2722418308258057,
+      "learning_rate": 4.4883350254737135e-06,
+      "loss": 0.8408,
+      "step": 4171
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3000179529190063,
+      "learning_rate": 4.486223802942979e-06,
+      "loss": 0.8491,
+      "step": 4172
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2202430963516235,
+      "learning_rate": 4.484112672987072e-06,
+      "loss": 0.8314,
+      "step": 4173
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2749775648117065,
+      "learning_rate": 4.482001635986387e-06,
+      "loss": 0.8393,
+      "step": 4174
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2631909847259521,
+      "learning_rate": 4.4798906923213e-06,
+      "loss": 0.8578,
+      "step": 4175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2928049564361572,
+      "learning_rate": 4.477779842372173e-06,
+      "loss": 0.8097,
+      "step": 4176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2142305374145508,
+      "learning_rate": 4.475669086519349e-06,
+      "loss": 0.7961,
+      "step": 4177
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.783581554889679,
+      "learning_rate": 4.473558425143156e-06,
+      "loss": 1.0461,
+      "step": 4178
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2764873504638672,
+      "learning_rate": 4.4714478586239014e-06,
+      "loss": 0.7607,
+      "step": 4179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2335608005523682,
+      "learning_rate": 4.46933738734188e-06,
+      "loss": 0.8021,
+      "step": 4180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1858774423599243,
+      "learning_rate": 4.467227011677367e-06,
+      "loss": 0.8276,
+      "step": 4181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4914056062698364,
+      "learning_rate": 4.465116732010622e-06,
+      "loss": 0.8214,
+      "step": 4182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.283502221107483,
+      "learning_rate": 4.463006548721881e-06,
+      "loss": 0.8053,
+      "step": 4183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3101060390472412,
+      "learning_rate": 4.460896462191373e-06,
+      "loss": 0.7941,
+      "step": 4184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.287718415260315,
+      "learning_rate": 4.458786472799302e-06,
+      "loss": 0.7714,
+      "step": 4185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6512855291366577,
+      "learning_rate": 4.456676580925858e-06,
+      "loss": 1.0228,
+      "step": 4186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2477569580078125,
+      "learning_rate": 4.454566786951209e-06,
+      "loss": 0.8171,
+      "step": 4187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6531429886817932,
+      "learning_rate": 4.452457091255511e-06,
+      "loss": 1.0248,
+      "step": 4188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2681691646575928,
+      "learning_rate": 4.4503474942189005e-06,
+      "loss": 0.7971,
+      "step": 4189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6109451055526733,
+      "learning_rate": 4.448237996221494e-06,
+      "loss": 1.0258,
+      "step": 4190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3201149702072144,
+      "learning_rate": 4.446128597643391e-06,
+      "loss": 0.8384,
+      "step": 4191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.252536654472351,
+      "learning_rate": 4.4440192988646754e-06,
+      "loss": 0.8366,
+      "step": 4192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2647651433944702,
+      "learning_rate": 4.441910100265411e-06,
+      "loss": 0.7953,
+      "step": 4193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3225598335266113,
+      "learning_rate": 4.439801002225645e-06,
+      "loss": 0.7826,
+      "step": 4194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2238796949386597,
+      "learning_rate": 4.437692005125401e-06,
+      "loss": 0.7729,
+      "step": 4195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2450381517410278,
+      "learning_rate": 4.435583109344694e-06,
+      "loss": 0.8151,
+      "step": 4196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.202049970626831,
+      "learning_rate": 4.4334743152635154e-06,
+      "loss": 0.7686,
+      "step": 4197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2372212409973145,
+      "learning_rate": 4.4313656232618365e-06,
+      "loss": 0.8006,
+      "step": 4198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.741284966468811,
+      "learning_rate": 4.42925703371961e-06,
+      "loss": 1.0412,
+      "step": 4199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2383397817611694,
+      "learning_rate": 4.427148547016777e-06,
+      "loss": 0.8079,
+      "step": 4200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2939491271972656,
+      "learning_rate": 4.425040163533254e-06,
+      "loss": 0.8165,
+      "step": 4201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.221356987953186,
+      "learning_rate": 4.422931883648939e-06,
+      "loss": 0.8053,
+      "step": 4202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.629709780216217,
+      "learning_rate": 4.4208237077437125e-06,
+      "loss": 1.0295,
+      "step": 4203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3103396892547607,
+      "learning_rate": 4.418715636197438e-06,
+      "loss": 0.8008,
+      "step": 4204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2253828048706055,
+      "learning_rate": 4.4166076693899595e-06,
+      "loss": 0.7863,
+      "step": 4205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6347380876541138,
+      "learning_rate": 4.414499807701097e-06,
+      "loss": 1.0514,
+      "step": 4206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6282178163528442,
+      "learning_rate": 4.412392051510662e-06,
+      "loss": 1.0262,
+      "step": 4207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5440342426300049,
+      "learning_rate": 4.410284401198436e-06,
+      "loss": 0.8899,
+      "step": 4208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3328315019607544,
+      "learning_rate": 4.408176857144188e-06,
+      "loss": 0.8309,
+      "step": 4209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3231829404830933,
+      "learning_rate": 4.406069419727664e-06,
+      "loss": 0.8566,
+      "step": 4210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.230281114578247,
+      "learning_rate": 4.403962089328598e-06,
+      "loss": 0.8039,
+      "step": 4211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2764084339141846,
+      "learning_rate": 4.401854866326693e-06,
+      "loss": 0.8203,
+      "step": 4212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7618656754493713,
+      "learning_rate": 4.3997477511016445e-06,
+      "loss": 1.0323,
+      "step": 4213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7040987014770508,
+      "learning_rate": 4.3976407440331194e-06,
+      "loss": 1.023,
+      "step": 4214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2823307514190674,
+      "learning_rate": 4.395533845500772e-06,
+      "loss": 0.8251,
+      "step": 4215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2378889322280884,
+      "learning_rate": 4.3934270558842335e-06,
+      "loss": 0.8169,
+      "step": 4216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6231555938720703,
+      "learning_rate": 4.391320375563116e-06,
+      "loss": 1.0354,
+      "step": 4217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.26235032081604,
+      "learning_rate": 4.389213804917009e-06,
+      "loss": 0.828,
+      "step": 4218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3578156232833862,
+      "learning_rate": 4.38710734432549e-06,
+      "loss": 0.8545,
+      "step": 4219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2629287242889404,
+      "learning_rate": 4.3850009941681085e-06,
+      "loss": 0.8364,
+      "step": 4220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2474006414413452,
+      "learning_rate": 4.382894754824399e-06,
+      "loss": 0.8692,
+      "step": 4221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7045938968658447,
+      "learning_rate": 4.380788626673872e-06,
+      "loss": 1.0121,
+      "step": 4222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.262954592704773,
+      "learning_rate": 4.378682610096025e-06,
+      "loss": 0.8343,
+      "step": 4223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1930795907974243,
+      "learning_rate": 4.376576705470327e-06,
+      "loss": 0.8105,
+      "step": 4224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6185070276260376,
+      "learning_rate": 4.374470913176233e-06,
+      "loss": 1.036,
+      "step": 4225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2693079710006714,
+      "learning_rate": 4.372365233593172e-06,
+      "loss": 0.8674,
+      "step": 4226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.225431203842163,
+      "learning_rate": 4.370259667100561e-06,
+      "loss": 0.8038,
+      "step": 4227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.206274151802063,
+      "learning_rate": 4.368154214077789e-06,
+      "loss": 0.7888,
+      "step": 4228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.22981595993042,
+      "learning_rate": 4.366048874904228e-06,
+      "loss": 0.7642,
+      "step": 4229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2949814796447754,
+      "learning_rate": 4.363943649959226e-06,
+      "loss": 0.8424,
+      "step": 4230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.205763339996338,
+      "learning_rate": 4.361838539622118e-06,
+      "loss": 0.8016,
+      "step": 4231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.204531192779541,
+      "learning_rate": 4.35973354427221e-06,
+      "loss": 0.8158,
+      "step": 4232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.140438437461853,
+      "learning_rate": 4.3576286642887936e-06,
+      "loss": 0.753,
+      "step": 4233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2436325550079346,
+      "learning_rate": 4.355523900051133e-06,
+      "loss": 0.8059,
+      "step": 4234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2035127878189087,
+      "learning_rate": 4.35341925193848e-06,
+      "loss": 0.7988,
+      "step": 4235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7735454440116882,
+      "learning_rate": 4.351314720330057e-06,
+      "loss": 1.037,
+      "step": 4236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2272177934646606,
+      "learning_rate": 4.349210305605072e-06,
+      "loss": 0.7972,
+      "step": 4237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3004884719848633,
+      "learning_rate": 4.3471060081427065e-06,
+      "loss": 0.8392,
+      "step": 4238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2231252193450928,
+      "learning_rate": 4.345001828322127e-06,
+      "loss": 0.8421,
+      "step": 4239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2444145679473877,
+      "learning_rate": 4.342897766522472e-06,
+      "loss": 0.8167,
+      "step": 4240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2406537532806396,
+      "learning_rate": 4.340793823122863e-06,
+      "loss": 0.8334,
+      "step": 4241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6304423809051514,
+      "learning_rate": 4.338689998502402e-06,
+      "loss": 1.0176,
+      "step": 4242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2668302059173584,
+      "learning_rate": 4.336586293040166e-06,
+      "loss": 0.7989,
+      "step": 4243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6221131086349487,
+      "learning_rate": 4.334482707115208e-06,
+      "loss": 1.0233,
+      "step": 4244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2200459241867065,
+      "learning_rate": 4.332379241106565e-06,
+      "loss": 0.747,
+      "step": 4245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2602856159210205,
+      "learning_rate": 4.330275895393252e-06,
+      "loss": 0.8059,
+      "step": 4246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2563467025756836,
+      "learning_rate": 4.32817267035426e-06,
+      "loss": 0.8036,
+      "step": 4247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.248520851135254,
+      "learning_rate": 4.326069566368557e-06,
+      "loss": 0.7925,
+      "step": 4248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2573788166046143,
+      "learning_rate": 4.3239665838150905e-06,
+      "loss": 0.8598,
+      "step": 4249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3116042613983154,
+      "learning_rate": 4.32186372307279e-06,
+      "loss": 0.7925,
+      "step": 4250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2381436824798584,
+      "learning_rate": 4.319760984520558e-06,
+      "loss": 0.7997,
+      "step": 4251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6852154731750488,
+      "learning_rate": 4.3176583685372755e-06,
+      "loss": 0.9988,
+      "step": 4252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2759921550750732,
+      "learning_rate": 4.3155558755018034e-06,
+      "loss": 0.834,
+      "step": 4253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.315014362335205,
+      "learning_rate": 4.313453505792982e-06,
+      "loss": 0.8174,
+      "step": 4254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.64686518907547,
+      "learning_rate": 4.311351259789625e-06,
+      "loss": 1.0231,
+      "step": 4255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2729318141937256,
+      "learning_rate": 4.3092491378705246e-06,
+      "loss": 0.8887,
+      "step": 4256
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2268867492675781,
+      "learning_rate": 4.307147140414452e-06,
+      "loss": 0.8454,
+      "step": 4257
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.604340136051178,
+      "learning_rate": 4.305045267800159e-06,
+      "loss": 1.0065,
+      "step": 4258
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6122445464134216,
+      "learning_rate": 4.302943520406371e-06,
+      "loss": 1.011,
+      "step": 4259
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3301773071289062,
+      "learning_rate": 4.300841898611789e-06,
+      "loss": 0.8018,
+      "step": 4260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2309972047805786,
+      "learning_rate": 4.2987404027950936e-06,
+      "loss": 0.7863,
+      "step": 4261
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2604584693908691,
+      "learning_rate": 4.2966390333349474e-06,
+      "loss": 0.8325,
+      "step": 4262
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2508612871170044,
+      "learning_rate": 4.294537790609984e-06,
+      "loss": 0.81,
+      "step": 4263
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.237605094909668,
+      "learning_rate": 4.292436674998814e-06,
+      "loss": 0.8354,
+      "step": 4264
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2258861064910889,
+      "learning_rate": 4.290335686880028e-06,
+      "loss": 0.7762,
+      "step": 4265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2979199886322021,
+      "learning_rate": 4.288234826632195e-06,
+      "loss": 0.8087,
+      "step": 4266
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3745791912078857,
+      "learning_rate": 4.286134094633858e-06,
+      "loss": 0.8455,
+      "step": 4267
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7092097997665405,
+      "learning_rate": 4.284033491263535e-06,
+      "loss": 0.9876,
+      "step": 4268
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2746174335479736,
+      "learning_rate": 4.281933016899725e-06,
+      "loss": 0.7889,
+      "step": 4269
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6514129638671875,
+      "learning_rate": 4.279832671920905e-06,
+      "loss": 1.0259,
+      "step": 4270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6068682074546814,
+      "learning_rate": 4.277732456705523e-06,
+      "loss": 1.0446,
+      "step": 4271
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6154870390892029,
+      "learning_rate": 4.275632371632007e-06,
+      "loss": 1.0287,
+      "step": 4272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2502092123031616,
+      "learning_rate": 4.273532417078759e-06,
+      "loss": 0.8116,
+      "step": 4273
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3312958478927612,
+      "learning_rate": 4.271432593424164e-06,
+      "loss": 0.8274,
+      "step": 4274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3000167608261108,
+      "learning_rate": 4.269332901046577e-06,
+      "loss": 0.8212,
+      "step": 4275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1798442602157593,
+      "learning_rate": 4.2672333403243296e-06,
+      "loss": 0.8148,
+      "step": 4276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2329729795455933,
+      "learning_rate": 4.265133911635733e-06,
+      "loss": 0.8048,
+      "step": 4277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2849228382110596,
+      "learning_rate": 4.263034615359074e-06,
+      "loss": 0.7435,
+      "step": 4278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7146936655044556,
+      "learning_rate": 4.2609354518726134e-06,
+      "loss": 1.02,
+      "step": 4279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2161263227462769,
+      "learning_rate": 4.258836421554588e-06,
+      "loss": 0.8494,
+      "step": 4280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2408461570739746,
+      "learning_rate": 4.256737524783213e-06,
+      "loss": 0.79,
+      "step": 4281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3006542921066284,
+      "learning_rate": 4.254638761936681e-06,
+      "loss": 0.8444,
+      "step": 4282
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2124186754226685,
+      "learning_rate": 4.252540133393154e-06,
+      "loss": 0.8291,
+      "step": 4283
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2803231477737427,
+      "learning_rate": 4.250441639530774e-06,
+      "loss": 0.8129,
+      "step": 4284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2579333782196045,
+      "learning_rate": 4.248343280727659e-06,
+      "loss": 0.8896,
+      "step": 4285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.23189377784729,
+      "learning_rate": 4.246245057361904e-06,
+      "loss": 0.7937,
+      "step": 4286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2251583337783813,
+      "learning_rate": 4.244146969811576e-06,
+      "loss": 0.79,
+      "step": 4287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6609206199645996,
+      "learning_rate": 4.242049018454718e-06,
+      "loss": 1.0209,
+      "step": 4288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2857097387313843,
+      "learning_rate": 4.23995120366935e-06,
+      "loss": 0.9011,
+      "step": 4289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6265982389450073,
+      "learning_rate": 4.237853525833471e-06,
+      "loss": 1.0367,
+      "step": 4290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3116250038146973,
+      "learning_rate": 4.235755985325046e-06,
+      "loss": 0.7848,
+      "step": 4291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6126999258995056,
+      "learning_rate": 4.233658582522022e-06,
+      "loss": 1.0292,
+      "step": 4292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2184518575668335,
+      "learning_rate": 4.231561317802322e-06,
+      "loss": 0.8058,
+      "step": 4293
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6119198203086853,
+      "learning_rate": 4.22946419154384e-06,
+      "loss": 0.9993,
+      "step": 4294
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1813629865646362,
+      "learning_rate": 4.227367204124448e-06,
+      "loss": 0.7943,
+      "step": 4295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2470042705535889,
+      "learning_rate": 4.22527035592199e-06,
+      "loss": 0.7914,
+      "step": 4296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2453947067260742,
+      "learning_rate": 4.2231736473142885e-06,
+      "loss": 0.7818,
+      "step": 4297
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3230613470077515,
+      "learning_rate": 4.22107707867914e-06,
+      "loss": 0.8329,
+      "step": 4298
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6676796078681946,
+      "learning_rate": 4.218980650394315e-06,
+      "loss": 1.0343,
+      "step": 4299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.632899284362793,
+      "learning_rate": 4.216884362837553e-06,
+      "loss": 1.034,
+      "step": 4300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2363016605377197,
+      "learning_rate": 4.214788216386582e-06,
+      "loss": 0.7521,
+      "step": 4301
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3049907684326172,
+      "learning_rate": 4.212692211419092e-06,
+      "loss": 0.828,
+      "step": 4302
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2381055355072021,
+      "learning_rate": 4.210596348312753e-06,
+      "loss": 0.8296,
+      "step": 4303
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2833397388458252,
+      "learning_rate": 4.208500627445207e-06,
+      "loss": 0.8318,
+      "step": 4304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3123924732208252,
+      "learning_rate": 4.206405049194072e-06,
+      "loss": 0.8574,
+      "step": 4305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2826282978057861,
+      "learning_rate": 4.2043096139369435e-06,
+      "loss": 0.8313,
+      "step": 4306
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6611954569816589,
+      "learning_rate": 4.202214322051385e-06,
+      "loss": 1.0095,
+      "step": 4307
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3075032234191895,
+      "learning_rate": 4.2001191739149335e-06,
+      "loss": 0.8476,
+      "step": 4308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.222068190574646,
+      "learning_rate": 4.198024169905109e-06,
+      "loss": 0.759,
+      "step": 4309
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3899468183517456,
+      "learning_rate": 4.1959293103993995e-06,
+      "loss": 0.8475,
+      "step": 4310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2494484186172485,
+      "learning_rate": 4.193834595775265e-06,
+      "loss": 0.8109,
+      "step": 4311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2238479852676392,
+      "learning_rate": 4.191740026410141e-06,
+      "loss": 0.8026,
+      "step": 4312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6375740766525269,
+      "learning_rate": 4.189645602681441e-06,
+      "loss": 1.0141,
+      "step": 4313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6288204789161682,
+      "learning_rate": 4.18755132496655e-06,
+      "loss": 1.0356,
+      "step": 4314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2517424821853638,
+      "learning_rate": 4.185457193642819e-06,
+      "loss": 0.8026,
+      "step": 4315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6064184308052063,
+      "learning_rate": 4.183363209087588e-06,
+      "loss": 1.0279,
+      "step": 4316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3133153915405273,
+      "learning_rate": 4.181269371678155e-06,
+      "loss": 0.8235,
+      "step": 4317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2505254745483398,
+      "learning_rate": 4.179175681791803e-06,
+      "loss": 0.8055,
+      "step": 4318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2373785972595215,
+      "learning_rate": 4.177082139805779e-06,
+      "loss": 0.7775,
+      "step": 4319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2540770769119263,
+      "learning_rate": 4.1749887460973125e-06,
+      "loss": 0.8156,
+      "step": 4320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2880492210388184,
+      "learning_rate": 4.1728955010436e-06,
+      "loss": 0.801,
+      "step": 4321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2546995878219604,
+      "learning_rate": 4.170802405021815e-06,
+      "loss": 0.8317,
+      "step": 4322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2472591400146484,
+      "learning_rate": 4.168709458409097e-06,
+      "loss": 0.7766,
+      "step": 4323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1557825803756714,
+      "learning_rate": 4.16661666158257e-06,
+      "loss": 0.8304,
+      "step": 4324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.260359525680542,
+      "learning_rate": 4.164524014919322e-06,
+      "loss": 0.849,
+      "step": 4325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1883594989776611,
+      "learning_rate": 4.162431518796417e-06,
+      "loss": 0.7831,
+      "step": 4326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3769102096557617,
+      "learning_rate": 4.1603391735908905e-06,
+      "loss": 0.8568,
+      "step": 4327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3617157936096191,
+      "learning_rate": 4.158246979679755e-06,
+      "loss": 0.8097,
+      "step": 4328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2954801321029663,
+      "learning_rate": 4.15615493743999e-06,
+      "loss": 0.8551,
+      "step": 4329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6748530864715576,
+      "learning_rate": 4.154063047248552e-06,
+      "loss": 1.0089,
+      "step": 4330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2920039892196655,
+      "learning_rate": 4.151971309482367e-06,
+      "loss": 0.8485,
+      "step": 4331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2719368934631348,
+      "learning_rate": 4.149879724518336e-06,
+      "loss": 0.8238,
+      "step": 4332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.342431902885437,
+      "learning_rate": 4.147788292733332e-06,
+      "loss": 0.7924,
+      "step": 4333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.26565682888031,
+      "learning_rate": 4.1456970145042e-06,
+      "loss": 0.8113,
+      "step": 4334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2834006547927856,
+      "learning_rate": 4.143605890207754e-06,
+      "loss": 0.8683,
+      "step": 4335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2176612615585327,
+      "learning_rate": 4.1415149202207884e-06,
+      "loss": 0.7646,
+      "step": 4336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3073310852050781,
+      "learning_rate": 4.139424104920061e-06,
+      "loss": 0.8484,
+      "step": 4337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2431532144546509,
+      "learning_rate": 4.137333444682308e-06,
+      "loss": 0.8436,
+      "step": 4338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2739372253417969,
+      "learning_rate": 4.135242939884233e-06,
+      "loss": 0.8062,
+      "step": 4339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.227687120437622,
+      "learning_rate": 4.133152590902516e-06,
+      "loss": 0.7824,
+      "step": 4340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2488012313842773,
+      "learning_rate": 4.131062398113807e-06,
+      "loss": 0.8401,
+      "step": 4341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2406039237976074,
+      "learning_rate": 4.128972361894725e-06,
+      "loss": 0.8156,
+      "step": 4342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6288854479789734,
+      "learning_rate": 4.126882482621865e-06,
+      "loss": 1.0407,
+      "step": 4343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2141033411026,
+      "learning_rate": 4.1247927606717936e-06,
+      "loss": 0.7691,
+      "step": 4344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3675227165222168,
+      "learning_rate": 4.122703196421046e-06,
+      "loss": 0.8177,
+      "step": 4345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.265896201133728,
+      "learning_rate": 4.120613790246131e-06,
+      "loss": 0.8275,
+      "step": 4346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.19873046875,
+      "learning_rate": 4.1185245425235256e-06,
+      "loss": 0.7653,
+      "step": 4347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2165870666503906,
+      "learning_rate": 4.116435453629687e-06,
+      "loss": 0.7696,
+      "step": 4348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2792717218399048,
+      "learning_rate": 4.114346523941035e-06,
+      "loss": 0.8092,
+      "step": 4349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3286734819412231,
+      "learning_rate": 4.112257753833961e-06,
+      "loss": 0.8594,
+      "step": 4350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2775447368621826,
+      "learning_rate": 4.110169143684833e-06,
+      "loss": 0.8697,
+      "step": 4351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.261393666267395,
+      "learning_rate": 4.10808069386999e-06,
+      "loss": 0.7474,
+      "step": 4352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2931748628616333,
+      "learning_rate": 4.105992404765736e-06,
+      "loss": 0.809,
+      "step": 4353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2414768934249878,
+      "learning_rate": 4.103904276748348e-06,
+      "loss": 0.8474,
+      "step": 4354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2411224842071533,
+      "learning_rate": 4.101816310194079e-06,
+      "loss": 0.8017,
+      "step": 4355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.271645426750183,
+      "learning_rate": 4.099728505479151e-06,
+      "loss": 0.866,
+      "step": 4356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6591108441352844,
+      "learning_rate": 4.097640862979752e-06,
+      "loss": 1.0064,
+      "step": 4357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2210168838500977,
+      "learning_rate": 4.095553383072043e-06,
+      "loss": 0.8685,
+      "step": 4358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6010392904281616,
+      "learning_rate": 4.093466066132159e-06,
+      "loss": 1.0219,
+      "step": 4359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.628093957901001,
+      "learning_rate": 4.091378912536206e-06,
+      "loss": 0.9994,
+      "step": 4360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2627160549163818,
+      "learning_rate": 4.089291922660255e-06,
+      "loss": 0.8209,
+      "step": 4361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.304437279701233,
+      "learning_rate": 4.087205096880348e-06,
+      "loss": 0.7522,
+      "step": 4362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2645999193191528,
+      "learning_rate": 4.085118435572505e-06,
+      "loss": 0.8448,
+      "step": 4363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.242073655128479,
+      "learning_rate": 4.083031939112708e-06,
+      "loss": 0.8092,
+      "step": 4364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.639585018157959,
+      "learning_rate": 4.080945607876916e-06,
+      "loss": 1.0231,
+      "step": 4365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2090173959732056,
+      "learning_rate": 4.0788594422410496e-06,
+      "loss": 0.7867,
+      "step": 4366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.210017442703247,
+      "learning_rate": 4.076773442581009e-06,
+      "loss": 0.7836,
+      "step": 4367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3066728115081787,
+      "learning_rate": 4.074687609272661e-06,
+      "loss": 0.8395,
+      "step": 4368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2861348390579224,
+      "learning_rate": 4.07260194269184e-06,
+      "loss": 0.785,
+      "step": 4369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2747879028320312,
+      "learning_rate": 4.070516443214348e-06,
+      "loss": 0.7716,
+      "step": 4370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3001803159713745,
+      "learning_rate": 4.068431111215968e-06,
+      "loss": 0.8337,
+      "step": 4371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2367002964019775,
+      "learning_rate": 4.066345947072443e-06,
+      "loss": 0.7808,
+      "step": 4372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2190614938735962,
+      "learning_rate": 4.0642609511594875e-06,
+      "loss": 0.8148,
+      "step": 4373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1942354440689087,
+      "learning_rate": 4.0621761238527866e-06,
+      "loss": 0.8324,
+      "step": 4374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6790394186973572,
+      "learning_rate": 4.060091465527997e-06,
+      "loss": 1.0396,
+      "step": 4375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1857222318649292,
+      "learning_rate": 4.058006976560743e-06,
+      "loss": 0.7645,
+      "step": 4376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2509191036224365,
+      "learning_rate": 4.055922657326618e-06,
+      "loss": 0.7921,
+      "step": 4377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1681632995605469,
+      "learning_rate": 4.0538385082011834e-06,
+      "loss": 0.8162,
+      "step": 4378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2933732271194458,
+      "learning_rate": 4.051754529559974e-06,
+      "loss": 0.7869,
+      "step": 4379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3076328039169312,
+      "learning_rate": 4.049670721778492e-06,
+      "loss": 0.8536,
+      "step": 4380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2323002815246582,
+      "learning_rate": 4.047587085232209e-06,
+      "loss": 0.8239,
+      "step": 4381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6500477194786072,
+      "learning_rate": 4.045503620296561e-06,
+      "loss": 1.0217,
+      "step": 4382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.175573468208313,
+      "learning_rate": 4.0434203273469615e-06,
+      "loss": 0.7808,
+      "step": 4383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3180153369903564,
+      "learning_rate": 4.041337206758789e-06,
+      "loss": 0.8627,
+      "step": 4384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6284903287887573,
+      "learning_rate": 4.039254258907391e-06,
+      "loss": 1.0451,
+      "step": 4385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.612859308719635,
+      "learning_rate": 4.037171484168079e-06,
+      "loss": 1.0344,
+      "step": 4386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6045985817909241,
+      "learning_rate": 4.035088882916143e-06,
+      "loss": 1.0224,
+      "step": 4387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.215358853340149,
+      "learning_rate": 4.033006455526836e-06,
+      "loss": 0.7628,
+      "step": 4388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.32438325881958,
+      "learning_rate": 4.030924202375379e-06,
+      "loss": 0.7995,
+      "step": 4389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2565983533859253,
+      "learning_rate": 4.028842123836965e-06,
+      "loss": 0.8208,
+      "step": 4390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6408143639564514,
+      "learning_rate": 4.0267602202867506e-06,
+      "loss": 0.9935,
+      "step": 4391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6295559406280518,
+      "learning_rate": 4.024678492099867e-06,
+      "loss": 0.9939,
+      "step": 4392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2907140254974365,
+      "learning_rate": 4.0225969396514076e-06,
+      "loss": 0.8322,
+      "step": 4393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6241195797920227,
+      "learning_rate": 4.02051556331644e-06,
+      "loss": 1.0349,
+      "step": 4394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.324623465538025,
+      "learning_rate": 4.0184343634699955e-06,
+      "loss": 0.8183,
+      "step": 4395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.272612452507019,
+      "learning_rate": 4.016353340487078e-06,
+      "loss": 0.8277,
+      "step": 4396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2316983938217163,
+      "learning_rate": 4.014272494742651e-06,
+      "loss": 0.7879,
+      "step": 4397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3419877290725708,
+      "learning_rate": 4.012191826611659e-06,
+      "loss": 0.8403,
+      "step": 4398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2890114784240723,
+      "learning_rate": 4.0101113364690025e-06,
+      "loss": 0.8115,
+      "step": 4399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2663120031356812,
+      "learning_rate": 4.008031024689558e-06,
+      "loss": 0.8889,
+      "step": 4400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2398658990859985,
+      "learning_rate": 4.0059508916481624e-06,
+      "loss": 0.8662,
+      "step": 4401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2595102787017822,
+      "learning_rate": 4.0038709377196305e-06,
+      "loss": 0.8096,
+      "step": 4402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6854233741760254,
+      "learning_rate": 4.001791163278735e-06,
+      "loss": 1.0088,
+      "step": 4403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6478098034858704,
+      "learning_rate": 3.999711568700222e-06,
+      "loss": 1.0137,
+      "step": 4404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4249145984649658,
+      "learning_rate": 3.9976321543588015e-06,
+      "loss": 0.8708,
+      "step": 4405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3267203569412231,
+      "learning_rate": 3.995552920629156e-06,
+      "loss": 0.776,
+      "step": 4406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6403699517250061,
+      "learning_rate": 3.9934738678859305e-06,
+      "loss": 1.0282,
+      "step": 4407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.306689977645874,
+      "learning_rate": 3.991394996503741e-06,
+      "loss": 0.8099,
+      "step": 4408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2164289951324463,
+      "learning_rate": 3.989316306857166e-06,
+      "loss": 0.8354,
+      "step": 4409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.246890664100647,
+      "learning_rate": 3.987237799320758e-06,
+      "loss": 0.8222,
+      "step": 4410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2095091342926025,
+      "learning_rate": 3.985159474269031e-06,
+      "loss": 0.8124,
+      "step": 4411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2713879346847534,
+      "learning_rate": 3.98308133207647e-06,
+      "loss": 0.7887,
+      "step": 4412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.228501558303833,
+      "learning_rate": 3.9810033731175215e-06,
+      "loss": 0.8141,
+      "step": 4413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2887389659881592,
+      "learning_rate": 3.9789255977666086e-06,
+      "loss": 0.816,
+      "step": 4414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.66269451379776,
+      "learning_rate": 3.976848006398111e-06,
+      "loss": 1.0327,
+      "step": 4415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2491357326507568,
+      "learning_rate": 3.974770599386382e-06,
+      "loss": 0.836,
+      "step": 4416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2504388093948364,
+      "learning_rate": 3.972693377105736e-06,
+      "loss": 0.8672,
+      "step": 4417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2779850959777832,
+      "learning_rate": 3.970616339930463e-06,
+      "loss": 0.8136,
+      "step": 4418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.260345220565796,
+      "learning_rate": 3.9685394882348105e-06,
+      "loss": 0.7912,
+      "step": 4419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6353626251220703,
+      "learning_rate": 3.966462822392998e-06,
+      "loss": 1.0314,
+      "step": 4420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2341934442520142,
+      "learning_rate": 3.964386342779206e-06,
+      "loss": 0.8516,
+      "step": 4421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59923255443573,
+      "learning_rate": 3.962310049767591e-06,
+      "loss": 1.0191,
+      "step": 4422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2347145080566406,
+      "learning_rate": 3.9602339437322665e-06,
+      "loss": 0.7669,
+      "step": 4423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2872170209884644,
+      "learning_rate": 3.958158025047315e-06,
+      "loss": 0.8021,
+      "step": 4424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2654203176498413,
+      "learning_rate": 3.956082294086791e-06,
+      "loss": 0.8766,
+      "step": 4425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.315756916999817,
+      "learning_rate": 3.954006751224706e-06,
+      "loss": 0.8391,
+      "step": 4426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2307796478271484,
+      "learning_rate": 3.9519313968350425e-06,
+      "loss": 0.7915,
+      "step": 4427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.645741879940033,
+      "learning_rate": 3.949856231291748e-06,
+      "loss": 1.0225,
+      "step": 4428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2355669736862183,
+      "learning_rate": 3.94778125496874e-06,
+      "loss": 0.8245,
+      "step": 4429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2106701135635376,
+      "learning_rate": 3.945706468239895e-06,
+      "loss": 0.7869,
+      "step": 4430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2952345609664917,
+      "learning_rate": 3.94363187147906e-06,
+      "loss": 0.8446,
+      "step": 4431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3584622144699097,
+      "learning_rate": 3.941557465060044e-06,
+      "loss": 0.8634,
+      "step": 4432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.232316493988037,
+      "learning_rate": 3.939483249356629e-06,
+      "loss": 0.8442,
+      "step": 4433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3600748777389526,
+      "learning_rate": 3.937409224742555e-06,
+      "loss": 0.8305,
+      "step": 4434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6286686658859253,
+      "learning_rate": 3.935335391591529e-06,
+      "loss": 1.016,
+      "step": 4435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2754923105239868,
+      "learning_rate": 3.933261750277226e-06,
+      "loss": 0.7962,
+      "step": 4436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6550734043121338,
+      "learning_rate": 3.931188301173287e-06,
+      "loss": 1.0208,
+      "step": 4437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2312183380126953,
+      "learning_rate": 3.929115044653316e-06,
+      "loss": 0.8145,
+      "step": 4438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.292770504951477,
+      "learning_rate": 3.927041981090881e-06,
+      "loss": 0.8209,
+      "step": 4439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6412322521209717,
+      "learning_rate": 3.924969110859518e-06,
+      "loss": 1.0463,
+      "step": 4440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2378548383712769,
+      "learning_rate": 3.9228964343327295e-06,
+      "loss": 0.7549,
+      "step": 4441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6096159219741821,
+      "learning_rate": 3.92082395188398e-06,
+      "loss": 1.0186,
+      "step": 4442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6199464201927185,
+      "learning_rate": 3.918751663886697e-06,
+      "loss": 1.0368,
+      "step": 4443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2552344799041748,
+      "learning_rate": 3.916679570714279e-06,
+      "loss": 0.8197,
+      "step": 4444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6304754614830017,
+      "learning_rate": 3.914607672740087e-06,
+      "loss": 1.0244,
+      "step": 4445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3901238441467285,
+      "learning_rate": 3.912535970337445e-06,
+      "loss": 0.8227,
+      "step": 4446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.24294114112854,
+      "learning_rate": 3.910464463879642e-06,
+      "loss": 0.7473,
+      "step": 4447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6014816164970398,
+      "learning_rate": 3.908393153739932e-06,
+      "loss": 1.0175,
+      "step": 4448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.300952672958374,
+      "learning_rate": 3.906322040291538e-06,
+      "loss": 0.791,
+      "step": 4449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2361897230148315,
+      "learning_rate": 3.904251123907642e-06,
+      "loss": 0.8126,
+      "step": 4450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7011151909828186,
+      "learning_rate": 3.902180404961391e-06,
+      "loss": 1.0289,
+      "step": 4451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.655212938785553,
+      "learning_rate": 3.9001098838258975e-06,
+      "loss": 1.0211,
+      "step": 4452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6668226718902588,
+      "learning_rate": 3.8980395608742415e-06,
+      "loss": 1.0334,
+      "step": 4453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2232675552368164,
+      "learning_rate": 3.895969436479464e-06,
+      "loss": 0.832,
+      "step": 4454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2029587030410767,
+      "learning_rate": 3.893899511014567e-06,
+      "loss": 0.9109,
+      "step": 4455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2765791416168213,
+      "learning_rate": 3.891829784852522e-06,
+      "loss": 0.8336,
+      "step": 4456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6755621433258057,
+      "learning_rate": 3.8897602583662656e-06,
+      "loss": 1.0274,
+      "step": 4457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3137000799179077,
+      "learning_rate": 3.887690931928693e-06,
+      "loss": 0.841,
+      "step": 4458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.391974925994873,
+      "learning_rate": 3.885621805912665e-06,
+      "loss": 0.8597,
+      "step": 4459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7013073563575745,
+      "learning_rate": 3.883552880691011e-06,
+      "loss": 1.0399,
+      "step": 4460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6856892108917236,
+      "learning_rate": 3.881484156636518e-06,
+      "loss": 1.0055,
+      "step": 4461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.30654776096344,
+      "learning_rate": 3.87941563412194e-06,
+      "loss": 0.8146,
+      "step": 4462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2264072895050049,
+      "learning_rate": 3.877347313519992e-06,
+      "loss": 0.849,
+      "step": 4463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6032237410545349,
+      "learning_rate": 3.8752791952033565e-06,
+      "loss": 1.0324,
+      "step": 4464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.241317868232727,
+      "learning_rate": 3.87321127954468e-06,
+      "loss": 0.7481,
+      "step": 4465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2649297714233398,
+      "learning_rate": 3.871143566916567e-06,
+      "loss": 0.8041,
+      "step": 4466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.336201548576355,
+      "learning_rate": 3.869076057691587e-06,
+      "loss": 0.7612,
+      "step": 4467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2602720260620117,
+      "learning_rate": 3.867008752242277e-06,
+      "loss": 0.7603,
+      "step": 4468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2532200813293457,
+      "learning_rate": 3.864941650941137e-06,
+      "loss": 0.8441,
+      "step": 4469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2135969400405884,
+      "learning_rate": 3.862874754160624e-06,
+      "loss": 0.857,
+      "step": 4470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2188589572906494,
+      "learning_rate": 3.8608080622731615e-06,
+      "loss": 0.8179,
+      "step": 4471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7013084888458252,
+      "learning_rate": 3.8587415756511385e-06,
+      "loss": 1.0046,
+      "step": 4472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3504966497421265,
+      "learning_rate": 3.856675294666907e-06,
+      "loss": 0.8639,
+      "step": 4473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.328002691268921,
+      "learning_rate": 3.854609219692778e-06,
+      "loss": 0.84,
+      "step": 4474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2595857381820679,
+      "learning_rate": 3.852543351101026e-06,
+      "loss": 0.8415,
+      "step": 4475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6254869103431702,
+      "learning_rate": 3.850477689263891e-06,
+      "loss": 1.0158,
+      "step": 4476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.291003942489624,
+      "learning_rate": 3.848412234553577e-06,
+      "loss": 0.8109,
+      "step": 4477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2684123516082764,
+      "learning_rate": 3.8463469873422456e-06,
+      "loss": 0.8487,
+      "step": 4478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1865808963775635,
+      "learning_rate": 3.844281948002023e-06,
+      "loss": 0.8457,
+      "step": 4479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2997890710830688,
+      "learning_rate": 3.842217116905e-06,
+      "loss": 0.8253,
+      "step": 4480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2098300457000732,
+      "learning_rate": 3.840152494423228e-06,
+      "loss": 0.8053,
+      "step": 4481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6900601387023926,
+      "learning_rate": 3.838088080928724e-06,
+      "loss": 1.0125,
+      "step": 4482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6532071232795715,
+      "learning_rate": 3.836023876793458e-06,
+      "loss": 1.0376,
+      "step": 4483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.229946494102478,
+      "learning_rate": 3.833959882389374e-06,
+      "loss": 0.8295,
+      "step": 4484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3148366212844849,
+      "learning_rate": 3.831896098088374e-06,
+      "loss": 0.8627,
+      "step": 4485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2106176614761353,
+      "learning_rate": 3.8298325242623185e-06,
+      "loss": 0.7558,
+      "step": 4486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6384183764457703,
+      "learning_rate": 3.827769161283031e-06,
+      "loss": 1.0178,
+      "step": 4487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1996872425079346,
+      "learning_rate": 3.825706009522304e-06,
+      "loss": 0.7555,
+      "step": 4488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3104242086410522,
+      "learning_rate": 3.823643069351883e-06,
+      "loss": 0.8423,
+      "step": 4489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.284827709197998,
+      "learning_rate": 3.821580341143481e-06,
+      "loss": 0.8127,
+      "step": 4490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2353578805923462,
+      "learning_rate": 3.819517825268769e-06,
+      "loss": 0.8018,
+      "step": 4491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2615306377410889,
+      "learning_rate": 3.817455522099383e-06,
+      "loss": 0.8172,
+      "step": 4492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2816849946975708,
+      "learning_rate": 3.81539343200692e-06,
+      "loss": 0.8743,
+      "step": 4493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.322094202041626,
+      "learning_rate": 3.813331555362938e-06,
+      "loss": 0.8183,
+      "step": 4494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2777519226074219,
+      "learning_rate": 3.8112698925389525e-06,
+      "loss": 0.8465,
+      "step": 4495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3069958686828613,
+      "learning_rate": 3.809208443906449e-06,
+      "loss": 0.796,
+      "step": 4496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.222934603691101,
+      "learning_rate": 3.8071472098368694e-06,
+      "loss": 0.7708,
+      "step": 4497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3050134181976318,
+      "learning_rate": 3.8050861907016146e-06,
+      "loss": 0.8187,
+      "step": 4498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7033875584602356,
+      "learning_rate": 3.8030253868720533e-06,
+      "loss": 1.0395,
+      "step": 4499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.241831660270691,
+      "learning_rate": 3.8009647987195084e-06,
+      "loss": 0.8328,
+      "step": 4500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2553355693817139,
+      "learning_rate": 3.7989044266152707e-06,
+      "loss": 0.7706,
+      "step": 4501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3481887578964233,
+      "learning_rate": 3.7968442709305844e-06,
+      "loss": 0.8771,
+      "step": 4502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6085338592529297,
+      "learning_rate": 3.794784332036662e-06,
+      "loss": 1.0079,
+      "step": 4503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.213266372680664,
+      "learning_rate": 3.792724610304673e-06,
+      "loss": 0.844,
+      "step": 4504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2860554456710815,
+      "learning_rate": 3.7906651061057485e-06,
+      "loss": 0.8381,
+      "step": 4505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2791486978530884,
+      "learning_rate": 3.78860581981098e-06,
+      "loss": 0.8167,
+      "step": 4506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.21354079246521,
+      "learning_rate": 3.786546751791422e-06,
+      "loss": 0.8254,
+      "step": 4507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2532802820205688,
+      "learning_rate": 3.784487902418085e-06,
+      "loss": 0.8415,
+      "step": 4508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3080912828445435,
+      "learning_rate": 3.782429272061947e-06,
+      "loss": 0.8151,
+      "step": 4509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3252109289169312,
+      "learning_rate": 3.7803708610939376e-06,
+      "loss": 0.7981,
+      "step": 4510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.281925082206726,
+      "learning_rate": 3.778312669884957e-06,
+      "loss": 0.8691,
+      "step": 4511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2585666179656982,
+      "learning_rate": 3.776254698805857e-06,
+      "loss": 0.8638,
+      "step": 4512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.218280553817749,
+      "learning_rate": 3.7741969482274556e-06,
+      "loss": 0.8302,
+      "step": 4513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.19737708568573,
+      "learning_rate": 3.772139418520525e-06,
+      "loss": 0.8109,
+      "step": 4514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.276649832725525,
+      "learning_rate": 3.770082110055807e-06,
+      "loss": 0.8715,
+      "step": 4515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1923645734786987,
+      "learning_rate": 3.7680250232039943e-06,
+      "loss": 0.7699,
+      "step": 4516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2477391958236694,
+      "learning_rate": 3.7659681583357447e-06,
+      "loss": 0.8268,
+      "step": 4517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6856231689453125,
+      "learning_rate": 3.7639115158216732e-06,
+      "loss": 1.0175,
+      "step": 4518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2527415752410889,
+      "learning_rate": 3.7618550960323584e-06,
+      "loss": 0.813,
+      "step": 4519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2457289695739746,
+      "learning_rate": 3.7597988993383345e-06,
+      "loss": 0.7958,
+      "step": 4520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.22726309299469,
+      "learning_rate": 3.7577429261100996e-06,
+      "loss": 0.8197,
+      "step": 4521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2505731582641602,
+      "learning_rate": 3.7556871767181064e-06,
+      "loss": 0.8033,
+      "step": 4522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2654980421066284,
+      "learning_rate": 3.753631651532774e-06,
+      "loss": 0.8365,
+      "step": 4523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2581963539123535,
+      "learning_rate": 3.7515763509244744e-06,
+      "loss": 0.7927,
+      "step": 4524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.232268214225769,
+      "learning_rate": 3.749521275263545e-06,
+      "loss": 0.7666,
+      "step": 4525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2910521030426025,
+      "learning_rate": 3.747466424920276e-06,
+      "loss": 0.8457,
+      "step": 4526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2539424896240234,
+      "learning_rate": 3.745411800264926e-06,
+      "loss": 0.7606,
+      "step": 4527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2766504287719727,
+      "learning_rate": 3.7433574016677033e-06,
+      "loss": 0.7821,
+      "step": 4528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.23460853099823,
+      "learning_rate": 3.741303229498783e-06,
+      "loss": 0.8177,
+      "step": 4529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.659340500831604,
+      "learning_rate": 3.7392492841282922e-06,
+      "loss": 1.0354,
+      "step": 4530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2919176816940308,
+      "learning_rate": 3.737195565926327e-06,
+      "loss": 0.7996,
+      "step": 4531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.353601098060608,
+      "learning_rate": 3.735142075262933e-06,
+      "loss": 0.8437,
+      "step": 4532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3276011943817139,
+      "learning_rate": 3.733088812508119e-06,
+      "loss": 0.8487,
+      "step": 4533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2752668857574463,
+      "learning_rate": 3.7310357780318547e-06,
+      "loss": 0.8033,
+      "step": 4534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3138158321380615,
+      "learning_rate": 3.728982972204065e-06,
+      "loss": 0.7706,
+      "step": 4535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3106509447097778,
+      "learning_rate": 3.7269303953946336e-06,
+      "loss": 0.7637,
+      "step": 4536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2755405902862549,
+      "learning_rate": 3.7248780479734037e-06,
+      "loss": 0.8064,
+      "step": 4537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.248266577720642,
+      "learning_rate": 3.722825930310182e-06,
+      "loss": 0.8593,
+      "step": 4538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6356272101402283,
+      "learning_rate": 3.7207740427747273e-06,
+      "loss": 1.0173,
+      "step": 4539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2844018936157227,
+      "learning_rate": 3.7187223857367584e-06,
+      "loss": 0.8301,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.229256510734558,
+      "learning_rate": 3.7166709595659516e-06,
+      "loss": 0.8046,
+      "step": 4541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2065513134002686,
+      "learning_rate": 3.714619764631949e-06,
+      "loss": 0.7744,
+      "step": 4542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6271182894706726,
+      "learning_rate": 3.7125688013043415e-06,
+      "loss": 1.0255,
+      "step": 4543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2482575178146362,
+      "learning_rate": 3.710518069952682e-06,
+      "loss": 0.7989,
+      "step": 4544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5990399718284607,
+      "learning_rate": 3.708467570946482e-06,
+      "loss": 0.9987,
+      "step": 4545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2470521926879883,
+      "learning_rate": 3.7064173046552133e-06,
+      "loss": 0.8423,
+      "step": 4546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5923970937728882,
+      "learning_rate": 3.7043672714483015e-06,
+      "loss": 1.019,
+      "step": 4547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.328834891319275,
+      "learning_rate": 3.7023174716951304e-06,
+      "loss": 0.8382,
+      "step": 4548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2759242057800293,
+      "learning_rate": 3.7002679057650456e-06,
+      "loss": 0.8188,
+      "step": 4549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2437392473220825,
+      "learning_rate": 3.6982185740273493e-06,
+      "loss": 0.8508,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2226641178131104,
+      "learning_rate": 3.696169476851299e-06,
+      "loss": 0.7812,
+      "step": 4551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3775835037231445,
+      "learning_rate": 3.6941206146061113e-06,
+      "loss": 0.795,
+      "step": 4552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2894258499145508,
+      "learning_rate": 3.69207198766096e-06,
+      "loss": 0.7904,
+      "step": 4553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3624697923660278,
+      "learning_rate": 3.6900235963849806e-06,
+      "loss": 0.8754,
+      "step": 4554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2673016786575317,
+      "learning_rate": 3.687975441147261e-06,
+      "loss": 0.8163,
+      "step": 4555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1671278476715088,
+      "learning_rate": 3.685927522316846e-06,
+      "loss": 0.7824,
+      "step": 4556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3566190004348755,
+      "learning_rate": 3.6838798402627417e-06,
+      "loss": 0.7948,
+      "step": 4557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2247974872589111,
+      "learning_rate": 3.6818323953539115e-06,
+      "loss": 0.7894,
+      "step": 4558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2545344829559326,
+      "learning_rate": 3.6797851879592738e-06,
+      "loss": 0.8603,
+      "step": 4559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2176662683486938,
+      "learning_rate": 3.677738218447703e-06,
+      "loss": 0.8193,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2515250444412231,
+      "learning_rate": 3.675691487188033e-06,
+      "loss": 0.7959,
+      "step": 4561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2349798679351807,
+      "learning_rate": 3.673644994549057e-06,
+      "loss": 0.7667,
+      "step": 4562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2467949390411377,
+      "learning_rate": 3.6715987408995213e-06,
+      "loss": 0.8483,
+      "step": 4563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2951722145080566,
+      "learning_rate": 3.6695527266081284e-06,
+      "loss": 0.8211,
+      "step": 4564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2317781448364258,
+      "learning_rate": 3.66750695204354e-06,
+      "loss": 0.8186,
+      "step": 4565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2107505798339844,
+      "learning_rate": 3.6654614175743773e-06,
+      "loss": 0.8675,
+      "step": 4566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.181704044342041,
+      "learning_rate": 3.6634161235692134e-06,
+      "loss": 0.7964,
+      "step": 4567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3091106414794922,
+      "learning_rate": 3.6613710703965767e-06,
+      "loss": 0.8084,
+      "step": 4568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.211680293083191,
+      "learning_rate": 3.65932625842496e-06,
+      "loss": 0.8484,
+      "step": 4569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6546837687492371,
+      "learning_rate": 3.6572816880228074e-06,
+      "loss": 0.999,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6253174543380737,
+      "learning_rate": 3.655237359558519e-06,
+      "loss": 1.0237,
+      "step": 4571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1934279203414917,
+      "learning_rate": 3.6531932734004497e-06,
+      "loss": 0.786,
+      "step": 4572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2089786529541016,
+      "learning_rate": 3.6511494299169175e-06,
+      "loss": 0.7654,
+      "step": 4573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2461588382720947,
+      "learning_rate": 3.6491058294761924e-06,
+      "loss": 0.8195,
+      "step": 4574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2619048357009888,
+      "learning_rate": 3.6470624724464986e-06,
+      "loss": 0.8537,
+      "step": 4575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6387355923652649,
+      "learning_rate": 3.645019359196018e-06,
+      "loss": 1.0186,
+      "step": 4576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.633800745010376,
+      "learning_rate": 3.6429764900928934e-06,
+      "loss": 0.9943,
+      "step": 4577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.229786992073059,
+      "learning_rate": 3.640933865505215e-06,
+      "loss": 0.7686,
+      "step": 4578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6331978440284729,
+      "learning_rate": 3.638891485801036e-06,
+      "loss": 1.0372,
+      "step": 4579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3343188762664795,
+      "learning_rate": 3.6368493513483606e-06,
+      "loss": 0.7988,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2423651218414307,
+      "learning_rate": 3.634807462515154e-06,
+      "loss": 0.8559,
+      "step": 4581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2243311405181885,
+      "learning_rate": 3.632765819669332e-06,
+      "loss": 0.8102,
+      "step": 4582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2109837532043457,
+      "learning_rate": 3.6307244231787697e-06,
+      "loss": 0.8235,
+      "step": 4583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2146072387695312,
+      "learning_rate": 3.6286832734112937e-06,
+      "loss": 0.7855,
+      "step": 4584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2390726804733276,
+      "learning_rate": 3.6266423707346926e-06,
+      "loss": 0.7879,
+      "step": 4585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1750751733779907,
+      "learning_rate": 3.6246017155167036e-06,
+      "loss": 0.8142,
+      "step": 4586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6496126055717468,
+      "learning_rate": 3.6225613081250243e-06,
+      "loss": 1.0076,
+      "step": 4587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.293938398361206,
+      "learning_rate": 3.620521148927304e-06,
+      "loss": 0.8801,
+      "step": 4588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2845851182937622,
+      "learning_rate": 3.6184812382911516e-06,
+      "loss": 0.8324,
+      "step": 4589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2739475965499878,
+      "learning_rate": 3.616441576584126e-06,
+      "loss": 0.8416,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3012616634368896,
+      "learning_rate": 3.614402164173746e-06,
+      "loss": 0.8539,
+      "step": 4591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6187114119529724,
+      "learning_rate": 3.612363001427481e-06,
+      "loss": 1.0153,
+      "step": 4592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.232221245765686,
+      "learning_rate": 3.6103240887127605e-06,
+      "loss": 0.7701,
+      "step": 4593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.253893256187439,
+      "learning_rate": 3.6082854263969642e-06,
+      "loss": 0.7931,
+      "step": 4594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6221067309379578,
+      "learning_rate": 3.6062470148474306e-06,
+      "loss": 1.0141,
+      "step": 4595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5973473191261292,
+      "learning_rate": 3.604208854431448e-06,
+      "loss": 1.0244,
+      "step": 4596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.284206509590149,
+      "learning_rate": 3.602170945516266e-06,
+      "loss": 0.8662,
+      "step": 4597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2804661989212036,
+      "learning_rate": 3.600133288469083e-06,
+      "loss": 0.8601,
+      "step": 4598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2383720874786377,
+      "learning_rate": 3.5980958836570578e-06,
+      "loss": 0.83,
+      "step": 4599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2480465173721313,
+      "learning_rate": 3.596058731447295e-06,
+      "loss": 0.8233,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6012210845947266,
+      "learning_rate": 3.5940218322068647e-06,
+      "loss": 1.024,
+      "step": 4601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2674508094787598,
+      "learning_rate": 3.591985186302783e-06,
+      "loss": 0.8043,
+      "step": 4602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6312143802642822,
+      "learning_rate": 3.5899487941020237e-06,
+      "loss": 1.0189,
+      "step": 4603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2451730966567993,
+      "learning_rate": 3.5879126559715127e-06,
+      "loss": 0.7883,
+      "step": 4604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.225619912147522,
+      "learning_rate": 3.585876772278135e-06,
+      "loss": 0.7702,
+      "step": 4605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2440401315689087,
+      "learning_rate": 3.5838411433887243e-06,
+      "loss": 0.7959,
+      "step": 4606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3096343278884888,
+      "learning_rate": 3.58180576967007e-06,
+      "loss": 0.7862,
+      "step": 4607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2352315187454224,
+      "learning_rate": 3.57977065148892e-06,
+      "loss": 0.8352,
+      "step": 4608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.693739652633667,
+      "learning_rate": 3.57773578921197e-06,
+      "loss": 1.0211,
+      "step": 4609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2181097269058228,
+      "learning_rate": 3.5757011832058697e-06,
+      "loss": 0.8031,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1728143692016602,
+      "learning_rate": 3.5736668338372265e-06,
+      "loss": 0.7458,
+      "step": 4611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.243321180343628,
+      "learning_rate": 3.5716327414726028e-06,
+      "loss": 0.8525,
+      "step": 4612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2486557960510254,
+      "learning_rate": 3.569598906478508e-06,
+      "loss": 0.7735,
+      "step": 4613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1891505718231201,
+      "learning_rate": 3.5675653292214096e-06,
+      "loss": 0.754,
+      "step": 4614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2247767448425293,
+      "learning_rate": 3.565532010067727e-06,
+      "loss": 0.8231,
+      "step": 4615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6472101211547852,
+      "learning_rate": 3.5634989493838387e-06,
+      "loss": 1.0239,
+      "step": 4616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.278808355331421,
+      "learning_rate": 3.561466147536068e-06,
+      "loss": 0.8298,
+      "step": 4617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2146539688110352,
+      "learning_rate": 3.559433604890695e-06,
+      "loss": 0.7895,
+      "step": 4618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2849440574645996,
+      "learning_rate": 3.5574013218139547e-06,
+      "loss": 0.8165,
+      "step": 4619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2957170009613037,
+      "learning_rate": 3.555369298672037e-06,
+      "loss": 0.8422,
+      "step": 4620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6507517695426941,
+      "learning_rate": 3.553337535831079e-06,
+      "loss": 1.0353,
+      "step": 4621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.620877206325531,
+      "learning_rate": 3.5513060336571743e-06,
+      "loss": 1.0255,
+      "step": 4622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2297050952911377,
+      "learning_rate": 3.5492747925163683e-06,
+      "loss": 0.8517,
+      "step": 4623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2514206171035767,
+      "learning_rate": 3.5472438127746656e-06,
+      "loss": 0.8037,
+      "step": 4624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6178726553916931,
+      "learning_rate": 3.5452130947980145e-06,
+      "loss": 1.0275,
+      "step": 4625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2563328742980957,
+      "learning_rate": 3.5431826389523193e-06,
+      "loss": 0.7661,
+      "step": 4626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3350173234939575,
+      "learning_rate": 3.5411524456034384e-06,
+      "loss": 0.7736,
+      "step": 4627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2667624950408936,
+      "learning_rate": 3.539122515117186e-06,
+      "loss": 0.7996,
+      "step": 4628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.230038046836853,
+      "learning_rate": 3.537092847859323e-06,
+      "loss": 0.8502,
+      "step": 4629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2727707624435425,
+      "learning_rate": 3.535063444195564e-06,
+      "loss": 0.838,
+      "step": 4630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6630803346633911,
+      "learning_rate": 3.533034304491577e-06,
+      "loss": 1.0259,
+      "step": 4631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2860618829727173,
+      "learning_rate": 3.5310054291129864e-06,
+      "loss": 0.8495,
+      "step": 4632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.220719575881958,
+      "learning_rate": 3.528976818425363e-06,
+      "loss": 0.8125,
+      "step": 4633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2943663597106934,
+      "learning_rate": 3.5269484727942316e-06,
+      "loss": 0.8899,
+      "step": 4634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.309919834136963,
+      "learning_rate": 3.52492039258507e-06,
+      "loss": 0.8105,
+      "step": 4635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2817336320877075,
+      "learning_rate": 3.5228925781633106e-06,
+      "loss": 0.7834,
+      "step": 4636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3097411394119263,
+      "learning_rate": 3.5208650298943335e-06,
+      "loss": 0.7595,
+      "step": 4637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2456815242767334,
+      "learning_rate": 3.518837748143472e-06,
+      "loss": 0.8099,
+      "step": 4638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.22903311252594,
+      "learning_rate": 3.5168107332760114e-06,
+      "loss": 0.7992,
+      "step": 4639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2804566621780396,
+      "learning_rate": 3.5147839856571943e-06,
+      "loss": 0.8492,
+      "step": 4640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2370548248291016,
+      "learning_rate": 3.512757505652207e-06,
+      "loss": 0.7629,
+      "step": 4641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7562258243560791,
+      "learning_rate": 3.510731293626189e-06,
+      "loss": 0.9819,
+      "step": 4642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2014042139053345,
+      "learning_rate": 3.508705349944237e-06,
+      "loss": 0.8125,
+      "step": 4643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.285132646560669,
+      "learning_rate": 3.506679674971396e-06,
+      "loss": 0.7981,
+      "step": 4644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2333528995513916,
+      "learning_rate": 3.5046542690726616e-06,
+      "loss": 0.8121,
+      "step": 4645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2560749053955078,
+      "learning_rate": 3.502629132612979e-06,
+      "loss": 0.8346,
+      "step": 4646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.328824758529663,
+      "learning_rate": 3.5006042659572515e-06,
+      "loss": 0.8865,
+      "step": 4647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2927846908569336,
+      "learning_rate": 3.49857966947033e-06,
+      "loss": 0.8667,
+      "step": 4648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2501825094223022,
+      "learning_rate": 3.4965553435170154e-06,
+      "loss": 0.8425,
+      "step": 4649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.650646448135376,
+      "learning_rate": 3.4945312884620588e-06,
+      "loss": 1.019,
+      "step": 4650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6371724009513855,
+      "learning_rate": 3.492507504670168e-06,
+      "loss": 1.0205,
+      "step": 4651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.597967267036438,
+      "learning_rate": 3.490483992505999e-06,
+      "loss": 1.0382,
+      "step": 4652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5918112993240356,
+      "learning_rate": 3.4884607523341575e-06,
+      "loss": 1.0127,
+      "step": 4653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5978596210479736,
+      "learning_rate": 3.4864377845191986e-06,
+      "loss": 1.0203,
+      "step": 4654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6314107179641724,
+      "learning_rate": 3.484415089425636e-06,
+      "loss": 1.0156,
+      "step": 4655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.426561713218689,
+      "learning_rate": 3.4823926674179266e-06,
+      "loss": 0.858,
+      "step": 4656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.255280613899231,
+      "learning_rate": 3.480370518860481e-06,
+      "loss": 0.8539,
+      "step": 4657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.245728850364685,
+      "learning_rate": 3.4783486441176585e-06,
+      "loss": 0.8297,
+      "step": 4658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2186942100524902,
+      "learning_rate": 3.4763270435537736e-06,
+      "loss": 0.7575,
+      "step": 4659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.321421504020691,
+      "learning_rate": 3.4743057175330893e-06,
+      "loss": 0.7834,
+      "step": 4660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2467330694198608,
+      "learning_rate": 3.472284666419816e-06,
+      "loss": 0.8047,
+      "step": 4661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2516292333602905,
+      "learning_rate": 3.4702638905781166e-06,
+      "loss": 0.7776,
+      "step": 4662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2820208072662354,
+      "learning_rate": 3.468243390372107e-06,
+      "loss": 0.7551,
+      "step": 4663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6431647539138794,
+      "learning_rate": 3.4662231661658515e-06,
+      "loss": 1.0176,
+      "step": 4664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6585355997085571,
+      "learning_rate": 3.464203218323363e-06,
+      "loss": 1.0179,
+      "step": 4665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6182594895362854,
+      "learning_rate": 3.4621835472086054e-06,
+      "loss": 1.0226,
+      "step": 4666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3116099834442139,
+      "learning_rate": 3.460164153185494e-06,
+      "loss": 0.8528,
+      "step": 4667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5911046862602234,
+      "learning_rate": 3.458145036617896e-06,
+      "loss": 1.0312,
+      "step": 4668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.231176495552063,
+      "learning_rate": 3.4561261978696236e-06,
+      "loss": 0.7882,
+      "step": 4669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2863413095474243,
+      "learning_rate": 3.4541076373044403e-06,
+      "loss": 0.7915,
+      "step": 4670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4529589414596558,
+      "learning_rate": 3.4520893552860633e-06,
+      "loss": 0.852,
+      "step": 4671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6249333024024963,
+      "learning_rate": 3.4500713521781578e-06,
+      "loss": 1.0232,
+      "step": 4672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2884019613265991,
+      "learning_rate": 3.4480536283443354e-06,
+      "loss": 0.8202,
+      "step": 4673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6087340712547302,
+      "learning_rate": 3.4460361841481593e-06,
+      "loss": 1.021,
+      "step": 4674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6279198527336121,
+      "learning_rate": 3.4440190199531455e-06,
+      "loss": 1.006,
+      "step": 4675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2227613925933838,
+      "learning_rate": 3.442002136122756e-06,
+      "loss": 0.791,
+      "step": 4676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2498939037322998,
+      "learning_rate": 3.4399855330204025e-06,
+      "loss": 0.8184,
+      "step": 4677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.597917377948761,
+      "learning_rate": 3.437969211009449e-06,
+      "loss": 1.0471,
+      "step": 4678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3101094961166382,
+      "learning_rate": 3.4359531704532032e-06,
+      "loss": 0.8002,
+      "step": 4679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2501606941223145,
+      "learning_rate": 3.4339374117149303e-06,
+      "loss": 0.8235,
+      "step": 4680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1908867359161377,
+      "learning_rate": 3.4319219351578346e-06,
+      "loss": 0.7132,
+      "step": 4681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6339069604873657,
+      "learning_rate": 3.4299067411450804e-06,
+      "loss": 1.0064,
+      "step": 4682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2743388414382935,
+      "learning_rate": 3.427891830039771e-06,
+      "loss": 0.7906,
+      "step": 4683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3040934801101685,
+      "learning_rate": 3.4258772022049673e-06,
+      "loss": 0.8137,
+      "step": 4684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.275375247001648,
+      "learning_rate": 3.4238628580036716e-06,
+      "loss": 0.8344,
+      "step": 4685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6317038536071777,
+      "learning_rate": 3.4218487977988426e-06,
+      "loss": 1.0179,
+      "step": 4686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6091936230659485,
+      "learning_rate": 3.4198350219533806e-06,
+      "loss": 1.0029,
+      "step": 4687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2494345903396606,
+      "learning_rate": 3.4178215308301407e-06,
+      "loss": 0.8178,
+      "step": 4688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1856701374053955,
+      "learning_rate": 3.4158083247919204e-06,
+      "loss": 0.8313,
+      "step": 4689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6131260991096497,
+      "learning_rate": 3.413795404201474e-06,
+      "loss": 1.0287,
+      "step": 4690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.280561089515686,
+      "learning_rate": 3.4117827694214973e-06,
+      "loss": 0.8115,
+      "step": 4691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2796645164489746,
+      "learning_rate": 3.409770420814638e-06,
+      "loss": 0.8536,
+      "step": 4692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6083937287330627,
+      "learning_rate": 3.4077583587434892e-06,
+      "loss": 1.0102,
+      "step": 4693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.232988953590393,
+      "learning_rate": 3.405746583570598e-06,
+      "loss": 0.8179,
+      "step": 4694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6194283366203308,
+      "learning_rate": 3.4037350956584542e-06,
+      "loss": 1.0222,
+      "step": 4695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.599122166633606,
+      "learning_rate": 3.401723895369499e-06,
+      "loss": 1.0087,
+      "step": 4696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2468509674072266,
+      "learning_rate": 3.3997129830661186e-06,
+      "loss": 0.8086,
+      "step": 4697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.201095700263977,
+      "learning_rate": 3.397702359110653e-06,
+      "loss": 0.7621,
+      "step": 4698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2438477277755737,
+      "learning_rate": 3.3956920238653834e-06,
+      "loss": 0.8302,
+      "step": 4699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2754640579223633,
+      "learning_rate": 3.393681977692545e-06,
+      "loss": 0.8611,
+      "step": 4700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6156188249588013,
+      "learning_rate": 3.3916722209543153e-06,
+      "loss": 1.0056,
+      "step": 4701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2350499629974365,
+      "learning_rate": 3.389662754012826e-06,
+      "loss": 0.8483,
+      "step": 4702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5983719229698181,
+      "learning_rate": 3.38765357723015e-06,
+      "loss": 1.019,
+      "step": 4703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.249796986579895,
+      "learning_rate": 3.385644690968314e-06,
+      "loss": 0.8847,
+      "step": 4704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2254754304885864,
+      "learning_rate": 3.3836360955892853e-06,
+      "loss": 0.8341,
+      "step": 4705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.258568286895752,
+      "learning_rate": 3.3816277914549876e-06,
+      "loss": 0.7818,
+      "step": 4706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2458959817886353,
+      "learning_rate": 3.3796197789272844e-06,
+      "loss": 0.8359,
+      "step": 4707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3015027046203613,
+      "learning_rate": 3.3776120583679917e-06,
+      "loss": 0.8306,
+      "step": 4708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2102534770965576,
+      "learning_rate": 3.3756046301388672e-06,
+      "loss": 0.8051,
+      "step": 4709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6589135527610779,
+      "learning_rate": 3.373597494601625e-06,
+      "loss": 1.0205,
+      "step": 4710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3030672073364258,
+      "learning_rate": 3.3715906521179163e-06,
+      "loss": 0.8246,
+      "step": 4711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6092220544815063,
+      "learning_rate": 3.369584103049348e-06,
+      "loss": 0.9934,
+      "step": 4712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2698261737823486,
+      "learning_rate": 3.367577847757466e-06,
+      "loss": 0.79,
+      "step": 4713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6092221736907959,
+      "learning_rate": 3.365571886603772e-06,
+      "loss": 1.0244,
+      "step": 4714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2396776676177979,
+      "learning_rate": 3.3635662199497084e-06,
+      "loss": 0.7654,
+      "step": 4715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2957810163497925,
+      "learning_rate": 3.361560848156665e-06,
+      "loss": 0.768,
+      "step": 4716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1967673301696777,
+      "learning_rate": 3.359555771585984e-06,
+      "loss": 0.8082,
+      "step": 4717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.271452784538269,
+      "learning_rate": 3.3575509905989477e-06,
+      "loss": 0.8105,
+      "step": 4718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2186259031295776,
+      "learning_rate": 3.355546505556787e-06,
+      "loss": 0.8131,
+      "step": 4719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2371795177459717,
+      "learning_rate": 3.3535423168206804e-06,
+      "loss": 0.8017,
+      "step": 4720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6423107385635376,
+      "learning_rate": 3.351538424751757e-06,
+      "loss": 1.0235,
+      "step": 4721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2678501605987549,
+      "learning_rate": 3.349534829711084e-06,
+      "loss": 0.8184,
+      "step": 4722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1941323280334473,
+      "learning_rate": 3.3475315320596796e-06,
+      "loss": 0.8136,
+      "step": 4723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6369463205337524,
+      "learning_rate": 3.3455285321585084e-06,
+      "loss": 0.9976,
+      "step": 4724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5967705845832825,
+      "learning_rate": 3.343525830368485e-06,
+      "loss": 1.0093,
+      "step": 4725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2601159811019897,
+      "learning_rate": 3.3415234270504625e-06,
+      "loss": 0.7989,
+      "step": 4726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2901816368103027,
+      "learning_rate": 3.3395213225652446e-06,
+      "loss": 0.8302,
+      "step": 4727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3255562782287598,
+      "learning_rate": 3.3375195172735796e-06,
+      "loss": 0.8329,
+      "step": 4728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2728729248046875,
+      "learning_rate": 3.335518011536168e-06,
+      "loss": 0.8214,
+      "step": 4729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6313348412513733,
+      "learning_rate": 3.3335168057136475e-06,
+      "loss": 1.0114,
+      "step": 4730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.276485562324524,
+      "learning_rate": 3.3315159001666053e-06,
+      "loss": 0.7911,
+      "step": 4731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2984838485717773,
+      "learning_rate": 3.329515295255575e-06,
+      "loss": 0.8337,
+      "step": 4732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2960246801376343,
+      "learning_rate": 3.327514991341038e-06,
+      "loss": 0.8153,
+      "step": 4733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2662287950515747,
+      "learning_rate": 3.3255149887834186e-06,
+      "loss": 0.8578,
+      "step": 4734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2023813724517822,
+      "learning_rate": 3.3235152879430853e-06,
+      "loss": 0.7904,
+      "step": 4735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2198255062103271,
+      "learning_rate": 3.3215158891803556e-06,
+      "loss": 0.8397,
+      "step": 4736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2983148097991943,
+      "learning_rate": 3.319516792855493e-06,
+      "loss": 0.8303,
+      "step": 4737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2295979261398315,
+      "learning_rate": 3.3175179993287044e-06,
+      "loss": 0.7824,
+      "step": 4738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3252049684524536,
+      "learning_rate": 3.315519508960141e-06,
+      "loss": 0.7628,
+      "step": 4739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2365920543670654,
+      "learning_rate": 3.3135213221099006e-06,
+      "loss": 0.798,
+      "step": 4740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2339684963226318,
+      "learning_rate": 3.3115234391380303e-06,
+      "loss": 0.8377,
+      "step": 4741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6259206533432007,
+      "learning_rate": 3.3095258604045167e-06,
+      "loss": 0.998,
+      "step": 4742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.216759204864502,
+      "learning_rate": 3.3075285862692923e-06,
+      "loss": 0.7793,
+      "step": 4743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3101561069488525,
+      "learning_rate": 3.305531617092237e-06,
+      "loss": 0.775,
+      "step": 4744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6176878809928894,
+      "learning_rate": 3.303534953233178e-06,
+      "loss": 1.0169,
+      "step": 4745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2139664888381958,
+      "learning_rate": 3.301538595051882e-06,
+      "loss": 0.7909,
+      "step": 4746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2845265865325928,
+      "learning_rate": 3.299542542908062e-06,
+      "loss": 0.8034,
+      "step": 4747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.620025098323822,
+      "learning_rate": 3.2975467971613763e-06,
+      "loss": 1.0335,
+      "step": 4748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5900619626045227,
+      "learning_rate": 3.2955513581714325e-06,
+      "loss": 0.9816,
+      "step": 4749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2526745796203613,
+      "learning_rate": 3.2935562262977773e-06,
+      "loss": 0.8071,
+      "step": 4750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2866638898849487,
+      "learning_rate": 3.291561401899901e-06,
+      "loss": 0.7775,
+      "step": 4751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6026549935340881,
+      "learning_rate": 3.2895668853372448e-06,
+      "loss": 0.9896,
+      "step": 4752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5985314249992371,
+      "learning_rate": 3.287572676969191e-06,
+      "loss": 1.03,
+      "step": 4753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6098494529724121,
+      "learning_rate": 3.285578777155064e-06,
+      "loss": 1.0161,
+      "step": 4754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2375012636184692,
+      "learning_rate": 3.2835851862541345e-06,
+      "loss": 0.8505,
+      "step": 4755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2471375465393066,
+      "learning_rate": 3.2815919046256193e-06,
+      "loss": 0.8196,
+      "step": 4756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.357895851135254,
+      "learning_rate": 3.2795989326286804e-06,
+      "loss": 0.7992,
+      "step": 4757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.266472578048706,
+      "learning_rate": 3.2776062706224187e-06,
+      "loss": 0.8355,
+      "step": 4758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6365193724632263,
+      "learning_rate": 3.275613918965881e-06,
+      "loss": 1.0038,
+      "step": 4759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2008517980575562,
+      "learning_rate": 3.2736218780180615e-06,
+      "loss": 0.7549,
+      "step": 4760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2372976541519165,
+      "learning_rate": 3.271630148137899e-06,
+      "loss": 0.8251,
+      "step": 4761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2436786890029907,
+      "learning_rate": 3.269638729684269e-06,
+      "loss": 0.7885,
+      "step": 4762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.274701714515686,
+      "learning_rate": 3.2676476230159957e-06,
+      "loss": 0.8744,
+      "step": 4763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2801774740219116,
+      "learning_rate": 3.26565682849185e-06,
+      "loss": 0.8482,
+      "step": 4764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2250372171401978,
+      "learning_rate": 3.2636663464705428e-06,
+      "loss": 0.8362,
+      "step": 4765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6421998143196106,
+      "learning_rate": 3.261676177310728e-06,
+      "loss": 1.0126,
+      "step": 4766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1996995210647583,
+      "learning_rate": 3.2596863213710027e-06,
+      "loss": 0.7423,
+      "step": 4767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2129720449447632,
+      "learning_rate": 3.2576967790099133e-06,
+      "loss": 0.8108,
+      "step": 4768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2749288082122803,
+      "learning_rate": 3.255707550585945e-06,
+      "loss": 0.8145,
+      "step": 4769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2215960025787354,
+      "learning_rate": 3.253718636457526e-06,
+      "loss": 0.7868,
+      "step": 4770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2522518634796143,
+      "learning_rate": 3.251730036983027e-06,
+      "loss": 0.8018,
+      "step": 4771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6467268466949463,
+      "learning_rate": 3.249741752520767e-06,
+      "loss": 1.0217,
+      "step": 4772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2762624025344849,
+      "learning_rate": 3.2477537834290063e-06,
+      "loss": 0.7688,
+      "step": 4773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3031368255615234,
+      "learning_rate": 3.245766130065945e-06,
+      "loss": 0.806,
+      "step": 4774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5941973328590393,
+      "learning_rate": 3.243778792789727e-06,
+      "loss": 1.0163,
+      "step": 4775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2311770915985107,
+      "learning_rate": 3.241791771958445e-06,
+      "loss": 0.8586,
+      "step": 4776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2116957902908325,
+      "learning_rate": 3.2398050679301283e-06,
+      "loss": 0.7839,
+      "step": 4777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2339560985565186,
+      "learning_rate": 3.2378186810627533e-06,
+      "loss": 0.7997,
+      "step": 4778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.327589988708496,
+      "learning_rate": 3.2358326117142337e-06,
+      "loss": 0.7933,
+      "step": 4779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6495524644851685,
+      "learning_rate": 3.2338468602424323e-06,
+      "loss": 1.0155,
+      "step": 4780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6655275821685791,
+      "learning_rate": 3.231861427005153e-06,
+      "loss": 1.0148,
+      "step": 4781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2519248723983765,
+      "learning_rate": 3.2298763123601396e-06,
+      "loss": 0.7833,
+      "step": 4782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1933097839355469,
+      "learning_rate": 3.227891516665078e-06,
+      "loss": 0.8022,
+      "step": 4783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2583346366882324,
+      "learning_rate": 3.2259070402776032e-06,
+      "loss": 0.763,
+      "step": 4784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5953554511070251,
+      "learning_rate": 3.223922883555287e-06,
+      "loss": 1.0117,
+      "step": 4785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2526787519454956,
+      "learning_rate": 3.2219390468556456e-06,
+      "loss": 0.7814,
+      "step": 4786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.177843689918518,
+      "learning_rate": 3.219955530536132e-06,
+      "loss": 0.7535,
+      "step": 4787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2379968166351318,
+      "learning_rate": 3.217972334954152e-06,
+      "loss": 0.7936,
+      "step": 4788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2289454936981201,
+      "learning_rate": 3.215989460467046e-06,
+      "loss": 0.7839,
+      "step": 4789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2574067115783691,
+      "learning_rate": 3.214006907432097e-06,
+      "loss": 0.8141,
+      "step": 4790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.294022798538208,
+      "learning_rate": 3.2120246762065355e-06,
+      "loss": 0.8746,
+      "step": 4791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6712297201156616,
+      "learning_rate": 3.210042767147526e-06,
+      "loss": 1.0137,
+      "step": 4792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2681150436401367,
+      "learning_rate": 3.208061180612182e-06,
+      "loss": 0.774,
+      "step": 4793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2691484689712524,
+      "learning_rate": 3.2060799169575518e-06,
+      "loss": 0.8097,
+      "step": 4794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2272523641586304,
+      "learning_rate": 3.204098976540635e-06,
+      "loss": 0.8003,
+      "step": 4795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2419483661651611,
+      "learning_rate": 3.2021183597183636e-06,
+      "loss": 0.7911,
+      "step": 4796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.223728060722351,
+      "learning_rate": 3.2001380668476173e-06,
+      "loss": 0.7997,
+      "step": 4797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2897740602493286,
+      "learning_rate": 3.198158098285213e-06,
+      "loss": 0.8276,
+      "step": 4798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2360320091247559,
+      "learning_rate": 3.196178454387915e-06,
+      "loss": 0.8049,
+      "step": 4799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2828606367111206,
+      "learning_rate": 3.1941991355124234e-06,
+      "loss": 0.8722,
+      "step": 4800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2283650636672974,
+      "learning_rate": 3.1922201420153835e-06,
+      "loss": 0.716,
+      "step": 4801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6303802132606506,
+      "learning_rate": 3.190241474253378e-06,
+      "loss": 1.0015,
+      "step": 4802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.315116286277771,
+      "learning_rate": 3.1882631325829362e-06,
+      "loss": 0.8409,
+      "step": 4803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2228999137878418,
+      "learning_rate": 3.1862851173605243e-06,
+      "loss": 0.8125,
+      "step": 4804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2455775737762451,
+      "learning_rate": 3.184307428942552e-06,
+      "loss": 0.7826,
+      "step": 4805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3801947832107544,
+      "learning_rate": 3.182330067685367e-06,
+      "loss": 0.8318,
+      "step": 4806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2999399900436401,
+      "learning_rate": 3.1803530339452645e-06,
+      "loss": 0.8157,
+      "step": 4807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6368336081504822,
+      "learning_rate": 3.178376328078473e-06,
+      "loss": 1.0314,
+      "step": 4808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3714369535446167,
+      "learning_rate": 3.176399950441168e-06,
+      "loss": 0.8296,
+      "step": 4809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2929270267486572,
+      "learning_rate": 3.1744239013894596e-06,
+      "loss": 0.8597,
+      "step": 4810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2451140880584717,
+      "learning_rate": 3.172448181279408e-06,
+      "loss": 0.7983,
+      "step": 4811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2311816215515137,
+      "learning_rate": 3.170472790467004e-06,
+      "loss": 0.807,
+      "step": 4812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2972265481948853,
+      "learning_rate": 3.1684977293081857e-06,
+      "loss": 0.8405,
+      "step": 4813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6200680732727051,
+      "learning_rate": 3.1665229981588274e-06,
+      "loss": 1.0168,
+      "step": 4814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2874836921691895,
+      "learning_rate": 3.164548597374751e-06,
+      "loss": 0.8067,
+      "step": 4815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.278631329536438,
+      "learning_rate": 3.162574527311709e-06,
+      "loss": 0.8432,
+      "step": 4816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3090717792510986,
+      "learning_rate": 3.160600788325404e-06,
+      "loss": 0.8007,
+      "step": 4817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6180094480514526,
+      "learning_rate": 3.1586273807714695e-06,
+      "loss": 1.0319,
+      "step": 4818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.271843671798706,
+      "learning_rate": 3.1566543050054887e-06,
+      "loss": 0.7773,
+      "step": 4819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2409709692001343,
+      "learning_rate": 3.154681561382978e-06,
+      "loss": 0.7982,
+      "step": 4820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2957661151885986,
+      "learning_rate": 3.1527091502593978e-06,
+      "loss": 0.8647,
+      "step": 4821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2469834089279175,
+      "learning_rate": 3.1507370719901443e-06,
+      "loss": 0.7755,
+      "step": 4822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5920112133026123,
+      "learning_rate": 3.1487653269305606e-06,
+      "loss": 0.9974,
+      "step": 4823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3253084421157837,
+      "learning_rate": 3.146793915435924e-06,
+      "loss": 0.8728,
+      "step": 4824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6092669367790222,
+      "learning_rate": 3.1448228378614534e-06,
+      "loss": 0.9875,
+      "step": 4825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3175736665725708,
+      "learning_rate": 3.142852094562307e-06,
+      "loss": 0.7641,
+      "step": 4826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.20430326461792,
+      "learning_rate": 3.1408816858935857e-06,
+      "loss": 0.8059,
+      "step": 4827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6076450943946838,
+      "learning_rate": 3.138911612210326e-06,
+      "loss": 1.0094,
+      "step": 4828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2061656713485718,
+      "learning_rate": 3.1369418738675055e-06,
+      "loss": 0.7456,
+      "step": 4829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2286040782928467,
+      "learning_rate": 3.134972471220043e-06,
+      "loss": 0.7949,
+      "step": 4830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2442272901535034,
+      "learning_rate": 3.1330034046227958e-06,
+      "loss": 0.8183,
+      "step": 4831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6089701652526855,
+      "learning_rate": 3.1310346744305587e-06,
+      "loss": 1.0196,
+      "step": 4832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2651768922805786,
+      "learning_rate": 3.1290662809980686e-06,
+      "loss": 0.8615,
+      "step": 4833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6133652329444885,
+      "learning_rate": 3.1270982246800006e-06,
+      "loss": 1.0073,
+      "step": 4834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.222933292388916,
+      "learning_rate": 3.125130505830971e-06,
+      "loss": 0.7869,
+      "step": 4835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2354950904846191,
+      "learning_rate": 3.1231631248055305e-06,
+      "loss": 0.8289,
+      "step": 4836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2493183612823486,
+      "learning_rate": 3.121196081958172e-06,
+      "loss": 0.8286,
+      "step": 4837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2156823873519897,
+      "learning_rate": 3.11922937764333e-06,
+      "loss": 0.8236,
+      "step": 4838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.243955135345459,
+      "learning_rate": 3.1172630122153737e-06,
+      "loss": 0.8695,
+      "step": 4839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2731611728668213,
+      "learning_rate": 3.1152969860286127e-06,
+      "loss": 0.7743,
+      "step": 4840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.301038384437561,
+      "learning_rate": 3.1133312994372956e-06,
+      "loss": 0.8585,
+      "step": 4841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6085789203643799,
+      "learning_rate": 3.111365952795611e-06,
+      "loss": 1.0364,
+      "step": 4842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2508625984191895,
+      "learning_rate": 3.1094009464576846e-06,
+      "loss": 0.7656,
+      "step": 4843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2059513330459595,
+      "learning_rate": 3.1074362807775805e-06,
+      "loss": 0.7661,
+      "step": 4844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2228813171386719,
+      "learning_rate": 3.105471956109302e-06,
+      "loss": 0.7927,
+      "step": 4845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.223561406135559,
+      "learning_rate": 3.103507972806793e-06,
+      "loss": 0.7605,
+      "step": 4846
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6052600145339966,
+      "learning_rate": 3.1015443312239333e-06,
+      "loss": 1.0029,
+      "step": 4847
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2146708965301514,
+      "learning_rate": 3.0995810317145416e-06,
+      "loss": 0.8361,
+      "step": 4848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1631698608398438,
+      "learning_rate": 3.097618074632375e-06,
+      "loss": 0.7893,
+      "step": 4849
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2677496671676636,
+      "learning_rate": 3.0956554603311294e-06,
+      "loss": 0.865,
+      "step": 4850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2094353437423706,
+      "learning_rate": 3.09369318916444e-06,
+      "loss": 0.8083,
+      "step": 4851
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6025097966194153,
+      "learning_rate": 3.091731261485877e-06,
+      "loss": 1.0073,
+      "step": 4852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2235729694366455,
+      "learning_rate": 3.0897696776489505e-06,
+      "loss": 0.8034,
+      "step": 4853
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2430036067962646,
+      "learning_rate": 3.0878084380071106e-06,
+      "loss": 0.8584,
+      "step": 4854
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3404568433761597,
+      "learning_rate": 3.0858475429137423e-06,
+      "loss": 0.8414,
+      "step": 4855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2541193962097168,
+      "learning_rate": 3.0838869927221693e-06,
+      "loss": 0.7988,
+      "step": 4856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1957484483718872,
+      "learning_rate": 3.081926787785654e-06,
+      "loss": 0.7866,
+      "step": 4857
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2173597812652588,
+      "learning_rate": 3.079966928457396e-06,
+      "loss": 0.8409,
+      "step": 4858
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.284885048866272,
+      "learning_rate": 3.0780074150905346e-06,
+      "loss": 0.8276,
+      "step": 4859
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2122001647949219,
+      "learning_rate": 3.0760482480381414e-06,
+      "loss": 0.8834,
+      "step": 4860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2883752584457397,
+      "learning_rate": 3.074089427653233e-06,
+      "loss": 0.766,
+      "step": 4861
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.274546504020691,
+      "learning_rate": 3.0721309542887565e-06,
+      "loss": 0.8438,
+      "step": 4862
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2192410230636597,
+      "learning_rate": 3.070172828297603e-06,
+      "loss": 0.7904,
+      "step": 4863
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1679631471633911,
+      "learning_rate": 3.0682150500325924e-06,
+      "loss": 0.8352,
+      "step": 4864
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6369791626930237,
+      "learning_rate": 3.0662576198464933e-06,
+      "loss": 1.0119,
+      "step": 4865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2624832391738892,
+      "learning_rate": 3.0643005380920023e-06,
+      "loss": 0.8223,
+      "step": 4866
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6326658725738525,
+      "learning_rate": 3.062343805121757e-06,
+      "loss": 1.0096,
+      "step": 4867
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.317689299583435,
+      "learning_rate": 3.06038742128833e-06,
+      "loss": 0.8692,
+      "step": 4868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3476570844650269,
+      "learning_rate": 3.058431386944235e-06,
+      "loss": 0.8613,
+      "step": 4869
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7349133491516113,
+      "learning_rate": 3.0564757024419187e-06,
+      "loss": 0.8423,
+      "step": 4870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3127127885818481,
+      "learning_rate": 3.054520368133769e-06,
+      "loss": 0.7644,
+      "step": 4871
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6154013872146606,
+      "learning_rate": 3.0525653843721024e-06,
+      "loss": 1.0172,
+      "step": 4872
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2563868761062622,
+      "learning_rate": 3.0506107515091843e-06,
+      "loss": 0.7418,
+      "step": 4873
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6674937605857849,
+      "learning_rate": 3.0486564698972066e-06,
+      "loss": 1.0364,
+      "step": 4874
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5895105600357056,
+      "learning_rate": 3.0467025398883033e-06,
+      "loss": 0.9911,
+      "step": 4875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2842735052108765,
+      "learning_rate": 3.0447489618345416e-06,
+      "loss": 0.7684,
+      "step": 4876
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.296642780303955,
+      "learning_rate": 3.04279573608793e-06,
+      "loss": 0.7632,
+      "step": 4877
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.257358431816101,
+      "learning_rate": 3.0408428630004095e-06,
+      "loss": 0.7926,
+      "step": 4878
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2364473342895508,
+      "learning_rate": 3.038890342923858e-06,
+      "loss": 0.7734,
+      "step": 4879
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.30104660987854,
+      "learning_rate": 3.03693817621009e-06,
+      "loss": 0.7878,
+      "step": 4880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.654697835445404,
+      "learning_rate": 3.0349863632108607e-06,
+      "loss": 1.0024,
+      "step": 4881
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2549751996994019,
+      "learning_rate": 3.0330349042778535e-06,
+      "loss": 0.7466,
+      "step": 4882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2451461553573608,
+      "learning_rate": 3.031083799762695e-06,
+      "loss": 0.7994,
+      "step": 4883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5982977747917175,
+      "learning_rate": 3.029133050016942e-06,
+      "loss": 1.0158,
+      "step": 4884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2311789989471436,
+      "learning_rate": 3.027182655392095e-06,
+      "loss": 0.7931,
+      "step": 4885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5858182311058044,
+      "learning_rate": 3.0252326162395824e-06,
+      "loss": 1.0115,
+      "step": 4886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2179795503616333,
+      "learning_rate": 3.023282932910775e-06,
+      "loss": 0.8097,
+      "step": 4887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.385314702987671,
+      "learning_rate": 3.021333605756973e-06,
+      "loss": 0.8591,
+      "step": 4888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6448004245758057,
+      "learning_rate": 3.019384635129421e-06,
+      "loss": 1.0289,
+      "step": 4889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5965030789375305,
+      "learning_rate": 3.01743602137929e-06,
+      "loss": 1.0064,
+      "step": 4890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2174921035766602,
+      "learning_rate": 3.0154877648576952e-06,
+      "loss": 0.8181,
+      "step": 4891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5921655297279358,
+      "learning_rate": 3.0135398659156796e-06,
+      "loss": 1.0073,
+      "step": 4892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2511180639266968,
+      "learning_rate": 3.011592324904229e-06,
+      "loss": 0.8086,
+      "step": 4893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2473435401916504,
+      "learning_rate": 3.009645142174259e-06,
+      "loss": 0.8098,
+      "step": 4894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2333811521530151,
+      "learning_rate": 3.007698318076625e-06,
+      "loss": 0.8702,
+      "step": 4895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6146270036697388,
+      "learning_rate": 3.005751852962112e-06,
+      "loss": 1.0221,
+      "step": 4896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6224642992019653,
+      "learning_rate": 3.003805747181449e-06,
+      "loss": 1.0354,
+      "step": 4897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6247946619987488,
+      "learning_rate": 3.0018600010852916e-06,
+      "loss": 1.0162,
+      "step": 4898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2227661609649658,
+      "learning_rate": 2.9999146150242353e-06,
+      "loss": 0.8025,
+      "step": 4899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.246415138244629,
+      "learning_rate": 2.9979695893488116e-06,
+      "loss": 0.8,
+      "step": 4900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2489084005355835,
+      "learning_rate": 2.996024924409483e-06,
+      "loss": 0.879,
+      "step": 4901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2256429195404053,
+      "learning_rate": 2.9940806205566496e-06,
+      "loss": 0.8449,
+      "step": 4902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2303813695907593,
+      "learning_rate": 2.9921366781406447e-06,
+      "loss": 0.8465,
+      "step": 4903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2298632860183716,
+      "learning_rate": 2.9901930975117418e-06,
+      "loss": 0.771,
+      "step": 4904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3129448890686035,
+      "learning_rate": 2.988249879020142e-06,
+      "loss": 0.847,
+      "step": 4905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2057726383209229,
+      "learning_rate": 2.9863070230159846e-06,
+      "loss": 0.7999,
+      "step": 4906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3192532062530518,
+      "learning_rate": 2.984364529849343e-06,
+      "loss": 0.8823,
+      "step": 4907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2548515796661377,
+      "learning_rate": 2.9824223998702286e-06,
+      "loss": 0.7682,
+      "step": 4908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.297148585319519,
+      "learning_rate": 2.9804806334285813e-06,
+      "loss": 0.8152,
+      "step": 4909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2253361940383911,
+      "learning_rate": 2.9785392308742788e-06,
+      "loss": 0.7984,
+      "step": 4910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7407810091972351,
+      "learning_rate": 2.9765981925571323e-06,
+      "loss": 1.0119,
+      "step": 4911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2234468460083008,
+      "learning_rate": 2.9746575188268917e-06,
+      "loss": 0.773,
+      "step": 4912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2317917346954346,
+      "learning_rate": 2.9727172100332345e-06,
+      "loss": 0.8189,
+      "step": 4913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2538325786590576,
+      "learning_rate": 2.970777266525776e-06,
+      "loss": 0.7777,
+      "step": 4914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.21487557888031,
+      "learning_rate": 2.9688376886540638e-06,
+      "loss": 0.8035,
+      "step": 4915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2755143642425537,
+      "learning_rate": 2.966898476767584e-06,
+      "loss": 0.8272,
+      "step": 4916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6240252256393433,
+      "learning_rate": 2.9649596312157526e-06,
+      "loss": 0.994,
+      "step": 4917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.25681471824646,
+      "learning_rate": 2.963021152347919e-06,
+      "loss": 0.8495,
+      "step": 4918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.292931079864502,
+      "learning_rate": 2.9610830405133683e-06,
+      "loss": 0.7858,
+      "step": 4919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2456045150756836,
+      "learning_rate": 2.9591452960613232e-06,
+      "loss": 0.7628,
+      "step": 4920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1746375560760498,
+      "learning_rate": 2.9572079193409336e-06,
+      "loss": 0.7768,
+      "step": 4921
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1851171255111694,
+      "learning_rate": 2.9552709107012857e-06,
+      "loss": 0.8013,
+      "step": 4922
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6583694219589233,
+      "learning_rate": 2.9533342704913983e-06,
+      "loss": 1.038,
+      "step": 4923
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2448763847351074,
+      "learning_rate": 2.9513979990602294e-06,
+      "loss": 0.8201,
+      "step": 4924
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2959345579147339,
+      "learning_rate": 2.9494620967566635e-06,
+      "loss": 0.7322,
+      "step": 4925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2633165121078491,
+      "learning_rate": 2.9475265639295213e-06,
+      "loss": 0.8109,
+      "step": 4926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2053083181381226,
+      "learning_rate": 2.9455914009275565e-06,
+      "loss": 0.8467,
+      "step": 4927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6172176599502563,
+      "learning_rate": 2.94365660809946e-06,
+      "loss": 1.0287,
+      "step": 4928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.292890191078186,
+      "learning_rate": 2.94172218579385e-06,
+      "loss": 0.7756,
+      "step": 4929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.274726152420044,
+      "learning_rate": 2.9397881343592806e-06,
+      "loss": 0.8561,
+      "step": 4930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5982641577720642,
+      "learning_rate": 2.937854454144238e-06,
+      "loss": 1.0221,
+      "step": 4931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3194772005081177,
+      "learning_rate": 2.9359211454971457e-06,
+      "loss": 0.7755,
+      "step": 4932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2615399360656738,
+      "learning_rate": 2.9339882087663562e-06,
+      "loss": 0.8066,
+      "step": 4933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2745505571365356,
+      "learning_rate": 2.9320556443001528e-06,
+      "loss": 0.7842,
+      "step": 4934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2219487428665161,
+      "learning_rate": 2.930123452446759e-06,
+      "loss": 0.8404,
+      "step": 4935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3125348091125488,
+      "learning_rate": 2.9281916335543257e-06,
+      "loss": 0.8323,
+      "step": 4936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6499187350273132,
+      "learning_rate": 2.9262601879709386e-06,
+      "loss": 1.0029,
+      "step": 4937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3105921745300293,
+      "learning_rate": 2.924329116044612e-06,
+      "loss": 0.8421,
+      "step": 4938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3303630352020264,
+      "learning_rate": 2.9223984181233e-06,
+      "loss": 0.832,
+      "step": 4939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3521562814712524,
+      "learning_rate": 2.9204680945548853e-06,
+      "loss": 0.8786,
+      "step": 4940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2258450984954834,
+      "learning_rate": 2.918538145687182e-06,
+      "loss": 0.7883,
+      "step": 4941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3571832180023193,
+      "learning_rate": 2.9166085718679373e-06,
+      "loss": 0.8789,
+      "step": 4942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6125993728637695,
+      "learning_rate": 2.9146793734448353e-06,
+      "loss": 1.0199,
+      "step": 4943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2670600414276123,
+      "learning_rate": 2.9127505507654872e-06,
+      "loss": 0.8079,
+      "step": 4944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2578634023666382,
+      "learning_rate": 2.9108221041774376e-06,
+      "loss": 0.8081,
+      "step": 4945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2365988492965698,
+      "learning_rate": 2.9088940340281613e-06,
+      "loss": 0.8213,
+      "step": 4946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.349663496017456,
+      "learning_rate": 2.906966340665074e-06,
+      "loss": 0.863,
+      "step": 4947
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6431282758712769,
+      "learning_rate": 2.9050390244355107e-06,
+      "loss": 1.0191,
+      "step": 4948
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6088172197341919,
+      "learning_rate": 2.9031120856867513e-06,
+      "loss": 1.0149,
+      "step": 4949
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.246419072151184,
+      "learning_rate": 2.901185524765996e-06,
+      "loss": 0.8389,
+      "step": 4950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3217816352844238,
+      "learning_rate": 2.899259342020387e-06,
+      "loss": 0.8557,
+      "step": 4951
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.227630615234375,
+      "learning_rate": 2.8973335377969913e-06,
+      "loss": 0.7823,
+      "step": 4952
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2333184480667114,
+      "learning_rate": 2.895408112442811e-06,
+      "loss": 0.8065,
+      "step": 4953
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2856465578079224,
+      "learning_rate": 2.8934830663047774e-06,
+      "loss": 0.7826,
+      "step": 4954
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2380229234695435,
+      "learning_rate": 2.8915583997297582e-06,
+      "loss": 0.803,
+      "step": 4955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6429518461227417,
+      "learning_rate": 2.8896341130645454e-06,
+      "loss": 0.9949,
+      "step": 4956
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2023894786834717,
+      "learning_rate": 2.8877102066558714e-06,
+      "loss": 0.833,
+      "step": 4957
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2265218496322632,
+      "learning_rate": 2.885786680850392e-06,
+      "loss": 0.7837,
+      "step": 4958
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1886744499206543,
+      "learning_rate": 2.883863535994702e-06,
+      "loss": 0.8619,
+      "step": 4959
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2102218866348267,
+      "learning_rate": 2.88194077243532e-06,
+      "loss": 0.747,
+      "step": 4960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6001368165016174,
+      "learning_rate": 2.8800183905187005e-06,
+      "loss": 1.0035,
+      "step": 4961
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.223306655883789,
+      "learning_rate": 2.8780963905912263e-06,
+      "loss": 0.8392,
+      "step": 4962
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.219224452972412,
+      "learning_rate": 2.876174772999216e-06,
+      "loss": 0.7803,
+      "step": 4963
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5731828808784485,
+      "learning_rate": 2.8742535380889147e-06,
+      "loss": 1.0098,
+      "step": 4964
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2135390043258667,
+      "learning_rate": 2.872332686206502e-06,
+      "loss": 0.7903,
+      "step": 4965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1880050897598267,
+      "learning_rate": 2.870412217698084e-06,
+      "loss": 0.7725,
+      "step": 4966
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3130567073822021,
+      "learning_rate": 2.868492132909705e-06,
+      "loss": 0.8558,
+      "step": 4967
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2339165210723877,
+      "learning_rate": 2.866572432187332e-06,
+      "loss": 0.8231,
+      "step": 4968
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2264080047607422,
+      "learning_rate": 2.8646531158768664e-06,
+      "loss": 0.7726,
+      "step": 4969
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2477741241455078,
+      "learning_rate": 2.862734184324144e-06,
+      "loss": 0.7014,
+      "step": 4970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.275848150253296,
+      "learning_rate": 2.8608156378749254e-06,
+      "loss": 0.8057,
+      "step": 4971
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6983612179756165,
+      "learning_rate": 2.858897476874901e-06,
+      "loss": 1.0182,
+      "step": 4972
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2601269483566284,
+      "learning_rate": 2.856979701669701e-06,
+      "loss": 0.8572,
+      "step": 4973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.237510323524475,
+      "learning_rate": 2.8550623126048743e-06,
+      "loss": 0.8264,
+      "step": 4974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.26045823097229,
+      "learning_rate": 2.853145310025911e-06,
+      "loss": 0.7973,
+      "step": 4975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2536513805389404,
+      "learning_rate": 2.8512286942782237e-06,
+      "loss": 0.7406,
+      "step": 4976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6151105761528015,
+      "learning_rate": 2.8493124657071562e-06,
+      "loss": 1.0307,
+      "step": 4977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2589097023010254,
+      "learning_rate": 2.847396624657989e-06,
+      "loss": 0.803,
+      "step": 4978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.227854609489441,
+      "learning_rate": 2.845481171475925e-06,
+      "loss": 0.7651,
+      "step": 4979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6279001235961914,
+      "learning_rate": 2.8435661065060995e-06,
+      "loss": 1.0162,
+      "step": 4980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5766288042068481,
+      "learning_rate": 2.8416514300935815e-06,
+      "loss": 1.0059,
+      "step": 4981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5741817951202393,
+      "learning_rate": 2.8397371425833655e-06,
+      "loss": 1.0255,
+      "step": 4982
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.182556390762329,
+      "learning_rate": 2.837823244320379e-06,
+      "loss": 0.8325,
+      "step": 4983
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.599308967590332,
+      "learning_rate": 2.835909735649477e-06,
+      "loss": 1.0082,
+      "step": 4984
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.24018394947052,
+      "learning_rate": 2.833996616915443e-06,
+      "loss": 0.8103,
+      "step": 4985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6015749573707581,
+      "learning_rate": 2.8320838884629975e-06,
+      "loss": 1.0158,
+      "step": 4986
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2084335088729858,
+      "learning_rate": 2.8301715506367832e-06,
+      "loss": 0.7922,
+      "step": 4987
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2639940977096558,
+      "learning_rate": 2.8282596037813724e-06,
+      "loss": 0.8285,
+      "step": 4988
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6221386790275574,
+      "learning_rate": 2.826348048241274e-06,
+      "loss": 1.0266,
+      "step": 4989
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2159267663955688,
+      "learning_rate": 2.824436884360918e-06,
+      "loss": 0.7892,
+      "step": 4990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6023252606391907,
+      "learning_rate": 2.822526112484672e-06,
+      "loss": 1.0017,
+      "step": 4991
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6003069281578064,
+      "learning_rate": 2.820615732956826e-06,
+      "loss": 1.0229,
+      "step": 4992
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2768313884735107,
+      "learning_rate": 2.8187057461215993e-06,
+      "loss": 0.8522,
+      "step": 4993
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2923792600631714,
+      "learning_rate": 2.8167961523231485e-06,
+      "loss": 0.8648,
+      "step": 4994
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5901111960411072,
+      "learning_rate": 2.814886951905551e-06,
+      "loss": 1.0082,
+      "step": 4995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.243783950805664,
+      "learning_rate": 2.812978145212815e-06,
+      "loss": 0.7908,
+      "step": 4996
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1858514547348022,
+      "learning_rate": 2.811069732588883e-06,
+      "loss": 0.8282,
+      "step": 4997
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1987215280532837,
+      "learning_rate": 2.8091617143776183e-06,
+      "loss": 0.7711,
+      "step": 4998
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2012354135513306,
+      "learning_rate": 2.807254090922821e-06,
+      "loss": 0.8388,
+      "step": 4999
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2693686485290527,
+      "learning_rate": 2.8053468625682155e-06,
+      "loss": 0.8221,
+      "step": 5000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2227360010147095,
+      "learning_rate": 2.8034400296574534e-06,
+      "loss": 0.7784,
+      "step": 5001
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6460988521575928,
+      "learning_rate": 2.8015335925341213e-06,
+      "loss": 1.0101,
+      "step": 5002
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.305837631225586,
+      "learning_rate": 2.799627551541729e-06,
+      "loss": 0.9027,
+      "step": 5003
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6278030276298523,
+      "learning_rate": 2.7977219070237155e-06,
+      "loss": 1.0076,
+      "step": 5004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2832688093185425,
+      "learning_rate": 2.7958166593234526e-06,
+      "loss": 0.8499,
+      "step": 5005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1894114017486572,
+      "learning_rate": 2.7939118087842336e-06,
+      "loss": 0.7457,
+      "step": 5006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2782784700393677,
+      "learning_rate": 2.792007355749289e-06,
+      "loss": 0.8049,
+      "step": 5007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2980120182037354,
+      "learning_rate": 2.7901033005617682e-06,
+      "loss": 0.8891,
+      "step": 5008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2387611865997314,
+      "learning_rate": 2.7881996435647583e-06,
+      "loss": 0.7429,
+      "step": 5009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2114646434783936,
+      "learning_rate": 2.7862963851012677e-06,
+      "loss": 0.8338,
+      "step": 5010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1633528470993042,
+      "learning_rate": 2.7843935255142346e-06,
+      "loss": 0.7382,
+      "step": 5011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2455220222473145,
+      "learning_rate": 2.782491065146525e-06,
+      "loss": 0.8048,
+      "step": 5012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.262868046760559,
+      "learning_rate": 2.7805890043409367e-06,
+      "loss": 0.7024,
+      "step": 5013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6615947484970093,
+      "learning_rate": 2.77868734344019e-06,
+      "loss": 1.0112,
+      "step": 5014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3130282163619995,
+      "learning_rate": 2.7767860827869397e-06,
+      "loss": 0.7735,
+      "step": 5015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6392685770988464,
+      "learning_rate": 2.77488522272376e-06,
+      "loss": 1.0318,
+      "step": 5016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2185345888137817,
+      "learning_rate": 2.772984763593162e-06,
+      "loss": 0.7877,
+      "step": 5017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.263053297996521,
+      "learning_rate": 2.771084705737579e-06,
+      "loss": 0.8138,
+      "step": 5018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5983603596687317,
+      "learning_rate": 2.7691850494993716e-06,
+      "loss": 1.0008,
+      "step": 5019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5943521857261658,
+      "learning_rate": 2.7672857952208287e-06,
+      "loss": 0.9958,
+      "step": 5020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.265223741531372,
+      "learning_rate": 2.7653869432441704e-06,
+      "loss": 0.8044,
+      "step": 5021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2079098224639893,
+      "learning_rate": 2.7634884939115396e-06,
+      "loss": 0.7933,
+      "step": 5022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2935174703598022,
+      "learning_rate": 2.7615904475650113e-06,
+      "loss": 0.7726,
+      "step": 5023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2686325311660767,
+      "learning_rate": 2.759692804546581e-06,
+      "loss": 0.8146,
+      "step": 5024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.378241062164307,
+      "learning_rate": 2.7577955651981803e-06,
+      "loss": 1.0404,
+      "step": 5025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.23456871509552,
+      "learning_rate": 2.7558987298616613e-06,
+      "loss": 0.7469,
+      "step": 5026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.196433424949646,
+      "learning_rate": 2.7540022988788063e-06,
+      "loss": 0.8175,
+      "step": 5027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2650885581970215,
+      "learning_rate": 2.7521062725913207e-06,
+      "loss": 0.829,
+      "step": 5028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2713005542755127,
+      "learning_rate": 2.7502106513408445e-06,
+      "loss": 0.8083,
+      "step": 5029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.231572151184082,
+      "learning_rate": 2.7483154354689367e-06,
+      "loss": 0.7913,
+      "step": 5030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7044644355773926,
+      "learning_rate": 2.7464206253170913e-06,
+      "loss": 0.9934,
+      "step": 5031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3267425298690796,
+      "learning_rate": 2.744526221226721e-06,
+      "loss": 0.8295,
+      "step": 5032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2653073072433472,
+      "learning_rate": 2.7426322235391724e-06,
+      "loss": 0.7901,
+      "step": 5033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3378069400787354,
+      "learning_rate": 2.7407386325957134e-06,
+      "loss": 0.7673,
+      "step": 5034
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1999262571334839,
+      "learning_rate": 2.7388454487375426e-06,
+      "loss": 0.883,
+      "step": 5035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2351487874984741,
+      "learning_rate": 2.7369526723057802e-06,
+      "loss": 0.7338,
+      "step": 5036
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6409409642219543,
+      "learning_rate": 2.7350603036414813e-06,
+      "loss": 1.0313,
+      "step": 5037
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2810924053192139,
+      "learning_rate": 2.733168343085618e-06,
+      "loss": 0.8306,
+      "step": 5038
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.243112564086914,
+      "learning_rate": 2.731276790979097e-06,
+      "loss": 0.7947,
+      "step": 5039
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.238248348236084,
+      "learning_rate": 2.7293856476627446e-06,
+      "loss": 0.7787,
+      "step": 5040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2957879304885864,
+      "learning_rate": 2.727494913477321e-06,
+      "loss": 0.8846,
+      "step": 5041
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2276798486709595,
+      "learning_rate": 2.7256045887635052e-06,
+      "loss": 0.8158,
+      "step": 5042
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2650578022003174,
+      "learning_rate": 2.7237146738619048e-06,
+      "loss": 0.8462,
+      "step": 5043
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2836756706237793,
+      "learning_rate": 2.721825169113057e-06,
+      "loss": 0.8058,
+      "step": 5044
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3793736696243286,
+      "learning_rate": 2.719936074857421e-06,
+      "loss": 0.821,
+      "step": 5045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1854336261749268,
+      "learning_rate": 2.7180473914353827e-06,
+      "loss": 0.7776,
+      "step": 5046
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.210761547088623,
+      "learning_rate": 2.716159119187255e-06,
+      "loss": 0.7936,
+      "step": 5047
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6195927858352661,
+      "learning_rate": 2.71427125845328e-06,
+      "loss": 1.007,
+      "step": 5048
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6218660473823547,
+      "learning_rate": 2.712383809573619e-06,
+      "loss": 1.0203,
+      "step": 5049
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2146292924880981,
+      "learning_rate": 2.710496772888362e-06,
+      "loss": 0.7634,
+      "step": 5050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2677088975906372,
+      "learning_rate": 2.7086101487375236e-06,
+      "loss": 0.7946,
+      "step": 5051
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5822989344596863,
+      "learning_rate": 2.70672393746105e-06,
+      "loss": 0.9818,
+      "step": 5052
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2748210430145264,
+      "learning_rate": 2.704838139398805e-06,
+      "loss": 0.8015,
+      "step": 5053
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2581459283828735,
+      "learning_rate": 2.70295275489058e-06,
+      "loss": 0.7927,
+      "step": 5054
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2410154342651367,
+      "learning_rate": 2.7010677842760958e-06,
+      "loss": 0.7577,
+      "step": 5055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2721996307373047,
+      "learning_rate": 2.6991832278949968e-06,
+      "loss": 0.7851,
+      "step": 5056
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2799915075302124,
+      "learning_rate": 2.697299086086852e-06,
+      "loss": 0.81,
+      "step": 5057
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.254318118095398,
+      "learning_rate": 2.6954153591911537e-06,
+      "loss": 0.7904,
+      "step": 5058
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.65220046043396,
+      "learning_rate": 2.6935320475473205e-06,
+      "loss": 0.9984,
+      "step": 5059
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6538282036781311,
+      "learning_rate": 2.6916491514947006e-06,
+      "loss": 0.9964,
+      "step": 5060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.265486717224121,
+      "learning_rate": 2.6897666713725623e-06,
+      "loss": 0.8518,
+      "step": 5061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.188706636428833,
+      "learning_rate": 2.687884607520098e-06,
+      "loss": 0.8298,
+      "step": 5062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2228559255599976,
+      "learning_rate": 2.6860029602764304e-06,
+      "loss": 0.8148,
+      "step": 5063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2034919261932373,
+      "learning_rate": 2.684121729980606e-06,
+      "loss": 0.7692,
+      "step": 5064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.224584698677063,
+      "learning_rate": 2.6822409169715922e-06,
+      "loss": 0.8914,
+      "step": 5065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6167585849761963,
+      "learning_rate": 2.680360521588284e-06,
+      "loss": 1.0125,
+      "step": 5066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3151307106018066,
+      "learning_rate": 2.6784805441694985e-06,
+      "loss": 0.7777,
+      "step": 5067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.218518853187561,
+      "learning_rate": 2.676600985053983e-06,
+      "loss": 0.7836,
+      "step": 5068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2498304843902588,
+      "learning_rate": 2.6747218445804052e-06,
+      "loss": 0.7625,
+      "step": 5069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5936225056648254,
+      "learning_rate": 2.6728431230873552e-06,
+      "loss": 1.0244,
+      "step": 5070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6412615776062012,
+      "learning_rate": 2.6709648209133537e-06,
+      "loss": 1.0173,
+      "step": 5071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2648643255233765,
+      "learning_rate": 2.669086938396844e-06,
+      "loss": 0.7319,
+      "step": 5072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.301914095878601,
+      "learning_rate": 2.6672094758761914e-06,
+      "loss": 0.8297,
+      "step": 5073
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2165292501449585,
+      "learning_rate": 2.665332433689686e-06,
+      "loss": 0.7885,
+      "step": 5074
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2602818012237549,
+      "learning_rate": 2.6634558121755406e-06,
+      "loss": 0.8517,
+      "step": 5075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6336770057678223,
+      "learning_rate": 2.6615796116718995e-06,
+      "loss": 1.0192,
+      "step": 5076
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2786370515823364,
+      "learning_rate": 2.6597038325168236e-06,
+      "loss": 0.8215,
+      "step": 5077
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2271857261657715,
+      "learning_rate": 2.657828475048298e-06,
+      "loss": 0.7987,
+      "step": 5078
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6286030411720276,
+      "learning_rate": 2.6559535396042367e-06,
+      "loss": 1.0302,
+      "step": 5079
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.173926830291748,
+      "learning_rate": 2.654079026522477e-06,
+      "loss": 0.7812,
+      "step": 5080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2350965738296509,
+      "learning_rate": 2.6522049361407764e-06,
+      "loss": 0.8627,
+      "step": 5081
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2558432817459106,
+      "learning_rate": 2.6503312687968164e-06,
+      "loss": 0.8734,
+      "step": 5082
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5969980359077454,
+      "learning_rate": 2.6484580248282076e-06,
+      "loss": 1.0094,
+      "step": 5083
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.296706199645996,
+      "learning_rate": 2.646585204572478e-06,
+      "loss": 0.8122,
+      "step": 5084
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2214195728302002,
+      "learning_rate": 2.6447128083670835e-06,
+      "loss": 0.7825,
+      "step": 5085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2620517015457153,
+      "learning_rate": 2.6428408365493985e-06,
+      "loss": 0.8089,
+      "step": 5086
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2231886386871338,
+      "learning_rate": 2.640969289456728e-06,
+      "loss": 0.7944,
+      "step": 5087
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1926909685134888,
+      "learning_rate": 2.639098167426298e-06,
+      "loss": 0.7949,
+      "step": 5088
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1863107681274414,
+      "learning_rate": 2.6372274707952545e-06,
+      "loss": 0.7664,
+      "step": 5089
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2476038932800293,
+      "learning_rate": 2.6353571999006676e-06,
+      "loss": 0.8588,
+      "step": 5090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2481151819229126,
+      "learning_rate": 2.633487355079537e-06,
+      "loss": 0.8054,
+      "step": 5091
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6130709648132324,
+      "learning_rate": 2.6316179366687776e-06,
+      "loss": 1.0271,
+      "step": 5092
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3028993606567383,
+      "learning_rate": 2.6297489450052317e-06,
+      "loss": 0.837,
+      "step": 5093
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2740082740783691,
+      "learning_rate": 2.627880380425662e-06,
+      "loss": 0.799,
+      "step": 5094
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6206666231155396,
+      "learning_rate": 2.6260122432667573e-06,
+      "loss": 1.008,
+      "step": 5095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1911784410476685,
+      "learning_rate": 2.6241445338651307e-06,
+      "loss": 0.7972,
+      "step": 5096
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2342294454574585,
+      "learning_rate": 2.6222772525573125e-06,
+      "loss": 0.7429,
+      "step": 5097
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5847752690315247,
+      "learning_rate": 2.6204103996797593e-06,
+      "loss": 1.019,
+      "step": 5098
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2796255350112915,
+      "learning_rate": 2.618543975568851e-06,
+      "loss": 0.7561,
+      "step": 5099
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2982569932937622,
+      "learning_rate": 2.616677980560891e-06,
+      "loss": 0.8189,
+      "step": 5100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.365362286567688,
+      "learning_rate": 2.614812414992101e-06,
+      "loss": 0.8067,
+      "step": 5101
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.203333854675293,
+      "learning_rate": 2.6129472791986276e-06,
+      "loss": 0.8345,
+      "step": 5102
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.25666344165802,
+      "learning_rate": 2.6110825735165424e-06,
+      "loss": 0.8441,
+      "step": 5103
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1913999319076538,
+      "learning_rate": 2.6092182982818392e-06,
+      "loss": 0.7486,
+      "step": 5104
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2480331659317017,
+      "learning_rate": 2.607354453830431e-06,
+      "loss": 0.8155,
+      "step": 5105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2540247440338135,
+      "learning_rate": 2.6054910404981528e-06,
+      "loss": 0.7851,
+      "step": 5106
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.217093825340271,
+      "learning_rate": 2.603628058620768e-06,
+      "loss": 0.7953,
+      "step": 5107
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2791125774383545,
+      "learning_rate": 2.6017655085339556e-06,
+      "loss": 0.8035,
+      "step": 5108
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6137209534645081,
+      "learning_rate": 2.5999033905733207e-06,
+      "loss": 1.0152,
+      "step": 5109
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2374368906021118,
+      "learning_rate": 2.5980417050743855e-06,
+      "loss": 0.7532,
+      "step": 5110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2626975774765015,
+      "learning_rate": 2.5961804523726014e-06,
+      "loss": 0.8096,
+      "step": 5111
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6283187866210938,
+      "learning_rate": 2.5943196328033404e-06,
+      "loss": 1.0302,
+      "step": 5112
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.285610556602478,
+      "learning_rate": 2.5924592467018917e-06,
+      "loss": 0.7817,
+      "step": 5113
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5902180671691895,
+      "learning_rate": 2.5905992944034676e-06,
+      "loss": 1.0198,
+      "step": 5114
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5887099504470825,
+      "learning_rate": 2.5887397762432083e-06,
+      "loss": 0.9956,
+      "step": 5115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2008030414581299,
+      "learning_rate": 2.5868806925561686e-06,
+      "loss": 0.831,
+      "step": 5116
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2499370574951172,
+      "learning_rate": 2.585022043677326e-06,
+      "loss": 0.8094,
+      "step": 5117
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2340251207351685,
+      "learning_rate": 2.583163829941585e-06,
+      "loss": 0.7877,
+      "step": 5118
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3149056434631348,
+      "learning_rate": 2.581306051683765e-06,
+      "loss": 0.8197,
+      "step": 5119
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2892918586730957,
+      "learning_rate": 2.5794487092386132e-06,
+      "loss": 0.8502,
+      "step": 5120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6172120571136475,
+      "learning_rate": 2.577591802940792e-06,
+      "loss": 1.0248,
+      "step": 5121
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2208750247955322,
+      "learning_rate": 2.5757353331248915e-06,
+      "loss": 0.7918,
+      "step": 5122
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2397655248641968,
+      "learning_rate": 2.573879300125417e-06,
+      "loss": 0.8594,
+      "step": 5123
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2822425365447998,
+      "learning_rate": 2.5720237042768003e-06,
+      "loss": 0.8232,
+      "step": 5124
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2517573833465576,
+      "learning_rate": 2.5701685459133884e-06,
+      "loss": 0.8295,
+      "step": 5125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5930043458938599,
+      "learning_rate": 2.5683138253694577e-06,
+      "loss": 1.0362,
+      "step": 5126
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2762047052383423,
+      "learning_rate": 2.566459542979198e-06,
+      "loss": 0.8654,
+      "step": 5127
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3331499099731445,
+      "learning_rate": 2.564605699076726e-06,
+      "loss": 0.8245,
+      "step": 5128
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6131367683410645,
+      "learning_rate": 2.562752293996074e-06,
+      "loss": 1.0093,
+      "step": 5129
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2963566780090332,
+      "learning_rate": 2.560899328071201e-06,
+      "loss": 0.8114,
+      "step": 5130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2642669677734375,
+      "learning_rate": 2.559046801635983e-06,
+      "loss": 0.7798,
+      "step": 5131
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2705167531967163,
+      "learning_rate": 2.557194715024217e-06,
+      "loss": 0.8371,
+      "step": 5132
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.27767813205719,
+      "learning_rate": 2.55534306856962e-06,
+      "loss": 0.7959,
+      "step": 5133
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2548751831054688,
+      "learning_rate": 2.5534918626058345e-06,
+      "loss": 0.7547,
+      "step": 5134
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2950944900512695,
+      "learning_rate": 2.5516410974664163e-06,
+      "loss": 0.8222,
+      "step": 5135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2597211599349976,
+      "learning_rate": 2.5497907734848506e-06,
+      "loss": 0.8191,
+      "step": 5136
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2536407709121704,
+      "learning_rate": 2.5479408909945324e-06,
+      "loss": 0.8627,
+      "step": 5137
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.318956971168518,
+      "learning_rate": 2.546091450328789e-06,
+      "loss": 0.7564,
+      "step": 5138
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6461451649665833,
+      "learning_rate": 2.5442424518208597e-06,
+      "loss": 1.0237,
+      "step": 5139
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6530783772468567,
+      "learning_rate": 2.5423938958039062e-06,
+      "loss": 0.9989,
+      "step": 5140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5986214876174927,
+      "learning_rate": 2.540545782611008e-06,
+      "loss": 1.0308,
+      "step": 5141
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2382704019546509,
+      "learning_rate": 2.538698112575173e-06,
+      "loss": 0.827,
+      "step": 5142
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2498923540115356,
+      "learning_rate": 2.536850886029319e-06,
+      "loss": 0.7489,
+      "step": 5143
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6296568512916565,
+      "learning_rate": 2.5350041033062935e-06,
+      "loss": 1.0265,
+      "step": 5144
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6350117921829224,
+      "learning_rate": 2.533157764738854e-06,
+      "loss": 1.02,
+      "step": 5145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6278455853462219,
+      "learning_rate": 2.5313118706596873e-06,
+      "loss": 1.0142,
+      "step": 5146
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2772902250289917,
+      "learning_rate": 2.529466421401396e-06,
+      "loss": 0.8092,
+      "step": 5147
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2880183458328247,
+      "learning_rate": 2.5276214172965007e-06,
+      "loss": 0.7792,
+      "step": 5148
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2422409057617188,
+      "learning_rate": 2.5257768586774423e-06,
+      "loss": 0.8129,
+      "step": 5149
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2884397506713867,
+      "learning_rate": 2.5239327458765874e-06,
+      "loss": 0.8004,
+      "step": 5150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2121435403823853,
+      "learning_rate": 2.5220890792262127e-06,
+      "loss": 0.7817,
+      "step": 5151
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2408629655838013,
+      "learning_rate": 2.520245859058522e-06,
+      "loss": 0.8363,
+      "step": 5152
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2464337348937988,
+      "learning_rate": 2.518403085705638e-06,
+      "loss": 0.778,
+      "step": 5153
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.294818639755249,
+      "learning_rate": 2.516560759499599e-06,
+      "loss": 0.7861,
+      "step": 5154
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3617883920669556,
+      "learning_rate": 2.514718880772365e-06,
+      "loss": 0.8065,
+      "step": 5155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6539561152458191,
+      "learning_rate": 2.512877449855813e-06,
+      "loss": 1.0083,
+      "step": 5156
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2089649438858032,
+      "learning_rate": 2.5110364670817455e-06,
+      "loss": 0.7956,
+      "step": 5157
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2235604524612427,
+      "learning_rate": 2.5091959327818784e-06,
+      "loss": 0.7917,
+      "step": 5158
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6419789791107178,
+      "learning_rate": 2.5073558472878467e-06,
+      "loss": 1.0175,
+      "step": 5159
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.192136287689209,
+      "learning_rate": 2.505516210931207e-06,
+      "loss": 0.8123,
+      "step": 5160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2451220750808716,
+      "learning_rate": 2.5036770240434383e-06,
+      "loss": 0.8142,
+      "step": 5161
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2824711799621582,
+      "learning_rate": 2.5018382869559323e-06,
+      "loss": 0.7818,
+      "step": 5162
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2043105363845825,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.825,
+      "step": 5163
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6322386860847473,
+      "learning_rate": 2.498162163506876e-06,
+      "loss": 1.0075,
+      "step": 5164
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.295401692390442,
+      "learning_rate": 2.4963247778077104e-06,
+      "loss": 0.8505,
+      "step": 5165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2416390180587769,
+      "learning_rate": 2.4944878432335724e-06,
+      "loss": 0.813,
+      "step": 5166
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2318782806396484,
+      "learning_rate": 2.492651360115449e-06,
+      "loss": 0.8048,
+      "step": 5167
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2827858924865723,
+      "learning_rate": 2.4908153287842484e-06,
+      "loss": 0.8392,
+      "step": 5168
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2752079963684082,
+      "learning_rate": 2.488979749570798e-06,
+      "loss": 0.8101,
+      "step": 5169
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2289276123046875,
+      "learning_rate": 2.4871446228058406e-06,
+      "loss": 0.7786,
+      "step": 5170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2668284177780151,
+      "learning_rate": 2.485309948820038e-06,
+      "loss": 0.7714,
+      "step": 5171
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2652066946029663,
+      "learning_rate": 2.4834757279439693e-06,
+      "loss": 0.8779,
+      "step": 5172
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6147565841674805,
+      "learning_rate": 2.4816419605081383e-06,
+      "loss": 1.0319,
+      "step": 5173
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2154604196548462,
+      "learning_rate": 2.4798086468429598e-06,
+      "loss": 0.7597,
+      "step": 5174
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6072635054588318,
+      "learning_rate": 2.477975787278769e-06,
+      "loss": 1.025,
+      "step": 5175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.267674446105957,
+      "learning_rate": 2.4761433821458205e-06,
+      "loss": 0.8316,
+      "step": 5176
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2458171844482422,
+      "learning_rate": 2.4743114317742887e-06,
+      "loss": 0.8716,
+      "step": 5177
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2137861251831055,
+      "learning_rate": 2.472479936494263e-06,
+      "loss": 0.764,
+      "step": 5178
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2223153114318848,
+      "learning_rate": 2.47064889663575e-06,
+      "loss": 0.7216,
+      "step": 5179
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2745015621185303,
+      "learning_rate": 2.4688183125286747e-06,
+      "loss": 0.7932,
+      "step": 5180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6050894260406494,
+      "learning_rate": 2.4669881845028853e-06,
+      "loss": 1.0036,
+      "step": 5181
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2832961082458496,
+      "learning_rate": 2.4651585128881407e-06,
+      "loss": 0.8251,
+      "step": 5182
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2413314580917358,
+      "learning_rate": 2.46332929801412e-06,
+      "loss": 0.7816,
+      "step": 5183
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6232224106788635,
+      "learning_rate": 2.4615005402104207e-06,
+      "loss": 1.0225,
+      "step": 5184
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6107534170150757,
+      "learning_rate": 2.459672239806561e-06,
+      "loss": 1.0105,
+      "step": 5185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2417447566986084,
+      "learning_rate": 2.4578443971319716e-06,
+      "loss": 0.7802,
+      "step": 5186
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.299094319343567,
+      "learning_rate": 2.456017012516001e-06,
+      "loss": 0.7958,
+      "step": 5187
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2462067604064941,
+      "learning_rate": 2.454190086287917e-06,
+      "loss": 0.7616,
+      "step": 5188
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.253203272819519,
+      "learning_rate": 2.452363618776906e-06,
+      "loss": 0.7404,
+      "step": 5189
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.308508038520813,
+      "learning_rate": 2.450537610312071e-06,
+      "loss": 0.8554,
+      "step": 5190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2306711673736572,
+      "learning_rate": 2.4487120612224274e-06,
+      "loss": 0.8002,
+      "step": 5191
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2479572296142578,
+      "learning_rate": 2.446886971836915e-06,
+      "loss": 0.7934,
+      "step": 5192
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6545622944831848,
+      "learning_rate": 2.445062342484389e-06,
+      "loss": 1.0005,
+      "step": 5193
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2442717552185059,
+      "learning_rate": 2.4432381734936195e-06,
+      "loss": 0.7839,
+      "step": 5194
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3046115636825562,
+      "learning_rate": 2.441414465193292e-06,
+      "loss": 0.8171,
+      "step": 5195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2069765329360962,
+      "learning_rate": 2.4395912179120153e-06,
+      "loss": 0.8452,
+      "step": 5196
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2935328483581543,
+      "learning_rate": 2.43776843197831e-06,
+      "loss": 0.8662,
+      "step": 5197
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.270753026008606,
+      "learning_rate": 2.435946107720615e-06,
+      "loss": 0.8335,
+      "step": 5198
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.280914068222046,
+      "learning_rate": 2.434124245467284e-06,
+      "loss": 0.8669,
+      "step": 5199
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.22576105594635,
+      "learning_rate": 2.4323028455465904e-06,
+      "loss": 0.7946,
+      "step": 5200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6057894229888916,
+      "learning_rate": 2.430481908286727e-06,
+      "loss": 1.0405,
+      "step": 5201
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5947681069374084,
+      "learning_rate": 2.4286614340157965e-06,
+      "loss": 0.9887,
+      "step": 5202
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2893645763397217,
+      "learning_rate": 2.4268414230618205e-06,
+      "loss": 0.7926,
+      "step": 5203
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2931722402572632,
+      "learning_rate": 2.42502187575274e-06,
+      "loss": 0.8382,
+      "step": 5204
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6031596064567566,
+      "learning_rate": 2.4232027924164098e-06,
+      "loss": 1.0288,
+      "step": 5205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3374829292297363,
+      "learning_rate": 2.4213841733806014e-06,
+      "loss": 0.8312,
+      "step": 5206
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2197576761245728,
+      "learning_rate": 2.419566018973001e-06,
+      "loss": 0.8192,
+      "step": 5207
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2249583005905151,
+      "learning_rate": 2.4177483295212146e-06,
+      "loss": 0.8508,
+      "step": 5208
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2007569074630737,
+      "learning_rate": 2.4159311053527643e-06,
+      "loss": 0.7819,
+      "step": 5209
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2456834316253662,
+      "learning_rate": 2.414114346795086e-06,
+      "loss": 0.7905,
+      "step": 5210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2051211595535278,
+      "learning_rate": 2.4122980541755303e-06,
+      "loss": 0.8258,
+      "step": 5211
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2759770154953003,
+      "learning_rate": 2.4104822278213696e-06,
+      "loss": 0.8117,
+      "step": 5212
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2643916606903076,
+      "learning_rate": 2.4086668680597874e-06,
+      "loss": 0.8235,
+      "step": 5213
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2866193056106567,
+      "learning_rate": 2.4068519752178837e-06,
+      "loss": 0.7847,
+      "step": 5214
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2719993591308594,
+      "learning_rate": 2.4050375496226742e-06,
+      "loss": 0.8319,
+      "step": 5215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.225150465965271,
+      "learning_rate": 2.4032235916010925e-06,
+      "loss": 0.8104,
+      "step": 5216
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2227474451065063,
+      "learning_rate": 2.401410101479989e-06,
+      "loss": 0.8068,
+      "step": 5217
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2031643390655518,
+      "learning_rate": 2.399597079586126e-06,
+      "loss": 0.78,
+      "step": 5218
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6171124577522278,
+      "learning_rate": 2.3977845262461814e-06,
+      "loss": 0.9963,
+      "step": 5219
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3646385669708252,
+      "learning_rate": 2.3959724417867537e-06,
+      "loss": 0.8614,
+      "step": 5220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6187945604324341,
+      "learning_rate": 2.3941608265343514e-06,
+      "loss": 1.0081,
+      "step": 5221
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6370612978935242,
+      "learning_rate": 2.392349680815401e-06,
+      "loss": 1.0135,
+      "step": 5222
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6083166599273682,
+      "learning_rate": 2.390539004956243e-06,
+      "loss": 1.0378,
+      "step": 5223
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5905686616897583,
+      "learning_rate": 2.3887287992831344e-06,
+      "loss": 1.006,
+      "step": 5224
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2469468116760254,
+      "learning_rate": 2.3869190641222502e-06,
+      "loss": 0.8533,
+      "step": 5225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.358419418334961,
+      "learning_rate": 2.3851097997996735e-06,
+      "loss": 0.7901,
+      "step": 5226
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2534338235855103,
+      "learning_rate": 2.3833010066414108e-06,
+      "loss": 0.7999,
+      "step": 5227
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2691493034362793,
+      "learning_rate": 2.3814926849733786e-06,
+      "loss": 0.7844,
+      "step": 5228
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.279222011566162,
+      "learning_rate": 2.3796848351214086e-06,
+      "loss": 0.8363,
+      "step": 5229
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.337537169456482,
+      "learning_rate": 2.3778774574112468e-06,
+      "loss": 0.8374,
+      "step": 5230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2529988288879395,
+      "learning_rate": 2.37607055216856e-06,
+      "loss": 0.837,
+      "step": 5231
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6837344169616699,
+      "learning_rate": 2.3742641197189216e-06,
+      "loss": 1.0141,
+      "step": 5232
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3133090734481812,
+      "learning_rate": 2.3724581603878267e-06,
+      "loss": 0.8105,
+      "step": 5233
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2132216691970825,
+      "learning_rate": 2.3706526745006796e-06,
+      "loss": 0.8015,
+      "step": 5234
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2302308082580566,
+      "learning_rate": 2.368847662382806e-06,
+      "loss": 0.8175,
+      "step": 5235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2261691093444824,
+      "learning_rate": 2.3670431243594394e-06,
+      "loss": 0.86,
+      "step": 5236
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.619255781173706,
+      "learning_rate": 2.3652390607557313e-06,
+      "loss": 1.0291,
+      "step": 5237
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.259902000427246,
+      "learning_rate": 2.3634354718967454e-06,
+      "loss": 0.7678,
+      "step": 5238
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6008098125457764,
+      "learning_rate": 2.3616323581074646e-06,
+      "loss": 1.0331,
+      "step": 5239
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1971724033355713,
+      "learning_rate": 2.35982971971278e-06,
+      "loss": 0.7813,
+      "step": 5240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.19990074634552,
+      "learning_rate": 2.3580275570375045e-06,
+      "loss": 0.8305,
+      "step": 5241
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2250392436981201,
+      "learning_rate": 2.3562258704063556e-06,
+      "loss": 0.7956,
+      "step": 5242
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5987153649330139,
+      "learning_rate": 2.3544246601439756e-06,
+      "loss": 1.0121,
+      "step": 5243
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2092878818511963,
+      "learning_rate": 2.3526239265749133e-06,
+      "loss": 0.8176,
+      "step": 5244
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6367577910423279,
+      "learning_rate": 2.350823670023635e-06,
+      "loss": 1.0068,
+      "step": 5245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2581896781921387,
+      "learning_rate": 2.349023890814517e-06,
+      "loss": 0.8002,
+      "step": 5246
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2035744190216064,
+      "learning_rate": 2.3472245892718572e-06,
+      "loss": 0.7662,
+      "step": 5247
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.435524344444275,
+      "learning_rate": 2.3454257657198588e-06,
+      "loss": 0.841,
+      "step": 5248
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2413402795791626,
+      "learning_rate": 2.343627420482647e-06,
+      "loss": 0.7903,
+      "step": 5249
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.232972264289856,
+      "learning_rate": 2.3418295538842534e-06,
+      "loss": 0.8061,
+      "step": 5250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2010226249694824,
+      "learning_rate": 2.3400321662486305e-06,
+      "loss": 0.7354,
+      "step": 5251
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.250683307647705,
+      "learning_rate": 2.3382352578996386e-06,
+      "loss": 0.8042,
+      "step": 5252
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6526536345481873,
+      "learning_rate": 2.336438829161054e-06,
+      "loss": 1.0156,
+      "step": 5253
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3037205934524536,
+      "learning_rate": 2.3346428803565646e-06,
+      "loss": 0.8444,
+      "step": 5254
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2529951333999634,
+      "learning_rate": 2.332847411809777e-06,
+      "loss": 0.8288,
+      "step": 5255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.229341983795166,
+      "learning_rate": 2.3310524238442055e-06,
+      "loss": 0.7836,
+      "step": 5256
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2577303647994995,
+      "learning_rate": 2.3292579167832824e-06,
+      "loss": 0.8403,
+      "step": 5257
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.265604019165039,
+      "learning_rate": 2.3274638909503483e-06,
+      "loss": 0.7994,
+      "step": 5258
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.327687382698059,
+      "learning_rate": 2.3256703466686633e-06,
+      "loss": 0.8094,
+      "step": 5259
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5890331268310547,
+      "learning_rate": 2.3238772842613953e-06,
+      "loss": 1.0204,
+      "step": 5260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.223931074142456,
+      "learning_rate": 2.322084704051627e-06,
+      "loss": 0.8094,
+      "step": 5261
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1641539335250854,
+      "learning_rate": 2.3202926063623566e-06,
+      "loss": 0.7515,
+      "step": 5262
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6057655215263367,
+      "learning_rate": 2.318500991516493e-06,
+      "loss": 1.0228,
+      "step": 5263
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2475067377090454,
+      "learning_rate": 2.3167098598368555e-06,
+      "loss": 0.8051,
+      "step": 5264
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2214417457580566,
+      "learning_rate": 2.3149192116461836e-06,
+      "loss": 0.8287,
+      "step": 5265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.603440523147583,
+      "learning_rate": 2.3131290472671224e-06,
+      "loss": 1.0383,
+      "step": 5266
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2570767402648926,
+      "learning_rate": 2.3113393670222363e-06,
+      "loss": 0.8348,
+      "step": 5267
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5940830111503601,
+      "learning_rate": 2.3095501712339963e-06,
+      "loss": 0.9991,
+      "step": 5268
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.249849796295166,
+      "learning_rate": 2.307761460224788e-06,
+      "loss": 0.7559,
+      "step": 5269
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2098913192749023,
+      "learning_rate": 2.3059732343169133e-06,
+      "loss": 0.7688,
+      "step": 5270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6076472401618958,
+      "learning_rate": 2.304185493832583e-06,
+      "loss": 1.0102,
+      "step": 5271
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2450305223464966,
+      "learning_rate": 2.3023982390939193e-06,
+      "loss": 0.8186,
+      "step": 5272
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2683863639831543,
+      "learning_rate": 2.300611470422962e-06,
+      "loss": 0.8368,
+      "step": 5273
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3065820932388306,
+      "learning_rate": 2.2988251881416567e-06,
+      "loss": 0.8036,
+      "step": 5274
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2078129053115845,
+      "learning_rate": 2.29703939257187e-06,
+      "loss": 0.768,
+      "step": 5275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2443233728408813,
+      "learning_rate": 2.2952540840353714e-06,
+      "loss": 0.8289,
+      "step": 5276
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2009332180023193,
+      "learning_rate": 2.2934692628538467e-06,
+      "loss": 0.7652,
+      "step": 5277
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6194974184036255,
+      "learning_rate": 2.291684929348898e-06,
+      "loss": 1.0237,
+      "step": 5278
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6152050495147705,
+      "learning_rate": 2.2899010838420334e-06,
+      "loss": 1.0243,
+      "step": 5279
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2570782899856567,
+      "learning_rate": 2.288117726654673e-06,
+      "loss": 0.8079,
+      "step": 5280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2423491477966309,
+      "learning_rate": 2.286334858108155e-06,
+      "loss": 0.8235,
+      "step": 5281
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2400850057601929,
+      "learning_rate": 2.284552478523723e-06,
+      "loss": 0.8002,
+      "step": 5282
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5946785807609558,
+      "learning_rate": 2.2827705882225376e-06,
+      "loss": 1.0003,
+      "step": 5283
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3262475728988647,
+      "learning_rate": 2.2809891875256683e-06,
+      "loss": 0.8223,
+      "step": 5284
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.591230034828186,
+      "learning_rate": 2.2792082767540953e-06,
+      "loss": 0.9904,
+      "step": 5285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1982260942459106,
+      "learning_rate": 2.2774278562287144e-06,
+      "loss": 0.7665,
+      "step": 5286
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1627601385116577,
+      "learning_rate": 2.27564792627033e-06,
+      "loss": 0.8085,
+      "step": 5287
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5890676975250244,
+      "learning_rate": 2.2738684871996597e-06,
+      "loss": 1.0165,
+      "step": 5288
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.619069516658783,
+      "learning_rate": 2.272089539337329e-06,
+      "loss": 0.988,
+      "step": 5289
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2484135627746582,
+      "learning_rate": 2.27031108300388e-06,
+      "loss": 0.7938,
+      "step": 5290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2474229335784912,
+      "learning_rate": 2.2685331185197657e-06,
+      "loss": 0.8085,
+      "step": 5291
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3352739810943604,
+      "learning_rate": 2.266755646205348e-06,
+      "loss": 0.8886,
+      "step": 5292
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2321486473083496,
+      "learning_rate": 2.2649786663808974e-06,
+      "loss": 0.7587,
+      "step": 5293
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2165108919143677,
+      "learning_rate": 2.263202179366605e-06,
+      "loss": 0.7857,
+      "step": 5294
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2230039834976196,
+      "learning_rate": 2.2614261854825636e-06,
+      "loss": 0.836,
+      "step": 5295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2676156759262085,
+      "learning_rate": 2.259650685048782e-06,
+      "loss": 0.8394,
+      "step": 5296
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2861998081207275,
+      "learning_rate": 2.257875678385177e-06,
+      "loss": 0.8026,
+      "step": 5297
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6222677230834961,
+      "learning_rate": 2.25610116581158e-06,
+      "loss": 1.0303,
+      "step": 5298
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2739157676696777,
+      "learning_rate": 2.2543271476477334e-06,
+      "loss": 0.8007,
+      "step": 5299
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6484249234199524,
+      "learning_rate": 2.2525536242132856e-06,
+      "loss": 1.0155,
+      "step": 5300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2911629676818848,
+      "learning_rate": 2.250780595827803e-06,
+      "loss": 0.7989,
+      "step": 5301
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.22185480594635,
+      "learning_rate": 2.2490080628107556e-06,
+      "loss": 0.8563,
+      "step": 5302
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2689003944396973,
+      "learning_rate": 2.24723602548153e-06,
+      "loss": 0.8071,
+      "step": 5303
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2630218267440796,
+      "learning_rate": 2.245464484159417e-06,
+      "loss": 0.691,
+      "step": 5304
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2468355894088745,
+      "learning_rate": 2.2436934391636267e-06,
+      "loss": 0.7816,
+      "step": 5305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5897181034088135,
+      "learning_rate": 2.2419228908132714e-06,
+      "loss": 0.9982,
+      "step": 5306
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1891045570373535,
+      "learning_rate": 2.24015283942738e-06,
+      "loss": 0.7869,
+      "step": 5307
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.290454626083374,
+      "learning_rate": 2.238383285324888e-06,
+      "loss": 0.8262,
+      "step": 5308
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2624238729476929,
+      "learning_rate": 2.236614228824644e-06,
+      "loss": 0.8147,
+      "step": 5309
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6010916233062744,
+      "learning_rate": 2.2348456702454054e-06,
+      "loss": 1.0201,
+      "step": 5310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.202627420425415,
+      "learning_rate": 2.2330776099058397e-06,
+      "loss": 0.79,
+      "step": 5311
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4841101169586182,
+      "learning_rate": 2.2313100481245233e-06,
+      "loss": 0.803,
+      "step": 5312
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.271602749824524,
+      "learning_rate": 2.2295429852199477e-06,
+      "loss": 0.8153,
+      "step": 5313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2340798377990723,
+      "learning_rate": 2.2277764215105084e-06,
+      "loss": 0.7774,
+      "step": 5314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.272207498550415,
+      "learning_rate": 2.2260103573145175e-06,
+      "loss": 0.7972,
+      "step": 5315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4025777578353882,
+      "learning_rate": 2.22424479295019e-06,
+      "loss": 0.8564,
+      "step": 5316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6070831418037415,
+      "learning_rate": 2.222479728735657e-06,
+      "loss": 1.0301,
+      "step": 5317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.264503002166748,
+      "learning_rate": 2.220715164988955e-06,
+      "loss": 0.8328,
+      "step": 5318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2881594896316528,
+      "learning_rate": 2.2189511020280335e-06,
+      "loss": 0.88,
+      "step": 5319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6039979457855225,
+      "learning_rate": 2.2171875401707477e-06,
+      "loss": 0.7786,
+      "step": 5320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2279338836669922,
+      "learning_rate": 2.215424479734869e-06,
+      "loss": 0.7943,
+      "step": 5321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5837095975875854,
+      "learning_rate": 2.213661921038071e-06,
+      "loss": 1.0134,
+      "step": 5322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2870980501174927,
+      "learning_rate": 2.211899864397944e-06,
+      "loss": 0.8176,
+      "step": 5323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2456257343292236,
+      "learning_rate": 2.2101383101319805e-06,
+      "loss": 0.7934,
+      "step": 5324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3249047994613647,
+      "learning_rate": 2.2083772585575914e-06,
+      "loss": 0.8064,
+      "step": 5325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.264946699142456,
+      "learning_rate": 2.2066167099920883e-06,
+      "loss": 0.8062,
+      "step": 5326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2466999292373657,
+      "learning_rate": 2.2048566647526974e-06,
+      "loss": 0.7934,
+      "step": 5327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2191399335861206,
+      "learning_rate": 2.2030971231565503e-06,
+      "loss": 0.8246,
+      "step": 5328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6197518706321716,
+      "learning_rate": 2.2013380855206927e-06,
+      "loss": 1.0157,
+      "step": 5329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1903513669967651,
+      "learning_rate": 2.1995795521620755e-06,
+      "loss": 0.7936,
+      "step": 5330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2411096096038818,
+      "learning_rate": 2.1978215233975623e-06,
+      "loss": 0.8427,
+      "step": 5331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6005315780639648,
+      "learning_rate": 2.196063999543921e-06,
+      "loss": 1.0067,
+      "step": 5332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2144885063171387,
+      "learning_rate": 2.194306980917834e-06,
+      "loss": 0.7618,
+      "step": 5333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5956076383590698,
+      "learning_rate": 2.1925504678358894e-06,
+      "loss": 1.022,
+      "step": 5334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2579706907272339,
+      "learning_rate": 2.1907944606145813e-06,
+      "loss": 0.835,
+      "step": 5335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2478119134902954,
+      "learning_rate": 2.189038959570321e-06,
+      "loss": 0.8197,
+      "step": 5336
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6007034778594971,
+      "learning_rate": 2.1872839650194216e-06,
+      "loss": 1.0324,
+      "step": 5337
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6070613265037537,
+      "learning_rate": 2.1855294772781044e-06,
+      "loss": 1.0196,
+      "step": 5338
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5832043290138245,
+      "learning_rate": 2.1837754966625045e-06,
+      "loss": 1.0139,
+      "step": 5339
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2092615365982056,
+      "learning_rate": 2.182022023488665e-06,
+      "loss": 0.7376,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.248815894126892,
+      "learning_rate": 2.180269058072534e-06,
+      "loss": 0.7552,
+      "step": 5341
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2408385276794434,
+      "learning_rate": 2.1785166007299692e-06,
+      "loss": 0.7722,
+      "step": 5342
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6161674857139587,
+      "learning_rate": 2.1767646517767354e-06,
+      "loss": 1.0308,
+      "step": 5343
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6072075963020325,
+      "learning_rate": 2.1750132115285122e-06,
+      "loss": 1.0048,
+      "step": 5344
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5968378186225891,
+      "learning_rate": 2.17326228030088e-06,
+      "loss": 1.0321,
+      "step": 5345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1498841047286987,
+      "learning_rate": 2.1715118584093307e-06,
+      "loss": 0.7656,
+      "step": 5346
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1603549718856812,
+      "learning_rate": 2.1697619461692645e-06,
+      "loss": 0.7151,
+      "step": 5347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.274672031402588,
+      "learning_rate": 2.1680125438959916e-06,
+      "loss": 0.7835,
+      "step": 5348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3102610111236572,
+      "learning_rate": 2.1662636519047268e-06,
+      "loss": 0.8121,
+      "step": 5349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2791893482208252,
+      "learning_rate": 2.1645152705105947e-06,
+      "loss": 0.7888,
+      "step": 5350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2952027320861816,
+      "learning_rate": 2.162767400028625e-06,
+      "loss": 0.8514,
+      "step": 5351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6318916082382202,
+      "learning_rate": 2.161020040773762e-06,
+      "loss": 1.0352,
+      "step": 5352
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.229367971420288,
+      "learning_rate": 2.1592731930608524e-06,
+      "loss": 0.7779,
+      "step": 5353
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2192338705062866,
+      "learning_rate": 2.157526857204649e-06,
+      "loss": 0.8627,
+      "step": 5354
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2433949708938599,
+      "learning_rate": 2.1557810335198193e-06,
+      "loss": 0.854,
+      "step": 5355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2528512477874756,
+      "learning_rate": 2.154035722320935e-06,
+      "loss": 0.8604,
+      "step": 5356
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3035489320755005,
+      "learning_rate": 2.1522909239224747e-06,
+      "loss": 0.877,
+      "step": 5357
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2693032026290894,
+      "learning_rate": 2.150546638638824e-06,
+      "loss": 0.7973,
+      "step": 5358
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2484614849090576,
+      "learning_rate": 2.148802866784276e-06,
+      "loss": 0.7803,
+      "step": 5359
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6149657368659973,
+      "learning_rate": 2.1470596086730348e-06,
+      "loss": 1.0141,
+      "step": 5360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1962252855300903,
+      "learning_rate": 2.14531686461921e-06,
+      "loss": 0.7418,
+      "step": 5361
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2264584302902222,
+      "learning_rate": 2.143574634936815e-06,
+      "loss": 0.8769,
+      "step": 5362
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2211507558822632,
+      "learning_rate": 2.1418329199397746e-06,
+      "loss": 0.8071,
+      "step": 5363
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2009882926940918,
+      "learning_rate": 2.1400917199419237e-06,
+      "loss": 0.7644,
+      "step": 5364
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.262411117553711,
+      "learning_rate": 2.1383510352569975e-06,
+      "loss": 0.8432,
+      "step": 5365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2777267694473267,
+      "learning_rate": 2.136610866198642e-06,
+      "loss": 0.7752,
+      "step": 5366
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2774372100830078,
+      "learning_rate": 2.1348712130804077e-06,
+      "loss": 0.7923,
+      "step": 5367
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6057081818580627,
+      "learning_rate": 2.133132076215758e-06,
+      "loss": 0.9907,
+      "step": 5368
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.25336754322052,
+      "learning_rate": 2.1313934559180576e-06,
+      "loss": 0.7953,
+      "step": 5369
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2693496942520142,
+      "learning_rate": 2.1296553525005786e-06,
+      "loss": 0.8428,
+      "step": 5370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2775007486343384,
+      "learning_rate": 2.1279177662765027e-06,
+      "loss": 0.7613,
+      "step": 5371
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.358510971069336,
+      "learning_rate": 2.1261806975589188e-06,
+      "loss": 0.8043,
+      "step": 5372
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6543329358100891,
+      "learning_rate": 2.124444146660819e-06,
+      "loss": 1.023,
+      "step": 5373
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.259966492652893,
+      "learning_rate": 2.1227081138951026e-06,
+      "loss": 0.7388,
+      "step": 5374
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2618064880371094,
+      "learning_rate": 2.1209725995745805e-06,
+      "loss": 0.8212,
+      "step": 5375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5929039716720581,
+      "learning_rate": 2.1192376040119643e-06,
+      "loss": 1.0172,
+      "step": 5376
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.584104061126709,
+      "learning_rate": 2.117503127519875e-06,
+      "loss": 1.0022,
+      "step": 5377
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3353426456451416,
+      "learning_rate": 2.1157691704108367e-06,
+      "loss": 0.825,
+      "step": 5378
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5948266386985779,
+      "learning_rate": 2.1140357329972856e-06,
+      "loss": 1.0261,
+      "step": 5379
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3339149951934814,
+      "learning_rate": 2.1123028155915625e-06,
+      "loss": 0.8101,
+      "step": 5380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1930307149887085,
+      "learning_rate": 2.110570418505912e-06,
+      "loss": 0.7654,
+      "step": 5381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5920686721801758,
+      "learning_rate": 2.1088385420524833e-06,
+      "loss": 1.009,
+      "step": 5382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2500234842300415,
+      "learning_rate": 2.107107186543339e-06,
+      "loss": 0.7578,
+      "step": 5383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.268033742904663,
+      "learning_rate": 2.1053763522904425e-06,
+      "loss": 0.8339,
+      "step": 5384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2951830625534058,
+      "learning_rate": 2.103646039605664e-06,
+      "loss": 0.8162,
+      "step": 5385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.279148817062378,
+      "learning_rate": 2.1019162488007775e-06,
+      "loss": 0.7899,
+      "step": 5386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2270935773849487,
+      "learning_rate": 2.1001869801874687e-06,
+      "loss": 0.8001,
+      "step": 5387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2873059511184692,
+      "learning_rate": 2.0984582340773276e-06,
+      "loss": 0.8508,
+      "step": 5388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2432241439819336,
+      "learning_rate": 2.0967300107818457e-06,
+      "loss": 0.8345,
+      "step": 5389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.268900752067566,
+      "learning_rate": 2.0950023106124217e-06,
+      "loss": 0.832,
+      "step": 5390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2631871700286865,
+      "learning_rate": 2.0932751338803656e-06,
+      "loss": 0.8082,
+      "step": 5391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.223426103591919,
+      "learning_rate": 2.091548480896887e-06,
+      "loss": 0.8024,
+      "step": 5392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3099454641342163,
+      "learning_rate": 2.089822351973103e-06,
+      "loss": 0.8399,
+      "step": 5393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3066784143447876,
+      "learning_rate": 2.088096747420034e-06,
+      "loss": 0.7973,
+      "step": 5394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.611721932888031,
+      "learning_rate": 2.0863716675486112e-06,
+      "loss": 1.0193,
+      "step": 5395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2274911403656006,
+      "learning_rate": 2.0846471126696695e-06,
+      "loss": 0.8303,
+      "step": 5396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.289771318435669,
+      "learning_rate": 2.082923083093946e-06,
+      "loss": 0.787,
+      "step": 5397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.337041974067688,
+      "learning_rate": 2.0811995791320837e-06,
+      "loss": 0.8119,
+      "step": 5398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.217295527458191,
+      "learning_rate": 2.079476601094636e-06,
+      "loss": 0.7342,
+      "step": 5399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.264874815940857,
+      "learning_rate": 2.077754149292056e-06,
+      "loss": 0.8185,
+      "step": 5400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6099874377250671,
+      "learning_rate": 2.0760322240347043e-06,
+      "loss": 1.0238,
+      "step": 5401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3748376369476318,
+      "learning_rate": 2.0743108256328437e-06,
+      "loss": 0.8321,
+      "step": 5402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1945538520812988,
+      "learning_rate": 2.0725899543966465e-06,
+      "loss": 0.7957,
+      "step": 5403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1840752363204956,
+      "learning_rate": 2.0708696106361904e-06,
+      "loss": 0.8454,
+      "step": 5404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.198140025138855,
+      "learning_rate": 2.069149794661453e-06,
+      "loss": 0.839,
+      "step": 5405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2268192768096924,
+      "learning_rate": 2.067430506782319e-06,
+      "loss": 0.797,
+      "step": 5406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6158957481384277,
+      "learning_rate": 2.0657117473085812e-06,
+      "loss": 1.0333,
+      "step": 5407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.306520700454712,
+      "learning_rate": 2.0639935165499334e-06,
+      "loss": 0.8147,
+      "step": 5408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1938425302505493,
+      "learning_rate": 2.0622758148159725e-06,
+      "loss": 0.757,
+      "step": 5409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3180427551269531,
+      "learning_rate": 2.0605586424162073e-06,
+      "loss": 0.8097,
+      "step": 5410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5939632654190063,
+      "learning_rate": 2.058841999660043e-06,
+      "loss": 1.0239,
+      "step": 5411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3683092594146729,
+      "learning_rate": 2.0571258868567957e-06,
+      "loss": 0.8339,
+      "step": 5412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3421419858932495,
+      "learning_rate": 2.055410304315681e-06,
+      "loss": 0.8468,
+      "step": 5413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3616747856140137,
+      "learning_rate": 2.053695252345824e-06,
+      "loss": 0.8926,
+      "step": 5414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2677801847457886,
+      "learning_rate": 2.0519807312562504e-06,
+      "loss": 0.776,
+      "step": 5415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2840439081192017,
+      "learning_rate": 2.050266741355892e-06,
+      "loss": 0.8342,
+      "step": 5416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.273403525352478,
+      "learning_rate": 2.048553282953581e-06,
+      "loss": 0.7855,
+      "step": 5417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2942075729370117,
+      "learning_rate": 2.0468403563580615e-06,
+      "loss": 0.7944,
+      "step": 5418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6159888505935669,
+      "learning_rate": 2.0451279618779736e-06,
+      "loss": 1.0142,
+      "step": 5419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2287054061889648,
+      "learning_rate": 2.0434160998218687e-06,
+      "loss": 0.7908,
+      "step": 5420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2716379165649414,
+      "learning_rate": 2.0417047704981956e-06,
+      "loss": 0.7851,
+      "step": 5421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2842342853546143,
+      "learning_rate": 2.039993974215315e-06,
+      "loss": 0.8164,
+      "step": 5422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3026586771011353,
+      "learning_rate": 2.038283711281483e-06,
+      "loss": 0.8018,
+      "step": 5423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3517533540725708,
+      "learning_rate": 2.0365739820048645e-06,
+      "loss": 0.8313,
+      "step": 5424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2895019054412842,
+      "learning_rate": 2.034864786693526e-06,
+      "loss": 0.8569,
+      "step": 5425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6003949642181396,
+      "learning_rate": 2.033156125655443e-06,
+      "loss": 0.9961,
+      "step": 5426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2706832885742188,
+      "learning_rate": 2.0314479991984855e-06,
+      "loss": 0.8201,
+      "step": 5427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1811349391937256,
+      "learning_rate": 2.0297404076304373e-06,
+      "loss": 0.7786,
+      "step": 5428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.250373363494873,
+      "learning_rate": 2.0280333512589772e-06,
+      "loss": 0.7924,
+      "step": 5429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.300297737121582,
+      "learning_rate": 2.026326830391695e-06,
+      "loss": 0.8062,
+      "step": 5430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5928135514259338,
+      "learning_rate": 2.0246208453360783e-06,
+      "loss": 0.9897,
+      "step": 5431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2347396612167358,
+      "learning_rate": 2.0229153963995205e-06,
+      "loss": 0.7353,
+      "step": 5432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2192455530166626,
+      "learning_rate": 2.0212104838893167e-06,
+      "loss": 0.7903,
+      "step": 5433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.283850073814392,
+      "learning_rate": 2.0195061081126694e-06,
+      "loss": 0.803,
+      "step": 5434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5924354195594788,
+      "learning_rate": 2.0178022693766786e-06,
+      "loss": 1.0182,
+      "step": 5435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2337201833724976,
+      "learning_rate": 2.0160989679883547e-06,
+      "loss": 0.8176,
+      "step": 5436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5869206786155701,
+      "learning_rate": 2.0143962042546028e-06,
+      "loss": 1.0383,
+      "step": 5437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2583727836608887,
+      "learning_rate": 2.01269397848224e-06,
+      "loss": 0.7704,
+      "step": 5438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6113333702087402,
+      "learning_rate": 2.0109922909779805e-06,
+      "loss": 1.0001,
+      "step": 5439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2345514297485352,
+      "learning_rate": 2.0092911420484425e-06,
+      "loss": 0.8313,
+      "step": 5440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4441863298416138,
+      "learning_rate": 2.0075905320001453e-06,
+      "loss": 0.7723,
+      "step": 5441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6068952083587646,
+      "learning_rate": 2.005890461139518e-06,
+      "loss": 0.9983,
+      "step": 5442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.273959994316101,
+      "learning_rate": 2.0041909297728844e-06,
+      "loss": 0.8186,
+      "step": 5443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2856073379516602,
+      "learning_rate": 2.0024919382064766e-06,
+      "loss": 0.8277,
+      "step": 5444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2140928506851196,
+      "learning_rate": 2.000793486746429e-06,
+      "loss": 0.863,
+      "step": 5445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2626948356628418,
+      "learning_rate": 1.999095575698776e-06,
+      "loss": 0.8156,
+      "step": 5446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2427600622177124,
+      "learning_rate": 1.9973982053694557e-06,
+      "loss": 0.8616,
+      "step": 5447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2641531229019165,
+      "learning_rate": 1.995701376064307e-06,
+      "loss": 0.7274,
+      "step": 5448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2454781532287598,
+      "learning_rate": 1.994005088089077e-06,
+      "loss": 0.8148,
+      "step": 5449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2110737562179565,
+      "learning_rate": 1.9923093417494097e-06,
+      "loss": 0.8255,
+      "step": 5450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6171950697898865,
+      "learning_rate": 1.990614137350852e-06,
+      "loss": 0.9983,
+      "step": 5451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3140320777893066,
+      "learning_rate": 1.9889194751988555e-06,
+      "loss": 0.8445,
+      "step": 5452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6023060083389282,
+      "learning_rate": 1.987225355598776e-06,
+      "loss": 1.0001,
+      "step": 5453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1611332893371582,
+      "learning_rate": 1.985531778855866e-06,
+      "loss": 0.7659,
+      "step": 5454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2114735841751099,
+      "learning_rate": 1.9838387452752825e-06,
+      "loss": 0.8682,
+      "step": 5455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6025387048721313,
+      "learning_rate": 1.9821462551620846e-06,
+      "loss": 1.0103,
+      "step": 5456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2123513221740723,
+      "learning_rate": 1.980454308821236e-06,
+      "loss": 0.7829,
+      "step": 5457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3023602962493896,
+      "learning_rate": 1.978762906557599e-06,
+      "loss": 0.8199,
+      "step": 5458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5859225392341614,
+      "learning_rate": 1.9770720486759377e-06,
+      "loss": 1.0117,
+      "step": 5459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1650182008743286,
+      "learning_rate": 1.975381735480921e-06,
+      "loss": 0.8012,
+      "step": 5460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.19221830368042,
+      "learning_rate": 1.973691967277119e-06,
+      "loss": 0.8297,
+      "step": 5461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2920148372650146,
+      "learning_rate": 1.9720027443690033e-06,
+      "loss": 0.6939,
+      "step": 5462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.278448462486267,
+      "learning_rate": 1.9703140670609448e-06,
+      "loss": 0.8262,
+      "step": 5463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2014673948287964,
+      "learning_rate": 1.968625935657217e-06,
+      "loss": 0.7387,
+      "step": 5464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2848390340805054,
+      "learning_rate": 1.9669383504619995e-06,
+      "loss": 0.7873,
+      "step": 5465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2298824787139893,
+      "learning_rate": 1.9652513117793683e-06,
+      "loss": 0.7876,
+      "step": 5466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2798869609832764,
+      "learning_rate": 1.9635648199133007e-06,
+      "loss": 0.7548,
+      "step": 5467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3354114294052124,
+      "learning_rate": 1.9618788751676794e-06,
+      "loss": 0.8655,
+      "step": 5468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2713420391082764,
+      "learning_rate": 1.960193477846288e-06,
+      "loss": 0.7909,
+      "step": 5469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2757662534713745,
+      "learning_rate": 1.9585086282528088e-06,
+      "loss": 0.8557,
+      "step": 5470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2498960494995117,
+      "learning_rate": 1.956824326690826e-06,
+      "loss": 0.7975,
+      "step": 5471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6376008987426758,
+      "learning_rate": 1.9551405734638253e-06,
+      "loss": 1.0219,
+      "step": 5472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2558010816574097,
+      "learning_rate": 1.9534573688751955e-06,
+      "loss": 0.7826,
+      "step": 5473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.309160828590393,
+      "learning_rate": 1.9517747132282243e-06,
+      "loss": 0.86,
+      "step": 5474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1811208724975586,
+      "learning_rate": 1.9500926068261e-06,
+      "loss": 0.7229,
+      "step": 5475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.336499810218811,
+      "learning_rate": 1.9484110499719144e-06,
+      "loss": 0.8568,
+      "step": 5476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2365341186523438,
+      "learning_rate": 1.9467300429686604e-06,
+      "loss": 0.8004,
+      "step": 5477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2349038124084473,
+      "learning_rate": 1.9450495861192285e-06,
+      "loss": 0.8338,
+      "step": 5478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1860004663467407,
+      "learning_rate": 1.9433696797264118e-06,
+      "loss": 0.7521,
+      "step": 5479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2114505767822266,
+      "learning_rate": 1.9416903240929064e-06,
+      "loss": 0.7593,
+      "step": 5480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3047369718551636,
+      "learning_rate": 1.940011519521306e-06,
+      "loss": 0.8659,
+      "step": 5481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.211894154548645,
+      "learning_rate": 1.9383332663141064e-06,
+      "loss": 0.7697,
+      "step": 5482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1875066757202148,
+      "learning_rate": 1.9366555647737013e-06,
+      "loss": 0.8119,
+      "step": 5483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1740566492080688,
+      "learning_rate": 1.9349784152023903e-06,
+      "loss": 0.8037,
+      "step": 5484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.25961434841156,
+      "learning_rate": 1.933301817902372e-06,
+      "loss": 0.7842,
+      "step": 5485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2944680452346802,
+      "learning_rate": 1.9316257731757433e-06,
+      "loss": 0.8125,
+      "step": 5486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2617018222808838,
+      "learning_rate": 1.9299502813244997e-06,
+      "loss": 0.7791,
+      "step": 5487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.609894871711731,
+      "learning_rate": 1.928275342650544e-06,
+      "loss": 1.0103,
+      "step": 5488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2785234451293945,
+      "learning_rate": 1.926600957455673e-06,
+      "loss": 0.8286,
+      "step": 5489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.222988486289978,
+      "learning_rate": 1.924927126041587e-06,
+      "loss": 0.7801,
+      "step": 5490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5834958553314209,
+      "learning_rate": 1.923253848709883e-06,
+      "loss": 1.0326,
+      "step": 5491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2462297677993774,
+      "learning_rate": 1.921581125762062e-06,
+      "loss": 0.7725,
+      "step": 5492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5912604928016663,
+      "learning_rate": 1.9199089574995265e-06,
+      "loss": 1.0018,
+      "step": 5493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2497316598892212,
+      "learning_rate": 1.918237344223574e-06,
+      "loss": 0.8009,
+      "step": 5494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2381854057312012,
+      "learning_rate": 1.9165662862354035e-06,
+      "loss": 0.8876,
+      "step": 5495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3683300018310547,
+      "learning_rate": 1.9148957838361165e-06,
+      "loss": 0.8253,
+      "step": 5496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.242226004600525,
+      "learning_rate": 1.9132258373267127e-06,
+      "loss": 0.829,
+      "step": 5497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2719542980194092,
+      "learning_rate": 1.911556447008091e-06,
+      "loss": 0.8043,
+      "step": 5498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3199893236160278,
+      "learning_rate": 1.9098876131810486e-06,
+      "loss": 0.8296,
+      "step": 5499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2424037456512451,
+      "learning_rate": 1.9082193361462865e-06,
+      "loss": 0.8497,
+      "step": 5500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6340615749359131,
+      "learning_rate": 1.906551616204405e-06,
+      "loss": 1.0179,
+      "step": 5501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.21268892288208,
+      "learning_rate": 1.9048844536559014e-06,
+      "loss": 0.7465,
+      "step": 5502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3038992881774902,
+      "learning_rate": 1.903217848801171e-06,
+      "loss": 0.792,
+      "step": 5503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6353519558906555,
+      "learning_rate": 1.9015518019405154e-06,
+      "loss": 0.9948,
+      "step": 5504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3063750267028809,
+      "learning_rate": 1.8998863133741292e-06,
+      "loss": 0.8088,
+      "step": 5505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2313108444213867,
+      "learning_rate": 1.8982213834021084e-06,
+      "loss": 0.8276,
+      "step": 5506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2569756507873535,
+      "learning_rate": 1.8965570123244476e-06,
+      "loss": 0.7637,
+      "step": 5507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.295475721359253,
+      "learning_rate": 1.894893200441043e-06,
+      "loss": 0.8282,
+      "step": 5508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6128140091896057,
+      "learning_rate": 1.8932299480516908e-06,
+      "loss": 1.0165,
+      "step": 5509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2474377155303955,
+      "learning_rate": 1.891567255456082e-06,
+      "loss": 0.8059,
+      "step": 5510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.621105968952179,
+      "learning_rate": 1.8899051229538079e-06,
+      "loss": 1.0097,
+      "step": 5511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3722442388534546,
+      "learning_rate": 1.8882435508443637e-06,
+      "loss": 0.7629,
+      "step": 5512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2332087755203247,
+      "learning_rate": 1.8865825394271375e-06,
+      "loss": 0.7429,
+      "step": 5513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.190191388130188,
+      "learning_rate": 1.8849220890014191e-06,
+      "loss": 0.7903,
+      "step": 5514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6178674101829529,
+      "learning_rate": 1.8832621998663957e-06,
+      "loss": 1.0135,
+      "step": 5515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3315387964248657,
+      "learning_rate": 1.8816028723211578e-06,
+      "loss": 0.7762,
+      "step": 5516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1852766275405884,
+      "learning_rate": 1.8799441066646878e-06,
+      "loss": 0.7666,
+      "step": 5517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.275854229927063,
+      "learning_rate": 1.8782859031958729e-06,
+      "loss": 0.7766,
+      "step": 5518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2318207025527954,
+      "learning_rate": 1.8766282622134978e-06,
+      "loss": 0.8244,
+      "step": 5519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.297316312789917,
+      "learning_rate": 1.8749711840162433e-06,
+      "loss": 0.8476,
+      "step": 5520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.308329463005066,
+      "learning_rate": 1.8733146689026904e-06,
+      "loss": 0.8157,
+      "step": 5521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.244674801826477,
+      "learning_rate": 1.8716587171713163e-06,
+      "loss": 0.7902,
+      "step": 5522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.296795129776001,
+      "learning_rate": 1.8700033291205027e-06,
+      "loss": 0.8418,
+      "step": 5523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2518589496612549,
+      "learning_rate": 1.8683485050485245e-06,
+      "loss": 0.7943,
+      "step": 5524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1965571641921997,
+      "learning_rate": 1.8666942452535536e-06,
+      "loss": 0.7863,
+      "step": 5525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2194064855575562,
+      "learning_rate": 1.8650405500336654e-06,
+      "loss": 0.8179,
+      "step": 5526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2027534246444702,
+      "learning_rate": 1.8633874196868334e-06,
+      "loss": 0.7975,
+      "step": 5527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3019115924835205,
+      "learning_rate": 1.8617348545109249e-06,
+      "loss": 0.8302,
+      "step": 5528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2208672761917114,
+      "learning_rate": 1.860082854803707e-06,
+      "loss": 0.8137,
+      "step": 5529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3062574863433838,
+      "learning_rate": 1.8584314208628446e-06,
+      "loss": 0.8164,
+      "step": 5530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2942087650299072,
+      "learning_rate": 1.856780552985904e-06,
+      "loss": 0.8313,
+      "step": 5531
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2037009000778198,
+      "learning_rate": 1.8551302514703468e-06,
+      "loss": 0.7822,
+      "step": 5532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.254338026046753,
+      "learning_rate": 1.8534805166135294e-06,
+      "loss": 0.8513,
+      "step": 5533
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2627716064453125,
+      "learning_rate": 1.8518313487127126e-06,
+      "loss": 0.802,
+      "step": 5534
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2568607330322266,
+      "learning_rate": 1.8501827480650525e-06,
+      "loss": 0.8468,
+      "step": 5535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6109738945960999,
+      "learning_rate": 1.8485347149676014e-06,
+      "loss": 1.0283,
+      "step": 5536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5868591070175171,
+      "learning_rate": 1.8468872497173108e-06,
+      "loss": 1.0059,
+      "step": 5537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1990435123443604,
+      "learning_rate": 1.8452403526110263e-06,
+      "loss": 0.7344,
+      "step": 5538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.260017991065979,
+      "learning_rate": 1.8435940239454987e-06,
+      "loss": 0.8264,
+      "step": 5539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6164412498474121,
+      "learning_rate": 1.8419482640173697e-06,
+      "loss": 1.0161,
+      "step": 5540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5912904739379883,
+      "learning_rate": 1.8403030731231798e-06,
+      "loss": 1.0028,
+      "step": 5541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2922863960266113,
+      "learning_rate": 1.8386584515593691e-06,
+      "loss": 0.7927,
+      "step": 5542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.211430311203003,
+      "learning_rate": 1.8370143996222755e-06,
+      "loss": 0.7638,
+      "step": 5543
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.16140615940094,
+      "learning_rate": 1.8353709176081314e-06,
+      "loss": 0.7005,
+      "step": 5544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.214897632598877,
+      "learning_rate": 1.8337280058130675e-06,
+      "loss": 0.8043,
+      "step": 5545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2198526859283447,
+      "learning_rate": 1.8320856645331103e-06,
+      "loss": 0.8069,
+      "step": 5546
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2879230976104736,
+      "learning_rate": 1.8304438940641889e-06,
+      "loss": 0.8293,
+      "step": 5547
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2628294229507446,
+      "learning_rate": 1.8288026947021242e-06,
+      "loss": 0.7737,
+      "step": 5548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2664611339569092,
+      "learning_rate": 1.8271620667426343e-06,
+      "loss": 0.8238,
+      "step": 5549
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2334606647491455,
+      "learning_rate": 1.8255220104813376e-06,
+      "loss": 0.8071,
+      "step": 5550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2818948030471802,
+      "learning_rate": 1.8238825262137493e-06,
+      "loss": 0.7734,
+      "step": 5551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6066529750823975,
+      "learning_rate": 1.8222436142352784e-06,
+      "loss": 1.0036,
+      "step": 5552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1948434114456177,
+      "learning_rate": 1.8206052748412313e-06,
+      "loss": 0.7985,
+      "step": 5553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2697478532791138,
+      "learning_rate": 1.818967508326815e-06,
+      "loss": 0.7777,
+      "step": 5554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3260042667388916,
+      "learning_rate": 1.817330314987129e-06,
+      "loss": 0.8215,
+      "step": 5555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1666591167449951,
+      "learning_rate": 1.8156936951171721e-06,
+      "loss": 0.7804,
+      "step": 5556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3264586925506592,
+      "learning_rate": 1.8140576490118361e-06,
+      "loss": 0.7621,
+      "step": 5557
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5923706293106079,
+      "learning_rate": 1.8124221769659144e-06,
+      "loss": 1.0378,
+      "step": 5558
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2884207963943481,
+      "learning_rate": 1.8107872792740971e-06,
+      "loss": 0.8083,
+      "step": 5559
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2877638339996338,
+      "learning_rate": 1.8091529562309651e-06,
+      "loss": 0.8171,
+      "step": 5560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2506667375564575,
+      "learning_rate": 1.807519208130999e-06,
+      "loss": 0.7692,
+      "step": 5561
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2402421236038208,
+      "learning_rate": 1.805886035268578e-06,
+      "loss": 0.7788,
+      "step": 5562
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2025691270828247,
+      "learning_rate": 1.804253437937975e-06,
+      "loss": 0.8175,
+      "step": 5563
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2261219024658203,
+      "learning_rate": 1.8026214164333589e-06,
+      "loss": 0.8177,
+      "step": 5564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.240106225013733,
+      "learning_rate": 1.8009899710487943e-06,
+      "loss": 0.7524,
+      "step": 5565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2167404890060425,
+      "learning_rate": 1.7993591020782453e-06,
+      "loss": 0.811,
+      "step": 5566
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5984556674957275,
+      "learning_rate": 1.7977288098155716e-06,
+      "loss": 0.9912,
+      "step": 5567
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5972461700439453,
+      "learning_rate": 1.796099094554526e-06,
+      "loss": 1.01,
+      "step": 5568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.238343596458435,
+      "learning_rate": 1.794469956588757e-06,
+      "loss": 0.7622,
+      "step": 5569
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2679855823516846,
+      "learning_rate": 1.7928413962118152e-06,
+      "loss": 0.7818,
+      "step": 5570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2809998989105225,
+      "learning_rate": 1.7912134137171406e-06,
+      "loss": 0.8335,
+      "step": 5571
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5811687707901001,
+      "learning_rate": 1.7895860093980716e-06,
+      "loss": 1.0026,
+      "step": 5572
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2798244953155518,
+      "learning_rate": 1.7879591835478399e-06,
+      "loss": 0.8416,
+      "step": 5573
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2500202655792236,
+      "learning_rate": 1.7863329364595783e-06,
+      "loss": 0.8553,
+      "step": 5574
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2509589195251465,
+      "learning_rate": 1.7847072684263133e-06,
+      "loss": 0.7844,
+      "step": 5575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5899243354797363,
+      "learning_rate": 1.7830821797409647e-06,
+      "loss": 0.9982,
+      "step": 5576
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2331256866455078,
+      "learning_rate": 1.781457670696347e-06,
+      "loss": 0.8488,
+      "step": 5577
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2372552156448364,
+      "learning_rate": 1.7798337415851763e-06,
+      "loss": 0.8411,
+      "step": 5578
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6033875346183777,
+      "learning_rate": 1.7782103927000587e-06,
+      "loss": 0.9892,
+      "step": 5579
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2781695127487183,
+      "learning_rate": 1.776587624333498e-06,
+      "loss": 0.8035,
+      "step": 5580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2346237897872925,
+      "learning_rate": 1.7749654367778906e-06,
+      "loss": 0.8015,
+      "step": 5581
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5943603515625,
+      "learning_rate": 1.773343830325533e-06,
+      "loss": 0.982,
+      "step": 5582
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2483086585998535,
+      "learning_rate": 1.7717228052686153e-06,
+      "loss": 0.7761,
+      "step": 5583
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.196338176727295,
+      "learning_rate": 1.7701023618992208e-06,
+      "loss": 0.7942,
+      "step": 5584
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5943436622619629,
+      "learning_rate": 1.7684825005093276e-06,
+      "loss": 0.9893,
+      "step": 5585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2844953536987305,
+      "learning_rate": 1.7668632213908143e-06,
+      "loss": 0.7881,
+      "step": 5586
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.590785026550293,
+      "learning_rate": 1.765244524835449e-06,
+      "loss": 1.0168,
+      "step": 5587
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.281472086906433,
+      "learning_rate": 1.7636264111348967e-06,
+      "loss": 0.828,
+      "step": 5588
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2415571212768555,
+      "learning_rate": 1.762008880580715e-06,
+      "loss": 0.8232,
+      "step": 5589
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2419402599334717,
+      "learning_rate": 1.760391933464362e-06,
+      "loss": 0.7469,
+      "step": 5590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3150676488876343,
+      "learning_rate": 1.7587755700771885e-06,
+      "loss": 0.7995,
+      "step": 5591
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2504191398620605,
+      "learning_rate": 1.7571597907104359e-06,
+      "loss": 0.7987,
+      "step": 5592
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.204003095626831,
+      "learning_rate": 1.7555445956552463e-06,
+      "loss": 0.7841,
+      "step": 5593
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2314600944519043,
+      "learning_rate": 1.7539299852026531e-06,
+      "loss": 0.7641,
+      "step": 5594
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.26714289188385,
+      "learning_rate": 1.752315959643584e-06,
+      "loss": 0.8126,
+      "step": 5595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3033746480941772,
+      "learning_rate": 1.750702519268862e-06,
+      "loss": 0.7365,
+      "step": 5596
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6321671605110168,
+      "learning_rate": 1.7490896643692074e-06,
+      "loss": 0.9961,
+      "step": 5597
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.237533688545227,
+      "learning_rate": 1.7474773952352298e-06,
+      "loss": 0.7675,
+      "step": 5598
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1704007387161255,
+      "learning_rate": 1.7458657121574395e-06,
+      "loss": 0.7506,
+      "step": 5599
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3007402420043945,
+      "learning_rate": 1.7442546154262342e-06,
+      "loss": 0.8099,
+      "step": 5600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6186404824256897,
+      "learning_rate": 1.7426441053319127e-06,
+      "loss": 1.0259,
+      "step": 5601
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1991937160491943,
+      "learning_rate": 1.7410341821646637e-06,
+      "loss": 0.8373,
+      "step": 5602
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.216766357421875,
+      "learning_rate": 1.7394248462145714e-06,
+      "loss": 0.802,
+      "step": 5603
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5917977094650269,
+      "learning_rate": 1.7378160977716125e-06,
+      "loss": 1.0035,
+      "step": 5604
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2309457063674927,
+      "learning_rate": 1.7362079371256623e-06,
+      "loss": 0.7733,
+      "step": 5605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3040902614593506,
+      "learning_rate": 1.734600364566485e-06,
+      "loss": 0.8489,
+      "step": 5606
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.349713921546936,
+      "learning_rate": 1.732993380383744e-06,
+      "loss": 0.773,
+      "step": 5607
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2689709663391113,
+      "learning_rate": 1.7313869848669906e-06,
+      "loss": 0.8014,
+      "step": 5608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6238599419593811,
+      "learning_rate": 1.7297811783056773e-06,
+      "loss": 1.0062,
+      "step": 5609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4037487506866455,
+      "learning_rate": 1.7281759609891446e-06,
+      "loss": 0.8776,
+      "step": 5610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5869238376617432,
+      "learning_rate": 1.726571333206628e-06,
+      "loss": 1.0022,
+      "step": 5611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5871005654335022,
+      "learning_rate": 1.7249672952472568e-06,
+      "loss": 1.0307,
+      "step": 5612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2872687578201294,
+      "learning_rate": 1.723363847400058e-06,
+      "loss": 0.8194,
+      "step": 5613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2003464698791504,
+      "learning_rate": 1.7217609899539457e-06,
+      "loss": 0.7344,
+      "step": 5614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.283765196800232,
+      "learning_rate": 1.720158723197734e-06,
+      "loss": 0.8106,
+      "step": 5615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2501001358032227,
+      "learning_rate": 1.718557047420124e-06,
+      "loss": 0.7894,
+      "step": 5616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3368091583251953,
+      "learning_rate": 1.7169559629097182e-06,
+      "loss": 0.8101,
+      "step": 5617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2128981351852417,
+      "learning_rate": 1.7153554699550058e-06,
+      "loss": 0.7835,
+      "step": 5618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.224574327468872,
+      "learning_rate": 1.7137555688443713e-06,
+      "loss": 0.836,
+      "step": 5619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.200358271598816,
+      "learning_rate": 1.7121562598660924e-06,
+      "loss": 0.7483,
+      "step": 5620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.649232804775238,
+      "learning_rate": 1.7105575433083437e-06,
+      "loss": 0.9996,
+      "step": 5621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2651901245117188,
+      "learning_rate": 1.7089594194591869e-06,
+      "loss": 0.8031,
+      "step": 5622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6143496036529541,
+      "learning_rate": 1.707361888606583e-06,
+      "loss": 1.0177,
+      "step": 5623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2755353450775146,
+      "learning_rate": 1.7057649510383806e-06,
+      "loss": 0.8138,
+      "step": 5624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2693688869476318,
+      "learning_rate": 1.7041686070423274e-06,
+      "loss": 0.7831,
+      "step": 5625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6017403602600098,
+      "learning_rate": 1.702572856906059e-06,
+      "loss": 0.986,
+      "step": 5626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.252625584602356,
+      "learning_rate": 1.700977700917104e-06,
+      "loss": 0.8192,
+      "step": 5627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6319291591644287,
+      "learning_rate": 1.699383139362889e-06,
+      "loss": 0.7488,
+      "step": 5628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.586107075214386,
+      "learning_rate": 1.6977891725307294e-06,
+      "loss": 0.9962,
+      "step": 5629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.615411639213562,
+      "learning_rate": 1.6961958007078317e-06,
+      "loss": 1.0171,
+      "step": 5630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2837547063827515,
+      "learning_rate": 1.6946030241813e-06,
+      "loss": 0.8121,
+      "step": 5631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5980183482170105,
+      "learning_rate": 1.6930108432381304e-06,
+      "loss": 1.0135,
+      "step": 5632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2295660972595215,
+      "learning_rate": 1.691419258165209e-06,
+      "loss": 0.7983,
+      "step": 5633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.272697925567627,
+      "learning_rate": 1.6898282692493145e-06,
+      "loss": 0.796,
+      "step": 5634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3047058582305908,
+      "learning_rate": 1.6882378767771196e-06,
+      "loss": 0.7751,
+      "step": 5635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2155946493148804,
+      "learning_rate": 1.686648081035191e-06,
+      "loss": 0.8341,
+      "step": 5636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2586047649383545,
+      "learning_rate": 1.6850588823099856e-06,
+      "loss": 0.7789,
+      "step": 5637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.601947009563446,
+      "learning_rate": 1.6834702808878517e-06,
+      "loss": 1.014,
+      "step": 5638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2871817350387573,
+      "learning_rate": 1.6818822770550325e-06,
+      "loss": 0.7804,
+      "step": 5639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.242430567741394,
+      "learning_rate": 1.6802948710976652e-06,
+      "loss": 0.8014,
+      "step": 5640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5999130606651306,
+      "learning_rate": 1.6787080633017744e-06,
+      "loss": 1.0049,
+      "step": 5641
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2665257453918457,
+      "learning_rate": 1.6771218539532796e-06,
+      "loss": 0.8086,
+      "step": 5642
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2091234922409058,
+      "learning_rate": 1.6755362433379907e-06,
+      "loss": 0.7283,
+      "step": 5643
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3551771640777588,
+      "learning_rate": 1.673951231741614e-06,
+      "loss": 0.8456,
+      "step": 5644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2152646780014038,
+      "learning_rate": 1.6723668194497439e-06,
+      "loss": 0.8189,
+      "step": 5645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2193673849105835,
+      "learning_rate": 1.6707830067478652e-06,
+      "loss": 0.8202,
+      "step": 5646
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.224439024925232,
+      "learning_rate": 1.6691997939213605e-06,
+      "loss": 0.7451,
+      "step": 5647
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5962003469467163,
+      "learning_rate": 1.667617181255502e-06,
+      "loss": 1.0003,
+      "step": 5648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2653205394744873,
+      "learning_rate": 1.6660351690354509e-06,
+      "loss": 0.7877,
+      "step": 5649
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5926564931869507,
+      "learning_rate": 1.6644537575462632e-06,
+      "loss": 1.0132,
+      "step": 5650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6028006672859192,
+      "learning_rate": 1.6628729470728838e-06,
+      "loss": 1.0124,
+      "step": 5651
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.280683159828186,
+      "learning_rate": 1.6612927379001543e-06,
+      "loss": 0.8068,
+      "step": 5652
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.263116717338562,
+      "learning_rate": 1.6597131303128033e-06,
+      "loss": 0.8449,
+      "step": 5653
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3498656749725342,
+      "learning_rate": 1.6581341245954508e-06,
+      "loss": 0.7994,
+      "step": 5654
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5995937585830688,
+      "learning_rate": 1.6565557210326122e-06,
+      "loss": 0.9916,
+      "step": 5655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3064745664596558,
+      "learning_rate": 1.6549779199086934e-06,
+      "loss": 0.819,
+      "step": 5656
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5142639875411987,
+      "learning_rate": 1.6534007215079895e-06,
+      "loss": 0.7971,
+      "step": 5657
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2592954635620117,
+      "learning_rate": 1.6518241261146883e-06,
+      "loss": 0.7522,
+      "step": 5658
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2796051502227783,
+      "learning_rate": 1.6502481340128669e-06,
+      "loss": 0.7677,
+      "step": 5659
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2307037115097046,
+      "learning_rate": 1.6486727454864988e-06,
+      "loss": 0.795,
+      "step": 5660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2637430429458618,
+      "learning_rate": 1.6470979608194443e-06,
+      "loss": 0.8116,
+      "step": 5661
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2678707838058472,
+      "learning_rate": 1.6455237802954538e-06,
+      "loss": 0.8248,
+      "step": 5662
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6007549166679382,
+      "learning_rate": 1.6439502041981742e-06,
+      "loss": 1.0386,
+      "step": 5663
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2647395133972168,
+      "learning_rate": 1.642377232811141e-06,
+      "loss": 0.8349,
+      "step": 5664
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5881412625312805,
+      "learning_rate": 1.6408048664177784e-06,
+      "loss": 1.0193,
+      "step": 5665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.274959683418274,
+      "learning_rate": 1.6392331053014032e-06,
+      "loss": 0.8,
+      "step": 5666
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2542757987976074,
+      "learning_rate": 1.637661949745225e-06,
+      "loss": 0.7554,
+      "step": 5667
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5930153727531433,
+      "learning_rate": 1.6360914000323425e-06,
+      "loss": 0.9966,
+      "step": 5668
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5639510750770569,
+      "learning_rate": 1.634521456445744e-06,
+      "loss": 1.0273,
+      "step": 5669
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1904473304748535,
+      "learning_rate": 1.6329521192683095e-06,
+      "loss": 0.7526,
+      "step": 5670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3485430479049683,
+      "learning_rate": 1.6313833887828112e-06,
+      "loss": 0.7955,
+      "step": 5671
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2599806785583496,
+      "learning_rate": 1.6298152652719129e-06,
+      "loss": 0.7942,
+      "step": 5672
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5874489545822144,
+      "learning_rate": 1.6282477490181653e-06,
+      "loss": 1.0125,
+      "step": 5673
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.581601619720459,
+      "learning_rate": 1.6266808403040103e-06,
+      "loss": 1.0125,
+      "step": 5674
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5709757804870605,
+      "learning_rate": 1.625114539411784e-06,
+      "loss": 0.9663,
+      "step": 5675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3707294464111328,
+      "learning_rate": 1.62354884662371e-06,
+      "loss": 0.8279,
+      "step": 5676
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3068639039993286,
+      "learning_rate": 1.6219837622219015e-06,
+      "loss": 0.8467,
+      "step": 5677
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5785022377967834,
+      "learning_rate": 1.6204192864883628e-06,
+      "loss": 1.0057,
+      "step": 5678
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2392752170562744,
+      "learning_rate": 1.6188554197049904e-06,
+      "loss": 0.7351,
+      "step": 5679
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2141674757003784,
+      "learning_rate": 1.617292162153572e-06,
+      "loss": 0.7592,
+      "step": 5680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2977006435394287,
+      "learning_rate": 1.6157295141157809e-06,
+      "loss": 0.7753,
+      "step": 5681
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3181476593017578,
+      "learning_rate": 1.6141674758731817e-06,
+      "loss": 0.8323,
+      "step": 5682
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6012328863143921,
+      "learning_rate": 1.6126060477072336e-06,
+      "loss": 1.0099,
+      "step": 5683
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5676648020744324,
+      "learning_rate": 1.6110452298992817e-06,
+      "loss": 1.0367,
+      "step": 5684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3316155672073364,
+      "learning_rate": 1.609485022730562e-06,
+      "loss": 0.7559,
+      "step": 5685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.241847038269043,
+      "learning_rate": 1.607925426482198e-06,
+      "loss": 0.8116,
+      "step": 5686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3243967294692993,
+      "learning_rate": 1.6063664414352086e-06,
+      "loss": 0.8339,
+      "step": 5687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2309452295303345,
+      "learning_rate": 1.6048080678705009e-06,
+      "loss": 0.7701,
+      "step": 5688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1917283535003662,
+      "learning_rate": 1.6032503060688681e-06,
+      "loss": 0.8278,
+      "step": 5689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1781609058380127,
+      "learning_rate": 1.6016931563109956e-06,
+      "loss": 0.7642,
+      "step": 5690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3283085823059082,
+      "learning_rate": 1.6001366188774613e-06,
+      "loss": 0.7772,
+      "step": 5691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2423272132873535,
+      "learning_rate": 1.598580694048728e-06,
+      "loss": 0.8119,
+      "step": 5692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2317909002304077,
+      "learning_rate": 1.5970253821051508e-06,
+      "loss": 0.7966,
+      "step": 5693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2394647598266602,
+      "learning_rate": 1.5954706833269717e-06,
+      "loss": 0.7685,
+      "step": 5694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2214901447296143,
+      "learning_rate": 1.5939165979943265e-06,
+      "loss": 0.7788,
+      "step": 5695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2168536186218262,
+      "learning_rate": 1.5923631263872392e-06,
+      "loss": 0.7841,
+      "step": 5696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2313281297683716,
+      "learning_rate": 1.5908102687856208e-06,
+      "loss": 0.7451,
+      "step": 5697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5935384631156921,
+      "learning_rate": 1.5892580254692718e-06,
+      "loss": 0.9954,
+      "step": 5698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6025569438934326,
+      "learning_rate": 1.5877063967178869e-06,
+      "loss": 1.0296,
+      "step": 5699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6304629445075989,
+      "learning_rate": 1.5861553828110437e-06,
+      "loss": 1.0203,
+      "step": 5700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5848435163497925,
+      "learning_rate": 1.5846049840282112e-06,
+      "loss": 1.0128,
+      "step": 5701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.219828724861145,
+      "learning_rate": 1.5830552006487515e-06,
+      "loss": 0.8326,
+      "step": 5702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.210870385169983,
+      "learning_rate": 1.5815060329519088e-06,
+      "loss": 0.803,
+      "step": 5703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5902683138847351,
+      "learning_rate": 1.5799574812168234e-06,
+      "loss": 1.0208,
+      "step": 5704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2105525732040405,
+      "learning_rate": 1.5784095457225185e-06,
+      "loss": 0.8044,
+      "step": 5705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5861756801605225,
+      "learning_rate": 1.5768622267479118e-06,
+      "loss": 1.0095,
+      "step": 5706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5993788838386536,
+      "learning_rate": 1.5753155245718054e-06,
+      "loss": 1.0247,
+      "step": 5707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2392632961273193,
+      "learning_rate": 1.5737694394728924e-06,
+      "loss": 0.787,
+      "step": 5708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6002811193466187,
+      "learning_rate": 1.5722239717297527e-06,
+      "loss": 0.9963,
+      "step": 5709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2445154190063477,
+      "learning_rate": 1.5706791216208595e-06,
+      "loss": 0.788,
+      "step": 5710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5899356007575989,
+      "learning_rate": 1.5691348894245689e-06,
+      "loss": 1.0144,
+      "step": 5711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3330422639846802,
+      "learning_rate": 1.5675912754191314e-06,
+      "loss": 0.8891,
+      "step": 5712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2262067794799805,
+      "learning_rate": 1.566048279882681e-06,
+      "loss": 0.7775,
+      "step": 5713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2632888555526733,
+      "learning_rate": 1.5645059030932447e-06,
+      "loss": 0.7657,
+      "step": 5714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.520643711090088,
+      "learning_rate": 1.5629641453287347e-06,
+      "loss": 0.8307,
+      "step": 5715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3227769136428833,
+      "learning_rate": 1.5614230068669528e-06,
+      "loss": 0.8166,
+      "step": 5716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4151718616485596,
+      "learning_rate": 1.5598824879855872e-06,
+      "loss": 0.7745,
+      "step": 5717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2213939428329468,
+      "learning_rate": 1.5583425889622206e-06,
+      "loss": 0.8183,
+      "step": 5718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5969546437263489,
+      "learning_rate": 1.5568033100743163e-06,
+      "loss": 1.0058,
+      "step": 5719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2092368602752686,
+      "learning_rate": 1.555264651599232e-06,
+      "loss": 0.8001,
+      "step": 5720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6167859435081482,
+      "learning_rate": 1.5537266138142086e-06,
+      "loss": 1.0268,
+      "step": 5721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2543070316314697,
+      "learning_rate": 1.5521891969963809e-06,
+      "loss": 0.7905,
+      "step": 5722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.238558053970337,
+      "learning_rate": 1.5506524014227665e-06,
+      "loss": 0.7938,
+      "step": 5723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2334494590759277,
+      "learning_rate": 1.549116227370273e-06,
+      "loss": 0.837,
+      "step": 5724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1747597455978394,
+      "learning_rate": 1.5475806751156946e-06,
+      "loss": 0.8088,
+      "step": 5725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6341240406036377,
+      "learning_rate": 1.5460457449357192e-06,
+      "loss": 1.0101,
+      "step": 5726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.314561128616333,
+      "learning_rate": 1.5445114371069136e-06,
+      "loss": 0.7996,
+      "step": 5727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1959978342056274,
+      "learning_rate": 1.5429777519057416e-06,
+      "loss": 0.7776,
+      "step": 5728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.294748306274414,
+      "learning_rate": 1.5414446896085461e-06,
+      "loss": 0.8501,
+      "step": 5729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2796010971069336,
+      "learning_rate": 1.5399122504915665e-06,
+      "loss": 0.8411,
+      "step": 5730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5984624624252319,
+      "learning_rate": 1.538380434830924e-06,
+      "loss": 1.0278,
+      "step": 5731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2580537796020508,
+      "learning_rate": 1.5368492429026277e-06,
+      "loss": 0.7693,
+      "step": 5732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6188507080078125,
+      "learning_rate": 1.5353186749825749e-06,
+      "loss": 1.0134,
+      "step": 5733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2274680137634277,
+      "learning_rate": 1.5337887313465543e-06,
+      "loss": 0.7423,
+      "step": 5734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1692081689834595,
+      "learning_rate": 1.5322594122702362e-06,
+      "loss": 0.7527,
+      "step": 5735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2876821756362915,
+      "learning_rate": 1.5307307180291814e-06,
+      "loss": 0.8082,
+      "step": 5736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2404475212097168,
+      "learning_rate": 1.529202648898841e-06,
+      "loss": 0.7917,
+      "step": 5737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3345201015472412,
+      "learning_rate": 1.5276752051545479e-06,
+      "loss": 0.8902,
+      "step": 5738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2992658615112305,
+      "learning_rate": 1.526148387071525e-06,
+      "loss": 0.7961,
+      "step": 5739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2007100582122803,
+      "learning_rate": 1.5246221949248802e-06,
+      "loss": 0.803,
+      "step": 5740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2264429330825806,
+      "learning_rate": 1.5230966289896143e-06,
+      "loss": 0.8023,
+      "step": 5741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5969757437705994,
+      "learning_rate": 1.5215716895406095e-06,
+      "loss": 1.0197,
+      "step": 5742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.604171097278595,
+      "learning_rate": 1.5200473768526365e-06,
+      "loss": 1.0058,
+      "step": 5743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6724482774734497,
+      "learning_rate": 1.5185236912003542e-06,
+      "loss": 0.777,
+      "step": 5744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3047775030136108,
+      "learning_rate": 1.5170006328583104e-06,
+      "loss": 0.8182,
+      "step": 5745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2814868688583374,
+      "learning_rate": 1.5154782021009357e-06,
+      "loss": 0.8531,
+      "step": 5746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.186157464981079,
+      "learning_rate": 1.5139563992025503e-06,
+      "loss": 0.8331,
+      "step": 5747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2833539247512817,
+      "learning_rate": 1.5124352244373574e-06,
+      "loss": 0.8108,
+      "step": 5748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3207459449768066,
+      "learning_rate": 1.5109146780794543e-06,
+      "loss": 0.8269,
+      "step": 5749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2844793796539307,
+      "learning_rate": 1.5093947604028185e-06,
+      "loss": 0.8117,
+      "step": 5750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2621155977249146,
+      "learning_rate": 1.5078754716813159e-06,
+      "loss": 0.8304,
+      "step": 5751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2492135763168335,
+      "learning_rate": 1.5063568121887023e-06,
+      "loss": 0.7967,
+      "step": 5752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5983306765556335,
+      "learning_rate": 1.5048387821986142e-06,
+      "loss": 1.0259,
+      "step": 5753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2543874979019165,
+      "learning_rate": 1.5033213819845816e-06,
+      "loss": 0.8443,
+      "step": 5754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5976306200027466,
+      "learning_rate": 1.5018046118200158e-06,
+      "loss": 1.0109,
+      "step": 5755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2593438625335693,
+      "learning_rate": 1.500288471978214e-06,
+      "loss": 0.8163,
+      "step": 5756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2699612379074097,
+      "learning_rate": 1.4987729627323666e-06,
+      "loss": 0.7988,
+      "step": 5757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2487543821334839,
+      "learning_rate": 1.497258084355543e-06,
+      "loss": 0.7562,
+      "step": 5758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5965562462806702,
+      "learning_rate": 1.4957438371207005e-06,
+      "loss": 0.9885,
+      "step": 5759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2504210472106934,
+      "learning_rate": 1.494230221300687e-06,
+      "loss": 0.7797,
+      "step": 5760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2296019792556763,
+      "learning_rate": 1.4927172371682308e-06,
+      "loss": 0.776,
+      "step": 5761
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3460023403167725,
+      "learning_rate": 1.491204884995952e-06,
+      "loss": 0.8343,
+      "step": 5762
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1934703588485718,
+      "learning_rate": 1.4896931650563529e-06,
+      "loss": 0.7568,
+      "step": 5763
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2234916687011719,
+      "learning_rate": 1.4881820776218203e-06,
+      "loss": 0.7804,
+      "step": 5764
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3154258728027344,
+      "learning_rate": 1.4866716229646334e-06,
+      "loss": 0.7911,
+      "step": 5765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2106105089187622,
+      "learning_rate": 1.4851618013569518e-06,
+      "loss": 0.778,
+      "step": 5766
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.172018051147461,
+      "learning_rate": 1.4836526130708224e-06,
+      "loss": 0.8104,
+      "step": 5767
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.247493028640747,
+      "learning_rate": 1.4821440583781804e-06,
+      "loss": 0.8052,
+      "step": 5768
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.256687045097351,
+      "learning_rate": 1.4806361375508422e-06,
+      "loss": 0.8148,
+      "step": 5769
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5781628489494324,
+      "learning_rate": 1.4791288508605156e-06,
+      "loss": 1.0106,
+      "step": 5770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6151180267333984,
+      "learning_rate": 1.4776221985787887e-06,
+      "loss": 0.8207,
+      "step": 5771
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2315834760665894,
+      "learning_rate": 1.4761161809771408e-06,
+      "loss": 0.8233,
+      "step": 5772
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5950877666473389,
+      "learning_rate": 1.4746107983269315e-06,
+      "loss": 1.0253,
+      "step": 5773
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.274150013923645,
+      "learning_rate": 1.4731060508994093e-06,
+      "loss": 0.7854,
+      "step": 5774
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.226152777671814,
+      "learning_rate": 1.4716019389657055e-06,
+      "loss": 0.7872,
+      "step": 5775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3289248943328857,
+      "learning_rate": 1.4700984627968412e-06,
+      "loss": 0.8345,
+      "step": 5776
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3115731477737427,
+      "learning_rate": 1.4685956226637183e-06,
+      "loss": 0.8571,
+      "step": 5777
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3012735843658447,
+      "learning_rate": 1.467093418837129e-06,
+      "loss": 0.8219,
+      "step": 5778
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5804508924484253,
+      "learning_rate": 1.4655918515877443e-06,
+      "loss": 1.0162,
+      "step": 5779
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.290719747543335,
+      "learning_rate": 1.4640909211861288e-06,
+      "loss": 0.8307,
+      "step": 5780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2288990020751953,
+      "learning_rate": 1.4625906279027247e-06,
+      "loss": 0.778,
+      "step": 5781
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.411199688911438,
+      "learning_rate": 1.4610909720078632e-06,
+      "loss": 0.8585,
+      "step": 5782
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2674797773361206,
+      "learning_rate": 1.459591953771759e-06,
+      "loss": 0.8493,
+      "step": 5783
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3654135465621948,
+      "learning_rate": 1.4580935734645151e-06,
+      "loss": 0.8041,
+      "step": 5784
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5968053340911865,
+      "learning_rate": 1.4565958313561146e-06,
+      "loss": 1.0071,
+      "step": 5785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.239083170890808,
+      "learning_rate": 1.4550987277164313e-06,
+      "loss": 0.8283,
+      "step": 5786
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2307525873184204,
+      "learning_rate": 1.4536022628152185e-06,
+      "loss": 0.7882,
+      "step": 5787
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2082027196884155,
+      "learning_rate": 1.4521064369221193e-06,
+      "loss": 0.8159,
+      "step": 5788
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.266205072402954,
+      "learning_rate": 1.4506112503066582e-06,
+      "loss": 0.8306,
+      "step": 5789
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2732858657836914,
+      "learning_rate": 1.4491167032382453e-06,
+      "loss": 0.81,
+      "step": 5790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2416120767593384,
+      "learning_rate": 1.447622795986174e-06,
+      "loss": 0.8751,
+      "step": 5791
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2621382474899292,
+      "learning_rate": 1.4461295288196276e-06,
+      "loss": 0.8125,
+      "step": 5792
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6026510000228882,
+      "learning_rate": 1.444636902007668e-06,
+      "loss": 1.0121,
+      "step": 5793
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1916009187698364,
+      "learning_rate": 1.4431449158192462e-06,
+      "loss": 0.7509,
+      "step": 5794
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2261592149734497,
+      "learning_rate": 1.4416535705231939e-06,
+      "loss": 0.7597,
+      "step": 5795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2553586959838867,
+      "learning_rate": 1.4401628663882316e-06,
+      "loss": 0.836,
+      "step": 5796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.242614984512329,
+      "learning_rate": 1.4386728036829612e-06,
+      "loss": 0.7679,
+      "step": 5797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2583143711090088,
+      "learning_rate": 1.4371833826758697e-06,
+      "loss": 0.7865,
+      "step": 5798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5838809013366699,
+      "learning_rate": 1.4356946036353269e-06,
+      "loss": 1.0214,
+      "step": 5799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5821447968482971,
+      "learning_rate": 1.4342064668295914e-06,
+      "loss": 1.0223,
+      "step": 5800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2817858457565308,
+      "learning_rate": 1.432718972526801e-06,
+      "loss": 0.8186,
+      "step": 5801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2828034162521362,
+      "learning_rate": 1.4312321209949825e-06,
+      "loss": 0.8256,
+      "step": 5802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.589566707611084,
+      "learning_rate": 1.4297459125020412e-06,
+      "loss": 0.9968,
+      "step": 5803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2494031190872192,
+      "learning_rate": 1.4282603473157737e-06,
+      "loss": 0.822,
+      "step": 5804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.216991662979126,
+      "learning_rate": 1.426775425703854e-06,
+      "loss": 0.808,
+      "step": 5805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5790331363677979,
+      "learning_rate": 1.4252911479338444e-06,
+      "loss": 1.0046,
+      "step": 5806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2274168729782104,
+      "learning_rate": 1.4238075142731867e-06,
+      "loss": 0.8321,
+      "step": 5807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2103391885757446,
+      "learning_rate": 1.4223245249892137e-06,
+      "loss": 0.7996,
+      "step": 5808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1897211074829102,
+      "learning_rate": 1.420842180349134e-06,
+      "loss": 0.7759,
+      "step": 5809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2420750856399536,
+      "learning_rate": 1.4193604806200462e-06,
+      "loss": 0.7711,
+      "step": 5810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5794219970703125,
+      "learning_rate": 1.417879426068932e-06,
+      "loss": 0.9923,
+      "step": 5811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2762582302093506,
+      "learning_rate": 1.4163990169626546e-06,
+      "loss": 0.8079,
+      "step": 5812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3268531560897827,
+      "learning_rate": 1.4149192535679602e-06,
+      "loss": 0.7871,
+      "step": 5813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.604062020778656,
+      "learning_rate": 1.4134401361514793e-06,
+      "loss": 1.0389,
+      "step": 5814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6570265889167786,
+      "learning_rate": 1.4119616649797302e-06,
+      "loss": 1.021,
+      "step": 5815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2602696418762207,
+      "learning_rate": 1.4104838403191096e-06,
+      "loss": 0.8513,
+      "step": 5816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5785011649131775,
+      "learning_rate": 1.4090066624358984e-06,
+      "loss": 0.9951,
+      "step": 5817
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3149082660675049,
+      "learning_rate": 1.4075301315962625e-06,
+      "loss": 0.8348,
+      "step": 5818
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2646125555038452,
+      "learning_rate": 1.4060542480662537e-06,
+      "loss": 0.8398,
+      "step": 5819
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.26462984085083,
+      "learning_rate": 1.404579012111802e-06,
+      "loss": 0.8096,
+      "step": 5820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2636045217514038,
+      "learning_rate": 1.403104423998723e-06,
+      "loss": 0.8234,
+      "step": 5821
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5873101949691772,
+      "learning_rate": 1.4016304839927141e-06,
+      "loss": 1.0295,
+      "step": 5822
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.225889801979065,
+      "learning_rate": 1.4001571923593604e-06,
+      "loss": 0.7717,
+      "step": 5823
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2489045858383179,
+      "learning_rate": 1.3986845493641255e-06,
+      "loss": 0.824,
+      "step": 5824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2505537271499634,
+      "learning_rate": 1.3972125552723565e-06,
+      "loss": 0.8067,
+      "step": 5825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2452272176742554,
+      "learning_rate": 1.3957412103492857e-06,
+      "loss": 0.7905,
+      "step": 5826
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6373071074485779,
+      "learning_rate": 1.39427051486003e-06,
+      "loss": 1.0194,
+      "step": 5827
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.291451096534729,
+      "learning_rate": 1.3928004690695852e-06,
+      "loss": 0.8524,
+      "step": 5828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2031664848327637,
+      "learning_rate": 1.3913310732428309e-06,
+      "loss": 0.8354,
+      "step": 5829
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2119925022125244,
+      "learning_rate": 1.3898623276445294e-06,
+      "loss": 0.7728,
+      "step": 5830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2808446884155273,
+      "learning_rate": 1.3883942325393302e-06,
+      "loss": 0.7941,
+      "step": 5831
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.285585641860962,
+      "learning_rate": 1.3869267881917598e-06,
+      "loss": 0.808,
+      "step": 5832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2559391260147095,
+      "learning_rate": 1.3854599948662289e-06,
+      "loss": 0.7498,
+      "step": 5833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.221504807472229,
+      "learning_rate": 1.3839938528270336e-06,
+      "loss": 0.8379,
+      "step": 5834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2577060461044312,
+      "learning_rate": 1.3825283623383518e-06,
+      "loss": 0.8245,
+      "step": 5835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2040460109710693,
+      "learning_rate": 1.3810635236642417e-06,
+      "loss": 0.7849,
+      "step": 5836
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2540684938430786,
+      "learning_rate": 1.3795993370686456e-06,
+      "loss": 0.7777,
+      "step": 5837
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5890868902206421,
+      "learning_rate": 1.3781358028153864e-06,
+      "loss": 1.0182,
+      "step": 5838
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2843390703201294,
+      "learning_rate": 1.3766729211681746e-06,
+      "loss": 0.8303,
+      "step": 5839
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.251376748085022,
+      "learning_rate": 1.3752106923905978e-06,
+      "loss": 0.8268,
+      "step": 5840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5841009616851807,
+      "learning_rate": 1.373749116746127e-06,
+      "loss": 1.0076,
+      "step": 5841
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2995787858963013,
+      "learning_rate": 1.372288194498117e-06,
+      "loss": 0.811,
+      "step": 5842
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2603468894958496,
+      "learning_rate": 1.3708279259098068e-06,
+      "loss": 0.7742,
+      "step": 5843
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2640868425369263,
+      "learning_rate": 1.3693683112443124e-06,
+      "loss": 0.8464,
+      "step": 5844
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2407456636428833,
+      "learning_rate": 1.3679093507646341e-06,
+      "loss": 0.7573,
+      "step": 5845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2354376316070557,
+      "learning_rate": 1.366451044733657e-06,
+      "loss": 0.7896,
+      "step": 5846
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3556522130966187,
+      "learning_rate": 1.3649933934141457e-06,
+      "loss": 0.779,
+      "step": 5847
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5840878486633301,
+      "learning_rate": 1.3635363970687466e-06,
+      "loss": 0.9902,
+      "step": 5848
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3302044868469238,
+      "learning_rate": 1.3620800559599874e-06,
+      "loss": 0.7935,
+      "step": 5849
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2307981252670288,
+      "learning_rate": 1.3606243703502804e-06,
+      "loss": 0.7975,
+      "step": 5850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.234232783317566,
+      "learning_rate": 1.3591693405019202e-06,
+      "loss": 0.7409,
+      "step": 5851
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2564383745193481,
+      "learning_rate": 1.35771496667708e-06,
+      "loss": 0.7062,
+      "step": 5852
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2758697271347046,
+      "learning_rate": 1.3562612491378152e-06,
+      "loss": 0.7769,
+      "step": 5853
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2487480640411377,
+      "learning_rate": 1.3548081881460668e-06,
+      "loss": 0.7733,
+      "step": 5854
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2993261814117432,
+      "learning_rate": 1.353355783963653e-06,
+      "loss": 0.7851,
+      "step": 5855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6165887117385864,
+      "learning_rate": 1.3519040368522756e-06,
+      "loss": 1.0072,
+      "step": 5856
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1759555339813232,
+      "learning_rate": 1.3504529470735162e-06,
+      "loss": 0.7187,
+      "step": 5857
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2252315282821655,
+      "learning_rate": 1.3490025148888414e-06,
+      "loss": 0.8522,
+      "step": 5858
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3240500688552856,
+      "learning_rate": 1.347552740559599e-06,
+      "loss": 0.8578,
+      "step": 5859
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2760413885116577,
+      "learning_rate": 1.346103624347015e-06,
+      "loss": 0.7866,
+      "step": 5860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2372220754623413,
+      "learning_rate": 1.344655166512197e-06,
+      "loss": 0.7855,
+      "step": 5861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2253435850143433,
+      "learning_rate": 1.3432073673161394e-06,
+      "loss": 0.8266,
+      "step": 5862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.192969799041748,
+      "learning_rate": 1.3417602270197111e-06,
+      "loss": 0.8063,
+      "step": 5863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1851203441619873,
+      "learning_rate": 1.3403137458836668e-06,
+      "loss": 0.8266,
+      "step": 5864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.566316545009613,
+      "learning_rate": 1.338867924168638e-06,
+      "loss": 1.0106,
+      "step": 5865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3363306522369385,
+      "learning_rate": 1.3374227621351431e-06,
+      "loss": 0.8412,
+      "step": 5866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6001700162887573,
+      "learning_rate": 1.3359782600435795e-06,
+      "loss": 1.0155,
+      "step": 5867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5918979048728943,
+      "learning_rate": 1.3345344181542231e-06,
+      "loss": 1.0007,
+      "step": 5868
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5919071435928345,
+      "learning_rate": 1.333091236727232e-06,
+      "loss": 1.0316,
+      "step": 5869
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2402281761169434,
+      "learning_rate": 1.3316487160226488e-06,
+      "loss": 0.8501,
+      "step": 5870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2176406383514404,
+      "learning_rate": 1.330206856300393e-06,
+      "loss": 0.7748,
+      "step": 5871
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6043719053268433,
+      "learning_rate": 1.3287656578202656e-06,
+      "loss": 0.9732,
+      "step": 5872
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2109016180038452,
+      "learning_rate": 1.3273251208419485e-06,
+      "loss": 0.8478,
+      "step": 5873
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.210853099822998,
+      "learning_rate": 1.3258852456250055e-06,
+      "loss": 0.7609,
+      "step": 5874
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2823635339736938,
+      "learning_rate": 1.3244460324288832e-06,
+      "loss": 0.8142,
+      "step": 5875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.228342890739441,
+      "learning_rate": 1.323007481512904e-06,
+      "loss": 0.8329,
+      "step": 5876
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.290907859802246,
+      "learning_rate": 1.3215695931362727e-06,
+      "loss": 0.8148,
+      "step": 5877
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6017650961875916,
+      "learning_rate": 1.3201323675580779e-06,
+      "loss": 1.0038,
+      "step": 5878
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.238552212715149,
+      "learning_rate": 1.318695805037285e-06,
+      "loss": 0.7499,
+      "step": 5879
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2636988162994385,
+      "learning_rate": 1.3172599058327412e-06,
+      "loss": 0.7697,
+      "step": 5880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.306602954864502,
+      "learning_rate": 1.3158246702031724e-06,
+      "loss": 0.7895,
+      "step": 5881
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2962415218353271,
+      "learning_rate": 1.3143900984071884e-06,
+      "loss": 0.8286,
+      "step": 5882
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3069825172424316,
+      "learning_rate": 1.3129561907032795e-06,
+      "loss": 0.7916,
+      "step": 5883
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2399225234985352,
+      "learning_rate": 1.3115229473498108e-06,
+      "loss": 0.8371,
+      "step": 5884
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2562215328216553,
+      "learning_rate": 1.3100903686050347e-06,
+      "loss": 0.788,
+      "step": 5885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3187766075134277,
+      "learning_rate": 1.3086584547270797e-06,
+      "loss": 0.7206,
+      "step": 5886
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6284576058387756,
+      "learning_rate": 1.307227205973955e-06,
+      "loss": 1.0388,
+      "step": 5887
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2453207969665527,
+      "learning_rate": 1.305796622603548e-06,
+      "loss": 0.8396,
+      "step": 5888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5783333778381348,
+      "learning_rate": 1.3043667048736324e-06,
+      "loss": 1.007,
+      "step": 5889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2847908735275269,
+      "learning_rate": 1.3029374530418554e-06,
+      "loss": 0.8446,
+      "step": 5890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2370407581329346,
+      "learning_rate": 1.3015088673657495e-06,
+      "loss": 0.8529,
+      "step": 5891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2437925338745117,
+      "learning_rate": 1.3000809481027216e-06,
+      "loss": 0.7834,
+      "step": 5892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3325248956680298,
+      "learning_rate": 1.2986536955100642e-06,
+      "loss": 0.807,
+      "step": 5893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6129140257835388,
+      "learning_rate": 1.2972271098449457e-06,
+      "loss": 0.9981,
+      "step": 5894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.205405354499817,
+      "learning_rate": 1.295801191364416e-06,
+      "loss": 0.7478,
+      "step": 5895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.279668927192688,
+      "learning_rate": 1.294375940325402e-06,
+      "loss": 0.8058,
+      "step": 5896
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2260043621063232,
+      "learning_rate": 1.2929513569847168e-06,
+      "loss": 0.8079,
+      "step": 5897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2755101919174194,
+      "learning_rate": 1.2915274415990454e-06,
+      "loss": 0.7755,
+      "step": 5898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2784762382507324,
+      "learning_rate": 1.2901041944249597e-06,
+      "loss": 0.8386,
+      "step": 5899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3202524185180664,
+      "learning_rate": 1.2886816157189041e-06,
+      "loss": 0.8237,
+      "step": 5900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1847442388534546,
+      "learning_rate": 1.2872597057372094e-06,
+      "loss": 0.7819,
+      "step": 5901
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3253474235534668,
+      "learning_rate": 1.2858384647360816e-06,
+      "loss": 0.7663,
+      "step": 5902
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2558703422546387,
+      "learning_rate": 1.2844178929716061e-06,
+      "loss": 0.8261,
+      "step": 5903
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6269530653953552,
+      "learning_rate": 1.282997990699748e-06,
+      "loss": 1.0262,
+      "step": 5904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2930114269256592,
+      "learning_rate": 1.2815787581763556e-06,
+      "loss": 0.8238,
+      "step": 5905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5954931378364563,
+      "learning_rate": 1.2801601956571501e-06,
+      "loss": 1.0083,
+      "step": 5906
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2474855184555054,
+      "learning_rate": 1.2787423033977391e-06,
+      "loss": 0.8467,
+      "step": 5907
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2951706647872925,
+      "learning_rate": 1.277325081653602e-06,
+      "loss": 0.8215,
+      "step": 5908
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2287918329238892,
+      "learning_rate": 1.2759085306801039e-06,
+      "loss": 0.8417,
+      "step": 5909
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2565696239471436,
+      "learning_rate": 1.2744926507324856e-06,
+      "loss": 0.7699,
+      "step": 5910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.262244701385498,
+      "learning_rate": 1.273077442065867e-06,
+      "loss": 0.8143,
+      "step": 5911
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5839703679084778,
+      "learning_rate": 1.2716629049352459e-06,
+      "loss": 1.0222,
+      "step": 5912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6175693869590759,
+      "learning_rate": 1.2702490395955042e-06,
+      "loss": 1.0102,
+      "step": 5913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2511297464370728,
+      "learning_rate": 1.2688358463013966e-06,
+      "loss": 0.8093,
+      "step": 5914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3156001567840576,
+      "learning_rate": 1.2674233253075625e-06,
+      "loss": 0.8088,
+      "step": 5915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2562265396118164,
+      "learning_rate": 1.2660114768685139e-06,
+      "loss": 0.8472,
+      "step": 5916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2565608024597168,
+      "learning_rate": 1.2646003012386476e-06,
+      "loss": 0.737,
+      "step": 5917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3443803787231445,
+      "learning_rate": 1.2631897986722353e-06,
+      "loss": 0.8264,
+      "step": 5918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2493022680282593,
+      "learning_rate": 1.2617799694234273e-06,
+      "loss": 0.7973,
+      "step": 5919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.285093903541565,
+      "learning_rate": 1.2603708137462567e-06,
+      "loss": 0.8114,
+      "step": 5920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2541300058364868,
+      "learning_rate": 1.2589623318946302e-06,
+      "loss": 0.7927,
+      "step": 5921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2733638286590576,
+      "learning_rate": 1.2575545241223353e-06,
+      "loss": 0.7936,
+      "step": 5922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2028915882110596,
+      "learning_rate": 1.256147390683038e-06,
+      "loss": 0.7131,
+      "step": 5923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3083869218826294,
+      "learning_rate": 1.2547409318302856e-06,
+      "loss": 0.8243,
+      "step": 5924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6299377679824829,
+      "learning_rate": 1.2533351478174982e-06,
+      "loss": 0.9865,
+      "step": 5925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.257972240447998,
+      "learning_rate": 1.2519300388979777e-06,
+      "loss": 0.7799,
+      "step": 5926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3071273565292358,
+      "learning_rate": 1.2505256053249032e-06,
+      "loss": 0.8013,
+      "step": 5927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6090672016143799,
+      "learning_rate": 1.2491218473513344e-06,
+      "loss": 1.0126,
+      "step": 5928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2612559795379639,
+      "learning_rate": 1.247718765230207e-06,
+      "loss": 0.7708,
+      "step": 5929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.575807511806488,
+      "learning_rate": 1.246316359214334e-06,
+      "loss": 1.0011,
+      "step": 5930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5792487859725952,
+      "learning_rate": 1.2449146295564091e-06,
+      "loss": 1.0172,
+      "step": 5931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5902169346809387,
+      "learning_rate": 1.2435135765090045e-06,
+      "loss": 1.0031,
+      "step": 5932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.214338779449463,
+      "learning_rate": 1.2421132003245684e-06,
+      "loss": 0.7941,
+      "step": 5933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2916675806045532,
+      "learning_rate": 1.2407135012554272e-06,
+      "loss": 0.7785,
+      "step": 5934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2764406204223633,
+      "learning_rate": 1.2393144795537849e-06,
+      "loss": 0.7588,
+      "step": 5935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2577166557312012,
+      "learning_rate": 1.2379161354717262e-06,
+      "loss": 0.7821,
+      "step": 5936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5135849714279175,
+      "learning_rate": 1.236518469261212e-06,
+      "loss": 0.8108,
+      "step": 5937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5820087194442749,
+      "learning_rate": 1.2351214811740781e-06,
+      "loss": 1.0083,
+      "step": 5938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2175612449645996,
+      "learning_rate": 1.2337251714620434e-06,
+      "loss": 0.8194,
+      "step": 5939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2463730573654175,
+      "learning_rate": 1.2323295403767027e-06,
+      "loss": 0.7921,
+      "step": 5940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2590471506118774,
+      "learning_rate": 1.2309345881695273e-06,
+      "loss": 0.8679,
+      "step": 5941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3407701253890991,
+      "learning_rate": 1.2295403150918667e-06,
+      "loss": 0.8538,
+      "step": 5942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.270958662033081,
+      "learning_rate": 1.2281467213949466e-06,
+      "loss": 0.86,
+      "step": 5943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6035025119781494,
+      "learning_rate": 1.2267538073298745e-06,
+      "loss": 1.0166,
+      "step": 5944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5925148129463196,
+      "learning_rate": 1.2253615731476315e-06,
+      "loss": 1.0297,
+      "step": 5945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3101195096969604,
+      "learning_rate": 1.223970019099076e-06,
+      "loss": 0.8296,
+      "step": 5946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.255118727684021,
+      "learning_rate": 1.2225791454349472e-06,
+      "loss": 0.7924,
+      "step": 5947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1727513074874878,
+      "learning_rate": 1.2211889524058611e-06,
+      "loss": 0.7658,
+      "step": 5948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5703648328781128,
+      "learning_rate": 1.2197994402623081e-06,
+      "loss": 1.0226,
+      "step": 5949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2790522575378418,
+      "learning_rate": 1.2184106092546578e-06,
+      "loss": 0.8211,
+      "step": 5950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.583304226398468,
+      "learning_rate": 1.2170224596331548e-06,
+      "loss": 1.0299,
+      "step": 5951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3040975332260132,
+      "learning_rate": 1.215634991647927e-06,
+      "loss": 0.8384,
+      "step": 5952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.234650731086731,
+      "learning_rate": 1.2142482055489735e-06,
+      "loss": 0.7516,
+      "step": 5953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2862298488616943,
+      "learning_rate": 1.2128621015861709e-06,
+      "loss": 0.8097,
+      "step": 5954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.228794813156128,
+      "learning_rate": 1.2114766800092764e-06,
+      "loss": 0.7802,
+      "step": 5955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5931674838066101,
+      "learning_rate": 1.2100919410679235e-06,
+      "loss": 1.0184,
+      "step": 5956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.576276957988739,
+      "learning_rate": 1.2087078850116201e-06,
+      "loss": 1.0342,
+      "step": 5957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.223283052444458,
+      "learning_rate": 1.2073245120897514e-06,
+      "loss": 0.8154,
+      "step": 5958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5831313729286194,
+      "learning_rate": 1.2059418225515828e-06,
+      "loss": 0.9858,
+      "step": 5959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5828909277915955,
+      "learning_rate": 1.2045598166462535e-06,
+      "loss": 1.0429,
+      "step": 5960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2111080884933472,
+      "learning_rate": 1.2031784946227804e-06,
+      "loss": 0.8165,
+      "step": 5961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.252633810043335,
+      "learning_rate": 1.201797856730056e-06,
+      "loss": 0.783,
+      "step": 5962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.588675320148468,
+      "learning_rate": 1.2004179032168512e-06,
+      "loss": 1.01,
+      "step": 5963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1973761320114136,
+      "learning_rate": 1.1990386343318155e-06,
+      "loss": 0.7561,
+      "step": 5964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2723519802093506,
+      "learning_rate": 1.1976600503234708e-06,
+      "loss": 0.8004,
+      "step": 5965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3010308742523193,
+      "learning_rate": 1.1962821514402156e-06,
+      "loss": 0.7928,
+      "step": 5966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2409117221832275,
+      "learning_rate": 1.1949049379303306e-06,
+      "loss": 0.7657,
+      "step": 5967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5833601355552673,
+      "learning_rate": 1.193528410041967e-06,
+      "loss": 1.0067,
+      "step": 5968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2770633697509766,
+      "learning_rate": 1.1921525680231544e-06,
+      "loss": 0.808,
+      "step": 5969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5942878127098083,
+      "learning_rate": 1.1907774121217986e-06,
+      "loss": 1.0155,
+      "step": 5970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2873932123184204,
+      "learning_rate": 1.189402942585683e-06,
+      "loss": 0.8322,
+      "step": 5971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3485709428787231,
+      "learning_rate": 1.1880291596624682e-06,
+      "loss": 0.7924,
+      "step": 5972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5743290781974792,
+      "learning_rate": 1.1866560635996883e-06,
+      "loss": 1.007,
+      "step": 5973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.285155177116394,
+      "learning_rate": 1.185283654644752e-06,
+      "loss": 0.8019,
+      "step": 5974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2601318359375,
+      "learning_rate": 1.1839119330449516e-06,
+      "loss": 0.7797,
+      "step": 5975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2400946617126465,
+      "learning_rate": 1.182540899047449e-06,
+      "loss": 0.7541,
+      "step": 5976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.208492398262024,
+      "learning_rate": 1.1811705528992834e-06,
+      "loss": 0.7309,
+      "step": 5977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.231923222541809,
+      "learning_rate": 1.1798008948473698e-06,
+      "loss": 0.7802,
+      "step": 5978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2420071363449097,
+      "learning_rate": 1.1784319251385018e-06,
+      "loss": 0.7615,
+      "step": 5979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2130537033081055,
+      "learning_rate": 1.177063644019349e-06,
+      "loss": 0.8361,
+      "step": 5980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5796589255332947,
+      "learning_rate": 1.175696051736453e-06,
+      "loss": 0.9948,
+      "step": 5981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1955955028533936,
+      "learning_rate": 1.1743291485362328e-06,
+      "loss": 0.8037,
+      "step": 5982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3140904903411865,
+      "learning_rate": 1.1729629346649868e-06,
+      "loss": 0.7782,
+      "step": 5983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.212172269821167,
+      "learning_rate": 1.1715974103688855e-06,
+      "loss": 0.7905,
+      "step": 5984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.221090316772461,
+      "learning_rate": 1.1702325758939748e-06,
+      "loss": 0.791,
+      "step": 5985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5936660170555115,
+      "learning_rate": 1.168868431486177e-06,
+      "loss": 0.9976,
+      "step": 5986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2646692991256714,
+      "learning_rate": 1.1675049773912922e-06,
+      "loss": 0.7969,
+      "step": 5987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57491534948349,
+      "learning_rate": 1.1661422138549956e-06,
+      "loss": 0.9961,
+      "step": 5988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2211270332336426,
+      "learning_rate": 1.1647801411228354e-06,
+      "loss": 0.8114,
+      "step": 5989
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2217555046081543,
+      "learning_rate": 1.1634187594402358e-06,
+      "loss": 0.7831,
+      "step": 5990
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.240155577659607,
+      "learning_rate": 1.1620580690525003e-06,
+      "loss": 0.7722,
+      "step": 5991
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3126484155654907,
+      "learning_rate": 1.1606980702048032e-06,
+      "loss": 0.8499,
+      "step": 5992
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2771133184432983,
+      "learning_rate": 1.1593387631421948e-06,
+      "loss": 0.797,
+      "step": 5993
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1999667882919312,
+      "learning_rate": 1.1579801481096054e-06,
+      "loss": 0.7366,
+      "step": 5994
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5712062120437622,
+      "learning_rate": 1.1566222253518339e-06,
+      "loss": 1.0197,
+      "step": 5995
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5955098867416382,
+      "learning_rate": 1.1552649951135608e-06,
+      "loss": 0.9971,
+      "step": 5996
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5807937383651733,
+      "learning_rate": 1.1539084576393355e-06,
+      "loss": 0.9957,
+      "step": 5997
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.232595682144165,
+      "learning_rate": 1.1525526131735892e-06,
+      "loss": 0.7747,
+      "step": 5998
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5922974348068237,
+      "learning_rate": 1.1511974619606238e-06,
+      "loss": 1.0254,
+      "step": 5999
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.288550853729248,
+      "learning_rate": 1.1498430042446163e-06,
+      "loss": 0.8029,
+      "step": 6000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2212443351745605,
+      "learning_rate": 1.148489240269619e-06,
+      "loss": 0.7961,
+      "step": 6001
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3297737836837769,
+      "learning_rate": 1.1471361702795636e-06,
+      "loss": 0.7996,
+      "step": 6002
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2421727180480957,
+      "learning_rate": 1.1457837945182493e-06,
+      "loss": 0.815,
+      "step": 6003
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3184300661087036,
+      "learning_rate": 1.1444321132293574e-06,
+      "loss": 0.8316,
+      "step": 6004
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3305341005325317,
+      "learning_rate": 1.1430811266564372e-06,
+      "loss": 0.8174,
+      "step": 6005
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.27985417842865,
+      "learning_rate": 1.1417308350429208e-06,
+      "loss": 0.8088,
+      "step": 6006
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1942245960235596,
+      "learning_rate": 1.1403812386321073e-06,
+      "loss": 0.744,
+      "step": 6007
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2698382139205933,
+      "learning_rate": 1.1390323376671751e-06,
+      "loss": 0.8688,
+      "step": 6008
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.204166293144226,
+      "learning_rate": 1.1376841323911752e-06,
+      "loss": 0.8099,
+      "step": 6009
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2373119592666626,
+      "learning_rate": 1.1363366230470357e-06,
+      "loss": 0.819,
+      "step": 6010
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3028607368469238,
+      "learning_rate": 1.1349898098775553e-06,
+      "loss": 0.8118,
+      "step": 6011
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6215481162071228,
+      "learning_rate": 1.1336436931254124e-06,
+      "loss": 1.0142,
+      "step": 6012
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5933640003204346,
+      "learning_rate": 1.1322982730331555e-06,
+      "loss": 1.0217,
+      "step": 6013
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6034465432167053,
+      "learning_rate": 1.1309535498432105e-06,
+      "loss": 1.0161,
+      "step": 6014
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2427589893341064,
+      "learning_rate": 1.1296095237978761e-06,
+      "loss": 0.8318,
+      "step": 6015
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5875406861305237,
+      "learning_rate": 1.1282661951393254e-06,
+      "loss": 1.0148,
+      "step": 6016
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2812496423721313,
+      "learning_rate": 1.1269235641096049e-06,
+      "loss": 0.832,
+      "step": 6017
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58179771900177,
+      "learning_rate": 1.1255816309506396e-06,
+      "loss": 1.0211,
+      "step": 6018
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.244803547859192,
+      "learning_rate": 1.124240395904223e-06,
+      "loss": 0.7996,
+      "step": 6019
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1994708776474,
+      "learning_rate": 1.1228998592120294e-06,
+      "loss": 0.7829,
+      "step": 6020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5971725583076477,
+      "learning_rate": 1.1215600211155992e-06,
+      "loss": 1.0054,
+      "step": 6021
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.212875485420227,
+      "learning_rate": 1.1202208818563549e-06,
+      "loss": 0.8747,
+      "step": 6022
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.276401400566101,
+      "learning_rate": 1.1188824416755883e-06,
+      "loss": 0.8119,
+      "step": 6023
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2748565673828125,
+      "learning_rate": 1.1175447008144658e-06,
+      "loss": 0.8504,
+      "step": 6024
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2367961406707764,
+      "learning_rate": 1.1162076595140276e-06,
+      "loss": 0.7906,
+      "step": 6025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6171417832374573,
+      "learning_rate": 1.114871318015191e-06,
+      "loss": 1.0099,
+      "step": 6026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2998625040054321,
+      "learning_rate": 1.1135356765587424e-06,
+      "loss": 0.8156,
+      "step": 6027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5786226987838745,
+      "learning_rate": 1.1122007353853476e-06,
+      "loss": 0.9948,
+      "step": 6028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2583496570587158,
+      "learning_rate": 1.110866494735539e-06,
+      "loss": 0.814,
+      "step": 6029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2683438062667847,
+      "learning_rate": 1.1095329548497314e-06,
+      "loss": 0.8715,
+      "step": 6030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2126412391662598,
+      "learning_rate": 1.1082001159682064e-06,
+      "loss": 0.8195,
+      "step": 6031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2381346225738525,
+      "learning_rate": 1.1068679783311205e-06,
+      "loss": 0.7774,
+      "step": 6032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.330183744430542,
+      "learning_rate": 1.105536542178508e-06,
+      "loss": 0.8651,
+      "step": 6033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2296271324157715,
+      "learning_rate": 1.1042058077502727e-06,
+      "loss": 0.8485,
+      "step": 6034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5807052850723267,
+      "learning_rate": 1.1028757752861912e-06,
+      "loss": 1.0145,
+      "step": 6035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5945307612419128,
+      "learning_rate": 1.101546445025919e-06,
+      "loss": 1.0168,
+      "step": 6036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2705967426300049,
+      "learning_rate": 1.1002178172089783e-06,
+      "loss": 0.8117,
+      "step": 6037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.242220401763916,
+      "learning_rate": 1.0988898920747716e-06,
+      "loss": 0.7915,
+      "step": 6038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.581731915473938,
+      "learning_rate": 1.0975626698625685e-06,
+      "loss": 1.0016,
+      "step": 6039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2774293422698975,
+      "learning_rate": 1.096236150811515e-06,
+      "loss": 0.8096,
+      "step": 6040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5849329829216003,
+      "learning_rate": 1.0949103351606315e-06,
+      "loss": 1.0234,
+      "step": 6041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3278043270111084,
+      "learning_rate": 1.0935852231488092e-06,
+      "loss": 0.7865,
+      "step": 6042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1791574954986572,
+      "learning_rate": 1.0922608150148129e-06,
+      "loss": 0.7767,
+      "step": 6043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2430070638656616,
+      "learning_rate": 1.0909371109972832e-06,
+      "loss": 0.7697,
+      "step": 6044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2039978504180908,
+      "learning_rate": 1.0896141113347298e-06,
+      "loss": 0.7944,
+      "step": 6045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1880881786346436,
+      "learning_rate": 1.0882918162655397e-06,
+      "loss": 0.7612,
+      "step": 6046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2743804454803467,
+      "learning_rate": 1.0869702260279696e-06,
+      "loss": 0.8421,
+      "step": 6047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.286025881767273,
+      "learning_rate": 1.0856493408601487e-06,
+      "loss": 0.8623,
+      "step": 6048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5945095419883728,
+      "learning_rate": 1.0843291610000844e-06,
+      "loss": 1.0364,
+      "step": 6049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3029717206954956,
+      "learning_rate": 1.0830096866856516e-06,
+      "loss": 0.8414,
+      "step": 6050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5839473009109497,
+      "learning_rate": 1.0816909181545988e-06,
+      "loss": 1.0076,
+      "step": 6051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.160442590713501,
+      "learning_rate": 1.080372855644551e-06,
+      "loss": 0.737,
+      "step": 6052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2750017642974854,
+      "learning_rate": 1.0790554993930002e-06,
+      "loss": 0.8288,
+      "step": 6053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5835795998573303,
+      "learning_rate": 1.077738849637318e-06,
+      "loss": 1.0116,
+      "step": 6054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5882957577705383,
+      "learning_rate": 1.0764229066147442e-06,
+      "loss": 0.985,
+      "step": 6055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2819191217422485,
+      "learning_rate": 1.0751076705623892e-06,
+      "loss": 0.8536,
+      "step": 6056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2858850955963135,
+      "learning_rate": 1.0737931417172425e-06,
+      "loss": 0.7958,
+      "step": 6057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.236484408378601,
+      "learning_rate": 1.0724793203161615e-06,
+      "loss": 0.7861,
+      "step": 6058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.269472360610962,
+      "learning_rate": 1.071166206595876e-06,
+      "loss": 0.8575,
+      "step": 6059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.366142749786377,
+      "learning_rate": 1.0698538007929914e-06,
+      "loss": 0.7981,
+      "step": 6060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2440540790557861,
+      "learning_rate": 1.0685421031439818e-06,
+      "loss": 0.8356,
+      "step": 6061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2964932918548584,
+      "learning_rate": 1.0672311138851982e-06,
+      "loss": 0.8264,
+      "step": 6062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5953989624977112,
+      "learning_rate": 1.0659208332528587e-06,
+      "loss": 1.0011,
+      "step": 6063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.347146987915039,
+      "learning_rate": 1.0646112614830589e-06,
+      "loss": 0.7816,
+      "step": 6064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.208044409751892,
+      "learning_rate": 1.0633023988117624e-06,
+      "loss": 0.7659,
+      "step": 6065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2711378335952759,
+      "learning_rate": 1.0619942454748072e-06,
+      "loss": 0.7717,
+      "step": 6066
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2445170879364014,
+      "learning_rate": 1.060686801707902e-06,
+      "loss": 0.819,
+      "step": 6067
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6063838005065918,
+      "learning_rate": 1.0593800677466316e-06,
+      "loss": 1.0016,
+      "step": 6068
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3063539266586304,
+      "learning_rate": 1.0580740438264459e-06,
+      "loss": 0.77,
+      "step": 6069
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2448906898498535,
+      "learning_rate": 1.0567687301826752e-06,
+      "loss": 0.8115,
+      "step": 6070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2841780185699463,
+      "learning_rate": 1.055464127050514e-06,
+      "loss": 0.7934,
+      "step": 6071
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.215964436531067,
+      "learning_rate": 1.054160234665036e-06,
+      "loss": 0.7982,
+      "step": 6072
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2830119132995605,
+      "learning_rate": 1.0528570532611804e-06,
+      "loss": 0.7944,
+      "step": 6073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2512938976287842,
+      "learning_rate": 1.0515545830737627e-06,
+      "loss": 0.7484,
+      "step": 6074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5936152935028076,
+      "learning_rate": 1.050252824337466e-06,
+      "loss": 1.0024,
+      "step": 6075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2580434083938599,
+      "learning_rate": 1.0489517772868519e-06,
+      "loss": 0.81,
+      "step": 6076
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5780141353607178,
+      "learning_rate": 1.0476514421563455e-06,
+      "loss": 1.0096,
+      "step": 6077
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5929876565933228,
+      "learning_rate": 1.0463518191802519e-06,
+      "loss": 1.0219,
+      "step": 6078
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2443615198135376,
+      "learning_rate": 1.0450529085927401e-06,
+      "loss": 0.7924,
+      "step": 6079
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2446998357772827,
+      "learning_rate": 1.0437547106278578e-06,
+      "loss": 0.7783,
+      "step": 6080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2346055507659912,
+      "learning_rate": 1.0424572255195193e-06,
+      "loss": 0.8051,
+      "step": 6081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2157478332519531,
+      "learning_rate": 1.0411604535015118e-06,
+      "loss": 0.8167,
+      "step": 6082
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.220712661743164,
+      "learning_rate": 1.0398643948074939e-06,
+      "loss": 0.8089,
+      "step": 6083
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.265342354774475,
+      "learning_rate": 1.0385690496709983e-06,
+      "loss": 0.8299,
+      "step": 6084
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2989001274108887,
+      "learning_rate": 1.037274418325424e-06,
+      "loss": 0.8275,
+      "step": 6085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2374093532562256,
+      "learning_rate": 1.0359805010040476e-06,
+      "loss": 0.8329,
+      "step": 6086
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5872758030891418,
+      "learning_rate": 1.0346872979400103e-06,
+      "loss": 1.01,
+      "step": 6087
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1964726448059082,
+      "learning_rate": 1.033394809366331e-06,
+      "loss": 0.7518,
+      "step": 6088
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5854173898696899,
+      "learning_rate": 1.0321030355158957e-06,
+      "loss": 1.0088,
+      "step": 6089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5973572134971619,
+      "learning_rate": 1.030811976621463e-06,
+      "loss": 1.0155,
+      "step": 6090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6032508015632629,
+      "learning_rate": 1.029521632915661e-06,
+      "loss": 1.032,
+      "step": 6091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2741209268569946,
+      "learning_rate": 1.028232004630993e-06,
+      "loss": 0.8117,
+      "step": 6092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2327616214752197,
+      "learning_rate": 1.026943091999828e-06,
+      "loss": 0.7575,
+      "step": 6093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1946091651916504,
+      "learning_rate": 1.025654895254412e-06,
+      "loss": 0.7932,
+      "step": 6094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2415610551834106,
+      "learning_rate": 1.024367414626856e-06,
+      "loss": 0.7771,
+      "step": 6095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2085832357406616,
+      "learning_rate": 1.0230806503491481e-06,
+      "loss": 0.8462,
+      "step": 6096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.584795355796814,
+      "learning_rate": 1.0217946026531416e-06,
+      "loss": 1.033,
+      "step": 6097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.231074571609497,
+      "learning_rate": 1.020509271770564e-06,
+      "loss": 0.8152,
+      "step": 6098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3031768798828125,
+      "learning_rate": 1.0192246579330111e-06,
+      "loss": 0.8337,
+      "step": 6099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6000784635543823,
+      "learning_rate": 1.017940761371954e-06,
+      "loss": 1.0247,
+      "step": 6100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5928666591644287,
+      "learning_rate": 1.0166575823187292e-06,
+      "loss": 0.9969,
+      "step": 6101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.212632179260254,
+      "learning_rate": 1.015375121004547e-06,
+      "loss": 0.8233,
+      "step": 6102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2273242473602295,
+      "learning_rate": 1.0140933776604906e-06,
+      "loss": 0.7969,
+      "step": 6103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2396339178085327,
+      "learning_rate": 1.0128123525175087e-06,
+      "loss": 0.7909,
+      "step": 6104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1850703954696655,
+      "learning_rate": 1.0115320458064237e-06,
+      "loss": 0.7543,
+      "step": 6105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2424795627593994,
+      "learning_rate": 1.0102524577579258e-06,
+      "loss": 0.7908,
+      "step": 6106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2076908349990845,
+      "learning_rate": 1.0089735886025808e-06,
+      "loss": 0.7736,
+      "step": 6107
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.200533151626587,
+      "learning_rate": 1.0076954385708205e-06,
+      "loss": 0.7583,
+      "step": 6108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2320499420166016,
+      "learning_rate": 1.006418007892947e-06,
+      "loss": 0.819,
+      "step": 6109
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1610487699508667,
+      "learning_rate": 1.0051412967991358e-06,
+      "loss": 0.7633,
+      "step": 6110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.26139235496521,
+      "learning_rate": 1.0038653055194324e-06,
+      "loss": 0.8392,
+      "step": 6111
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2563556432724,
+      "learning_rate": 1.0025900342837503e-06,
+      "loss": 0.7982,
+      "step": 6112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6050691604614258,
+      "learning_rate": 1.0013154833218746e-06,
+      "loss": 0.9929,
+      "step": 6113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2583298683166504,
+      "learning_rate": 1.0000416528634587e-06,
+      "loss": 0.8539,
+      "step": 6114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.254934549331665,
+      "learning_rate": 9.987685431380306e-07,
+      "loss": 0.8106,
+      "step": 6115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.307407021522522,
+      "learning_rate": 9.974961543749845e-07,
+      "loss": 0.8492,
+      "step": 6116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5712260603904724,
+      "learning_rate": 9.962244868035848e-07,
+      "loss": 1.011,
+      "step": 6117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.570229709148407,
+      "learning_rate": 9.949535406529682e-07,
+      "loss": 1.0088,
+      "step": 6118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.278688669204712,
+      "learning_rate": 9.936833161521415e-07,
+      "loss": 0.8066,
+      "step": 6119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5846341252326965,
+      "learning_rate": 9.924138135299792e-07,
+      "loss": 0.997,
+      "step": 6120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5785278677940369,
+      "learning_rate": 9.911450330152266e-07,
+      "loss": 1.006,
+      "step": 6121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.237647294998169,
+      "learning_rate": 9.898769748364977e-07,
+      "loss": 0.7717,
+      "step": 6122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2660051584243774,
+      "learning_rate": 9.88609639222281e-07,
+      "loss": 0.8166,
+      "step": 6123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5702605843544006,
+      "learning_rate": 9.873430264009294e-07,
+      "loss": 0.9926,
+      "step": 6124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3334238529205322,
+      "learning_rate": 9.86077136600666e-07,
+      "loss": 0.8333,
+      "step": 6125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2775096893310547,
+      "learning_rate": 9.84811970049588e-07,
+      "loss": 0.7943,
+      "step": 6126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2133089303970337,
+      "learning_rate": 9.8354752697566e-07,
+      "loss": 0.7303,
+      "step": 6127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2657676935195923,
+      "learning_rate": 9.822838076067144e-07,
+      "loss": 0.7702,
+      "step": 6128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2640204429626465,
+      "learning_rate": 9.810208121704551e-07,
+      "loss": 0.7741,
+      "step": 6129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5889837741851807,
+      "learning_rate": 9.797585408944537e-07,
+      "loss": 1.0062,
+      "step": 6130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2241789102554321,
+      "learning_rate": 9.78496994006155e-07,
+      "loss": 0.8315,
+      "step": 6131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2564955949783325,
+      "learning_rate": 9.772361717328704e-07,
+      "loss": 0.7786,
+      "step": 6132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1880179643630981,
+      "learning_rate": 9.759760743017789e-07,
+      "loss": 0.7727,
+      "step": 6133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.224674105644226,
+      "learning_rate": 9.747167019399335e-07,
+      "loss": 0.8317,
+      "step": 6134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5851430296897888,
+      "learning_rate": 9.73458054874255e-07,
+      "loss": 1.0111,
+      "step": 6135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2341150045394897,
+      "learning_rate": 9.722001333315329e-07,
+      "loss": 0.823,
+      "step": 6136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2852760553359985,
+      "learning_rate": 9.709429375384227e-07,
+      "loss": 0.8208,
+      "step": 6137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.189984679222107,
+      "learning_rate": 9.696864677214562e-07,
+      "loss": 0.7601,
+      "step": 6138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.268266201019287,
+      "learning_rate": 9.684307241070296e-07,
+      "loss": 0.7767,
+      "step": 6139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5692620873451233,
+      "learning_rate": 9.671757069214084e-07,
+      "loss": 1.0152,
+      "step": 6140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2164576053619385,
+      "learning_rate": 9.659214163907265e-07,
+      "loss": 0.8125,
+      "step": 6141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5992798209190369,
+      "learning_rate": 9.646678527409909e-07,
+      "loss": 1.0023,
+      "step": 6142
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2256699800491333,
+      "learning_rate": 9.634150161980755e-07,
+      "loss": 0.7956,
+      "step": 6143
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2425835132598877,
+      "learning_rate": 9.621629069877215e-07,
+      "loss": 0.7555,
+      "step": 6144
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2984391450881958,
+      "learning_rate": 9.60911525335539e-07,
+      "loss": 0.7678,
+      "step": 6145
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6005594730377197,
+      "learning_rate": 9.59660871467012e-07,
+      "loss": 1.0049,
+      "step": 6146
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2258331775665283,
+      "learning_rate": 9.58410945607487e-07,
+      "loss": 0.7911,
+      "step": 6147
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5877885222434998,
+      "learning_rate": 9.571617479821827e-07,
+      "loss": 1.0501,
+      "step": 6148
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5734853744506836,
+      "learning_rate": 9.559132788161852e-07,
+      "loss": 1.0013,
+      "step": 6149
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.296999454498291,
+      "learning_rate": 9.546655383344506e-07,
+      "loss": 0.7762,
+      "step": 6150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2963061332702637,
+      "learning_rate": 9.53418526761804e-07,
+      "loss": 0.7848,
+      "step": 6151
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.320507526397705,
+      "learning_rate": 9.521722443229381e-07,
+      "loss": 0.8399,
+      "step": 6152
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.282181739807129,
+      "learning_rate": 9.509266912424131e-07,
+      "loss": 0.8321,
+      "step": 6153
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2796381711959839,
+      "learning_rate": 9.496818677446612e-07,
+      "loss": 0.8381,
+      "step": 6154
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2063891887664795,
+      "learning_rate": 9.484377740539796e-07,
+      "loss": 0.7745,
+      "step": 6155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3401775360107422,
+      "learning_rate": 9.471944103945363e-07,
+      "loss": 0.8007,
+      "step": 6156
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2246983051300049,
+      "learning_rate": 9.459517769903648e-07,
+      "loss": 0.7951,
+      "step": 6157
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2292184829711914,
+      "learning_rate": 9.44709874065371e-07,
+      "loss": 0.7912,
+      "step": 6158
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2355045080184937,
+      "learning_rate": 9.434687018433286e-07,
+      "loss": 0.786,
+      "step": 6159
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2313754558563232,
+      "learning_rate": 9.422282605478767e-07,
+      "loss": 0.7517,
+      "step": 6160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2174160480499268,
+      "learning_rate": 9.409885504025234e-07,
+      "loss": 0.8454,
+      "step": 6161
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.273503065109253,
+      "learning_rate": 9.397495716306481e-07,
+      "loss": 0.8122,
+      "step": 6162
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.212339162826538,
+      "learning_rate": 9.385113244554961e-07,
+      "loss": 0.8114,
+      "step": 6163
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2685424089431763,
+      "learning_rate": 9.372738091001804e-07,
+      "loss": 0.7951,
+      "step": 6164
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2728896141052246,
+      "learning_rate": 9.360370257876811e-07,
+      "loss": 0.8602,
+      "step": 6165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2297563552856445,
+      "learning_rate": 9.348009747408505e-07,
+      "loss": 0.8082,
+      "step": 6166
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2900599241256714,
+      "learning_rate": 9.335656561824069e-07,
+      "loss": 0.8617,
+      "step": 6167
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3722747564315796,
+      "learning_rate": 9.323310703349359e-07,
+      "loss": 0.8912,
+      "step": 6168
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3060648441314697,
+      "learning_rate": 9.310972174208899e-07,
+      "loss": 0.8037,
+      "step": 6169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5164892673492432,
+      "learning_rate": 9.298640976625933e-07,
+      "loss": 0.8273,
+      "step": 6170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.619987964630127,
+      "learning_rate": 9.286317112822346e-07,
+      "loss": 1.0205,
+      "step": 6171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.367548942565918,
+      "learning_rate": 9.2740005850187e-07,
+      "loss": 0.8599,
+      "step": 6172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2406002283096313,
+      "learning_rate": 9.261691395434286e-07,
+      "loss": 0.7784,
+      "step": 6173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4373557567596436,
+      "learning_rate": 9.249389546286996e-07,
+      "loss": 0.872,
+      "step": 6174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5670666098594666,
+      "learning_rate": 9.237095039793481e-07,
+      "loss": 1.0211,
+      "step": 6175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5841541290283203,
+      "learning_rate": 9.224807878168989e-07,
+      "loss": 1.0093,
+      "step": 6176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.571835994720459,
+      "learning_rate": 9.212528063627518e-07,
+      "loss": 0.9878,
+      "step": 6177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5866872072219849,
+      "learning_rate": 9.200255598381696e-07,
+      "loss": 1.0265,
+      "step": 6178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2912334203720093,
+      "learning_rate": 9.187990484642834e-07,
+      "loss": 0.7936,
+      "step": 6179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.267980933189392,
+      "learning_rate": 9.175732724620906e-07,
+      "loss": 0.7911,
+      "step": 6180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2093397378921509,
+      "learning_rate": 9.163482320524614e-07,
+      "loss": 0.8121,
+      "step": 6181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2715436220169067,
+      "learning_rate": 9.151239274561269e-07,
+      "loss": 0.8241,
+      "step": 6182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5832650065422058,
+      "learning_rate": 9.139003588936912e-07,
+      "loss": 0.9999,
+      "step": 6183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2180204391479492,
+      "learning_rate": 9.126775265856202e-07,
+      "loss": 0.7805,
+      "step": 6184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.252397060394287,
+      "learning_rate": 9.114554307522527e-07,
+      "loss": 0.8081,
+      "step": 6185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.279968023300171,
+      "learning_rate": 9.102340716137914e-07,
+      "loss": 0.7911,
+      "step": 6186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.274613857269287,
+      "learning_rate": 9.090134493903063e-07,
+      "loss": 0.7929,
+      "step": 6187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6087394952774048,
+      "learning_rate": 9.077935643017349e-07,
+      "loss": 1.0146,
+      "step": 6188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2911863327026367,
+      "learning_rate": 9.065744165678847e-07,
+      "loss": 0.8211,
+      "step": 6189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5942144989967346,
+      "learning_rate": 9.053560064084254e-07,
+      "loss": 0.9981,
+      "step": 6190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1289188861846924,
+      "learning_rate": 9.041383340428983e-07,
+      "loss": 0.7896,
+      "step": 6191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2441610097885132,
+      "learning_rate": 9.029213996907082e-07,
+      "loss": 0.7729,
+      "step": 6192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2412383556365967,
+      "learning_rate": 9.017052035711305e-07,
+      "loss": 0.7781,
+      "step": 6193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5861518979072571,
+      "learning_rate": 9.004897459033046e-07,
+      "loss": 1.0056,
+      "step": 6194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.224413275718689,
+      "learning_rate": 8.992750269062378e-07,
+      "loss": 0.8074,
+      "step": 6195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2858810424804688,
+      "learning_rate": 8.980610467988033e-07,
+      "loss": 0.7412,
+      "step": 6196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2288037538528442,
+      "learning_rate": 8.968478057997448e-07,
+      "loss": 0.7351,
+      "step": 6197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.605623185634613,
+      "learning_rate": 8.956353041276678e-07,
+      "loss": 0.9973,
+      "step": 6198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1828571557998657,
+      "learning_rate": 8.944235420010489e-07,
+      "loss": 0.7755,
+      "step": 6199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.250314712524414,
+      "learning_rate": 8.93212519638228e-07,
+      "loss": 0.7959,
+      "step": 6200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6258211731910706,
+      "learning_rate": 8.920022372574155e-07,
+      "loss": 1.0078,
+      "step": 6201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2214298248291016,
+      "learning_rate": 8.907926950766848e-07,
+      "loss": 0.853,
+      "step": 6202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6049996018409729,
+      "learning_rate": 8.895838933139773e-07,
+      "loss": 0.9937,
+      "step": 6203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2733221054077148,
+      "learning_rate": 8.883758321871005e-07,
+      "loss": 0.8127,
+      "step": 6204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5780828595161438,
+      "learning_rate": 8.871685119137314e-07,
+      "loss": 1.0001,
+      "step": 6205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2490872144699097,
+      "learning_rate": 8.859619327114078e-07,
+      "loss": 0.7753,
+      "step": 6206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2618167400360107,
+      "learning_rate": 8.847560947975409e-07,
+      "loss": 0.8416,
+      "step": 6207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.236948013305664,
+      "learning_rate": 8.835509983894014e-07,
+      "loss": 0.8408,
+      "step": 6208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2391549348831177,
+      "learning_rate": 8.823466437041328e-07,
+      "loss": 0.7681,
+      "step": 6209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2039114236831665,
+      "learning_rate": 8.811430309587404e-07,
+      "loss": 0.7546,
+      "step": 6210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2659142017364502,
+      "learning_rate": 8.799401603700964e-07,
+      "loss": 0.8498,
+      "step": 6211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2209956645965576,
+      "learning_rate": 8.787380321549422e-07,
+      "loss": 0.7738,
+      "step": 6212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.260565161705017,
+      "learning_rate": 8.775366465298829e-07,
+      "loss": 0.8169,
+      "step": 6213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.571347713470459,
+      "learning_rate": 8.763360037113877e-07,
+      "loss": 0.9947,
+      "step": 6214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.203884482383728,
+      "learning_rate": 8.751361039157974e-07,
+      "loss": 0.8363,
+      "step": 6215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2146505117416382,
+      "learning_rate": 8.739369473593162e-07,
+      "loss": 0.8176,
+      "step": 6216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2832385301589966,
+      "learning_rate": 8.727385342580136e-07,
+      "loss": 0.844,
+      "step": 6217
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2620528936386108,
+      "learning_rate": 8.715408648278257e-07,
+      "loss": 0.8264,
+      "step": 6218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.585114598274231,
+      "learning_rate": 8.703439392845536e-07,
+      "loss": 1.0004,
+      "step": 6219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.261293888092041,
+      "learning_rate": 8.691477578438679e-07,
+      "loss": 0.7919,
+      "step": 6220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.326144814491272,
+      "learning_rate": 8.679523207213009e-07,
+      "loss": 0.8128,
+      "step": 6221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.278884768486023,
+      "learning_rate": 8.667576281322521e-07,
+      "loss": 0.8056,
+      "step": 6222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3089770078659058,
+      "learning_rate": 8.655636802919876e-07,
+      "loss": 0.8099,
+      "step": 6223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.221651315689087,
+      "learning_rate": 8.643704774156419e-07,
+      "loss": 0.7564,
+      "step": 6224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.20480477809906,
+      "learning_rate": 8.631780197182093e-07,
+      "loss": 0.8016,
+      "step": 6225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2458906173706055,
+      "learning_rate": 8.619863074145546e-07,
+      "loss": 0.8526,
+      "step": 6226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5976096391677856,
+      "learning_rate": 8.607953407194042e-07,
+      "loss": 1.0079,
+      "step": 6227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2725682258605957,
+      "learning_rate": 8.596051198473559e-07,
+      "loss": 0.8055,
+      "step": 6228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.187304973602295,
+      "learning_rate": 8.584156450128678e-07,
+      "loss": 0.7713,
+      "step": 6229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3987808227539062,
+      "learning_rate": 8.572269164302649e-07,
+      "loss": 0.833,
+      "step": 6230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6111930012702942,
+      "learning_rate": 8.560389343137393e-07,
+      "loss": 1.0295,
+      "step": 6231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.585344672203064,
+      "learning_rate": 8.548516988773492e-07,
+      "loss": 1.0073,
+      "step": 6232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5755967497825623,
+      "learning_rate": 8.536652103350152e-07,
+      "loss": 1.0017,
+      "step": 6233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3182018995285034,
+      "learning_rate": 8.52479468900525e-07,
+      "loss": 0.7555,
+      "step": 6234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2595847845077515,
+      "learning_rate": 8.512944747875307e-07,
+      "loss": 0.8351,
+      "step": 6235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1822136640548706,
+      "learning_rate": 8.50110228209553e-07,
+      "loss": 0.7586,
+      "step": 6236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3117111921310425,
+      "learning_rate": 8.489267293799736e-07,
+      "loss": 0.8316,
+      "step": 6237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2916171550750732,
+      "learning_rate": 8.477439785120407e-07,
+      "loss": 0.8346,
+      "step": 6238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.574785590171814,
+      "learning_rate": 8.465619758188697e-07,
+      "loss": 0.9974,
+      "step": 6239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5654577016830444,
+      "learning_rate": 8.453807215134407e-07,
+      "loss": 1.0135,
+      "step": 6240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2358020544052124,
+      "learning_rate": 8.442002158085977e-07,
+      "loss": 0.8065,
+      "step": 6241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2652349472045898,
+      "learning_rate": 8.430204589170493e-07,
+      "loss": 0.8289,
+      "step": 6242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6084911227226257,
+      "learning_rate": 8.418414510513695e-07,
+      "loss": 1.0342,
+      "step": 6243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6130889058113098,
+      "learning_rate": 8.406631924240005e-07,
+      "loss": 0.9989,
+      "step": 6244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2270317077636719,
+      "learning_rate": 8.394856832472459e-07,
+      "loss": 0.8631,
+      "step": 6245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2400438785552979,
+      "learning_rate": 8.383089237332731e-07,
+      "loss": 0.7943,
+      "step": 6246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.290181279182434,
+      "learning_rate": 8.371329140941192e-07,
+      "loss": 0.7673,
+      "step": 6247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2578372955322266,
+      "learning_rate": 8.359576545416836e-07,
+      "loss": 0.7823,
+      "step": 6248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2201615571975708,
+      "learning_rate": 8.347831452877303e-07,
+      "loss": 0.7513,
+      "step": 6249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2701780796051025,
+      "learning_rate": 8.336093865438872e-07,
+      "loss": 0.8094,
+      "step": 6250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2190841436386108,
+      "learning_rate": 8.324363785216499e-07,
+      "loss": 0.7997,
+      "step": 6251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3259971141815186,
+      "learning_rate": 8.312641214323758e-07,
+      "loss": 0.847,
+      "step": 6252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2341291904449463,
+      "learning_rate": 8.300926154872891e-07,
+      "loss": 0.8157,
+      "step": 6253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5955546498298645,
+      "learning_rate": 8.289218608974759e-07,
+      "loss": 1.0043,
+      "step": 6254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2514991760253906,
+      "learning_rate": 8.277518578738897e-07,
+      "loss": 0.7753,
+      "step": 6255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.174780249595642,
+      "learning_rate": 8.265826066273497e-07,
+      "loss": 0.7983,
+      "step": 6256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.574390709400177,
+      "learning_rate": 8.254141073685357e-07,
+      "loss": 1.0081,
+      "step": 6257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.218487024307251,
+      "learning_rate": 8.242463603079925e-07,
+      "loss": 0.781,
+      "step": 6258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.186623215675354,
+      "learning_rate": 8.230793656561336e-07,
+      "loss": 0.7903,
+      "step": 6259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2622674703598022,
+      "learning_rate": 8.219131236232331e-07,
+      "loss": 0.7933,
+      "step": 6260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1632736921310425,
+      "learning_rate": 8.207476344194299e-07,
+      "loss": 0.7531,
+      "step": 6261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3212931156158447,
+      "learning_rate": 8.195828982547266e-07,
+      "loss": 0.753,
+      "step": 6262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2286514043807983,
+      "learning_rate": 8.184189153389933e-07,
+      "loss": 0.733,
+      "step": 6263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2318545579910278,
+      "learning_rate": 8.172556858819624e-07,
+      "loss": 0.802,
+      "step": 6264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3360915184020996,
+      "learning_rate": 8.160932100932306e-07,
+      "loss": 0.8191,
+      "step": 6265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2181544303894043,
+      "learning_rate": 8.149314881822567e-07,
+      "loss": 0.7665,
+      "step": 6266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2534282207489014,
+      "learning_rate": 8.137705203583685e-07,
+      "loss": 0.8123,
+      "step": 6267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2026405334472656,
+      "learning_rate": 8.126103068307539e-07,
+      "loss": 0.7812,
+      "step": 6268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2148277759552002,
+      "learning_rate": 8.114508478084659e-07,
+      "loss": 0.7902,
+      "step": 6269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1902300119400024,
+      "learning_rate": 8.102921435004202e-07,
+      "loss": 0.8075,
+      "step": 6270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2662516832351685,
+      "learning_rate": 8.091341941154002e-07,
+      "loss": 0.7898,
+      "step": 6271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3033490180969238,
+      "learning_rate": 8.079769998620518e-07,
+      "loss": 0.774,
+      "step": 6272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3062752485275269,
+      "learning_rate": 8.068205609488832e-07,
+      "loss": 0.8772,
+      "step": 6273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2072588205337524,
+      "learning_rate": 8.056648775842662e-07,
+      "loss": 0.7433,
+      "step": 6274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2527278661727905,
+      "learning_rate": 8.045099499764403e-07,
+      "loss": 0.8343,
+      "step": 6275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2933231592178345,
+      "learning_rate": 8.033557783335049e-07,
+      "loss": 0.8452,
+      "step": 6276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2666667699813843,
+      "learning_rate": 8.022023628634245e-07,
+      "loss": 0.8093,
+      "step": 6277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3207511901855469,
+      "learning_rate": 8.010497037740261e-07,
+      "loss": 0.892,
+      "step": 6278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2163548469543457,
+      "learning_rate": 7.99897801273003e-07,
+      "loss": 0.8243,
+      "step": 6279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.577150821685791,
+      "learning_rate": 7.987466555679124e-07,
+      "loss": 1.011,
+      "step": 6280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2965741157531738,
+      "learning_rate": 7.975962668661724e-07,
+      "loss": 0.7878,
+      "step": 6281
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5931695103645325,
+      "learning_rate": 7.964466353750644e-07,
+      "loss": 1.0129,
+      "step": 6282
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2744413614273071,
+      "learning_rate": 7.952977613017371e-07,
+      "loss": 0.8024,
+      "step": 6283
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3084379434585571,
+      "learning_rate": 7.941496448531999e-07,
+      "loss": 0.7955,
+      "step": 6284
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3422857522964478,
+      "learning_rate": 7.930022862363252e-07,
+      "loss": 0.7631,
+      "step": 6285
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2261892557144165,
+      "learning_rate": 7.918556856578513e-07,
+      "loss": 0.7876,
+      "step": 6286
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.20757257938385,
+      "learning_rate": 7.907098433243777e-07,
+      "loss": 0.8076,
+      "step": 6287
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2288343906402588,
+      "learning_rate": 7.895647594423694e-07,
+      "loss": 0.7588,
+      "step": 6288
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2602834701538086,
+      "learning_rate": 7.884204342181512e-07,
+      "loss": 0.8042,
+      "step": 6289
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1864978075027466,
+      "learning_rate": 7.872768678579163e-07,
+      "loss": 0.7939,
+      "step": 6290
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3073698282241821,
+      "learning_rate": 7.861340605677165e-07,
+      "loss": 0.7976,
+      "step": 6291
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6075767278671265,
+      "learning_rate": 7.849920125534694e-07,
+      "loss": 1.0,
+      "step": 6292
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1966538429260254,
+      "learning_rate": 7.838507240209536e-07,
+      "loss": 0.8354,
+      "step": 6293
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2341877222061157,
+      "learning_rate": 7.827101951758148e-07,
+      "loss": 0.8601,
+      "step": 6294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5989691019058228,
+      "learning_rate": 7.815704262235568e-07,
+      "loss": 1.0164,
+      "step": 6295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5800666213035583,
+      "learning_rate": 7.804314173695515e-07,
+      "loss": 1.0109,
+      "step": 6296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.270938754081726,
+      "learning_rate": 7.792931688190287e-07,
+      "loss": 0.8334,
+      "step": 6297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5835976004600525,
+      "learning_rate": 7.781556807770862e-07,
+      "loss": 0.9878,
+      "step": 6298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.28618586063385,
+      "learning_rate": 7.770189534486816e-07,
+      "loss": 0.7856,
+      "step": 6299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5813612341880798,
+      "learning_rate": 7.758829870386364e-07,
+      "loss": 0.9981,
+      "step": 6300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.313490390777588,
+      "learning_rate": 7.74747781751633e-07,
+      "loss": 0.8058,
+      "step": 6301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2794392108917236,
+      "learning_rate": 7.736133377922211e-07,
+      "loss": 0.803,
+      "step": 6302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2763935327529907,
+      "learning_rate": 7.72479655364809e-07,
+      "loss": 0.8183,
+      "step": 6303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3471803665161133,
+      "learning_rate": 7.713467346736709e-07,
+      "loss": 0.8438,
+      "step": 6304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2575628757476807,
+      "learning_rate": 7.702145759229401e-07,
+      "loss": 0.8136,
+      "step": 6305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5761956572532654,
+      "learning_rate": 7.69083179316617e-07,
+      "loss": 1.0221,
+      "step": 6306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5777075290679932,
+      "learning_rate": 7.679525450585618e-07,
+      "loss": 1.0036,
+      "step": 6307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5731942653656006,
+      "learning_rate": 7.66822673352497e-07,
+      "loss": 1.0085,
+      "step": 6308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2268061637878418,
+      "learning_rate": 7.656935644020086e-07,
+      "loss": 0.7607,
+      "step": 6309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2637939453125,
+      "learning_rate": 7.645652184105468e-07,
+      "loss": 0.8405,
+      "step": 6310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.374704122543335,
+      "learning_rate": 7.63437635581421e-07,
+      "loss": 0.753,
+      "step": 6311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2601622343063354,
+      "learning_rate": 7.623108161178067e-07,
+      "loss": 0.7985,
+      "step": 6312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5829699039459229,
+      "learning_rate": 7.611847602227379e-07,
+      "loss": 1.0035,
+      "step": 6313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5969942808151245,
+      "learning_rate": 7.600594680991158e-07,
+      "loss": 0.9939,
+      "step": 6314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2428711652755737,
+      "learning_rate": 7.589349399496992e-07,
+      "loss": 0.8295,
+      "step": 6315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2775238752365112,
+      "learning_rate": 7.578111759771123e-07,
+      "loss": 0.8121,
+      "step": 6316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5781016945838928,
+      "learning_rate": 7.566881763838391e-07,
+      "loss": 1.0044,
+      "step": 6317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.265175461769104,
+      "learning_rate": 7.555659413722294e-07,
+      "loss": 0.7981,
+      "step": 6318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2129803895950317,
+      "learning_rate": 7.544444711444915e-07,
+      "loss": 0.8278,
+      "step": 6319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5794740915298462,
+      "learning_rate": 7.533237659026993e-07,
+      "loss": 1.0099,
+      "step": 6320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2942588329315186,
+      "learning_rate": 7.522038258487858e-07,
+      "loss": 0.886,
+      "step": 6321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.187067985534668,
+      "learning_rate": 7.510846511845493e-07,
+      "loss": 0.7982,
+      "step": 6322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.246727466583252,
+      "learning_rate": 7.499662421116472e-07,
+      "loss": 0.795,
+      "step": 6323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2715502977371216,
+      "learning_rate": 7.488485988315991e-07,
+      "loss": 0.866,
+      "step": 6324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3371453285217285,
+      "learning_rate": 7.477317215457902e-07,
+      "loss": 0.8556,
+      "step": 6325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2703837156295776,
+      "learning_rate": 7.46615610455464e-07,
+      "loss": 0.8065,
+      "step": 6326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2727266550064087,
+      "learning_rate": 7.455002657617255e-07,
+      "loss": 0.849,
+      "step": 6327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.244978666305542,
+      "learning_rate": 7.443856876655464e-07,
+      "loss": 0.8484,
+      "step": 6328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3244071006774902,
+      "learning_rate": 7.432718763677537e-07,
+      "loss": 0.8158,
+      "step": 6329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2870746850967407,
+      "learning_rate": 7.421588320690432e-07,
+      "loss": 0.8474,
+      "step": 6330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5820946097373962,
+      "learning_rate": 7.410465549699663e-07,
+      "loss": 1.0203,
+      "step": 6331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.198525071144104,
+      "learning_rate": 7.399350452709386e-07,
+      "loss": 0.738,
+      "step": 6332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5791910290718079,
+      "learning_rate": 7.388243031722398e-07,
+      "loss": 1.0207,
+      "step": 6333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2877064943313599,
+      "learning_rate": 7.377143288740079e-07,
+      "loss": 0.8041,
+      "step": 6334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5981134176254272,
+      "learning_rate": 7.366051225762428e-07,
+      "loss": 1.0085,
+      "step": 6335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.220349907875061,
+      "learning_rate": 7.354966844788086e-07,
+      "loss": 0.7921,
+      "step": 6336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2169649600982666,
+      "learning_rate": 7.343890147814276e-07,
+      "loss": 0.7881,
+      "step": 6337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5836965441703796,
+      "learning_rate": 7.332821136836877e-07,
+      "loss": 1.0154,
+      "step": 6338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2323511838912964,
+      "learning_rate": 7.321759813850344e-07,
+      "loss": 0.8266,
+      "step": 6339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5827897191047668,
+      "learning_rate": 7.310706180847759e-07,
+      "loss": 1.0121,
+      "step": 6340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2455803155899048,
+      "learning_rate": 7.299660239820839e-07,
+      "loss": 0.8007,
+      "step": 6341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5600418448448181,
+      "learning_rate": 7.288621992759887e-07,
+      "loss": 1.0017,
+      "step": 6342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5908542275428772,
+      "learning_rate": 7.277591441653836e-07,
+      "loss": 0.9987,
+      "step": 6343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.248295783996582,
+      "learning_rate": 7.266568588490214e-07,
+      "loss": 0.8018,
+      "step": 6344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2667956352233887,
+      "learning_rate": 7.255553435255181e-07,
+      "loss": 0.8032,
+      "step": 6345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2768796682357788,
+      "learning_rate": 7.244545983933521e-07,
+      "loss": 0.8476,
+      "step": 6346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3153138160705566,
+      "learning_rate": 7.233546236508599e-07,
+      "loss": 0.8309,
+      "step": 6347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5838391184806824,
+      "learning_rate": 7.222554194962395e-07,
+      "loss": 1.0123,
+      "step": 6348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2192655801773071,
+      "learning_rate": 7.211569861275535e-07,
+      "loss": 0.8011,
+      "step": 6349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.212873935699463,
+      "learning_rate": 7.200593237427223e-07,
+      "loss": 0.7305,
+      "step": 6350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.275606393814087,
+      "learning_rate": 7.189624325395278e-07,
+      "loss": 0.8098,
+      "step": 6351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5584097504615784,
+      "learning_rate": 7.178663127156127e-07,
+      "loss": 0.9947,
+      "step": 6352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5783238410949707,
+      "learning_rate": 7.167709644684823e-07,
+      "loss": 1.0125,
+      "step": 6353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5880258679389954,
+      "learning_rate": 7.156763879955036e-07,
+      "loss": 1.0223,
+      "step": 6354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.207017183303833,
+      "learning_rate": 7.145825834939007e-07,
+      "loss": 0.8435,
+      "step": 6355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.292365550994873,
+      "learning_rate": 7.134895511607632e-07,
+      "loss": 0.7448,
+      "step": 6356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5763599276542664,
+      "learning_rate": 7.123972911930377e-07,
+      "loss": 1.0256,
+      "step": 6357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.228047490119934,
+      "learning_rate": 7.113058037875331e-07,
+      "loss": 0.7833,
+      "step": 6358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2200108766555786,
+      "learning_rate": 7.102150891409187e-07,
+      "loss": 0.8384,
+      "step": 6359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2030166387557983,
+      "learning_rate": 7.091251474497273e-07,
+      "loss": 0.8051,
+      "step": 6360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6076105833053589,
+      "learning_rate": 7.080359789103475e-07,
+      "loss": 1.0251,
+      "step": 6361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.234567642211914,
+      "learning_rate": 7.069475837190338e-07,
+      "loss": 0.8137,
+      "step": 6362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2910735607147217,
+      "learning_rate": 7.058599620718959e-07,
+      "loss": 0.8209,
+      "step": 6363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2850570678710938,
+      "learning_rate": 7.047731141649105e-07,
+      "loss": 0.8276,
+      "step": 6364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3002748489379883,
+      "learning_rate": 7.036870401939094e-07,
+      "loss": 0.8056,
+      "step": 6365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2272353172302246,
+      "learning_rate": 7.026017403545871e-07,
+      "loss": 0.8087,
+      "step": 6366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2820287942886353,
+      "learning_rate": 7.015172148424976e-07,
+      "loss": 0.7978,
+      "step": 6367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1925421953201294,
+      "learning_rate": 7.00433463853058e-07,
+      "loss": 0.7638,
+      "step": 6368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.196811556816101,
+      "learning_rate": 6.993504875815426e-07,
+      "loss": 0.7908,
+      "step": 6369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.308045506477356,
+      "learning_rate": 6.982682862230894e-07,
+      "loss": 0.8335,
+      "step": 6370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2397823333740234,
+      "learning_rate": 6.971868599726933e-07,
+      "loss": 0.7912,
+      "step": 6371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5821560025215149,
+      "learning_rate": 6.961062090252124e-07,
+      "loss": 1.0258,
+      "step": 6372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2743092775344849,
+      "learning_rate": 6.950263335753638e-07,
+      "loss": 0.7908,
+      "step": 6373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2500864267349243,
+      "learning_rate": 6.939472338177239e-07,
+      "loss": 0.814,
+      "step": 6374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5947352051734924,
+      "learning_rate": 6.928689099467306e-07,
+      "loss": 1.0019,
+      "step": 6375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1763774156570435,
+      "learning_rate": 6.917913621566835e-07,
+      "loss": 0.8067,
+      "step": 6376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2804968357086182,
+      "learning_rate": 6.907145906417379e-07,
+      "loss": 0.8192,
+      "step": 6377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2308387756347656,
+      "learning_rate": 6.896385955959151e-07,
+      "loss": 0.7991,
+      "step": 6378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5831558704376221,
+      "learning_rate": 6.885633772130906e-07,
+      "loss": 0.9884,
+      "step": 6379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1871050596237183,
+      "learning_rate": 6.874889356870057e-07,
+      "loss": 0.8192,
+      "step": 6380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2232035398483276,
+      "learning_rate": 6.864152712112565e-07,
+      "loss": 0.8444,
+      "step": 6381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2475991249084473,
+      "learning_rate": 6.853423839793028e-07,
+      "loss": 0.7939,
+      "step": 6382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2785980701446533,
+      "learning_rate": 6.842702741844604e-07,
+      "loss": 0.7601,
+      "step": 6383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5947389602661133,
+      "learning_rate": 6.831989420199103e-07,
+      "loss": 1.0089,
+      "step": 6384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5723752379417419,
+      "learning_rate": 6.821283876786888e-07,
+      "loss": 0.9904,
+      "step": 6385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.107632637023926,
+      "learning_rate": 6.81058611353696e-07,
+      "loss": 0.7978,
+      "step": 6386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.218501091003418,
+      "learning_rate": 6.799896132376865e-07,
+      "loss": 0.8111,
+      "step": 6387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2769038677215576,
+      "learning_rate": 6.789213935232814e-07,
+      "loss": 0.8114,
+      "step": 6388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2197004556655884,
+      "learning_rate": 6.778539524029565e-07,
+      "loss": 0.7904,
+      "step": 6389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4354093074798584,
+      "learning_rate": 6.767872900690481e-07,
+      "loss": 0.7828,
+      "step": 6390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2524127960205078,
+      "learning_rate": 6.757214067137524e-07,
+      "loss": 0.805,
+      "step": 6391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1964136362075806,
+      "learning_rate": 6.746563025291275e-07,
+      "loss": 0.7546,
+      "step": 6392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.238685131072998,
+      "learning_rate": 6.735919777070882e-07,
+      "loss": 0.7723,
+      "step": 6393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5794294476509094,
+      "learning_rate": 6.725284324394099e-07,
+      "loss": 1.0118,
+      "step": 6394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.393027663230896,
+      "learning_rate": 6.714656669177289e-07,
+      "loss": 0.7786,
+      "step": 6395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5285214185714722,
+      "learning_rate": 6.704036813335396e-07,
+      "loss": 0.8523,
+      "step": 6396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2255077362060547,
+      "learning_rate": 6.693424758781947e-07,
+      "loss": 0.8104,
+      "step": 6397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.279369592666626,
+      "learning_rate": 6.682820507429072e-07,
+      "loss": 0.7877,
+      "step": 6398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2264037132263184,
+      "learning_rate": 6.672224061187527e-07,
+      "loss": 0.7985,
+      "step": 6399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2664519548416138,
+      "learning_rate": 6.661635421966611e-07,
+      "loss": 0.7817,
+      "step": 6400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5764803290367126,
+      "learning_rate": 6.651054591674233e-07,
+      "loss": 0.9969,
+      "step": 6401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2025775909423828,
+      "learning_rate": 6.640481572216917e-07,
+      "loss": 0.8108,
+      "step": 6402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.237470030784607,
+      "learning_rate": 6.629916365499767e-07,
+      "loss": 0.7671,
+      "step": 6403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5783345699310303,
+      "learning_rate": 6.619358973426476e-07,
+      "loss": 0.9969,
+      "step": 6404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.28560209274292,
+      "learning_rate": 6.608809397899318e-07,
+      "loss": 0.8112,
+      "step": 6405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5866835117340088,
+      "learning_rate": 6.598267640819161e-07,
+      "loss": 1.0037,
+      "step": 6406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2992740869522095,
+      "learning_rate": 6.587733704085497e-07,
+      "loss": 0.7897,
+      "step": 6407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6007325053215027,
+      "learning_rate": 6.577207589596374e-07,
+      "loss": 1.0162,
+      "step": 6408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5887657403945923,
+      "learning_rate": 6.56668929924843e-07,
+      "loss": 1.0426,
+      "step": 6409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5654848217964172,
+      "learning_rate": 6.556178834936916e-07,
+      "loss": 1.0141,
+      "step": 6410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5961306691169739,
+      "learning_rate": 6.545676198555672e-07,
+      "loss": 1.0149,
+      "step": 6411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2376112937927246,
+      "learning_rate": 6.535181391997103e-07,
+      "loss": 0.8123,
+      "step": 6412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3246028423309326,
+      "learning_rate": 6.524694417152216e-07,
+      "loss": 0.8702,
+      "step": 6413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3136452436447144,
+      "learning_rate": 6.514215275910601e-07,
+      "loss": 0.7273,
+      "step": 6414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2129662036895752,
+      "learning_rate": 6.50374397016047e-07,
+      "loss": 0.8013,
+      "step": 6415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2908142805099487,
+      "learning_rate": 6.493280501788573e-07,
+      "loss": 0.7729,
+      "step": 6416
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.585561990737915,
+      "learning_rate": 6.482824872680272e-07,
+      "loss": 1.0088,
+      "step": 6417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2447776794433594,
+      "learning_rate": 6.472377084719516e-07,
+      "loss": 0.7658,
+      "step": 6418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5657902359962463,
+      "learning_rate": 6.461937139788866e-07,
+      "loss": 1.0098,
+      "step": 6419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2234787940979004,
+      "learning_rate": 6.451505039769418e-07,
+      "loss": 0.7498,
+      "step": 6420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2890894412994385,
+      "learning_rate": 6.441080786540893e-07,
+      "loss": 0.8478,
+      "step": 6421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.578273594379425,
+      "learning_rate": 6.430664381981572e-07,
+      "loss": 1.0127,
+      "step": 6422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2073215246200562,
+      "learning_rate": 6.420255827968352e-07,
+      "loss": 0.775,
+      "step": 6423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3199739456176758,
+      "learning_rate": 6.409855126376696e-07,
+      "loss": 0.8242,
+      "step": 6424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2581355571746826,
+      "learning_rate": 6.399462279080648e-07,
+      "loss": 0.8358,
+      "step": 6425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3002967834472656,
+      "learning_rate": 6.389077287952844e-07,
+      "loss": 0.8059,
+      "step": 6426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2181146144866943,
+      "learning_rate": 6.378700154864525e-07,
+      "loss": 0.7761,
+      "step": 6427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2666637897491455,
+      "learning_rate": 6.368330881685481e-07,
+      "loss": 0.787,
+      "step": 6428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2057411670684814,
+      "learning_rate": 6.357969470284087e-07,
+      "loss": 0.7728,
+      "step": 6429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1897164583206177,
+      "learning_rate": 6.347615922527345e-07,
+      "loss": 0.8501,
+      "step": 6430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.320770025253296,
+      "learning_rate": 6.337270240280796e-07,
+      "loss": 0.8314,
+      "step": 6431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2591558694839478,
+      "learning_rate": 6.326932425408577e-07,
+      "loss": 0.7911,
+      "step": 6432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6114093065261841,
+      "learning_rate": 6.316602479773393e-07,
+      "loss": 1.0065,
+      "step": 6433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6043328046798706,
+      "learning_rate": 6.306280405236564e-07,
+      "loss": 1.0315,
+      "step": 6434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5821831226348877,
+      "learning_rate": 6.295966203657988e-07,
+      "loss": 1.0113,
+      "step": 6435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5766149759292603,
+      "learning_rate": 6.285659876896111e-07,
+      "loss": 0.9988,
+      "step": 6436
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.239666223526001,
+      "learning_rate": 6.275361426807974e-07,
+      "loss": 0.8041,
+      "step": 6437
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5679900646209717,
+      "learning_rate": 6.265070855249228e-07,
+      "loss": 1.0089,
+      "step": 6438
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0203964710235596,
+      "learning_rate": 6.25478816407406e-07,
+      "loss": 0.7742,
+      "step": 6439
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5958799719810486,
+      "learning_rate": 6.244513355135274e-07,
+      "loss": 0.9859,
+      "step": 6440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.307889699935913,
+      "learning_rate": 6.234246430284213e-07,
+      "loss": 0.7833,
+      "step": 6441
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5732889175415039,
+      "learning_rate": 6.22398739137084e-07,
+      "loss": 1.02,
+      "step": 6442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2128182649612427,
+      "learning_rate": 6.213736240243695e-07,
+      "loss": 0.771,
+      "step": 6443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2075214385986328,
+      "learning_rate": 6.203492978749864e-07,
+      "loss": 0.8389,
+      "step": 6444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2345174551010132,
+      "learning_rate": 6.193257608735031e-07,
+      "loss": 0.8077,
+      "step": 6445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5801681876182556,
+      "learning_rate": 6.183030132043466e-07,
+      "loss": 1.0142,
+      "step": 6446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.574234127998352,
+      "learning_rate": 6.17281055051801e-07,
+      "loss": 1.0209,
+      "step": 6447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2924435138702393,
+      "learning_rate": 6.162598866000069e-07,
+      "loss": 0.8117,
+      "step": 6448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2597533464431763,
+      "learning_rate": 6.152395080329638e-07,
+      "loss": 0.8326,
+      "step": 6449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2233415842056274,
+      "learning_rate": 6.142199195345283e-07,
+      "loss": 0.7807,
+      "step": 6450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.199658989906311,
+      "learning_rate": 6.13201121288417e-07,
+      "loss": 0.8046,
+      "step": 6451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2736905813217163,
+      "learning_rate": 6.121831134782013e-07,
+      "loss": 0.83,
+      "step": 6452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1645077466964722,
+      "learning_rate": 6.111658962873096e-07,
+      "loss": 0.8343,
+      "step": 6453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4851182699203491,
+      "learning_rate": 6.101494698990317e-07,
+      "loss": 0.8157,
+      "step": 6454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2355138063430786,
+      "learning_rate": 6.091338344965108e-07,
+      "loss": 0.7844,
+      "step": 6455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2094793319702148,
+      "learning_rate": 6.081189902627505e-07,
+      "loss": 0.7476,
+      "step": 6456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.28327214717865,
+      "learning_rate": 6.071049373806087e-07,
+      "loss": 0.8245,
+      "step": 6457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5941285490989685,
+      "learning_rate": 6.060916760328034e-07,
+      "loss": 0.9912,
+      "step": 6458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2952402830123901,
+      "learning_rate": 6.050792064019112e-07,
+      "loss": 0.773,
+      "step": 6459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2670533657073975,
+      "learning_rate": 6.040675286703623e-07,
+      "loss": 0.7603,
+      "step": 6460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3050814867019653,
+      "learning_rate": 6.030566430204454e-07,
+      "loss": 0.8056,
+      "step": 6461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.23836088180542,
+      "learning_rate": 6.020465496343092e-07,
+      "loss": 0.7806,
+      "step": 6462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2595921754837036,
+      "learning_rate": 6.010372486939558e-07,
+      "loss": 0.8267,
+      "step": 6463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5963814854621887,
+      "learning_rate": 6.000287403812454e-07,
+      "loss": 1.0138,
+      "step": 6464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5770989060401917,
+      "learning_rate": 5.990210248778989e-07,
+      "loss": 1.0073,
+      "step": 6465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.236688256263733,
+      "learning_rate": 5.980141023654889e-07,
+      "loss": 0.794,
+      "step": 6466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5796517133712769,
+      "learning_rate": 5.970079730254502e-07,
+      "loss": 0.9988,
+      "step": 6467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5895466804504395,
+      "learning_rate": 5.960026370390698e-07,
+      "loss": 1.0204,
+      "step": 6468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1957712173461914,
+      "learning_rate": 5.949980945874973e-07,
+      "loss": 0.7477,
+      "step": 6469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.311026930809021,
+      "learning_rate": 5.939943458517345e-07,
+      "loss": 0.7986,
+      "step": 6470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2217981815338135,
+      "learning_rate": 5.929913910126423e-07,
+      "loss": 0.7643,
+      "step": 6471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2091643810272217,
+      "learning_rate": 5.919892302509372e-07,
+      "loss": 0.8029,
+      "step": 6472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2098257541656494,
+      "learning_rate": 5.909878637471956e-07,
+      "loss": 0.8004,
+      "step": 6473
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5760887265205383,
+      "learning_rate": 5.899872916818466e-07,
+      "loss": 1.0103,
+      "step": 6474
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.575436532497406,
+      "learning_rate": 5.889875142351808e-07,
+      "loss": 1.0315,
+      "step": 6475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2016375064849854,
+      "learning_rate": 5.879885315873413e-07,
+      "loss": 0.7927,
+      "step": 6476
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3102586269378662,
+      "learning_rate": 5.869903439183317e-07,
+      "loss": 0.8575,
+      "step": 6477
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2918654680252075,
+      "learning_rate": 5.859929514080092e-07,
+      "loss": 0.831,
+      "step": 6478
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2687788009643555,
+      "learning_rate": 5.849963542360898e-07,
+      "loss": 0.8159,
+      "step": 6479
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.275776743888855,
+      "learning_rate": 5.840005525821441e-07,
+      "loss": 0.7607,
+      "step": 6480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1816842555999756,
+      "learning_rate": 5.830055466256024e-07,
+      "loss": 0.7894,
+      "step": 6481
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2520188093185425,
+      "learning_rate": 5.820113365457486e-07,
+      "loss": 0.8137,
+      "step": 6482
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.301500678062439,
+      "learning_rate": 5.810179225217266e-07,
+      "loss": 0.7551,
+      "step": 6483
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2640537023544312,
+      "learning_rate": 5.800253047325327e-07,
+      "loss": 0.8309,
+      "step": 6484
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.267233967781067,
+      "learning_rate": 5.790334833570232e-07,
+      "loss": 0.8103,
+      "step": 6485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5856502056121826,
+      "learning_rate": 5.780424585739092e-07,
+      "loss": 1.0342,
+      "step": 6486
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5833908319473267,
+      "learning_rate": 5.770522305617587e-07,
+      "loss": 0.9994,
+      "step": 6487
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5837546586990356,
+      "learning_rate": 5.760627994989948e-07,
+      "loss": 1.0017,
+      "step": 6488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2569102048873901,
+      "learning_rate": 5.750741655639003e-07,
+      "loss": 0.8089,
+      "step": 6489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2884202003479004,
+      "learning_rate": 5.740863289346099e-07,
+      "loss": 0.7996,
+      "step": 6490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6038500666618347,
+      "learning_rate": 5.730992897891202e-07,
+      "loss": 0.9916,
+      "step": 6491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2149527072906494,
+      "learning_rate": 5.721130483052778e-07,
+      "loss": 0.8093,
+      "step": 6492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2321856021881104,
+      "learning_rate": 5.71127604660791e-07,
+      "loss": 0.7825,
+      "step": 6493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5757505893707275,
+      "learning_rate": 5.701429590332219e-07,
+      "loss": 1.0183,
+      "step": 6494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5581036806106567,
+      "learning_rate": 5.691591115999884e-07,
+      "loss": 1.019,
+      "step": 6495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2467855215072632,
+      "learning_rate": 5.681760625383637e-07,
+      "loss": 0.8354,
+      "step": 6496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5851644277572632,
+      "learning_rate": 5.67193812025481e-07,
+      "loss": 0.9844,
+      "step": 6497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5828741788864136,
+      "learning_rate": 5.662123602383257e-07,
+      "loss": 0.9952,
+      "step": 6498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5759245753288269,
+      "learning_rate": 5.652317073537422e-07,
+      "loss": 0.9941,
+      "step": 6499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.199633240699768,
+      "learning_rate": 5.642518535484282e-07,
+      "loss": 0.7991,
+      "step": 6500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5802640318870544,
+      "learning_rate": 5.6327279899894e-07,
+      "loss": 1.0206,
+      "step": 6501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.231591820716858,
+      "learning_rate": 5.622945438816879e-07,
+      "loss": 0.7701,
+      "step": 6502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.311056137084961,
+      "learning_rate": 5.61317088372938e-07,
+      "loss": 0.8121,
+      "step": 6503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.34149169921875,
+      "learning_rate": 5.603404326488155e-07,
+      "loss": 0.8291,
+      "step": 6504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2253824472427368,
+      "learning_rate": 5.593645768852979e-07,
+      "loss": 0.7977,
+      "step": 6505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2152057886123657,
+      "learning_rate": 5.583895212582191e-07,
+      "loss": 0.7898,
+      "step": 6506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5837469696998596,
+      "learning_rate": 5.574152659432708e-07,
+      "loss": 1.0086,
+      "step": 6507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5992687940597534,
+      "learning_rate": 5.564418111159997e-07,
+      "loss": 1.0207,
+      "step": 6508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3227622509002686,
+      "learning_rate": 5.554691569518073e-07,
+      "loss": 0.8898,
+      "step": 6509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2091448307037354,
+      "learning_rate": 5.54497303625951e-07,
+      "loss": 0.7873,
+      "step": 6510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.223151445388794,
+      "learning_rate": 5.535262513135437e-07,
+      "loss": 0.8092,
+      "step": 6511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2451483011245728,
+      "learning_rate": 5.525560001895569e-07,
+      "loss": 0.7876,
+      "step": 6512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2685166597366333,
+      "learning_rate": 5.515865504288137e-07,
+      "loss": 0.777,
+      "step": 6513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5910624265670776,
+      "learning_rate": 5.506179022059943e-07,
+      "loss": 1.0216,
+      "step": 6514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2503145933151245,
+      "learning_rate": 5.496500556956347e-07,
+      "loss": 0.7866,
+      "step": 6515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2142982482910156,
+      "learning_rate": 5.486830110721281e-07,
+      "loss": 0.8429,
+      "step": 6516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.303049087524414,
+      "learning_rate": 5.477167685097206e-07,
+      "loss": 0.8566,
+      "step": 6517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1972589492797852,
+      "learning_rate": 5.467513281825149e-07,
+      "loss": 0.8159,
+      "step": 6518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2917906045913696,
+      "learning_rate": 5.457866902644671e-07,
+      "loss": 0.7601,
+      "step": 6519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.567937433719635,
+      "learning_rate": 5.448228549293932e-07,
+      "loss": 0.9986,
+      "step": 6520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2676746845245361,
+      "learning_rate": 5.438598223509617e-07,
+      "loss": 0.7975,
+      "step": 6521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2285010814666748,
+      "learning_rate": 5.428975927026942e-07,
+      "loss": 0.7855,
+      "step": 6522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3546226024627686,
+      "learning_rate": 5.419361661579726e-07,
+      "loss": 0.8404,
+      "step": 6523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5755918622016907,
+      "learning_rate": 5.409755428900315e-07,
+      "loss": 1.033,
+      "step": 6524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2121175527572632,
+      "learning_rate": 5.400157230719605e-07,
+      "loss": 0.8211,
+      "step": 6525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2311416864395142,
+      "learning_rate": 5.39056706876705e-07,
+      "loss": 0.7553,
+      "step": 6526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2735120058059692,
+      "learning_rate": 5.38098494477064e-07,
+      "loss": 0.8408,
+      "step": 6527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.251589059829712,
+      "learning_rate": 5.371410860456955e-07,
+      "loss": 0.8365,
+      "step": 6528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1998672485351562,
+      "learning_rate": 5.361844817551093e-07,
+      "loss": 0.763,
+      "step": 6529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2330414056777954,
+      "learning_rate": 5.352286817776703e-07,
+      "loss": 0.8327,
+      "step": 6530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2417341470718384,
+      "learning_rate": 5.342736862855996e-07,
+      "loss": 0.8299,
+      "step": 6531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2640504837036133,
+      "learning_rate": 5.333194954509752e-07,
+      "loss": 0.797,
+      "step": 6532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2776020765304565,
+      "learning_rate": 5.323661094457266e-07,
+      "loss": 0.8427,
+      "step": 6533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2828696966171265,
+      "learning_rate": 5.3141352844164e-07,
+      "loss": 0.8002,
+      "step": 6534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.580664336681366,
+      "learning_rate": 5.30461752610355e-07,
+      "loss": 1.0221,
+      "step": 6535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2235032320022583,
+      "learning_rate": 5.295107821233697e-07,
+      "loss": 0.812,
+      "step": 6536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2467626333236694,
+      "learning_rate": 5.285606171520341e-07,
+      "loss": 0.7977,
+      "step": 6537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.249957799911499,
+      "learning_rate": 5.276112578675524e-07,
+      "loss": 0.7854,
+      "step": 6538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2755975723266602,
+      "learning_rate": 5.266627044409856e-07,
+      "loss": 0.8337,
+      "step": 6539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7220802307128906,
+      "learning_rate": 5.257149570432507e-07,
+      "loss": 0.7954,
+      "step": 6540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2737680673599243,
+      "learning_rate": 5.247680158451163e-07,
+      "loss": 0.7912,
+      "step": 6541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2844486236572266,
+      "learning_rate": 5.238218810172063e-07,
+      "loss": 0.8362,
+      "step": 6542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.134483814239502,
+      "learning_rate": 5.228765527300023e-07,
+      "loss": 0.7932,
+      "step": 6543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2363415956497192,
+      "learning_rate": 5.219320311538367e-07,
+      "loss": 0.8755,
+      "step": 6544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2895015478134155,
+      "learning_rate": 5.209883164588991e-07,
+      "loss": 0.7554,
+      "step": 6545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2034833431243896,
+      "learning_rate": 5.200454088152313e-07,
+      "loss": 0.765,
+      "step": 6546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2527713775634766,
+      "learning_rate": 5.191033083927322e-07,
+      "loss": 0.8177,
+      "step": 6547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2485395669937134,
+      "learning_rate": 5.18162015361155e-07,
+      "loss": 0.7707,
+      "step": 6548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2746031284332275,
+      "learning_rate": 5.172215298901068e-07,
+      "loss": 0.7617,
+      "step": 6549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5889440774917603,
+      "learning_rate": 5.16281852149047e-07,
+      "loss": 1.034,
+      "step": 6550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5772544145584106,
+      "learning_rate": 5.153429823072942e-07,
+      "loss": 1.019,
+      "step": 6551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2966692447662354,
+      "learning_rate": 5.144049205340174e-07,
+      "loss": 0.8047,
+      "step": 6552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2561922073364258,
+      "learning_rate": 5.134676669982413e-07,
+      "loss": 0.8213,
+      "step": 6553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5569530725479126,
+      "learning_rate": 5.125312218688444e-07,
+      "loss": 0.9931,
+      "step": 6554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2658677101135254,
+      "learning_rate": 5.11595585314561e-07,
+      "loss": 0.7711,
+      "step": 6555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2446296215057373,
+      "learning_rate": 5.106607575039801e-07,
+      "loss": 0.7511,
+      "step": 6556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2432376146316528,
+      "learning_rate": 5.097267386055427e-07,
+      "loss": 0.794,
+      "step": 6557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5700669884681702,
+      "learning_rate": 5.087935287875434e-07,
+      "loss": 1.0151,
+      "step": 6558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2290081977844238,
+      "learning_rate": 5.078611282181361e-07,
+      "loss": 0.7562,
+      "step": 6559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.254569411277771,
+      "learning_rate": 5.069295370653232e-07,
+      "loss": 0.8541,
+      "step": 6560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5775114893913269,
+      "learning_rate": 5.059987554969642e-07,
+      "loss": 1.0156,
+      "step": 6561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.270227074623108,
+      "learning_rate": 5.050687836807711e-07,
+      "loss": 0.8397,
+      "step": 6562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2089108228683472,
+      "learning_rate": 5.04139621784312e-07,
+      "loss": 0.7496,
+      "step": 6563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3438584804534912,
+      "learning_rate": 5.032112699750085e-07,
+      "loss": 0.797,
+      "step": 6564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2341082096099854,
+      "learning_rate": 5.022837284201354e-07,
+      "loss": 0.8102,
+      "step": 6565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.259832739830017,
+      "learning_rate": 5.01356997286821e-07,
+      "loss": 0.7657,
+      "step": 6566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2437719106674194,
+      "learning_rate": 5.004310767420495e-07,
+      "loss": 0.7948,
+      "step": 6567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2084568738937378,
+      "learning_rate": 4.99505966952658e-07,
+      "loss": 0.7597,
+      "step": 6568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2206742763519287,
+      "learning_rate": 4.985816680853373e-07,
+      "loss": 0.7938,
+      "step": 6569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.276503324508667,
+      "learning_rate": 4.976581803066316e-07,
+      "loss": 0.7754,
+      "step": 6570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2897241115570068,
+      "learning_rate": 4.967355037829396e-07,
+      "loss": 0.8405,
+      "step": 6571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2693003416061401,
+      "learning_rate": 4.958136386805157e-07,
+      "loss": 0.783,
+      "step": 6572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1667613983154297,
+      "learning_rate": 4.948925851654646e-07,
+      "loss": 0.7662,
+      "step": 6573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5866423845291138,
+      "learning_rate": 4.939723434037475e-07,
+      "loss": 1.0187,
+      "step": 6574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.180684208869934,
+      "learning_rate": 4.930529135611778e-07,
+      "loss": 0.8279,
+      "step": 6575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.214199185371399,
+      "learning_rate": 4.921342958034225e-07,
+      "loss": 0.7486,
+      "step": 6576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6024898886680603,
+      "learning_rate": 4.912164902960031e-07,
+      "loss": 1.0024,
+      "step": 6577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5772212743759155,
+      "learning_rate": 4.902994972042957e-07,
+      "loss": 1.0263,
+      "step": 6578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2549012899398804,
+      "learning_rate": 4.893833166935275e-07,
+      "loss": 0.8353,
+      "step": 6579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.244933009147644,
+      "learning_rate": 4.884679489287803e-07,
+      "loss": 0.8216,
+      "step": 6580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2071024179458618,
+      "learning_rate": 4.8755339407499e-07,
+      "loss": 0.7556,
+      "step": 6581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4386985301971436,
+      "learning_rate": 4.866396522969474e-07,
+      "loss": 0.7804,
+      "step": 6582
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2103121280670166,
+      "learning_rate": 4.857267237592944e-07,
+      "loss": 0.7582,
+      "step": 6583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5658465027809143,
+      "learning_rate": 4.848146086265265e-07,
+      "loss": 0.9994,
+      "step": 6584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3151403665542603,
+      "learning_rate": 4.839033070629922e-07,
+      "loss": 0.7946,
+      "step": 6585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1977665424346924,
+      "learning_rate": 4.82992819232897e-07,
+      "loss": 0.7741,
+      "step": 6586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2309216260910034,
+      "learning_rate": 4.820831453002961e-07,
+      "loss": 0.8232,
+      "step": 6587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2117937803268433,
+      "learning_rate": 4.811742854290979e-07,
+      "loss": 0.786,
+      "step": 6588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.228499412536621,
+      "learning_rate": 4.80266239783067e-07,
+      "loss": 0.835,
+      "step": 6589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2402855157852173,
+      "learning_rate": 4.793590085258205e-07,
+      "loss": 0.7942,
+      "step": 6590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2328381538391113,
+      "learning_rate": 4.784525918208266e-07,
+      "loss": 0.7558,
+      "step": 6591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3094125986099243,
+      "learning_rate": 4.775469898314083e-07,
+      "loss": 0.791,
+      "step": 6592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2631359100341797,
+      "learning_rate": 4.766422027207407e-07,
+      "loss": 0.7343,
+      "step": 6593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.245301365852356,
+      "learning_rate": 4.757382306518549e-07,
+      "loss": 0.79,
+      "step": 6594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2558233737945557,
+      "learning_rate": 4.748350737876323e-07,
+      "loss": 0.841,
+      "step": 6595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.213423252105713,
+      "learning_rate": 4.739327322908077e-07,
+      "loss": 0.7767,
+      "step": 6596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5946854948997498,
+      "learning_rate": 4.7303120632397e-07,
+      "loss": 0.9939,
+      "step": 6597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2081125974655151,
+      "learning_rate": 4.7213049604956205e-07,
+      "loss": 0.8653,
+      "step": 6598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2628915309906006,
+      "learning_rate": 4.7123060162987734e-07,
+      "loss": 0.7928,
+      "step": 6599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2847470045089722,
+      "learning_rate": 4.7033152322706335e-07,
+      "loss": 0.8061,
+      "step": 6600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3656123876571655,
+      "learning_rate": 4.694332610031199e-07,
+      "loss": 0.7716,
+      "step": 6601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5678070783615112,
+      "learning_rate": 4.685358151199021e-07,
+      "loss": 1.0142,
+      "step": 6602
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2517311573028564,
+      "learning_rate": 4.6763918573911606e-07,
+      "loss": 0.8116,
+      "step": 6603
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2442457675933838,
+      "learning_rate": 4.6674337302231923e-07,
+      "loss": 0.761,
+      "step": 6604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2088879346847534,
+      "learning_rate": 4.6584837713092524e-07,
+      "loss": 0.7808,
+      "step": 6605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7953107357025146,
+      "learning_rate": 4.649541982262001e-07,
+      "loss": 0.825,
+      "step": 6606
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2830613851547241,
+      "learning_rate": 4.6406083646926e-07,
+      "loss": 0.8009,
+      "step": 6607
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5570387244224548,
+      "learning_rate": 4.631682920210762e-07,
+      "loss": 0.9723,
+      "step": 6608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1687616109848022,
+      "learning_rate": 4.622765650424704e-07,
+      "loss": 0.7826,
+      "step": 6609
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2548577785491943,
+      "learning_rate": 4.613856556941204e-07,
+      "loss": 0.8521,
+      "step": 6610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.294318437576294,
+      "learning_rate": 4.604955641365544e-07,
+      "loss": 0.7982,
+      "step": 6611
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2392377853393555,
+      "learning_rate": 4.5960629053015185e-07,
+      "loss": 0.7202,
+      "step": 6612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1854957342147827,
+      "learning_rate": 4.5871783503514843e-07,
+      "loss": 0.733,
+      "step": 6613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2607399225234985,
+      "learning_rate": 4.578301978116312e-07,
+      "loss": 0.8031,
+      "step": 6614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5913032293319702,
+      "learning_rate": 4.5694337901953823e-07,
+      "loss": 1.0175,
+      "step": 6615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2566680908203125,
+      "learning_rate": 4.560573788186601e-07,
+      "loss": 0.8256,
+      "step": 6616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2339802980422974,
+      "learning_rate": 4.551721973686429e-07,
+      "loss": 0.7638,
+      "step": 6617
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.224705457687378,
+      "learning_rate": 4.54287834828982e-07,
+      "loss": 0.8282,
+      "step": 6618
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3172720670700073,
+      "learning_rate": 4.5340429135902696e-07,
+      "loss": 0.8227,
+      "step": 6619
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2134124040603638,
+      "learning_rate": 4.5252156711797724e-07,
+      "loss": 0.7644,
+      "step": 6620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1653311252593994,
+      "learning_rate": 4.516396622648883e-07,
+      "loss": 0.7641,
+      "step": 6621
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2459027767181396,
+      "learning_rate": 4.50758576958667e-07,
+      "loss": 0.8363,
+      "step": 6622
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.589400589466095,
+      "learning_rate": 4.498783113580707e-07,
+      "loss": 1.0065,
+      "step": 6623
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5824689865112305,
+      "learning_rate": 4.4899886562170924e-07,
+      "loss": 1.0081,
+      "step": 6624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2757376432418823,
+      "learning_rate": 4.481202399080481e-07,
+      "loss": 0.7804,
+      "step": 6625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5845754146575928,
+      "learning_rate": 4.472424343754006e-07,
+      "loss": 0.9801,
+      "step": 6626
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.255938172340393,
+      "learning_rate": 4.4636544918193527e-07,
+      "loss": 0.8224,
+      "step": 6627
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5887300968170166,
+      "learning_rate": 4.454892844856701e-07,
+      "loss": 1.0081,
+      "step": 6628
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5739373564720154,
+      "learning_rate": 4.446139404444788e-07,
+      "loss": 1.0038,
+      "step": 6629
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5679662823677063,
+      "learning_rate": 4.437394172160853e-07,
+      "loss": 1.0243,
+      "step": 6630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2718569040298462,
+      "learning_rate": 4.4286571495806505e-07,
+      "loss": 0.8037,
+      "step": 6631
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1930149793624878,
+      "learning_rate": 4.4199283382784507e-07,
+      "loss": 0.8049,
+      "step": 6632
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9557451009750366,
+      "learning_rate": 4.411207739827084e-07,
+      "loss": 0.8506,
+      "step": 6633
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2533283233642578,
+      "learning_rate": 4.402495355797848e-07,
+      "loss": 0.7672,
+      "step": 6634
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.583083987236023,
+      "learning_rate": 4.393791187760599e-07,
+      "loss": 1.007,
+      "step": 6635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2144012451171875,
+      "learning_rate": 4.385095237283682e-07,
+      "loss": 0.7868,
+      "step": 6636
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5916332602500916,
+      "learning_rate": 4.3764075059339884e-07,
+      "loss": 1.03,
+      "step": 6637
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.200591802597046,
+      "learning_rate": 4.3677279952769216e-07,
+      "loss": 0.8344,
+      "step": 6638
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5881144404411316,
+      "learning_rate": 4.359056706876402e-07,
+      "loss": 1.0268,
+      "step": 6639
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2738585472106934,
+      "learning_rate": 4.3503936422948476e-07,
+      "loss": 0.7981,
+      "step": 6640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.580919623374939,
+      "learning_rate": 4.341738803093243e-07,
+      "loss": 1.0029,
+      "step": 6641
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2202470302581787,
+      "learning_rate": 4.3330921908310397e-07,
+      "loss": 0.767,
+      "step": 6642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2573896646499634,
+      "learning_rate": 4.324453807066242e-07,
+      "loss": 0.8391,
+      "step": 6643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5923941135406494,
+      "learning_rate": 4.315823653355333e-07,
+      "loss": 0.986,
+      "step": 6644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3054933547973633,
+      "learning_rate": 4.3072017312533634e-07,
+      "loss": 0.8192,
+      "step": 6645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5750889778137207,
+      "learning_rate": 4.298588042313873e-07,
+      "loss": 0.9978,
+      "step": 6646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.242954134941101,
+      "learning_rate": 4.289982588088909e-07,
+      "loss": 0.7827,
+      "step": 6647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2614701986312866,
+      "learning_rate": 4.28138537012906e-07,
+      "loss": 0.7846,
+      "step": 6648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2593458890914917,
+      "learning_rate": 4.272796389983408e-07,
+      "loss": 0.7704,
+      "step": 6649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2351475954055786,
+      "learning_rate": 4.2642156491995654e-07,
+      "loss": 0.8168,
+      "step": 6650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2839291095733643,
+      "learning_rate": 4.255643149323635e-07,
+      "loss": 0.8512,
+      "step": 6651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2431493997573853,
+      "learning_rate": 4.2470788919002804e-07,
+      "loss": 0.7657,
+      "step": 6652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2431718111038208,
+      "learning_rate": 4.238522878472628e-07,
+      "loss": 0.8145,
+      "step": 6653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.271964192390442,
+      "learning_rate": 4.2299751105823737e-07,
+      "loss": 0.8344,
+      "step": 6654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5759539604187012,
+      "learning_rate": 4.221435589769668e-07,
+      "loss": 1.0114,
+      "step": 6655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2165517807006836,
+      "learning_rate": 4.2129043175732243e-07,
+      "loss": 0.8087,
+      "step": 6656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2250804901123047,
+      "learning_rate": 4.2043812955302523e-07,
+      "loss": 0.811,
+      "step": 6657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5710950493812561,
+      "learning_rate": 4.195866525176462e-07,
+      "loss": 1.0072,
+      "step": 6658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2586710453033447,
+      "learning_rate": 4.187360008046087e-07,
+      "loss": 0.8331,
+      "step": 6659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2821167707443237,
+      "learning_rate": 4.178861745671892e-07,
+      "loss": 0.7724,
+      "step": 6660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2186235189437866,
+      "learning_rate": 4.1703717395851217e-07,
+      "loss": 0.7801,
+      "step": 6661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2491101026535034,
+      "learning_rate": 4.1618899913155543e-07,
+      "loss": 0.7982,
+      "step": 6662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2526636123657227,
+      "learning_rate": 4.1534165023914717e-07,
+      "loss": 0.8014,
+      "step": 6663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1792100667953491,
+      "learning_rate": 4.1449512743396814e-07,
+      "loss": 0.7445,
+      "step": 6664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2108169794082642,
+      "learning_rate": 4.136494308685485e-07,
+      "loss": 0.809,
+      "step": 6665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5754523277282715,
+      "learning_rate": 4.1280456069526965e-07,
+      "loss": 1.0108,
+      "step": 6666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2438312768936157,
+      "learning_rate": 4.119605170663649e-07,
+      "loss": 0.7602,
+      "step": 6667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2322468757629395,
+      "learning_rate": 4.111173001339186e-07,
+      "loss": 0.7809,
+      "step": 6668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5747836828231812,
+      "learning_rate": 4.102749100498654e-07,
+      "loss": 1.0059,
+      "step": 6669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2288137674331665,
+      "learning_rate": 4.0943334696599226e-07,
+      "loss": 0.8001,
+      "step": 6670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3060904741287231,
+      "learning_rate": 4.0859261103393556e-07,
+      "loss": 0.8102,
+      "step": 6671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.57171630859375,
+      "learning_rate": 4.077527024051842e-07,
+      "loss": 0.9854,
+      "step": 6672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2300233840942383,
+      "learning_rate": 4.0691362123107704e-07,
+      "loss": 0.8594,
+      "step": 6673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5694521069526672,
+      "learning_rate": 4.060753676628038e-07,
+      "loss": 0.9757,
+      "step": 6674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2541496753692627,
+      "learning_rate": 4.052379418514041e-07,
+      "loss": 0.789,
+      "step": 6675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1978915929794312,
+      "learning_rate": 4.044013439477723e-07,
+      "loss": 0.7617,
+      "step": 6676
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5703970193862915,
+      "learning_rate": 4.0356557410264784e-07,
+      "loss": 1.0255,
+      "step": 6677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2283676862716675,
+      "learning_rate": 4.0273063246662693e-07,
+      "loss": 0.8253,
+      "step": 6678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.604314923286438,
+      "learning_rate": 4.0189651919015084e-07,
+      "loss": 1.0323,
+      "step": 6679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2313586473464966,
+      "learning_rate": 4.0106323442351716e-07,
+      "loss": 0.8254,
+      "step": 6680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2610801458358765,
+      "learning_rate": 4.002307783168696e-07,
+      "loss": 0.806,
+      "step": 6681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2189373970031738,
+      "learning_rate": 3.99399151020205e-07,
+      "loss": 0.8067,
+      "step": 6682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3178303241729736,
+      "learning_rate": 3.9856835268336946e-07,
+      "loss": 0.772,
+      "step": 6683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6003929972648621,
+      "learning_rate": 3.977383834560616e-07,
+      "loss": 1.0288,
+      "step": 6684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5616012215614319,
+      "learning_rate": 3.96909243487828e-07,
+      "loss": 0.9994,
+      "step": 6685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2894915342330933,
+      "learning_rate": 3.9608093292806846e-07,
+      "loss": 0.7701,
+      "step": 6686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58697110414505,
+      "learning_rate": 3.9525345192603314e-07,
+      "loss": 1.034,
+      "step": 6687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2793114185333252,
+      "learning_rate": 3.944268006308205e-07,
+      "loss": 0.7877,
+      "step": 6688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2275651693344116,
+      "learning_rate": 3.936009791913814e-07,
+      "loss": 0.764,
+      "step": 6689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2417625188827515,
+      "learning_rate": 3.9277598775651514e-07,
+      "loss": 0.804,
+      "step": 6690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2063474655151367,
+      "learning_rate": 3.919518264748745e-07,
+      "loss": 0.6967,
+      "step": 6691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2570834159851074,
+      "learning_rate": 3.9112849549496124e-07,
+      "loss": 0.8304,
+      "step": 6692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.197390079498291,
+      "learning_rate": 3.9030599496512557e-07,
+      "loss": 0.77,
+      "step": 6693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2427617311477661,
+      "learning_rate": 3.894843250335706e-07,
+      "loss": 0.7835,
+      "step": 6694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.290560007095337,
+      "learning_rate": 3.886634858483507e-07,
+      "loss": 0.7621,
+      "step": 6695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2209296226501465,
+      "learning_rate": 3.878434775573675e-07,
+      "loss": 0.7846,
+      "step": 6696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2156360149383545,
+      "learning_rate": 3.87024300308374e-07,
+      "loss": 0.7587,
+      "step": 6697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5786851644515991,
+      "learning_rate": 3.862059542489738e-07,
+      "loss": 0.9967,
+      "step": 6698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2683266401290894,
+      "learning_rate": 3.8538843952662165e-07,
+      "loss": 0.8275,
+      "step": 6699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5899977684020996,
+      "learning_rate": 3.845717562886209e-07,
+      "loss": 1.0461,
+      "step": 6700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.316285252571106,
+      "learning_rate": 3.8375590468212485e-07,
+      "loss": 0.8022,
+      "step": 6701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2565972805023193,
+      "learning_rate": 3.829408848541388e-07,
+      "loss": 0.8348,
+      "step": 6702
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2551236152648926,
+      "learning_rate": 3.8212669695151803e-07,
+      "loss": 0.8232,
+      "step": 6703
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3154816627502441,
+      "learning_rate": 3.8131334112096574e-07,
+      "loss": 0.8213,
+      "step": 6704
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.206716775894165,
+      "learning_rate": 3.8050081750903747e-07,
+      "loss": 0.7988,
+      "step": 6705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2667806148529053,
+      "learning_rate": 3.7968912626213604e-07,
+      "loss": 0.8518,
+      "step": 6706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5889447927474976,
+      "learning_rate": 3.7887826752651846e-07,
+      "loss": 1.0177,
+      "step": 6707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2535101175308228,
+      "learning_rate": 3.7806824144828834e-07,
+      "loss": 0.7653,
+      "step": 6708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2529070377349854,
+      "learning_rate": 3.772590481734001e-07,
+      "loss": 0.8181,
+      "step": 6709
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2459269762039185,
+      "learning_rate": 3.7645068784765813e-07,
+      "loss": 0.845,
+      "step": 6710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1997402906417847,
+      "learning_rate": 3.7564316061671825e-07,
+      "loss": 0.7044,
+      "step": 6711
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.334816336631775,
+      "learning_rate": 3.7483646662608454e-07,
+      "loss": 0.815,
+      "step": 6712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2002326250076294,
+      "learning_rate": 3.740306060211102e-07,
+      "loss": 0.7601,
+      "step": 6713
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2035974264144897,
+      "learning_rate": 3.7322557894699963e-07,
+      "loss": 0.7645,
+      "step": 6714
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2675408124923706,
+      "learning_rate": 3.7242138554880735e-07,
+      "loss": 0.7784,
+      "step": 6715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1775380373001099,
+      "learning_rate": 3.716180259714369e-07,
+      "loss": 0.8215,
+      "step": 6716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2039034366607666,
+      "learning_rate": 3.708155003596403e-07,
+      "loss": 0.8223,
+      "step": 6717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2392468452453613,
+      "learning_rate": 3.700138088580224e-07,
+      "loss": 0.7754,
+      "step": 6718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.244180679321289,
+      "learning_rate": 3.6921295161103656e-07,
+      "loss": 0.7779,
+      "step": 6719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5707463026046753,
+      "learning_rate": 3.6841292876298397e-07,
+      "loss": 1.0059,
+      "step": 6720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5914134979248047,
+      "learning_rate": 3.6761374045801656e-07,
+      "loss": 1.0132,
+      "step": 6721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2385720014572144,
+      "learning_rate": 3.668153868401381e-07,
+      "loss": 0.7951,
+      "step": 6722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3149771690368652,
+      "learning_rate": 3.6601786805319837e-07,
+      "loss": 0.8134,
+      "step": 6723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2782232761383057,
+      "learning_rate": 3.6522118424089923e-07,
+      "loss": 0.8588,
+      "step": 6724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2648487091064453,
+      "learning_rate": 3.6442533554678974e-07,
+      "loss": 0.7427,
+      "step": 6725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2791721820831299,
+      "learning_rate": 3.636303221142712e-07,
+      "loss": 0.7961,
+      "step": 6726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2083284854888916,
+      "learning_rate": 3.6283614408659406e-07,
+      "loss": 0.8163,
+      "step": 6727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2326487302780151,
+      "learning_rate": 3.620428016068561e-07,
+      "loss": 0.8339,
+      "step": 6728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.161415934562683,
+      "learning_rate": 3.6125029481800575e-07,
+      "loss": 0.7774,
+      "step": 6729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.585267961025238,
+      "learning_rate": 3.6045862386284157e-07,
+      "loss": 0.9906,
+      "step": 6730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.297783613204956,
+      "learning_rate": 3.5966778888401167e-07,
+      "loss": 0.8075,
+      "step": 6731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2173144817352295,
+      "learning_rate": 3.588777900240109e-07,
+      "loss": 0.7885,
+      "step": 6732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3407320976257324,
+      "learning_rate": 3.58088627425186e-07,
+      "loss": 0.7733,
+      "step": 6733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1400141716003418,
+      "learning_rate": 3.5730030122973215e-07,
+      "loss": 0.7295,
+      "step": 6734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.202078104019165,
+      "learning_rate": 3.565128115796951e-07,
+      "loss": 0.7523,
+      "step": 6735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.297566294670105,
+      "learning_rate": 3.55726158616968e-07,
+      "loss": 0.8538,
+      "step": 6736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2850720882415771,
+      "learning_rate": 3.5494034248329365e-07,
+      "loss": 0.8381,
+      "step": 6737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2969774007797241,
+      "learning_rate": 3.5415536332026544e-07,
+      "loss": 0.844,
+      "step": 6738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5891761779785156,
+      "learning_rate": 3.533712212693241e-07,
+      "loss": 1.0435,
+      "step": 6739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.233932375907898,
+      "learning_rate": 3.5258791647176047e-07,
+      "loss": 0.7278,
+      "step": 6740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5674914121627808,
+      "learning_rate": 3.5180544906871394e-07,
+      "loss": 1.0083,
+      "step": 6741
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3333187103271484,
+      "learning_rate": 3.5102381920117447e-07,
+      "loss": 0.7876,
+      "step": 6742
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1593005657196045,
+      "learning_rate": 3.502430270099799e-07,
+      "loss": 0.7974,
+      "step": 6743
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2600668668746948,
+      "learning_rate": 3.494630726358178e-07,
+      "loss": 0.8203,
+      "step": 6744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.247460126876831,
+      "learning_rate": 3.4868395621922237e-07,
+      "loss": 0.7728,
+      "step": 6745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.242389440536499,
+      "learning_rate": 3.479056779005813e-07,
+      "loss": 0.7885,
+      "step": 6746
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2814581394195557,
+      "learning_rate": 3.471282378201274e-07,
+      "loss": 0.8226,
+      "step": 6747
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.273982048034668,
+      "learning_rate": 3.463516361179442e-07,
+      "loss": 0.8032,
+      "step": 6748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5793243050575256,
+      "learning_rate": 3.455758729339631e-07,
+      "loss": 1.0222,
+      "step": 6749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2314574718475342,
+      "learning_rate": 3.4480094840796507e-07,
+      "loss": 0.7846,
+      "step": 6750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1359237432479858,
+      "learning_rate": 3.440268626795817e-07,
+      "loss": 0.7653,
+      "step": 6751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3520839214324951,
+      "learning_rate": 3.4325361588829043e-07,
+      "loss": 0.8684,
+      "step": 6752
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2622499465942383,
+      "learning_rate": 3.42481208173418e-07,
+      "loss": 0.8016,
+      "step": 6753
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3021680116653442,
+      "learning_rate": 3.4170963967414317e-07,
+      "loss": 0.8204,
+      "step": 6754
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5665681958198547,
+      "learning_rate": 3.4093891052948913e-07,
+      "loss": 1.015,
+      "step": 6755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6048325896263123,
+      "learning_rate": 3.401690208783293e-07,
+      "loss": 1.0103,
+      "step": 6756
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1991864442825317,
+      "learning_rate": 3.393999708593887e-07,
+      "loss": 0.727,
+      "step": 6757
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1703436374664307,
+      "learning_rate": 3.3863176061123606e-07,
+      "loss": 0.7423,
+      "step": 6758
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2318471670150757,
+      "learning_rate": 3.378643902722939e-07,
+      "loss": 0.7827,
+      "step": 6759
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3148682117462158,
+      "learning_rate": 3.370978599808289e-07,
+      "loss": 0.7762,
+      "step": 6760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2199265956878662,
+      "learning_rate": 3.3633216987496055e-07,
+      "loss": 0.8471,
+      "step": 6761
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2865734100341797,
+      "learning_rate": 3.355673200926529e-07,
+      "loss": 0.8071,
+      "step": 6762
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2685245275497437,
+      "learning_rate": 3.3480331077172134e-07,
+      "loss": 0.8128,
+      "step": 6763
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5717203617095947,
+      "learning_rate": 3.340401420498285e-07,
+      "loss": 1.0093,
+      "step": 6764
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.29947030544281,
+      "learning_rate": 3.3327781406448657e-07,
+      "loss": 0.7548,
+      "step": 6765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2298442125320435,
+      "learning_rate": 3.3251632695305514e-07,
+      "loss": 0.841,
+      "step": 6766
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2276514768600464,
+      "learning_rate": 3.317556808527439e-07,
+      "loss": 0.7833,
+      "step": 6767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3507734537124634,
+      "learning_rate": 3.309958759006082e-07,
+      "loss": 0.8359,
+      "step": 6768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5976834297180176,
+      "learning_rate": 3.3023691223355583e-07,
+      "loss": 1.0111,
+      "step": 6769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.297600507736206,
+      "learning_rate": 3.294787899883395e-07,
+      "loss": 0.843,
+      "step": 6770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2377455234527588,
+      "learning_rate": 3.2872150930156164e-07,
+      "loss": 0.798,
+      "step": 6771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2711260318756104,
+      "learning_rate": 3.2796507030967207e-07,
+      "loss": 0.7956,
+      "step": 6772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5893491506576538,
+      "learning_rate": 3.272094731489717e-07,
+      "loss": 1.0184,
+      "step": 6773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5669708251953125,
+      "learning_rate": 3.2645471795560603e-07,
+      "loss": 1.0108,
+      "step": 6774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2900639772415161,
+      "learning_rate": 3.2570080486557245e-07,
+      "loss": 0.8116,
+      "step": 6775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2158331871032715,
+      "learning_rate": 3.249477340147134e-07,
+      "loss": 0.7396,
+      "step": 6776
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3716306686401367,
+      "learning_rate": 3.241955055387219e-07,
+      "loss": 0.7957,
+      "step": 6777
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2336182594299316,
+      "learning_rate": 3.2344411957313913e-07,
+      "loss": 0.8789,
+      "step": 6778
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2572267055511475,
+      "learning_rate": 3.226935762533517e-07,
+      "loss": 0.776,
+      "step": 6779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5611056089401245,
+      "learning_rate": 3.219438757145971e-07,
+      "loss": 1.0127,
+      "step": 6780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5951408743858337,
+      "learning_rate": 3.2119501809196096e-07,
+      "loss": 0.9897,
+      "step": 6781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2866096496582031,
+      "learning_rate": 3.204470035203755e-07,
+      "loss": 0.8253,
+      "step": 6782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2424864768981934,
+      "learning_rate": 3.196998321346223e-07,
+      "loss": 0.8183,
+      "step": 6783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1912869215011597,
+      "learning_rate": 3.189535040693298e-07,
+      "loss": 0.7893,
+      "step": 6784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5824277997016907,
+      "learning_rate": 3.182080194589765e-07,
+      "loss": 1.0178,
+      "step": 6785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2765271663665771,
+      "learning_rate": 3.1746337843788666e-07,
+      "loss": 0.7656,
+      "step": 6786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.228468656539917,
+      "learning_rate": 3.167195811402335e-07,
+      "loss": 0.7953,
+      "step": 6787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2549537420272827,
+      "learning_rate": 3.159766277000381e-07,
+      "loss": 0.8077,
+      "step": 6788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.255455493927002,
+      "learning_rate": 3.152345182511707e-07,
+      "loss": 0.8158,
+      "step": 6789
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2413581609725952,
+      "learning_rate": 3.1449325292734644e-07,
+      "loss": 0.8353,
+      "step": 6790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2786893844604492,
+      "learning_rate": 3.1375283186213243e-07,
+      "loss": 0.8221,
+      "step": 6791
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2561501264572144,
+      "learning_rate": 3.130132551889403e-07,
+      "loss": 0.8313,
+      "step": 6792
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2494198083877563,
+      "learning_rate": 3.1227452304103125e-07,
+      "loss": 0.7657,
+      "step": 6793
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5674928426742554,
+      "learning_rate": 3.115366355515137e-07,
+      "loss": 1.0237,
+      "step": 6794
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2405232191085815,
+      "learning_rate": 3.107995928533436e-07,
+      "loss": 0.8794,
+      "step": 6795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3458456993103027,
+      "learning_rate": 3.100633950793258e-07,
+      "loss": 0.7905,
+      "step": 6796
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2434619665145874,
+      "learning_rate": 3.09328042362112e-07,
+      "loss": 0.7606,
+      "step": 6797
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3055952787399292,
+      "learning_rate": 3.085935348342006e-07,
+      "loss": 0.8437,
+      "step": 6798
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5868663787841797,
+      "learning_rate": 3.0785987262794027e-07,
+      "loss": 1.0109,
+      "step": 6799
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.277184009552002,
+      "learning_rate": 3.0712705587552627e-07,
+      "loss": 0.8216,
+      "step": 6800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2528493404388428,
+      "learning_rate": 3.0639508470900083e-07,
+      "loss": 0.8194,
+      "step": 6801
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2514960765838623,
+      "learning_rate": 3.0566395926025396e-07,
+      "loss": 0.743,
+      "step": 6802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2536700963974,
+      "learning_rate": 3.0493367966102305e-07,
+      "loss": 0.7928,
+      "step": 6803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2037872076034546,
+      "learning_rate": 3.042042460428951e-07,
+      "loss": 0.8144,
+      "step": 6804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2112189531326294,
+      "learning_rate": 3.0347565853730274e-07,
+      "loss": 0.8098,
+      "step": 6805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5844966769218445,
+      "learning_rate": 3.027479172755249e-07,
+      "loss": 1.0272,
+      "step": 6806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2176134586334229,
+      "learning_rate": 3.0202102238869156e-07,
+      "loss": 0.7887,
+      "step": 6807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5603477358818054,
+      "learning_rate": 3.0129497400777863e-07,
+      "loss": 1.0005,
+      "step": 6808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.263746738433838,
+      "learning_rate": 3.0056977226360863e-07,
+      "loss": 0.7922,
+      "step": 6809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2064064741134644,
+      "learning_rate": 2.99845417286852e-07,
+      "loss": 0.7931,
+      "step": 6810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2402722835540771,
+      "learning_rate": 2.991219092080261e-07,
+      "loss": 0.7847,
+      "step": 6811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2673519849777222,
+      "learning_rate": 2.983992481574977e-07,
+      "loss": 0.7629,
+      "step": 6812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.190598726272583,
+      "learning_rate": 2.9767743426547867e-07,
+      "loss": 0.7809,
+      "step": 6813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5807653665542603,
+      "learning_rate": 2.9695646766202843e-07,
+      "loss": 0.9979,
+      "step": 6814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.213835597038269,
+      "learning_rate": 2.962363484770564e-07,
+      "loss": 0.7852,
+      "step": 6815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2084797620773315,
+      "learning_rate": 2.9551707684031483e-07,
+      "loss": 0.8111,
+      "step": 6816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2354495525360107,
+      "learning_rate": 2.94798652881409e-07,
+      "loss": 0.7989,
+      "step": 6817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3044966459274292,
+      "learning_rate": 2.940810767297858e-07,
+      "loss": 0.8391,
+      "step": 6818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2508705854415894,
+      "learning_rate": 2.9336434851474136e-07,
+      "loss": 0.8331,
+      "step": 6819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.25771963596344,
+      "learning_rate": 2.9264846836542184e-07,
+      "loss": 0.7934,
+      "step": 6820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1642035245895386,
+      "learning_rate": 2.919334364108162e-07,
+      "loss": 0.7392,
+      "step": 6821
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5681125521659851,
+      "learning_rate": 2.9121925277976304e-07,
+      "loss": 1.0218,
+      "step": 6822
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2631736993789673,
+      "learning_rate": 2.9050591760094837e-07,
+      "loss": 0.8204,
+      "step": 6823
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5840922594070435,
+      "learning_rate": 2.8979343100290327e-07,
+      "loss": 0.9893,
+      "step": 6824
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.219592571258545,
+      "learning_rate": 2.890817931140083e-07,
+      "loss": 0.8057,
+      "step": 6825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3052664995193481,
+      "learning_rate": 2.883710040624899e-07,
+      "loss": 0.8218,
+      "step": 6826
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1662555932998657,
+      "learning_rate": 2.87661063976421e-07,
+      "loss": 0.7668,
+      "step": 6827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3043169975280762,
+      "learning_rate": 2.869519729837228e-07,
+      "loss": 0.8319,
+      "step": 6828
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6061702370643616,
+      "learning_rate": 2.862437312121635e-07,
+      "loss": 1.0242,
+      "step": 6829
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2830238342285156,
+      "learning_rate": 2.8553633878935615e-07,
+      "loss": 0.7975,
+      "step": 6830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.270377516746521,
+      "learning_rate": 2.8482979584276373e-07,
+      "loss": 0.7766,
+      "step": 6831
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1866248846054077,
+      "learning_rate": 2.841241024996938e-07,
+      "loss": 0.7745,
+      "step": 6832
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3023139238357544,
+      "learning_rate": 2.8341925888730247e-07,
+      "loss": 0.8016,
+      "step": 6833
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2370109558105469,
+      "learning_rate": 2.8271526513259153e-07,
+      "loss": 0.7789,
+      "step": 6834
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2524189949035645,
+      "learning_rate": 2.820121213624116e-07,
+      "loss": 0.8635,
+      "step": 6835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5752331018447876,
+      "learning_rate": 2.813098277034576e-07,
+      "loss": 0.9991,
+      "step": 6836
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.572629451751709,
+      "learning_rate": 2.806083842822721e-07,
+      "loss": 1.0129,
+      "step": 6837
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.232875108718872,
+      "learning_rate": 2.799077912252446e-07,
+      "loss": 0.736,
+      "step": 6838
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3806121349334717,
+      "learning_rate": 2.7920804865861294e-07,
+      "loss": 0.7956,
+      "step": 6839
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2497613430023193,
+      "learning_rate": 2.7850915670845855e-07,
+      "loss": 0.8103,
+      "step": 6840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2424944639205933,
+      "learning_rate": 2.7781111550071293e-07,
+      "loss": 0.8251,
+      "step": 6841
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1957145929336548,
+      "learning_rate": 2.7711392516115153e-07,
+      "loss": 0.7499,
+      "step": 6842
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5771256685256958,
+      "learning_rate": 2.764175858153989e-07,
+      "loss": 1.0092,
+      "step": 6843
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3076105117797852,
+      "learning_rate": 2.757220975889241e-07,
+      "loss": 0.8475,
+      "step": 6844
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2953910827636719,
+      "learning_rate": 2.750274606070441e-07,
+      "loss": 0.7529,
+      "step": 6845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5956050753593445,
+      "learning_rate": 2.743336749949216e-07,
+      "loss": 1.0054,
+      "step": 6846
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3018313646316528,
+      "learning_rate": 2.7364074087756763e-07,
+      "loss": 0.8038,
+      "step": 6847
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2383977174758911,
+      "learning_rate": 2.7294865837983686e-07,
+      "loss": 0.8014,
+      "step": 6848
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2235174179077148,
+      "learning_rate": 2.7225742762643445e-07,
+      "loss": 0.8056,
+      "step": 6849
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.21023690700531,
+      "learning_rate": 2.7156704874190756e-07,
+      "loss": 0.7862,
+      "step": 6850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5704277157783508,
+      "learning_rate": 2.708775218506543e-07,
+      "loss": 0.9965,
+      "step": 6851
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6116350293159485,
+      "learning_rate": 2.7018884707691595e-07,
+      "loss": 1.0164,
+      "step": 6852
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3120330572128296,
+      "learning_rate": 2.695010245447821e-07,
+      "loss": 0.8606,
+      "step": 6853
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3077346086502075,
+      "learning_rate": 2.6881405437818695e-07,
+      "loss": 0.7752,
+      "step": 6854
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2373415231704712,
+      "learning_rate": 2.6812793670091377e-07,
+      "loss": 0.8034,
+      "step": 6855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.252881646156311,
+      "learning_rate": 2.6744267163658964e-07,
+      "loss": 0.8424,
+      "step": 6856
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2118339538574219,
+      "learning_rate": 2.667582593086898e-07,
+      "loss": 0.7671,
+      "step": 6857
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2492488622665405,
+      "learning_rate": 2.6607469984053446e-07,
+      "loss": 0.7821,
+      "step": 6858
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.278499722480774,
+      "learning_rate": 2.653919933552923e-07,
+      "loss": 0.7789,
+      "step": 6859
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2514303922653198,
+      "learning_rate": 2.6471013997597605e-07,
+      "loss": 0.8595,
+      "step": 6860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1990290880203247,
+      "learning_rate": 2.640291398254452e-07,
+      "loss": 0.7891,
+      "step": 6861
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2214308977127075,
+      "learning_rate": 2.633489930264055e-07,
+      "loss": 0.7603,
+      "step": 6862
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2378261089324951,
+      "learning_rate": 2.6266969970141054e-07,
+      "loss": 0.7924,
+      "step": 6863
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.238477349281311,
+      "learning_rate": 2.6199125997285746e-07,
+      "loss": 0.8892,
+      "step": 6864
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.194852352142334,
+      "learning_rate": 2.613136739629918e-07,
+      "loss": 0.8519,
+      "step": 6865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3215395212173462,
+      "learning_rate": 2.6063694179390477e-07,
+      "loss": 0.7729,
+      "step": 6866
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5702834129333496,
+      "learning_rate": 2.5996106358753323e-07,
+      "loss": 1.0001,
+      "step": 6867
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2074052095413208,
+      "learning_rate": 2.5928603946566046e-07,
+      "loss": 0.7981,
+      "step": 6868
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1542049646377563,
+      "learning_rate": 2.5861186954991514e-07,
+      "loss": 0.7505,
+      "step": 6869
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2585539817810059,
+      "learning_rate": 2.579385539617735e-07,
+      "loss": 0.8043,
+      "step": 6870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5744489431381226,
+      "learning_rate": 2.572660928225562e-07,
+      "loss": 1.001,
+      "step": 6871
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5714431405067444,
+      "learning_rate": 2.5659448625343085e-07,
+      "loss": 1.0329,
+      "step": 6872
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2488974332809448,
+      "learning_rate": 2.5592373437541117e-07,
+      "loss": 0.7731,
+      "step": 6873
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2368134260177612,
+      "learning_rate": 2.5525383730935714e-07,
+      "loss": 0.7637,
+      "step": 6874
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2444967031478882,
+      "learning_rate": 2.5458479517597445e-07,
+      "loss": 0.8319,
+      "step": 6875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5698719024658203,
+      "learning_rate": 2.539166080958133e-07,
+      "loss": 0.9901,
+      "step": 6876
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2751811742782593,
+      "learning_rate": 2.5324927618927077e-07,
+      "loss": 0.8018,
+      "step": 6877
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2174721956253052,
+      "learning_rate": 2.5258279957659226e-07,
+      "loss": 0.7335,
+      "step": 6878
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5771949887275696,
+      "learning_rate": 2.5191717837786566e-07,
+      "loss": 1.0071,
+      "step": 6879
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5677880048751831,
+      "learning_rate": 2.51252412713025e-07,
+      "loss": 0.9829,
+      "step": 6880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5672284364700317,
+      "learning_rate": 2.5058850270185286e-07,
+      "loss": 1.0013,
+      "step": 6881
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5809547305107117,
+      "learning_rate": 2.4992544846397513e-07,
+      "loss": 0.9945,
+      "step": 6882
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2062625885009766,
+      "learning_rate": 2.492632501188652e-07,
+      "loss": 0.7634,
+      "step": 6883
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2399228811264038,
+      "learning_rate": 2.4860190778584035e-07,
+      "loss": 0.8216,
+      "step": 6884
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.272971749305725,
+      "learning_rate": 2.4794142158406474e-07,
+      "loss": 0.8227,
+      "step": 6885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2310540676116943,
+      "learning_rate": 2.472817916325482e-07,
+      "loss": 0.7974,
+      "step": 6886
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2920582294464111,
+      "learning_rate": 2.466230180501472e-07,
+      "loss": 0.8125,
+      "step": 6887
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2122057676315308,
+      "learning_rate": 2.4596510095556083e-07,
+      "loss": 0.8263,
+      "step": 6888
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1639411449432373,
+      "learning_rate": 2.4530804046733746e-07,
+      "loss": 0.712,
+      "step": 6889
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1717811822891235,
+      "learning_rate": 2.446518367038703e-07,
+      "loss": 0.7773,
+      "step": 6890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2006983757019043,
+      "learning_rate": 2.4399648978339683e-07,
+      "loss": 0.8145,
+      "step": 6891
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5818468332290649,
+      "learning_rate": 2.4334199982399996e-07,
+      "loss": 1.0075,
+      "step": 6892
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3326045274734497,
+      "learning_rate": 2.426883669436092e-07,
+      "loss": 0.8077,
+      "step": 6893
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2491893768310547,
+      "learning_rate": 2.4203559126000085e-07,
+      "loss": 0.7885,
+      "step": 6894
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2696192264556885,
+      "learning_rate": 2.413836728907948e-07,
+      "loss": 0.7509,
+      "step": 6895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3044513463974,
+      "learning_rate": 2.4073261195345534e-07,
+      "loss": 0.7875,
+      "step": 6896
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.268727421760559,
+      "learning_rate": 2.4008240856529583e-07,
+      "loss": 0.795,
+      "step": 6897
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2411869764328003,
+      "learning_rate": 2.3943306284347325e-07,
+      "loss": 0.8249,
+      "step": 6898
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.243158221244812,
+      "learning_rate": 2.3878457490498995e-07,
+      "loss": 0.8133,
+      "step": 6899
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2694785594940186,
+      "learning_rate": 2.3813694486669313e-07,
+      "loss": 0.8108,
+      "step": 6900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1673656702041626,
+      "learning_rate": 2.3749017284527555e-07,
+      "loss": 0.8016,
+      "step": 6901
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1724536418914795,
+      "learning_rate": 2.3684425895727782e-07,
+      "loss": 0.8223,
+      "step": 6902
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.201448678970337,
+      "learning_rate": 2.3619920331908242e-07,
+      "loss": 0.7818,
+      "step": 6903
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.236837387084961,
+      "learning_rate": 2.355550060469186e-07,
+      "loss": 0.8102,
+      "step": 6904
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2470349073410034,
+      "learning_rate": 2.3491166725686244e-07,
+      "loss": 0.8198,
+      "step": 6905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5754885673522949,
+      "learning_rate": 2.3426918706483337e-07,
+      "loss": 1.046,
+      "step": 6906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.249061107635498,
+      "learning_rate": 2.3362756558659717e-07,
+      "loss": 0.8341,
+      "step": 6907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5770266652107239,
+      "learning_rate": 2.3298680293776353e-07,
+      "loss": 1.012,
+      "step": 6908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2507743835449219,
+      "learning_rate": 2.323468992337896e-07,
+      "loss": 0.7797,
+      "step": 6909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1834981441497803,
+      "learning_rate": 2.3170785458997592e-07,
+      "loss": 0.7415,
+      "step": 6910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.345505952835083,
+      "learning_rate": 2.310696691214681e-07,
+      "loss": 0.8264,
+      "step": 6911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2638113498687744,
+      "learning_rate": 2.3043234294325812e-07,
+      "loss": 0.7949,
+      "step": 6912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.298439621925354,
+      "learning_rate": 2.2979587617018295e-07,
+      "loss": 0.7817,
+      "step": 6913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2337627410888672,
+      "learning_rate": 2.2916026891692477e-07,
+      "loss": 0.7922,
+      "step": 6914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2299765348434448,
+      "learning_rate": 2.285255212980103e-07,
+      "loss": 0.7752,
+      "step": 6915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5746579766273499,
+      "learning_rate": 2.278916334278103e-07,
+      "loss": 1.0166,
+      "step": 6916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2377092838287354,
+      "learning_rate": 2.2725860542054445e-07,
+      "loss": 0.812,
+      "step": 6917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.326365351676941,
+      "learning_rate": 2.2662643739027324e-07,
+      "loss": 0.8496,
+      "step": 6918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.191691517829895,
+      "learning_rate": 2.259951294509044e-07,
+      "loss": 0.7973,
+      "step": 6919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2208912372589111,
+      "learning_rate": 2.253646817161892e-07,
+      "loss": 0.7984,
+      "step": 6920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5701229572296143,
+      "learning_rate": 2.2473509429972619e-07,
+      "loss": 1.0015,
+      "step": 6921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.299767255783081,
+      "learning_rate": 2.2410636731495794e-07,
+      "loss": 0.8613,
+      "step": 6922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2321319580078125,
+      "learning_rate": 2.234785008751711e-07,
+      "loss": 0.8028,
+      "step": 6923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2266806364059448,
+      "learning_rate": 2.228514950934968e-07,
+      "loss": 0.7203,
+      "step": 6924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2891896963119507,
+      "learning_rate": 2.2222535008291412e-07,
+      "loss": 0.7972,
+      "step": 6925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2565563917160034,
+      "learning_rate": 2.216000659562445e-07,
+      "loss": 0.8327,
+      "step": 6926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.213374137878418,
+      "learning_rate": 2.2097564282615446e-07,
+      "loss": 0.8442,
+      "step": 6927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.301681399345398,
+      "learning_rate": 2.2035208080515512e-07,
+      "loss": 0.781,
+      "step": 6928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.313215732574463,
+      "learning_rate": 2.1972938000560328e-07,
+      "loss": 0.766,
+      "step": 6929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2521374225616455,
+      "learning_rate": 2.1910754053970195e-07,
+      "loss": 0.7697,
+      "step": 6930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5767236351966858,
+      "learning_rate": 2.1848656251949652e-07,
+      "loss": 1.0202,
+      "step": 6931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2307151556015015,
+      "learning_rate": 2.1786644605687633e-07,
+      "loss": 0.8297,
+      "step": 6932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2834575176239014,
+      "learning_rate": 2.1724719126357984e-07,
+      "loss": 0.8084,
+      "step": 6933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5685036778450012,
+      "learning_rate": 2.166287982511861e-07,
+      "loss": 1.0272,
+      "step": 6934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3201220035552979,
+      "learning_rate": 2.1601126713111986e-07,
+      "loss": 0.7163,
+      "step": 6935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2806073427200317,
+      "learning_rate": 2.1539459801465158e-07,
+      "loss": 0.8218,
+      "step": 6936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2226486206054688,
+      "learning_rate": 2.1477879101289624e-07,
+      "loss": 0.7829,
+      "step": 6937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.624463677406311,
+      "learning_rate": 2.1416384623681284e-07,
+      "loss": 0.7906,
+      "step": 6938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5849138498306274,
+      "learning_rate": 2.1354976379720503e-07,
+      "loss": 1.0251,
+      "step": 6939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2850226163864136,
+      "learning_rate": 2.1293654380472207e-07,
+      "loss": 0.7875,
+      "step": 6940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2711312770843506,
+      "learning_rate": 2.1232418636985607e-07,
+      "loss": 0.8223,
+      "step": 6941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2124216556549072,
+      "learning_rate": 2.117126916029455e-07,
+      "loss": 0.8003,
+      "step": 6942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5729374885559082,
+      "learning_rate": 2.111020596141722e-07,
+      "loss": 1.0285,
+      "step": 6943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.253027081489563,
+      "learning_rate": 2.104922905135631e-07,
+      "loss": 0.8074,
+      "step": 6944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.254747748374939,
+      "learning_rate": 2.0988338441098932e-07,
+      "loss": 0.8201,
+      "step": 6945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.230934739112854,
+      "learning_rate": 2.092753414161669e-07,
+      "loss": 0.7919,
+      "step": 6946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.557386040687561,
+      "learning_rate": 2.08668161638656e-07,
+      "loss": 1.0235,
+      "step": 6947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5790265798568726,
+      "learning_rate": 2.0806184518786187e-07,
+      "loss": 1.0078,
+      "step": 6948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.209312081336975,
+      "learning_rate": 2.0745639217303382e-07,
+      "loss": 0.7687,
+      "step": 6949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2354280948638916,
+      "learning_rate": 2.0685180270326454e-07,
+      "loss": 0.7608,
+      "step": 6950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2114921808242798,
+      "learning_rate": 2.0624807688749195e-07,
+      "loss": 0.8391,
+      "step": 6951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.568632960319519,
+      "learning_rate": 2.056452148344995e-07,
+      "loss": 1.0176,
+      "step": 6952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5698532462120056,
+      "learning_rate": 2.0504321665291315e-07,
+      "loss": 1.0093,
+      "step": 6953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5827435255050659,
+      "learning_rate": 2.04442082451205e-07,
+      "loss": 1.0196,
+      "step": 6954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2853368520736694,
+      "learning_rate": 2.0384181233768952e-07,
+      "loss": 0.8368,
+      "step": 6955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.234094500541687,
+      "learning_rate": 2.0324240642052685e-07,
+      "loss": 0.7647,
+      "step": 6956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2287592887878418,
+      "learning_rate": 2.0264386480772112e-07,
+      "loss": 0.7824,
+      "step": 6957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2890527248382568,
+      "learning_rate": 2.0204618760712058e-07,
+      "loss": 0.8373,
+      "step": 6958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1921885013580322,
+      "learning_rate": 2.014493749264168e-07,
+      "loss": 0.769,
+      "step": 6959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2324106693267822,
+      "learning_rate": 2.0085342687314823e-07,
+      "loss": 0.8232,
+      "step": 6960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.573251485824585,
+      "learning_rate": 2.0025834355469397e-07,
+      "loss": 0.9976,
+      "step": 6961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5747997760772705,
+      "learning_rate": 1.9966412507828158e-07,
+      "loss": 1.016,
+      "step": 6962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2635639905929565,
+      "learning_rate": 1.9907077155097766e-07,
+      "loss": 0.7603,
+      "step": 6963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2470149993896484,
+      "learning_rate": 1.984782830796983e-07,
+      "loss": 0.78,
+      "step": 6964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2214655876159668,
+      "learning_rate": 1.9788665977119926e-07,
+      "loss": 0.7491,
+      "step": 6965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2593897581100464,
+      "learning_rate": 1.9729590173208357e-07,
+      "loss": 0.8009,
+      "step": 6966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2644621133804321,
+      "learning_rate": 1.96706009068795e-07,
+      "loss": 0.7949,
+      "step": 6967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.578827977180481,
+      "learning_rate": 1.9611698188762572e-07,
+      "loss": 0.9991,
+      "step": 6968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2185992002487183,
+      "learning_rate": 1.955288202947081e-07,
+      "loss": 0.8199,
+      "step": 6969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.27882981300354,
+      "learning_rate": 1.9494152439602122e-07,
+      "loss": 0.7797,
+      "step": 6970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2043981552124023,
+      "learning_rate": 1.9435509429738598e-07,
+      "loss": 0.8247,
+      "step": 6971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2752784490585327,
+      "learning_rate": 1.9376953010446953e-07,
+      "loss": 0.7753,
+      "step": 6972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2182774543762207,
+      "learning_rate": 1.9318483192278137e-07,
+      "loss": 0.7878,
+      "step": 6973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1930710077285767,
+      "learning_rate": 1.92600999857675e-07,
+      "loss": 0.8179,
+      "step": 6974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.254098653793335,
+      "learning_rate": 1.9201803401434848e-07,
+      "loss": 0.7988,
+      "step": 6975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5649222731590271,
+      "learning_rate": 1.9143593449784393e-07,
+      "loss": 1.0071,
+      "step": 6976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2716392278671265,
+      "learning_rate": 1.908547014130463e-07,
+      "loss": 0.803,
+      "step": 6977
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5687278509140015,
+      "learning_rate": 1.9027433486468516e-07,
+      "loss": 1.0088,
+      "step": 6978
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2553269863128662,
+      "learning_rate": 1.896948349573352e-07,
+      "loss": 0.8081,
+      "step": 6979
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2561066150665283,
+      "learning_rate": 1.891162017954129e-07,
+      "loss": 0.7923,
+      "step": 6980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.24393630027771,
+      "learning_rate": 1.8853843548317928e-07,
+      "loss": 0.7389,
+      "step": 6981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.245293378829956,
+      "learning_rate": 1.8796153612473887e-07,
+      "loss": 0.7501,
+      "step": 6982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5495428442955017,
+      "learning_rate": 1.873855038240413e-07,
+      "loss": 1.0142,
+      "step": 6983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.240867257118225,
+      "learning_rate": 1.8681033868487796e-07,
+      "loss": 0.8308,
+      "step": 6984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2157691717147827,
+      "learning_rate": 1.8623604081088543e-07,
+      "loss": 0.7938,
+      "step": 6985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5723128318786621,
+      "learning_rate": 1.8566261030554366e-07,
+      "loss": 0.9965,
+      "step": 6986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5702294111251831,
+      "learning_rate": 1.850900472721767e-07,
+      "loss": 1.0291,
+      "step": 6987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.572680652141571,
+      "learning_rate": 1.8451835181395206e-07,
+      "loss": 1.0148,
+      "step": 6988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1912925243377686,
+      "learning_rate": 1.8394752403387949e-07,
+      "loss": 0.8327,
+      "step": 6989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.205457329750061,
+      "learning_rate": 1.8337756403481456e-07,
+      "loss": 0.8058,
+      "step": 6990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3002398014068604,
+      "learning_rate": 1.828084719194556e-07,
+      "loss": 0.8322,
+      "step": 6991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1940313577651978,
+      "learning_rate": 1.822402477903451e-07,
+      "loss": 0.8048,
+      "step": 6992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5664176940917969,
+      "learning_rate": 1.816728917498667e-07,
+      "loss": 1.0154,
+      "step": 6993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.416250467300415,
+      "learning_rate": 1.811064039002508e-07,
+      "loss": 0.7434,
+      "step": 6994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.248412847518921,
+      "learning_rate": 1.8054078434357082e-07,
+      "loss": 0.841,
+      "step": 6995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.280069351196289,
+      "learning_rate": 1.7997603318174183e-07,
+      "loss": 0.7862,
+      "step": 6996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2175836563110352,
+      "learning_rate": 1.7941215051652417e-07,
+      "loss": 0.7874,
+      "step": 6997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5743497610092163,
+      "learning_rate": 1.788491364495204e-07,
+      "loss": 0.9916,
+      "step": 6998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2713037729263306,
+      "learning_rate": 1.782869910821783e-07,
+      "loss": 0.7543,
+      "step": 6999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.251247525215149,
+      "learning_rate": 1.7772571451578735e-07,
+      "loss": 0.7406,
+      "step": 7000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5832956433296204,
+      "learning_rate": 1.7716530685148115e-07,
+      "loss": 1.0249,
+      "step": 7001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.244388461112976,
+      "learning_rate": 1.7660576819023668e-07,
+      "loss": 0.7962,
+      "step": 7002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2446473836898804,
+      "learning_rate": 1.7604709863287605e-07,
+      "loss": 0.8322,
+      "step": 7003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2390727996826172,
+      "learning_rate": 1.7548929828006156e-07,
+      "loss": 0.8333,
+      "step": 7004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.265989899635315,
+      "learning_rate": 1.7493236723230167e-07,
+      "loss": 0.8004,
+      "step": 7005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.283776044845581,
+      "learning_rate": 1.74376305589945e-07,
+      "loss": 0.797,
+      "step": 7006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2535589933395386,
+      "learning_rate": 1.7382111345318808e-07,
+      "loss": 0.802,
+      "step": 7007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2203092575073242,
+      "learning_rate": 1.7326679092206756e-07,
+      "loss": 0.8248,
+      "step": 7008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2494826316833496,
+      "learning_rate": 1.7271333809646295e-07,
+      "loss": 0.8322,
+      "step": 7009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2308465242385864,
+      "learning_rate": 1.7216075507609896e-07,
+      "loss": 0.8105,
+      "step": 7010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2348952293395996,
+      "learning_rate": 1.7160904196054317e-07,
+      "loss": 0.8543,
+      "step": 7011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3342347145080566,
+      "learning_rate": 1.7105819884920604e-07,
+      "loss": 0.8252,
+      "step": 7012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2839523553848267,
+      "learning_rate": 1.7050822584134097e-07,
+      "loss": 0.8017,
+      "step": 7013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2614668607711792,
+      "learning_rate": 1.6995912303604534e-07,
+      "loss": 0.7834,
+      "step": 7014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5797383785247803,
+      "learning_rate": 1.6941089053225833e-07,
+      "loss": 1.0038,
+      "step": 7015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1452603340148926,
+      "learning_rate": 1.6886352842876485e-07,
+      "loss": 0.7537,
+      "step": 7016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2364797592163086,
+      "learning_rate": 1.6831703682418932e-07,
+      "loss": 0.8279,
+      "step": 7017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.235158920288086,
+      "learning_rate": 1.6777141581700295e-07,
+      "loss": 0.8259,
+      "step": 7018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5805956721305847,
+      "learning_rate": 1.6722666550551881e-07,
+      "loss": 1.0161,
+      "step": 7019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2888524532318115,
+      "learning_rate": 1.6668278598789167e-07,
+      "loss": 0.8064,
+      "step": 7020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.266817569732666,
+      "learning_rate": 1.6613977736212095e-07,
+      "loss": 0.7955,
+      "step": 7021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.259655237197876,
+      "learning_rate": 1.6559763972604891e-07,
+      "loss": 0.799,
+      "step": 7022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.238930344581604,
+      "learning_rate": 1.6505637317736078e-07,
+      "loss": 0.7785,
+      "step": 7023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3018196821212769,
+      "learning_rate": 1.6451597781358463e-07,
+      "loss": 0.8039,
+      "step": 7024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.583430290222168,
+      "learning_rate": 1.6397645373209093e-07,
+      "loss": 1.0111,
+      "step": 7025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5783350467681885,
+      "learning_rate": 1.634378010300941e-07,
+      "loss": 0.9854,
+      "step": 7026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1911897659301758,
+      "learning_rate": 1.6290001980465263e-07,
+      "loss": 0.8256,
+      "step": 7027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2214738130569458,
+      "learning_rate": 1.6236311015266627e-07,
+      "loss": 0.7776,
+      "step": 7028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2608901262283325,
+      "learning_rate": 1.6182707217087646e-07,
+      "loss": 0.8314,
+      "step": 7029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5837475657463074,
+      "learning_rate": 1.6129190595587153e-07,
+      "loss": 1.0241,
+      "step": 7030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5837562680244446,
+      "learning_rate": 1.6075761160407932e-07,
+      "loss": 1.0076,
+      "step": 7031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2151938676834106,
+      "learning_rate": 1.6022418921177175e-07,
+      "loss": 0.8073,
+      "step": 7032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2485147714614868,
+      "learning_rate": 1.59691638875063e-07,
+      "loss": 0.7977,
+      "step": 7033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2041583061218262,
+      "learning_rate": 1.591599606899119e-07,
+      "loss": 0.8009,
+      "step": 7034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5786846876144409,
+      "learning_rate": 1.5862915475211905e-07,
+      "loss": 1.0155,
+      "step": 7035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5675244927406311,
+      "learning_rate": 1.580992211573268e-07,
+      "loss": 1.0112,
+      "step": 7036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5785964131355286,
+      "learning_rate": 1.57570160001021e-07,
+      "loss": 0.9927,
+      "step": 7037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5639241337776184,
+      "learning_rate": 1.570419713785326e-07,
+      "loss": 1.0417,
+      "step": 7038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2504756450653076,
+      "learning_rate": 1.5651465538503107e-07,
+      "loss": 0.8068,
+      "step": 7039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.246934175491333,
+      "learning_rate": 1.559882121155326e-07,
+      "loss": 0.8578,
+      "step": 7040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3662998676300049,
+      "learning_rate": 1.55462641664893e-07,
+      "loss": 0.7854,
+      "step": 7041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.302665114402771,
+      "learning_rate": 1.5493794412781316e-07,
+      "loss": 0.8277,
+      "step": 7042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.202092170715332,
+      "learning_rate": 1.5441411959883578e-07,
+      "loss": 0.7709,
+      "step": 7043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2787953615188599,
+      "learning_rate": 1.5389116817234594e-07,
+      "loss": 0.7643,
+      "step": 7044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.180902361869812,
+      "learning_rate": 1.533690899425716e-07,
+      "loss": 0.8361,
+      "step": 7045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.251002550125122,
+      "learning_rate": 1.5284788500358417e-07,
+      "loss": 0.8039,
+      "step": 7046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3199633359909058,
+      "learning_rate": 1.5232755344929627e-07,
+      "loss": 0.8254,
+      "step": 7047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5770930647850037,
+      "learning_rate": 1.5180809537346408e-07,
+      "loss": 1.0064,
+      "step": 7048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5754579305648804,
+      "learning_rate": 1.5128951086968657e-07,
+      "loss": 0.9899,
+      "step": 7049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2715699672698975,
+      "learning_rate": 1.5077180003140458e-07,
+      "loss": 0.8473,
+      "step": 7050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2020928859710693,
+      "learning_rate": 1.5025496295190234e-07,
+      "loss": 0.7899,
+      "step": 7051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5626668930053711,
+      "learning_rate": 1.4973899972430485e-07,
+      "loss": 1.0122,
+      "step": 7052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.201998233795166,
+      "learning_rate": 1.4922391044158324e-07,
+      "loss": 0.7922,
+      "step": 7053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2261775732040405,
+      "learning_rate": 1.487096951965472e-07,
+      "loss": 0.8179,
+      "step": 7054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2450426816940308,
+      "learning_rate": 1.4819635408185152e-07,
+      "loss": 0.8422,
+      "step": 7055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2052007913589478,
+      "learning_rate": 1.4768388718999216e-07,
+      "loss": 0.7599,
+      "step": 7056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2881929874420166,
+      "learning_rate": 1.4717229461330807e-07,
+      "loss": 0.771,
+      "step": 7057
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859953165054321,
+      "learning_rate": 1.466615764439805e-07,
+      "loss": 1.0164,
+      "step": 7058
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2191458940505981,
+      "learning_rate": 1.4615173277403417e-07,
+      "loss": 0.8258,
+      "step": 7059
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.222584843635559,
+      "learning_rate": 1.4564276369533392e-07,
+      "loss": 0.8407,
+      "step": 7060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.24533212184906,
+      "learning_rate": 1.4513466929958976e-07,
+      "loss": 0.7963,
+      "step": 7061
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2708420753479004,
+      "learning_rate": 1.4462744967835174e-07,
+      "loss": 0.8178,
+      "step": 7062
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3065118789672852,
+      "learning_rate": 1.44121104923014e-07,
+      "loss": 0.7707,
+      "step": 7063
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.232577919960022,
+      "learning_rate": 1.4361563512481193e-07,
+      "loss": 0.7695,
+      "step": 7064
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5713006258010864,
+      "learning_rate": 1.431110403748237e-07,
+      "loss": 1.0124,
+      "step": 7065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2418582439422607,
+      "learning_rate": 1.4260732076396943e-07,
+      "loss": 0.7591,
+      "step": 7066
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.164595365524292,
+      "learning_rate": 1.4210447638301318e-07,
+      "loss": 0.7396,
+      "step": 7067
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2428542375564575,
+      "learning_rate": 1.4160250732255797e-07,
+      "loss": 0.8043,
+      "step": 7068
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.201924204826355,
+      "learning_rate": 1.4110141367305318e-07,
+      "loss": 0.784,
+      "step": 7069
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5623791217803955,
+      "learning_rate": 1.4060119552478768e-07,
+      "loss": 1.0098,
+      "step": 7070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2041786909103394,
+      "learning_rate": 1.4010185296789326e-07,
+      "loss": 0.7867,
+      "step": 7071
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2879071235656738,
+      "learning_rate": 1.3960338609234349e-07,
+      "loss": 0.8232,
+      "step": 7072
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2693487405776978,
+      "learning_rate": 1.391057949879554e-07,
+      "loss": 0.8195,
+      "step": 7073
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.295578122138977,
+      "learning_rate": 1.3860907974438731e-07,
+      "loss": 0.8268,
+      "step": 7074
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2622859477996826,
+      "learning_rate": 1.3811324045113982e-07,
+      "loss": 0.7205,
+      "step": 7075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.247297763824463,
+      "learning_rate": 1.3761827719755594e-07,
+      "loss": 0.7587,
+      "step": 7076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7774195671081543,
+      "learning_rate": 1.371241900728204e-07,
+      "loss": 0.7729,
+      "step": 7077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3131591081619263,
+      "learning_rate": 1.3663097916596147e-07,
+      "loss": 0.8033,
+      "step": 7078
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5827656388282776,
+      "learning_rate": 1.3613864456584692e-07,
+      "loss": 1.0078,
+      "step": 7079
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5663161277770996,
+      "learning_rate": 1.3564718636118857e-07,
+      "loss": 0.9887,
+      "step": 7080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2696895599365234,
+      "learning_rate": 1.3515660464054004e-07,
+      "loss": 0.7683,
+      "step": 7081
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5737518668174744,
+      "learning_rate": 1.3466689949229672e-07,
+      "loss": 1.0179,
+      "step": 7082
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3904962539672852,
+      "learning_rate": 1.3417807100469692e-07,
+      "loss": 0.8033,
+      "step": 7083
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2225650548934937,
+      "learning_rate": 1.3369011926581854e-07,
+      "loss": 0.769,
+      "step": 7084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5696855783462524,
+      "learning_rate": 1.3320304436358567e-07,
+      "loss": 1.0173,
+      "step": 7085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2660095691680908,
+      "learning_rate": 1.3271684638576033e-07,
+      "loss": 0.7712,
+      "step": 7086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1161227226257324,
+      "learning_rate": 1.3223152541994743e-07,
+      "loss": 0.7376,
+      "step": 7087
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.242422103881836,
+      "learning_rate": 1.3174708155359707e-07,
+      "loss": 0.7704,
+      "step": 7088
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2605444192886353,
+      "learning_rate": 1.312635148739966e-07,
+      "loss": 0.8326,
+      "step": 7089
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5768649578094482,
+      "learning_rate": 1.3078082546827797e-07,
+      "loss": 1.0031,
+      "step": 7090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.293475866317749,
+      "learning_rate": 1.3029901342341612e-07,
+      "loss": 0.7787,
+      "step": 7091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.261337399482727,
+      "learning_rate": 1.2981807882622434e-07,
+      "loss": 0.767,
+      "step": 7092
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.229882001876831,
+      "learning_rate": 1.293380217633611e-07,
+      "loss": 0.7887,
+      "step": 7093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5618970394134521,
+      "learning_rate": 1.288588423213255e-07,
+      "loss": 1.0177,
+      "step": 7094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2989835739135742,
+      "learning_rate": 1.28380540586458e-07,
+      "loss": 0.8567,
+      "step": 7095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.229612112045288,
+      "learning_rate": 1.2790311664494182e-07,
+      "loss": 0.793,
+      "step": 7096
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2576173543930054,
+      "learning_rate": 1.2742657058280205e-07,
+      "loss": 0.856,
+      "step": 7097
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2250933647155762,
+      "learning_rate": 1.2695090248590392e-07,
+      "loss": 0.7257,
+      "step": 7098
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.246089220046997,
+      "learning_rate": 1.2647611243995715e-07,
+      "loss": 0.7849,
+      "step": 7099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5687388181686401,
+      "learning_rate": 1.260022005305106e-07,
+      "loss": 1.0157,
+      "step": 7100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5702821612358093,
+      "learning_rate": 1.2552916684295759e-07,
+      "loss": 1.0268,
+      "step": 7101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.179696798324585,
+      "learning_rate": 1.2505701146253102e-07,
+      "loss": 0.7941,
+      "step": 7102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2344393730163574,
+      "learning_rate": 1.2458573447430567e-07,
+      "loss": 0.7742,
+      "step": 7103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2715243101119995,
+      "learning_rate": 1.241153359631997e-07,
+      "loss": 0.8372,
+      "step": 7104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2417117357254028,
+      "learning_rate": 1.236458160139714e-07,
+      "loss": 0.7957,
+      "step": 7105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5714539289474487,
+      "learning_rate": 1.2317717471122092e-07,
+      "loss": 0.9761,
+      "step": 7106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5683950185775757,
+      "learning_rate": 1.2270941213939124e-07,
+      "loss": 0.9926,
+      "step": 7107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5755870938301086,
+      "learning_rate": 1.2224252838276495e-07,
+      "loss": 1.0172,
+      "step": 7108
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.308260202407837,
+      "learning_rate": 1.2177652352546976e-07,
+      "loss": 0.8561,
+      "step": 7109
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5727825164794922,
+      "learning_rate": 1.213113976514707e-07,
+      "loss": 1.0155,
+      "step": 7110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2438232898712158,
+      "learning_rate": 1.2084715084457688e-07,
+      "loss": 0.7548,
+      "step": 7111
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2329399585723877,
+      "learning_rate": 1.203837831884397e-07,
+      "loss": 0.7972,
+      "step": 7112
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.20693838596344,
+      "learning_rate": 1.1992129476655067e-07,
+      "loss": 0.7341,
+      "step": 7113
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5779739618301392,
+      "learning_rate": 1.1945968566224263e-07,
+      "loss": 0.983,
+      "step": 7114
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5560532808303833,
+      "learning_rate": 1.1899895595869182e-07,
+      "loss": 1.021,
+      "step": 7115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.282448649406433,
+      "learning_rate": 1.1853910573891348e-07,
+      "loss": 0.8266,
+      "step": 7116
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2285127639770508,
+      "learning_rate": 1.1808013508576743e-07,
+      "loss": 0.8253,
+      "step": 7117
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2579556703567505,
+      "learning_rate": 1.1762204408195199e-07,
+      "loss": 0.7672,
+      "step": 7118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5699810981750488,
+      "learning_rate": 1.1716483281000835e-07,
+      "loss": 1.0132,
+      "step": 7119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5743225812911987,
+      "learning_rate": 1.167085013523206e-07,
+      "loss": 0.9971,
+      "step": 7120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2777481079101562,
+      "learning_rate": 1.1625304979111185e-07,
+      "loss": 0.8124,
+      "step": 7121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.28805673122406,
+      "learning_rate": 1.1579847820844703e-07,
+      "loss": 0.7774,
+      "step": 7122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.322843074798584,
+      "learning_rate": 1.153447866862345e-07,
+      "loss": 0.858,
+      "step": 7123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5846237540245056,
+      "learning_rate": 1.148919753062222e-07,
+      "loss": 1.0271,
+      "step": 7124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1469130516052246,
+      "learning_rate": 1.1444004414999987e-07,
+      "loss": 0.7698,
+      "step": 7125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.183838129043579,
+      "learning_rate": 1.1398899329899849e-07,
+      "loss": 0.7928,
+      "step": 7126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2481929063796997,
+      "learning_rate": 1.1353882283449135e-07,
+      "loss": 0.7971,
+      "step": 7127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5774111151695251,
+      "learning_rate": 1.130895328375925e-07,
+      "loss": 0.9985,
+      "step": 7128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2155998945236206,
+      "learning_rate": 1.1264112338925715e-07,
+      "loss": 0.7329,
+      "step": 7129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5727185606956482,
+      "learning_rate": 1.1219359457028123e-07,
+      "loss": 1.0066,
+      "step": 7130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5763208270072937,
+      "learning_rate": 1.1174694646130413e-07,
+      "loss": 0.9968,
+      "step": 7131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3180593252182007,
+      "learning_rate": 1.1130117914280369e-07,
+      "loss": 0.7786,
+      "step": 7132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3010015487670898,
+      "learning_rate": 1.1085629269510234e-07,
+      "loss": 0.7439,
+      "step": 7133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2569572925567627,
+      "learning_rate": 1.1041228719836039e-07,
+      "loss": 0.7833,
+      "step": 7134
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.579408586025238,
+      "learning_rate": 1.0996916273258162e-07,
+      "loss": 1.0052,
+      "step": 7135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.291548728942871,
+      "learning_rate": 1.0952691937761107e-07,
+      "loss": 0.8048,
+      "step": 7136
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2539783716201782,
+      "learning_rate": 1.0908555721313385e-07,
+      "loss": 0.8015,
+      "step": 7137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5657061338424683,
+      "learning_rate": 1.0864507631867582e-07,
+      "loss": 0.9893,
+      "step": 7138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2088422775268555,
+      "learning_rate": 1.0820547677360738e-07,
+      "loss": 0.7655,
+      "step": 7139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2737207412719727,
+      "learning_rate": 1.077667586571357e-07,
+      "loss": 0.7674,
+      "step": 7140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5815380215644836,
+      "learning_rate": 1.0732892204831258e-07,
+      "loss": 1.0333,
+      "step": 7141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2991743087768555,
+      "learning_rate": 1.0689196702602823e-07,
+      "loss": 0.8554,
+      "step": 7142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5846630334854126,
+      "learning_rate": 1.0645589366901743e-07,
+      "loss": 0.9865,
+      "step": 7143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5674802660942078,
+      "learning_rate": 1.0602070205585347e-07,
+      "loss": 1.0004,
+      "step": 7144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.574319064617157,
+      "learning_rate": 1.0558639226495026e-07,
+      "loss": 1.0077,
+      "step": 7145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.194970965385437,
+      "learning_rate": 1.0515296437456468e-07,
+      "loss": 0.8439,
+      "step": 7146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.265079140663147,
+      "learning_rate": 1.0472041846279479e-07,
+      "loss": 0.81,
+      "step": 7147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5616117119789124,
+      "learning_rate": 1.0428875460757715e-07,
+      "loss": 1.0012,
+      "step": 7148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2301971912384033,
+      "learning_rate": 1.0385797288669286e-07,
+      "loss": 0.7955,
+      "step": 7149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.17631995677948,
+      "learning_rate": 1.034280733777615e-07,
+      "loss": 0.7929,
+      "step": 7150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3198875188827515,
+      "learning_rate": 1.0299905615824501e-07,
+      "loss": 0.7954,
+      "step": 7151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2841322422027588,
+      "learning_rate": 1.0257092130544599e-07,
+      "loss": 0.8061,
+      "step": 7152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2626068592071533,
+      "learning_rate": 1.0214366889650718e-07,
+      "loss": 0.7989,
+      "step": 7153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.214251160621643,
+      "learning_rate": 1.0171729900841365e-07,
+      "loss": 0.8072,
+      "step": 7154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2968263626098633,
+      "learning_rate": 1.0129181171799063e-07,
+      "loss": 0.8613,
+      "step": 7155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1995916366577148,
+      "learning_rate": 1.0086720710190511e-07,
+      "loss": 0.8099,
+      "step": 7156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3012282848358154,
+      "learning_rate": 1.004434852366637e-07,
+      "loss": 0.8213,
+      "step": 7157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.283403992652893,
+      "learning_rate": 1.0002064619861585e-07,
+      "loss": 0.8139,
+      "step": 7158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2892682552337646,
+      "learning_rate": 9.959869006395062e-08,
+      "loss": 0.8381,
+      "step": 7159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5753827095031738,
+      "learning_rate": 9.917761690869777e-08,
+      "loss": 0.9989,
+      "step": 7160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2528854608535767,
+      "learning_rate": 9.875742680872824e-08,
+      "loss": 0.7985,
+      "step": 7161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2162529230117798,
+      "learning_rate": 9.833811983975483e-08,
+      "loss": 0.7716,
+      "step": 7162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5883492231369019,
+      "learning_rate": 9.791969607732987e-08,
+      "loss": 0.9924,
+      "step": 7163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2359397411346436,
+      "learning_rate": 9.750215559684639e-08,
+      "loss": 0.7926,
+      "step": 7164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5605860948562622,
+      "learning_rate": 9.708549847354031e-08,
+      "loss": 0.9735,
+      "step": 7165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2556216716766357,
+      "learning_rate": 9.666972478248715e-08,
+      "loss": 0.7766,
+      "step": 7166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2020580768585205,
+      "learning_rate": 9.625483459860196e-08,
+      "loss": 0.8064,
+      "step": 7167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2146881818771362,
+      "learning_rate": 9.584082799664273e-08,
+      "loss": 0.7605,
+      "step": 7168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5801219940185547,
+      "learning_rate": 9.542770505120591e-08,
+      "loss": 0.999,
+      "step": 7169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2330830097198486,
+      "learning_rate": 9.501546583673194e-08,
+      "loss": 0.8079,
+      "step": 7170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2613060474395752,
+      "learning_rate": 9.460411042749918e-08,
+      "loss": 0.7756,
+      "step": 7171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.426759958267212,
+      "learning_rate": 9.419363889762779e-08,
+      "loss": 0.7907,
+      "step": 7172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2245656251907349,
+      "learning_rate": 9.378405132107804e-08,
+      "loss": 0.8578,
+      "step": 7173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3226288557052612,
+      "learning_rate": 9.337534777165313e-08,
+      "loss": 0.7987,
+      "step": 7174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2626415491104126,
+      "learning_rate": 9.296752832299472e-08,
+      "loss": 0.8115,
+      "step": 7175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5812157988548279,
+      "learning_rate": 9.256059304858512e-08,
+      "loss": 0.9956,
+      "step": 7176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2724014520645142,
+      "learning_rate": 9.215454202174844e-08,
+      "loss": 0.7884,
+      "step": 7177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2631239891052246,
+      "learning_rate": 9.174937531564898e-08,
+      "loss": 0.8173,
+      "step": 7178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.276317834854126,
+      "learning_rate": 9.13450930032922e-08,
+      "loss": 0.8471,
+      "step": 7179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.242375135421753,
+      "learning_rate": 9.094169515752315e-08,
+      "loss": 0.7673,
+      "step": 7180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.279054880142212,
+      "learning_rate": 9.053918185102762e-08,
+      "loss": 0.8103,
+      "step": 7181
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.301270842552185,
+      "learning_rate": 9.013755315633421e-08,
+      "loss": 0.8183,
+      "step": 7182
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5796676874160767,
+      "learning_rate": 8.973680914580951e-08,
+      "loss": 1.0025,
+      "step": 7183
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.25609290599823,
+      "learning_rate": 8.93369498916613e-08,
+      "loss": 0.7468,
+      "step": 7184
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5555849075317383,
+      "learning_rate": 8.893797546593808e-08,
+      "loss": 1.0068,
+      "step": 7185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2229411602020264,
+      "learning_rate": 8.853988594053009e-08,
+      "loss": 0.8204,
+      "step": 7186
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.235015869140625,
+      "learning_rate": 8.814268138716664e-08,
+      "loss": 0.7241,
+      "step": 7187
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3300455808639526,
+      "learning_rate": 8.77463618774177e-08,
+      "loss": 0.7943,
+      "step": 7188
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4680370092391968,
+      "learning_rate": 8.735092748269447e-08,
+      "loss": 0.8113,
+      "step": 7189
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5857223868370056,
+      "learning_rate": 8.695637827424886e-08,
+      "loss": 0.9823,
+      "step": 7190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2505749464035034,
+      "learning_rate": 8.656271432317232e-08,
+      "loss": 0.8015,
+      "step": 7191
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2193583250045776,
+      "learning_rate": 8.6169935700397e-08,
+      "loss": 0.7536,
+      "step": 7192
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4790253639221191,
+      "learning_rate": 8.577804247669574e-08,
+      "loss": 0.784,
+      "step": 7193
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5733036994934082,
+      "learning_rate": 8.538703472268262e-08,
+      "loss": 1.0519,
+      "step": 7194
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2515619993209839,
+      "learning_rate": 8.499691250881071e-08,
+      "loss": 0.7996,
+      "step": 7195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2323033809661865,
+      "learning_rate": 8.460767590537432e-08,
+      "loss": 0.7866,
+      "step": 7196
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3477396965026855,
+      "learning_rate": 8.421932498250796e-08,
+      "loss": 0.8419,
+      "step": 7197
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2708399295806885,
+      "learning_rate": 8.383185981018726e-08,
+      "loss": 0.7915,
+      "step": 7198
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3195909261703491,
+      "learning_rate": 8.344528045822753e-08,
+      "loss": 0.8019,
+      "step": 7199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5792005062103271,
+      "learning_rate": 8.305958699628413e-08,
+      "loss": 1.0153,
+      "step": 7200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2505463361740112,
+      "learning_rate": 8.267477949385372e-08,
+      "loss": 0.773,
+      "step": 7201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5734670162200928,
+      "learning_rate": 8.229085802027304e-08,
+      "loss": 1.0298,
+      "step": 7202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2525179386138916,
+      "learning_rate": 8.190782264471841e-08,
+      "loss": 0.7548,
+      "step": 7203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2162048816680908,
+      "learning_rate": 8.152567343620743e-08,
+      "loss": 0.7626,
+      "step": 7204
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2151029109954834,
+      "learning_rate": 8.114441046359778e-08,
+      "loss": 0.7344,
+      "step": 7205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2673426866531372,
+      "learning_rate": 8.076403379558728e-08,
+      "loss": 0.795,
+      "step": 7206
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.257444143295288,
+      "learning_rate": 8.038454350071445e-08,
+      "loss": 0.7763,
+      "step": 7207
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1726815700531006,
+      "learning_rate": 8.000593964735737e-08,
+      "loss": 0.7944,
+      "step": 7208
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5883839130401611,
+      "learning_rate": 7.962822230373479e-08,
+      "loss": 1.0002,
+      "step": 7209
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2563835382461548,
+      "learning_rate": 7.925139153790617e-08,
+      "loss": 0.8267,
+      "step": 7210
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5782341957092285,
+      "learning_rate": 7.887544741777109e-08,
+      "loss": 1.0066,
+      "step": 7211
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2439852952957153,
+      "learning_rate": 7.850039001106758e-08,
+      "loss": 0.7982,
+      "step": 7212
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2664237022399902,
+      "learning_rate": 7.812621938537657e-08,
+      "loss": 0.8449,
+      "step": 7213
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2019678354263306,
+      "learning_rate": 7.775293560811802e-08,
+      "loss": 0.7555,
+      "step": 7214
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.255826711654663,
+      "learning_rate": 7.73805387465526e-08,
+      "loss": 0.7922,
+      "step": 7215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3038368225097656,
+      "learning_rate": 7.700902886777939e-08,
+      "loss": 0.8195,
+      "step": 7216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2969248294830322,
+      "learning_rate": 7.663840603873984e-08,
+      "loss": 0.7545,
+      "step": 7217
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.237287998199463,
+      "learning_rate": 7.626867032621444e-08,
+      "loss": 0.7611,
+      "step": 7218
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.297176718711853,
+      "learning_rate": 7.589982179682431e-08,
+      "loss": 0.7947,
+      "step": 7219
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.568422794342041,
+      "learning_rate": 7.553186051702965e-08,
+      "loss": 0.9955,
+      "step": 7220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.586800217628479,
+      "learning_rate": 7.516478655313187e-08,
+      "loss": 1.0129,
+      "step": 7221
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2192249298095703,
+      "learning_rate": 7.479859997127359e-08,
+      "loss": 0.733,
+      "step": 7222
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2181516885757446,
+      "learning_rate": 7.443330083743483e-08,
+      "loss": 0.7849,
+      "step": 7223
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2502185106277466,
+      "learning_rate": 7.406888921743682e-08,
+      "loss": 0.8434,
+      "step": 7224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5659818649291992,
+      "learning_rate": 7.370536517694204e-08,
+      "loss": 1.0145,
+      "step": 7225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2169159650802612,
+      "learning_rate": 7.3342728781452e-08,
+      "loss": 0.8002,
+      "step": 7226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.26487398147583,
+      "learning_rate": 7.298098009630772e-08,
+      "loss": 0.755,
+      "step": 7227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2472271919250488,
+      "learning_rate": 7.262011918669153e-08,
+      "loss": 0.8004,
+      "step": 7228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.225090742111206,
+      "learning_rate": 7.226014611762478e-08,
+      "loss": 0.7827,
+      "step": 7229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5720840692520142,
+      "learning_rate": 7.190106095396942e-08,
+      "loss": 0.9857,
+      "step": 7230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.214021921157837,
+      "learning_rate": 7.154286376042763e-08,
+      "loss": 0.7847,
+      "step": 7231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.224353313446045,
+      "learning_rate": 7.118555460154109e-08,
+      "loss": 0.7961,
+      "step": 7232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.568952739238739,
+      "learning_rate": 7.082913354169108e-08,
+      "loss": 1.0139,
+      "step": 7233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.424709677696228,
+      "learning_rate": 7.047360064510011e-08,
+      "loss": 0.7753,
+      "step": 7234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1889317035675049,
+      "learning_rate": 7.011895597582862e-08,
+      "loss": 0.7874,
+      "step": 7235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2289395332336426,
+      "learning_rate": 6.976519959777995e-08,
+      "loss": 0.8017,
+      "step": 7236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.562727153301239,
+      "learning_rate": 6.941233157469418e-08,
+      "loss": 1.0004,
+      "step": 7237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5788621306419373,
+      "learning_rate": 6.906035197015437e-08,
+      "loss": 1.0063,
+      "step": 7238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2555654048919678,
+      "learning_rate": 6.870926084758145e-08,
+      "loss": 0.8202,
+      "step": 7239
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1879645586013794,
+      "learning_rate": 6.835905827023647e-08,
+      "loss": 0.7912,
+      "step": 7240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2070783376693726,
+      "learning_rate": 6.800974430122065e-08,
+      "loss": 0.7875,
+      "step": 7241
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3206452131271362,
+      "learning_rate": 6.766131900347583e-08,
+      "loss": 0.794,
+      "step": 7242
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3165377378463745,
+      "learning_rate": 6.731378243978237e-08,
+      "loss": 0.7788,
+      "step": 7243
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.584885835647583,
+      "learning_rate": 6.696713467276184e-08,
+      "loss": 1.0275,
+      "step": 7244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5837993025779724,
+      "learning_rate": 6.662137576487427e-08,
+      "loss": 0.9745,
+      "step": 7245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2463370561599731,
+      "learning_rate": 6.627650577842093e-08,
+      "loss": 0.7792,
+      "step": 7246
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.263026475906372,
+      "learning_rate": 6.593252477554213e-08,
+      "loss": 0.8072,
+      "step": 7247
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2078274488449097,
+      "learning_rate": 6.558943281821772e-08,
+      "loss": 0.7917,
+      "step": 7248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2592824697494507,
+      "learning_rate": 6.524722996826827e-08,
+      "loss": 0.7953,
+      "step": 7249
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1637213230133057,
+      "learning_rate": 6.490591628735387e-08,
+      "loss": 0.8483,
+      "step": 7250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2355198860168457,
+      "learning_rate": 6.456549183697258e-08,
+      "loss": 0.766,
+      "step": 7251
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5729495882987976,
+      "learning_rate": 6.422595667846587e-08,
+      "loss": 1.0335,
+      "step": 7252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2944834232330322,
+      "learning_rate": 6.388731087301148e-08,
+      "loss": 0.7751,
+      "step": 7253
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2524449825286865,
+      "learning_rate": 6.354955448162947e-08,
+      "loss": 0.8549,
+      "step": 7254
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2336288690567017,
+      "learning_rate": 6.321268756517728e-08,
+      "loss": 0.7872,
+      "step": 7255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5721073150634766,
+      "learning_rate": 6.28767101843547e-08,
+      "loss": 1.0091,
+      "step": 7256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5721392631530762,
+      "learning_rate": 6.254162239969886e-08,
+      "loss": 1.0109,
+      "step": 7257
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2951949834823608,
+      "learning_rate": 6.22074242715881e-08,
+      "loss": 0.8806,
+      "step": 7258
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5020374059677124,
+      "learning_rate": 6.187411586023873e-08,
+      "loss": 0.8076,
+      "step": 7259
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.200406551361084,
+      "learning_rate": 6.154169722570991e-08,
+      "loss": 0.785,
+      "step": 7260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5656630396842957,
+      "learning_rate": 6.121016842789707e-08,
+      "loss": 1.0136,
+      "step": 7261
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2117160558700562,
+      "learning_rate": 6.087952952653742e-08,
+      "loss": 0.77,
+      "step": 7262
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2708740234375,
+      "learning_rate": 6.05497805812072e-08,
+      "loss": 0.7945,
+      "step": 7263
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2393962144851685,
+      "learning_rate": 6.022092165132166e-08,
+      "loss": 0.7827,
+      "step": 7264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.321839690208435,
+      "learning_rate": 5.989295279613727e-08,
+      "loss": 0.734,
+      "step": 7265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3411576747894287,
+      "learning_rate": 5.9565874074747896e-08,
+      "loss": 0.774,
+      "step": 7266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1642755270004272,
+      "learning_rate": 5.923968554608972e-08,
+      "loss": 0.788,
+      "step": 7267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5827069282531738,
+      "learning_rate": 5.891438726893628e-08,
+      "loss": 1.0061,
+      "step": 7268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5768369436264038,
+      "learning_rate": 5.858997930190069e-08,
+      "loss": 1.0066,
+      "step": 7269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5765435099601746,
+      "learning_rate": 5.82664617034373e-08,
+      "loss": 1.0012,
+      "step": 7270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5723947882652283,
+      "learning_rate": 5.794383453183949e-08,
+      "loss": 0.9924,
+      "step": 7271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2468634843826294,
+      "learning_rate": 5.7622097845239645e-08,
+      "loss": 0.8703,
+      "step": 7272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.26534903049469,
+      "learning_rate": 5.7301251701609714e-08,
+      "loss": 0.7978,
+      "step": 7273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3015180826187134,
+      "learning_rate": 5.6981296158761246e-08,
+      "loss": 0.7784,
+      "step": 7274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.280724287033081,
+      "learning_rate": 5.666223127434589e-08,
+      "loss": 0.8178,
+      "step": 7275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5818914175033569,
+      "learning_rate": 5.6344057105854336e-08,
+      "loss": 0.9951,
+      "step": 7276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5612097382545471,
+      "learning_rate": 5.6026773710616266e-08,
+      "loss": 0.9745,
+      "step": 7277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.566035270690918,
+      "learning_rate": 5.571038114580207e-08,
+      "loss": 1.0363,
+      "step": 7278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2552244663238525,
+      "learning_rate": 5.5394879468420594e-08,
+      "loss": 0.7867,
+      "step": 7279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5847835540771484,
+      "learning_rate": 5.5080268735320796e-08,
+      "loss": 1.0136,
+      "step": 7280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5715364217758179,
+      "learning_rate": 5.4766549003191225e-08,
+      "loss": 1.0081,
+      "step": 7281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.198485016822815,
+      "learning_rate": 5.445372032855833e-08,
+      "loss": 0.765,
+      "step": 7282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2019026279449463,
+      "learning_rate": 5.414178276779036e-08,
+      "loss": 0.8184,
+      "step": 7283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5672779083251953,
+      "learning_rate": 5.383073637709346e-08,
+      "loss": 1.0295,
+      "step": 7284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2943271398544312,
+      "learning_rate": 5.35205812125128e-08,
+      "loss": 0.8188,
+      "step": 7285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5721132755279541,
+      "learning_rate": 5.321131732993479e-08,
+      "loss": 1.0075,
+      "step": 7286
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5862195491790771,
+      "learning_rate": 5.290294478508429e-08,
+      "loss": 0.991,
+      "step": 7287
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2273567914962769,
+      "learning_rate": 5.259546363352408e-08,
+      "loss": 0.7934,
+      "step": 7288
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.335013508796692,
+      "learning_rate": 5.228887393065929e-08,
+      "loss": 0.801,
+      "step": 7289
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5680358409881592,
+      "learning_rate": 5.198317573173184e-08,
+      "loss": 1.0089,
+      "step": 7290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2643166780471802,
+      "learning_rate": 5.167836909182433e-08,
+      "loss": 0.8544,
+      "step": 7291
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3052467107772827,
+      "learning_rate": 5.137445406585784e-08,
+      "loss": 0.8537,
+      "step": 7292
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2388968467712402,
+      "learning_rate": 5.1071430708594107e-08,
+      "loss": 0.7594,
+      "step": 7293
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.27638578414917,
+      "learning_rate": 5.0769299074632796e-08,
+      "loss": 0.7988,
+      "step": 7294
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2595248222351074,
+      "learning_rate": 5.046805921841424e-08,
+      "loss": 0.7833,
+      "step": 7295
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5660293102264404,
+      "learning_rate": 5.016771119421726e-08,
+      "loss": 0.9992,
+      "step": 7296
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2719192504882812,
+      "learning_rate": 4.986825505615911e-08,
+      "loss": 0.8776,
+      "step": 7297
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2108063697814941,
+      "learning_rate": 4.956969085819829e-08,
+      "loss": 0.7794,
+      "step": 7298
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5625219345092773,
+      "learning_rate": 4.927201865413178e-08,
+      "loss": 1.0199,
+      "step": 7299
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.345430850982666,
+      "learning_rate": 4.8975238497595e-08,
+      "loss": 0.7712,
+      "step": 7300
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2351785898208618,
+      "learning_rate": 4.867935044206351e-08,
+      "loss": 0.8169,
+      "step": 7301
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3607367277145386,
+      "learning_rate": 4.8384354540851883e-08,
+      "loss": 0.8127,
+      "step": 7302
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5845731496810913,
+      "learning_rate": 4.809025084711483e-08,
+      "loss": 1.0127,
+      "step": 7303
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.225340485572815,
+      "learning_rate": 4.779703941384439e-08,
+      "loss": 0.8098,
+      "step": 7304
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3248968124389648,
+      "learning_rate": 4.750472029387332e-08,
+      "loss": 0.7939,
+      "step": 7305
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.259049892425537,
+      "learning_rate": 4.7213293539872805e-08,
+      "loss": 0.7849,
+      "step": 7306
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5718547105789185,
+      "learning_rate": 4.692275920435474e-08,
+      "loss": 1.011,
+      "step": 7307
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.283123254776001,
+      "learning_rate": 4.66331173396678e-08,
+      "loss": 0.7763,
+      "step": 7308
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.168209195137024,
+      "learning_rate": 4.634436799800135e-08,
+      "loss": 0.8365,
+      "step": 7309
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2649507522583008,
+      "learning_rate": 4.605651123138433e-08,
+      "loss": 0.8074,
+      "step": 7310
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5692506432533264,
+      "learning_rate": 4.576954709168413e-08,
+      "loss": 1.0159,
+      "step": 7311
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3196855783462524,
+      "learning_rate": 4.5483475630607175e-08,
+      "loss": 0.8023,
+      "step": 7312
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5905637145042419,
+      "learning_rate": 4.51982968996989e-08,
+      "loss": 1.0059,
+      "step": 7313
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2219421863555908,
+      "learning_rate": 4.4914010950344865e-08,
+      "loss": 0.7784,
+      "step": 7314
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2536460161209106,
+      "learning_rate": 4.46306178337691e-08,
+      "loss": 0.8275,
+      "step": 7315
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.263623595237732,
+      "learning_rate": 4.434811760103519e-08,
+      "loss": 0.7916,
+      "step": 7316
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2171646356582642,
+      "learning_rate": 4.40665103030441e-08,
+      "loss": 0.7811,
+      "step": 7317
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2592313289642334,
+      "learning_rate": 4.378579599053856e-08,
+      "loss": 0.8405,
+      "step": 7318
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5745171308517456,
+      "learning_rate": 4.350597471409923e-08,
+      "loss": 1.0232,
+      "step": 7319
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5710151791572571,
+      "learning_rate": 4.3227046524144665e-08,
+      "loss": 0.9868,
+      "step": 7320
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3351151943206787,
+      "learning_rate": 4.2949011470934665e-08,
+      "loss": 0.8541,
+      "step": 7321
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2121100425720215,
+      "learning_rate": 4.267186960456637e-08,
+      "loss": 0.7652,
+      "step": 7322
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3070858716964722,
+      "learning_rate": 4.2395620974976515e-08,
+      "loss": 0.8289,
+      "step": 7323
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2594412565231323,
+      "learning_rate": 4.212026563194138e-08,
+      "loss": 0.8142,
+      "step": 7324
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2963780164718628,
+      "learning_rate": 4.184580362507573e-08,
+      "loss": 0.7856,
+      "step": 7325
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5825981497764587,
+      "learning_rate": 4.1572235003833336e-08,
+      "loss": 1.0065,
+      "step": 7326
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5716639161109924,
+      "learning_rate": 4.1299559817508106e-08,
+      "loss": 0.9902,
+      "step": 7327
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2267851829528809,
+      "learning_rate": 4.1027778115231286e-08,
+      "loss": 0.7852,
+      "step": 7328
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.588279128074646,
+      "learning_rate": 4.075688994597315e-08,
+      "loss": 1.0159,
+      "step": 7329
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5562119483947754,
+      "learning_rate": 4.048689535854522e-08,
+      "loss": 0.9895,
+      "step": 7330
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2997158765792847,
+      "learning_rate": 4.021779440159579e-08,
+      "loss": 0.8377,
+      "step": 7331
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5750969648361206,
+      "learning_rate": 3.994958712361274e-08,
+      "loss": 1.0236,
+      "step": 7332
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5763407945632935,
+      "learning_rate": 3.968227357292354e-08,
+      "loss": 1.0315,
+      "step": 7333
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2451001405715942,
+      "learning_rate": 3.941585379769297e-08,
+      "loss": 0.808,
+      "step": 7334
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2703036069869995,
+      "learning_rate": 3.915032784592765e-08,
+      "loss": 0.8036,
+      "step": 7335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.278901219367981,
+      "learning_rate": 3.8885695765469855e-08,
+      "loss": 0.827,
+      "step": 7336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2496615648269653,
+      "learning_rate": 3.862195760400311e-08,
+      "loss": 0.8023,
+      "step": 7337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2442626953125,
+      "learning_rate": 3.835911340904885e-08,
+      "loss": 0.8003,
+      "step": 7338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3258166313171387,
+      "learning_rate": 3.809716322796808e-08,
+      "loss": 0.8202,
+      "step": 7339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2172470092773438,
+      "learning_rate": 3.783610710795971e-08,
+      "loss": 0.7935,
+      "step": 7340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5670519471168518,
+      "learning_rate": 3.757594509606277e-08,
+      "loss": 1.0032,
+      "step": 7341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.222243070602417,
+      "learning_rate": 3.731667723915367e-08,
+      "loss": 0.7621,
+      "step": 7342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.245413899421692,
+      "learning_rate": 3.705830358395002e-08,
+      "loss": 0.8089,
+      "step": 7343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.223643183708191,
+      "learning_rate": 3.68008241770057e-08,
+      "loss": 0.7898,
+      "step": 7344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5764699578285217,
+      "learning_rate": 3.654423906471527e-08,
+      "loss": 1.0179,
+      "step": 7345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3365821838378906,
+      "learning_rate": 3.6288548293311743e-08,
+      "loss": 0.8161,
+      "step": 7346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.234889030456543,
+      "learning_rate": 3.603375190886604e-08,
+      "loss": 0.7874,
+      "step": 7347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2119256258010864,
+      "learning_rate": 3.577984995728922e-08,
+      "loss": 0.7611,
+      "step": 7348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.204322099685669,
+      "learning_rate": 3.552684248433025e-08,
+      "loss": 0.7793,
+      "step": 7349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1909763813018799,
+      "learning_rate": 3.527472953557765e-08,
+      "loss": 0.7846,
+      "step": 7350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2082481384277344,
+      "learning_rate": 3.502351115645841e-08,
+      "loss": 0.7881,
+      "step": 7351
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.210262417793274,
+      "learning_rate": 3.477318739223801e-08,
+      "loss": 0.7547,
+      "step": 7352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2921315431594849,
+      "learning_rate": 3.4523758288022014e-08,
+      "loss": 0.8031,
+      "step": 7353
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.257389783859253,
+      "learning_rate": 3.4275223888752816e-08,
+      "loss": 0.8094,
+      "step": 7354
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3279714584350586,
+      "learning_rate": 3.402758423921293e-08,
+      "loss": 0.777,
+      "step": 7355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1985458135604858,
+      "learning_rate": 3.3780839384023326e-08,
+      "loss": 0.8055,
+      "step": 7356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2864985466003418,
+      "learning_rate": 3.3534989367643436e-08,
+      "loss": 0.7942,
+      "step": 7357
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3613841533660889,
+      "learning_rate": 3.329003423437227e-08,
+      "loss": 0.7825,
+      "step": 7358
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5699459910392761,
+      "learning_rate": 3.30459740283473e-08,
+      "loss": 1.007,
+      "step": 7359
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.210830807685852,
+      "learning_rate": 3.280280879354392e-08,
+      "loss": 0.7737,
+      "step": 7360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2135844230651855,
+      "learning_rate": 3.256053857377761e-08,
+      "loss": 0.7685,
+      "step": 7361
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2904773950576782,
+      "learning_rate": 3.231916341270125e-08,
+      "loss": 0.8425,
+      "step": 7362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2959522008895874,
+      "learning_rate": 3.207868335380726e-08,
+      "loss": 0.8104,
+      "step": 7363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3152580261230469,
+      "learning_rate": 3.183909844042599e-08,
+      "loss": 0.8229,
+      "step": 7364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.257918119430542,
+      "learning_rate": 3.1600408715727894e-08,
+      "loss": 0.7901,
+      "step": 7365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2799403667449951,
+      "learning_rate": 3.13626142227208e-08,
+      "loss": 0.7701,
+      "step": 7366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5911410450935364,
+      "learning_rate": 3.112571500425266e-08,
+      "loss": 1.0286,
+      "step": 7367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2277125120162964,
+      "learning_rate": 3.088971110300765e-08,
+      "loss": 0.8182,
+      "step": 7368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5671989917755127,
+      "learning_rate": 3.065460256151176e-08,
+      "loss": 1.0264,
+      "step": 7369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5758010149002075,
+      "learning_rate": 3.042038942212722e-08,
+      "loss": 1.0277,
+      "step": 7370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.265508770942688,
+      "learning_rate": 3.0187071727055816e-08,
+      "loss": 0.8632,
+      "step": 7371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2592209577560425,
+      "learning_rate": 2.995464951833726e-08,
+      "loss": 0.7747,
+      "step": 7372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4253451824188232,
+      "learning_rate": 2.9723122837851947e-08,
+      "loss": 0.794,
+      "step": 7373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2508262395858765,
+      "learning_rate": 2.9492491727316498e-08,
+      "loss": 0.7531,
+      "step": 7374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2449071407318115,
+      "learning_rate": 2.9262756228287668e-08,
+      "loss": 0.7852,
+      "step": 7375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.328805923461914,
+      "learning_rate": 2.9033916382160133e-08,
+      "loss": 0.7973,
+      "step": 7376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5800228118896484,
+      "learning_rate": 2.8805972230168122e-08,
+      "loss": 1.0246,
+      "step": 7377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5580467581748962,
+      "learning_rate": 2.8578923813382677e-08,
+      "loss": 1.0009,
+      "step": 7378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5452970266342163,
+      "learning_rate": 2.835277117271551e-08,
+      "loss": 0.8338,
+      "step": 7379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2124810218811035,
+      "learning_rate": 2.812751434891514e-08,
+      "loss": 0.7775,
+      "step": 7380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.229527235031128,
+      "learning_rate": 2.7903153382570214e-08,
+      "loss": 0.8108,
+      "step": 7381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2639389038085938,
+      "learning_rate": 2.7679688314106724e-08,
+      "loss": 0.7924,
+      "step": 7382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5668310523033142,
+      "learning_rate": 2.745711918379024e-08,
+      "loss": 1.0022,
+      "step": 7383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2359939813613892,
+      "learning_rate": 2.723544603172368e-08,
+      "loss": 0.8376,
+      "step": 7384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1835312843322754,
+      "learning_rate": 2.7014668897850094e-08,
+      "loss": 0.7687,
+      "step": 7385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2553972005844116,
+      "learning_rate": 2.679478782194933e-08,
+      "loss": 0.7381,
+      "step": 7386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2205698490142822,
+      "learning_rate": 2.657580284364192e-08,
+      "loss": 0.7699,
+      "step": 7387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.567348301410675,
+      "learning_rate": 2.6357714002384628e-08,
+      "loss": 0.9972,
+      "step": 7388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2084953784942627,
+      "learning_rate": 2.61405213374738e-08,
+      "loss": 0.7526,
+      "step": 7389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4592617750167847,
+      "learning_rate": 2.5924224888044803e-08,
+      "loss": 0.8083,
+      "step": 7390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3300961256027222,
+      "learning_rate": 2.5708824693071453e-08,
+      "loss": 0.8607,
+      "step": 7391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1704037189483643,
+      "learning_rate": 2.549432079136438e-08,
+      "loss": 0.7365,
+      "step": 7392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2565025091171265,
+      "learning_rate": 2.5280713221575436e-08,
+      "loss": 0.8051,
+      "step": 7393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5701477527618408,
+      "learning_rate": 2.5068002022192727e-08,
+      "loss": 0.9984,
+      "step": 7394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2079792022705078,
+      "learning_rate": 2.4856187231543372e-08,
+      "loss": 0.8184,
+      "step": 7395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1921701431274414,
+      "learning_rate": 2.464526888779406e-08,
+      "loss": 0.8175,
+      "step": 7396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2189719676971436,
+      "learning_rate": 2.4435247028948283e-08,
+      "loss": 0.7767,
+      "step": 7397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.575565755367279,
+      "learning_rate": 2.4226121692849658e-08,
+      "loss": 1.0032,
+      "step": 7398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5884968638420105,
+      "learning_rate": 2.4017892917178597e-08,
+      "loss": 1.005,
+      "step": 7399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.197354793548584,
+      "learning_rate": 2.3810560739455646e-08,
+      "loss": 0.7573,
+      "step": 7400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2131190299987793,
+      "learning_rate": 2.3604125197038698e-08,
+      "loss": 0.7696,
+      "step": 7401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5698307156562805,
+      "learning_rate": 2.3398586327124108e-08,
+      "loss": 1.0251,
+      "step": 7402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2683204412460327,
+      "learning_rate": 2.319394416674725e-08,
+      "loss": 0.8478,
+      "step": 7403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.179505705833435,
+      "learning_rate": 2.299019875278141e-08,
+      "loss": 0.7779,
+      "step": 7404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5733248591423035,
+      "learning_rate": 2.278735012193889e-08,
+      "loss": 1.0024,
+      "step": 7405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.268797516822815,
+      "learning_rate": 2.2585398310769335e-08,
+      "loss": 0.7641,
+      "step": 7406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.18450129032135,
+      "learning_rate": 2.2384343355661975e-08,
+      "loss": 0.7574,
+      "step": 7407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.30922532081604,
+      "learning_rate": 2.2184185292843385e-08,
+      "loss": 0.8231,
+      "step": 7408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5644564628601074,
+      "learning_rate": 2.198492415837916e-08,
+      "loss": 1.0259,
+      "step": 7409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2594791650772095,
+      "learning_rate": 2.1786559988173915e-08,
+      "loss": 0.8685,
+      "step": 7410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2909656763076782,
+      "learning_rate": 2.1589092817969614e-08,
+      "loss": 0.8375,
+      "step": 7411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2395645380020142,
+      "learning_rate": 2.1392522683346684e-08,
+      "loss": 0.8456,
+      "step": 7412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2072017192840576,
+      "learning_rate": 2.1196849619724012e-08,
+      "loss": 0.8585,
+      "step": 7413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2762051820755005,
+      "learning_rate": 2.1002073662358958e-08,
+      "loss": 0.7742,
+      "step": 7414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.231446623802185,
+      "learning_rate": 2.080819484634733e-08,
+      "loss": 0.7921,
+      "step": 7415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2810715436935425,
+      "learning_rate": 2.061521320662341e-08,
+      "loss": 0.8154,
+      "step": 7416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2837553024291992,
+      "learning_rate": 2.0423128777959933e-08,
+      "loss": 0.7668,
+      "step": 7417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1607189178466797,
+      "learning_rate": 2.0231941594966996e-08,
+      "loss": 0.7588,
+      "step": 7418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2788032293319702,
+      "learning_rate": 2.0041651692094266e-08,
+      "loss": 0.8114,
+      "step": 7419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.577252209186554,
+      "learning_rate": 1.9852259103628202e-08,
+      "loss": 1.0074,
+      "step": 7420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.290231704711914,
+      "learning_rate": 1.9663763863695952e-08,
+      "loss": 0.768,
+      "step": 7421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2781413793563843,
+      "learning_rate": 1.947616600625979e-08,
+      "loss": 0.8021,
+      "step": 7422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2442502975463867,
+      "learning_rate": 1.9289465565123787e-08,
+      "loss": 0.8198,
+      "step": 7423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.264359474182129,
+      "learning_rate": 1.910366257392715e-08,
+      "loss": 0.7832,
+      "step": 7424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2508891820907593,
+      "learning_rate": 1.8918757066150316e-08,
+      "loss": 0.7621,
+      "step": 7425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5686482191085815,
+      "learning_rate": 1.873474907510886e-08,
+      "loss": 1.0026,
+      "step": 7426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.269606113433838,
+      "learning_rate": 1.855163863395959e-08,
+      "loss": 0.8002,
+      "step": 7427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2404139041900635,
+      "learning_rate": 1.8369425775696114e-08,
+      "loss": 0.8051,
+      "step": 7428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5652773380279541,
+      "learning_rate": 1.8188110533149393e-08,
+      "loss": 0.9843,
+      "step": 7429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3323811292648315,
+      "learning_rate": 1.8007692938991072e-08,
+      "loss": 0.7808,
+      "step": 7430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5836760997772217,
+      "learning_rate": 1.7828173025729034e-08,
+      "loss": 1.001,
+      "step": 7431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5675433874130249,
+      "learning_rate": 1.764955082570963e-08,
+      "loss": 0.9902,
+      "step": 7432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5650202035903931,
+      "learning_rate": 1.7471826371119327e-08,
+      "loss": 1.0025,
+      "step": 7433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1962382793426514,
+      "learning_rate": 1.7294999693980298e-08,
+      "loss": 0.8216,
+      "step": 7434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1810628175735474,
+      "learning_rate": 1.7119070826154272e-08,
+      "loss": 0.7411,
+      "step": 7435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.14455246925354,
+      "learning_rate": 1.694403979934145e-08,
+      "loss": 0.7668,
+      "step": 7436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2096831798553467,
+      "learning_rate": 1.6769906645078826e-08,
+      "loss": 0.7538,
+      "step": 7437
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1864794492721558,
+      "learning_rate": 1.659667139474408e-08,
+      "loss": 0.7527,
+      "step": 7438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1818760633468628,
+      "learning_rate": 1.642433407955002e-08,
+      "loss": 0.7528,
+      "step": 7439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3048797845840454,
+      "learning_rate": 1.6252894730550696e-08,
+      "loss": 0.7924,
+      "step": 7440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.25279700756073,
+      "learning_rate": 1.608235337863584e-08,
+      "loss": 0.8504,
+      "step": 7441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2581151723861694,
+      "learning_rate": 1.5912710054535318e-08,
+      "loss": 0.8638,
+      "step": 7442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3387492895126343,
+      "learning_rate": 1.574396478881579e-08,
+      "loss": 0.8247,
+      "step": 7443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.244526982307434,
+      "learning_rate": 1.5576117611882934e-08,
+      "loss": 0.7933,
+      "step": 7444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3225382566452026,
+      "learning_rate": 1.5409168553980336e-08,
+      "loss": 0.7554,
+      "step": 7445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.276891827583313,
+      "learning_rate": 1.524311764518893e-08,
+      "loss": 0.8694,
+      "step": 7446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.214391827583313,
+      "learning_rate": 1.507796491542979e-08,
+      "loss": 0.8055,
+      "step": 7447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1927378177642822,
+      "learning_rate": 1.4913710394460213e-08,
+      "loss": 0.8148,
+      "step": 7448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2657909393310547,
+      "learning_rate": 1.475035411187653e-08,
+      "loss": 0.8206,
+      "step": 7449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.56931072473526,
+      "learning_rate": 1.458789609711353e-08,
+      "loss": 1.026,
+      "step": 7450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5658748745918274,
+      "learning_rate": 1.442633637944335e-08,
+      "loss": 1.025,
+      "step": 7451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2679450511932373,
+      "learning_rate": 1.4265674987976596e-08,
+      "loss": 0.8079,
+      "step": 7452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.22740638256073,
+      "learning_rate": 1.4105911951662332e-08,
+      "loss": 0.7538,
+      "step": 7453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2885193824768066,
+      "learning_rate": 1.394704729928753e-08,
+      "loss": 0.7966,
+      "step": 7454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2516032457351685,
+      "learning_rate": 1.3789081059477072e-08,
+      "loss": 0.7733,
+      "step": 7455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2321501970291138,
+      "learning_rate": 1.3632013260693744e-08,
+      "loss": 0.8078,
+      "step": 7456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2785968780517578,
+      "learning_rate": 1.3475843931239352e-08,
+      "loss": 0.801,
+      "step": 7457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.184374451637268,
+      "learning_rate": 1.3320573099253053e-08,
+      "loss": 0.8257,
+      "step": 7458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.313239336013794,
+      "learning_rate": 1.3166200792712469e-08,
+      "loss": 0.8122,
+      "step": 7459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2361701726913452,
+      "learning_rate": 1.3012727039433127e-08,
+      "loss": 0.7878,
+      "step": 7460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2063536643981934,
+      "learning_rate": 1.2860151867069014e-08,
+      "loss": 0.7893,
+      "step": 7461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2659755945205688,
+      "learning_rate": 1.2708475303111478e-08,
+      "loss": 0.7644,
+      "step": 7462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3014696836471558,
+      "learning_rate": 1.255769737489032e-08,
+      "loss": 0.7887,
+      "step": 7463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2250374555587769,
+      "learning_rate": 1.2407818109573811e-08,
+      "loss": 0.8066,
+      "step": 7464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2107435464859009,
+      "learning_rate": 1.2258837534167567e-08,
+      "loss": 0.8199,
+      "step": 7465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3414275646209717,
+      "learning_rate": 1.2110755675516228e-08,
+      "loss": 0.8252,
+      "step": 7466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2429908514022827,
+      "learning_rate": 1.1963572560301784e-08,
+      "loss": 0.8172,
+      "step": 7467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2852352857589722,
+      "learning_rate": 1.181728821504413e-08,
+      "loss": 0.7862,
+      "step": 7468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2544819116592407,
+      "learning_rate": 1.1671902666101631e-08,
+      "loss": 0.8254,
+      "step": 7469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.150689959526062,
+      "learning_rate": 1.152741593967055e-08,
+      "loss": 0.8048,
+      "step": 7470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2525148391723633,
+      "learning_rate": 1.1383828061785618e-08,
+      "loss": 0.761,
+      "step": 7471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5618834495544434,
+      "learning_rate": 1.1241139058318917e-08,
+      "loss": 1.0227,
+      "step": 7472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5797435641288757,
+      "learning_rate": 1.1099348954980992e-08,
+      "loss": 1.0013,
+      "step": 7473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2470424175262451,
+      "learning_rate": 1.0958457777320297e-08,
+      "loss": 0.7706,
+      "step": 7474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5774914026260376,
+      "learning_rate": 1.0818465550723745e-08,
+      "loss": 1.0187,
+      "step": 7475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2433699369430542,
+      "learning_rate": 1.0679372300414492e-08,
+      "loss": 0.8901,
+      "step": 7476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.348412036895752,
+      "learning_rate": 1.054117805145638e-08,
+      "loss": 0.8471,
+      "step": 7477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.24238920211792,
+      "learning_rate": 1.0403882828750044e-08,
+      "loss": 0.8246,
+      "step": 7478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.238544225692749,
+      "learning_rate": 1.0267486657032922e-08,
+      "loss": 0.8149,
+      "step": 7479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5623152256011963,
+      "learning_rate": 1.0131989560882572e-08,
+      "loss": 0.9756,
+      "step": 7480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5637221336364746,
+      "learning_rate": 9.9973915647128e-09,
+      "loss": 0.9764,
+      "step": 7481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2750135660171509,
+      "learning_rate": 9.863692692776982e-09,
+      "loss": 0.8066,
+      "step": 7482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.210561990737915,
+      "learning_rate": 9.730892969165296e-09,
+      "loss": 0.806,
+      "step": 7483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2846828699111938,
+      "learning_rate": 9.598992417805818e-09,
+      "loss": 0.7478,
+      "step": 7484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5620244145393372,
+      "learning_rate": 9.467991062465654e-09,
+      "loss": 1.0121,
+      "step": 7485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.574501097202301,
+      "learning_rate": 9.33788892674925e-09,
+      "loss": 1.0098,
+      "step": 7486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2718497514724731,
+      "learning_rate": 9.208686034098968e-09,
+      "loss": 0.7533,
+      "step": 7487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.191428780555725,
+      "learning_rate": 9.080382407795074e-09,
+      "loss": 0.7385,
+      "step": 7488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2509323358535767,
+      "learning_rate": 8.952978070956298e-09,
+      "loss": 0.8217,
+      "step": 7489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1964298486709595,
+      "learning_rate": 8.826473046538719e-09,
+      "loss": 0.719,
+      "step": 7490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2217882871627808,
+      "learning_rate": 8.700867357336885e-09,
+      "loss": 0.7771,
+      "step": 7491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5657175779342651,
+      "learning_rate": 8.576161025982688e-09,
+      "loss": 1.0182,
+      "step": 7492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.250524640083313,
+      "learning_rate": 8.45235407494649e-09,
+      "loss": 0.7787,
+      "step": 7493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1967315673828125,
+      "learning_rate": 8.32944652653711e-09,
+      "loss": 0.7902,
+      "step": 7494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2557663917541504,
+      "learning_rate": 8.207438402900169e-09,
+      "loss": 0.8459,
+      "step": 7495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.349968671798706,
+      "learning_rate": 8.086329726019192e-09,
+      "loss": 0.8196,
+      "step": 7496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5670687556266785,
+      "learning_rate": 7.966120517716725e-09,
+      "loss": 1.0125,
+      "step": 7497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2346229553222656,
+      "learning_rate": 7.84681079965266e-09,
+      "loss": 0.7592,
+      "step": 7498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2200120687484741,
+      "learning_rate": 7.728400593325357e-09,
+      "loss": 0.7277,
+      "step": 7499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5854438543319702,
+      "learning_rate": 7.610889920069419e-09,
+      "loss": 0.9962,
+      "step": 7500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5811635851860046,
+      "learning_rate": 7.494278801059018e-09,
+      "loss": 1.0086,
+      "step": 7501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2432746887207031,
+      "learning_rate": 7.378567257306235e-09,
+      "loss": 0.829,
+      "step": 7502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5710585713386536,
+      "learning_rate": 7.263755309659948e-09,
+      "loss": 0.994,
+      "step": 7503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1949986219406128,
+      "learning_rate": 7.149842978808053e-09,
+      "loss": 0.7484,
+      "step": 7504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.250617504119873,
+      "learning_rate": 7.036830285275242e-09,
+      "loss": 0.7719,
+      "step": 7505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.219609260559082,
+      "learning_rate": 6.924717249425228e-09,
+      "loss": 0.7908,
+      "step": 7506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2866255044937134,
+      "learning_rate": 6.813503891459072e-09,
+      "loss": 0.836,
+      "step": 7507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7244688272476196,
+      "learning_rate": 6.7031902314151905e-09,
+      "loss": 0.7646,
+      "step": 7508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2227554321289062,
+      "learning_rate": 6.593776289171016e-09,
+      "loss": 0.7644,
+      "step": 7509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3165929317474365,
+      "learning_rate": 6.48526208444189e-09,
+      "loss": 0.8346,
+      "step": 7510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5780598521232605,
+      "learning_rate": 6.377647636779393e-09,
+      "loss": 0.9967,
+      "step": 7511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3057618141174316,
+      "learning_rate": 6.270932965574128e-09,
+      "loss": 0.8241,
+      "step": 7512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2261968851089478,
+      "learning_rate": 6.1651180900546e-09,
+      "loss": 0.7993,
+      "step": 7513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1430755853652954,
+      "learning_rate": 6.060203029287781e-09,
+      "loss": 0.7397,
+      "step": 7514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2238177061080933,
+      "learning_rate": 5.956187802177438e-09,
+      "loss": 0.8482,
+      "step": 7515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2287901639938354,
+      "learning_rate": 5.8530724274652454e-09,
+      "loss": 0.7912,
+      "step": 7516
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5530630946159363,
+      "learning_rate": 5.750856923731341e-09,
+      "loss": 1.0083,
+      "step": 7517
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2869454622268677,
+      "learning_rate": 5.6495413093932136e-09,
+      "loss": 0.8055,
+      "step": 7518
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2950029373168945,
+      "learning_rate": 5.549125602706262e-09,
+      "loss": 0.8367,
+      "step": 7519
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5764557123184204,
+      "learning_rate": 5.449609821764346e-09,
+      "loss": 1.003,
+      "step": 7520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5738521218299866,
+      "learning_rate": 5.350993984498676e-09,
+      "loss": 1.0015,
+      "step": 7521
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.281487226486206,
+      "learning_rate": 5.253278108678372e-09,
+      "loss": 0.7415,
+      "step": 7522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5628918409347534,
+      "learning_rate": 5.156462211909907e-09,
+      "loss": 1.0046,
+      "step": 7523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2662287950515747,
+      "learning_rate": 5.0605463116387695e-09,
+      "loss": 0.8141,
+      "step": 7524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3430439233779907,
+      "learning_rate": 4.965530425147247e-09,
+      "loss": 0.85,
+      "step": 7525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2064447402954102,
+      "learning_rate": 4.8714145695560874e-09,
+      "loss": 0.812,
+      "step": 7526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.262203574180603,
+      "learning_rate": 4.778198761822839e-09,
+      "loss": 0.808,
+      "step": 7527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2531652450561523,
+      "learning_rate": 4.685883018744064e-09,
+      "loss": 0.7956,
+      "step": 7528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.338187336921692,
+      "learning_rate": 4.594467356953125e-09,
+      "loss": 0.8621,
+      "step": 7529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5733357071876526,
+      "learning_rate": 4.503951792922956e-09,
+      "loss": 1.0098,
+      "step": 7530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5680468678474426,
+      "learning_rate": 4.414336342962177e-09,
+      "loss": 1.0076,
+      "step": 7531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2434539794921875,
+      "learning_rate": 4.325621023218429e-09,
+      "loss": 0.7238,
+      "step": 7532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2319949865341187,
+      "learning_rate": 4.237805849677257e-09,
+      "loss": 0.8231,
+      "step": 7533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.561350405216217,
+      "learning_rate": 4.150890838161003e-09,
+      "loss": 1.0071,
+      "step": 7534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2553085088729858,
+      "learning_rate": 4.0648760043304755e-09,
+      "loss": 0.7995,
+      "step": 7535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 30.697439193725586,
+      "learning_rate": 3.979761363684942e-09,
+      "loss": 0.9967,
+      "step": 7536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5783773064613342,
+      "learning_rate": 3.8955469315604676e-09,
+      "loss": 1.0223,
+      "step": 7537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2998121976852417,
+      "learning_rate": 3.8122327231310266e-09,
+      "loss": 0.8781,
+      "step": 7538
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2188820838928223,
+      "learning_rate": 3.729818753408498e-09,
+      "loss": 0.7425,
+      "step": 7539
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2292420864105225,
+      "learning_rate": 3.648305037242672e-09,
+      "loss": 0.7439,
+      "step": 7540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5783214569091797,
+      "learning_rate": 3.567691589321798e-09,
+      "loss": 1.0096,
+      "step": 7541
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2736599445343018,
+      "learning_rate": 3.4879784241709236e-09,
+      "loss": 0.8462,
+      "step": 7542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5767547488212585,
+      "learning_rate": 3.4091655561524494e-09,
+      "loss": 1.0088,
+      "step": 7543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5659704208374023,
+      "learning_rate": 3.3312529994677934e-09,
+      "loss": 1.0107,
+      "step": 7544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5961639881134033,
+      "learning_rate": 3.2542407681562806e-09,
+      "loss": 1.0093,
+      "step": 7545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2835214138031006,
+      "learning_rate": 3.178128876092923e-09,
+      "loss": 0.7859,
+      "step": 7546
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5777270793914795,
+      "learning_rate": 3.102917336993416e-09,
+      "loss": 1.0294,
+      "step": 7547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2592167854309082,
+      "learning_rate": 3.0286061644091425e-09,
+      "loss": 0.8351,
+      "step": 7548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2406539916992188,
+      "learning_rate": 2.955195371729391e-09,
+      "loss": 0.8071,
+      "step": 7549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2178093194961548,
+      "learning_rate": 2.88268497218247e-09,
+      "loss": 0.7893,
+      "step": 7550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5636703968048096,
+      "learning_rate": 2.8110749788334836e-09,
+      "loss": 1.0337,
+      "step": 7551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2442477941513062,
+      "learning_rate": 2.7403654045854434e-09,
+      "loss": 0.7777,
+      "step": 7552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2698073387145996,
+      "learning_rate": 2.670556262178714e-09,
+      "loss": 0.776,
+      "step": 7553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2102394104003906,
+      "learning_rate": 2.6016475641921223e-09,
+      "loss": 0.7931,
+      "step": 7554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2469764947891235,
+      "learning_rate": 2.5336393230424028e-09,
+      "loss": 0.7886,
+      "step": 7555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2338835000991821,
+      "learning_rate": 2.4665315509836417e-09,
+      "loss": 0.7967,
+      "step": 7556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2787489891052246,
+      "learning_rate": 2.4003242601067235e-09,
+      "loss": 0.81,
+      "step": 7557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3157681226730347,
+      "learning_rate": 2.335017462342104e-09,
+      "loss": 0.7979,
+      "step": 7558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2297477722167969,
+      "learning_rate": 2.2706111694570377e-09,
+      "loss": 0.7516,
+      "step": 7559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.563044548034668,
+      "learning_rate": 2.2071053930561302e-09,
+      "loss": 1.0171,
+      "step": 7560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2638825178146362,
+      "learning_rate": 2.144500144582451e-09,
+      "loss": 0.7166,
+      "step": 7561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2285542488098145,
+      "learning_rate": 2.0827954353169755e-09,
+      "loss": 0.8057,
+      "step": 7562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2385952472686768,
+      "learning_rate": 2.0219912763769223e-09,
+      "loss": 0.7694,
+      "step": 7563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2523411512374878,
+      "learning_rate": 1.9620876787190825e-09,
+      "loss": 0.7636,
+      "step": 7564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3171743154525757,
+      "learning_rate": 1.903084653137044e-09,
+      "loss": 0.8059,
+      "step": 7565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.213744878768921,
+      "learning_rate": 1.8449822102623028e-09,
+      "loss": 0.7844,
+      "step": 7566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2473663091659546,
+      "learning_rate": 1.7877803605637068e-09,
+      "loss": 0.8244,
+      "step": 7567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3118540048599243,
+      "learning_rate": 1.7314791143485665e-09,
+      "loss": 0.8024,
+      "step": 7568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2450971603393555,
+      "learning_rate": 1.676078481761545e-09,
+      "loss": 0.759,
+      "step": 7569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.244563102722168,
+      "learning_rate": 1.6215784727846573e-09,
+      "loss": 0.8561,
+      "step": 7570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5752342343330383,
+      "learning_rate": 1.5679790972383813e-09,
+      "loss": 0.9963,
+      "step": 7571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2363054752349854,
+      "learning_rate": 1.515280364780547e-09,
+      "loss": 0.8494,
+      "step": 7572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.293895959854126,
+      "learning_rate": 1.4634822849063368e-09,
+      "loss": 0.8462,
+      "step": 7573
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1304785013198853,
+      "learning_rate": 1.4125848669488406e-09,
+      "loss": 0.7376,
+      "step": 7574
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2341058254241943,
+      "learning_rate": 1.362588120079611e-09,
+      "loss": 0.8291,
+      "step": 7575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2667105197906494,
+      "learning_rate": 1.3134920533069972e-09,
+      "loss": 0.7907,
+      "step": 7576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5856973528862,
+      "learning_rate": 1.2652966754772567e-09,
+      "loss": 1.0063,
+      "step": 7577
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2880395650863647,
+      "learning_rate": 1.2180019952751087e-09,
+      "loss": 0.8473,
+      "step": 7578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2247790098190308,
+      "learning_rate": 1.1716080212215153e-09,
+      "loss": 0.8196,
+      "step": 7579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2127047777175903,
+      "learning_rate": 1.1261147616764557e-09,
+      "loss": 0.7697,
+      "step": 7580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2090981006622314,
+      "learning_rate": 1.0815222248367064e-09,
+      "loss": 0.7765,
+      "step": 7581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5716981291770935,
+      "learning_rate": 1.0378304187380617e-09,
+      "loss": 1.0108,
+      "step": 7582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5507031679153442,
+      "learning_rate": 9.95039351252558e-10,
+      "loss": 0.9916,
+      "step": 7583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5631796717643738,
+      "learning_rate": 9.53149030090139e-10,
+      "loss": 1.0147,
+      "step": 7584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2075209617614746,
+      "learning_rate": 9.121594627992114e-10,
+      "loss": 0.8007,
+      "step": 7585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2373733520507812,
+      "learning_rate": 8.720706567655335e-10,
+      "loss": 0.8086,
+      "step": 7586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.213586688041687,
+      "learning_rate": 8.328826192127715e-10,
+      "loss": 0.7494,
+      "step": 7587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.217940330505371,
+      "learning_rate": 7.945953572013887e-10,
+      "loss": 0.8387,
+      "step": 7588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3264856338500977,
+      "learning_rate": 7.572088776308661e-10,
+      "loss": 0.8143,
+      "step": 7589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5820049047470093,
+      "learning_rate": 7.207231872369269e-10,
+      "loss": 1.0122,
+      "step": 7590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.276576280593872,
+      "learning_rate": 6.851382925943118e-10,
+      "loss": 0.816,
+      "step": 7591
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5657370686531067,
+      "learning_rate": 6.504542001145586e-10,
+      "loss": 1.0059,
+      "step": 7592
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3205327987670898,
+      "learning_rate": 6.16670916047668e-10,
+      "loss": 0.811,
+      "step": 7593
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5688636302947998,
+      "learning_rate": 5.837884464804377e-10,
+      "loss": 1.0011,
+      "step": 7594
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.562196671962738,
+      "learning_rate": 5.51806797338128e-10,
+      "loss": 1.0208,
+      "step": 7595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.284239649772644,
+      "learning_rate": 5.207259743833515e-10,
+      "loss": 0.7793,
+      "step": 7596
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2191030979156494,
+      "learning_rate": 4.905459832160731e-10,
+      "loss": 0.8804,
+      "step": 7597
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.337931513786316,
+      "learning_rate": 4.612668292741651e-10,
+      "loss": 0.819,
+      "step": 7598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5710059404373169,
+      "learning_rate": 4.328885178339626e-10,
+      "loss": 1.0422,
+      "step": 7599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2516902685165405,
+      "learning_rate": 4.054110540085976e-10,
+      "loss": 0.8258,
+      "step": 7600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3089832067489624,
+      "learning_rate": 3.7883444274855466e-10,
+      "loss": 0.8113,
+      "step": 7601
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.326658844947815,
+      "learning_rate": 3.5315868884333583e-10,
+      "loss": 0.7669,
+      "step": 7602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.260116696357727,
+      "learning_rate": 3.283837969186854e-10,
+      "loss": 0.7838,
+      "step": 7603
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2010408639907837,
+      "learning_rate": 3.0450977143881013e-10,
+      "loss": 0.7991,
+      "step": 7604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2785991430282593,
+      "learning_rate": 2.815366167058242e-10,
+      "loss": 0.812,
+      "step": 7605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.249447226524353,
+      "learning_rate": 2.5946433685863914e-10,
+      "loss": 0.7922,
+      "step": 7606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5735354423522949,
+      "learning_rate": 2.382929358746289e-10,
+      "loss": 1.019,
+      "step": 7607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2942782640457153,
+      "learning_rate": 2.1802241756907483e-10,
+      "loss": 0.7687,
+      "step": 7608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5489591360092163,
+      "learning_rate": 1.9865278559350054e-10,
+      "loss": 1.001,
+      "step": 7609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5722007155418396,
+      "learning_rate": 1.801840434384472e-10,
+      "loss": 0.9909,
+      "step": 7610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2175540924072266,
+      "learning_rate": 1.6261619443125321e-10,
+      "loss": 0.8332,
+      "step": 7611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5746839046478271,
+      "learning_rate": 1.4594924173827464e-10,
+      "loss": 1.0352,
+      "step": 7612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3120050430297852,
+      "learning_rate": 1.3018318836210962e-10,
+      "loss": 0.7901,
+      "step": 7613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.234095811843872,
+      "learning_rate": 1.1531803714381895e-10,
+      "loss": 0.7869,
+      "step": 7614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.204640507698059,
+      "learning_rate": 1.013537907612605e-10,
+      "loss": 0.7552,
+      "step": 7615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2449787855148315,
+      "learning_rate": 8.82904517313099e-11,
+      "loss": 0.7962,
+      "step": 7616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2425694465637207,
+      "learning_rate": 7.612802240708484e-11,
+      "loss": 0.7641,
+      "step": 7617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2324742078781128,
+      "learning_rate": 6.486650498127578e-11,
+      "loss": 0.8314,
+      "step": 7618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2165147066116333,
+      "learning_rate": 5.4505901481705095e-11,
+      "loss": 0.7341,
+      "step": 7619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2414878606796265,
+      "learning_rate": 4.50462137757679e-11,
+      "loss": 0.7865,
+      "step": 7620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2318317890167236,
+      "learning_rate": 3.6487443567656546e-11,
+      "loss": 0.8309,
+      "step": 7621
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.192772626876831,
+      "learning_rate": 2.882959240058103e-11,
+      "loss": 0.7884,
+      "step": 7622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2602559328079224,
+      "learning_rate": 2.2072661652883243e-11,
+      "loss": 0.791,
+      "step": 7623
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2178747653961182,
+      "learning_rate": 1.621665254303295e-11,
+      "loss": 0.7941,
+      "step": 7624
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5797600150108337,
+      "learning_rate": 1.126156612629714e-11,
+      "loss": 0.9982,
+      "step": 7625
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.202384114265442,
+      "learning_rate": 7.207403294740012e-12,
+      "loss": 0.8293,
+      "step": 7626
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5844014883041382,
+      "learning_rate": 4.054164779443426e-12,
+      "loss": 1.0122,
+      "step": 7627
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2353763580322266,
+      "learning_rate": 1.8018511482864597e-12,
+      "loss": 0.8197,
+      "step": 7628
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2734051942825317,
+      "learning_rate": 4.504628076107409e-13,
+      "loss": 0.8427,
+      "step": 7629
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9137070178985596,
+      "learning_rate": 0.0,
+      "loss": 0.9041,
+      "step": 7630
+    },
+    {
+      "epoch": 1.0,
+      "step": 7630,
+      "total_flos": 8.506348550093799e+18,
+      "train_loss": 0.5919686249246134,
+      "train_runtime": 142308.7793,
+      "train_samples_per_second": 27.451,
+      "train_steps_per_second": 0.054
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7630,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1200,
+  "total_flos": 8.506348550093799e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}